diff --git a/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa_shuffled-streaming_baseline_tmppos_lora_agg_fc4_s2048/README.md b/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa_shuffled-streaming_baseline_tmppos_lora_agg_fc4_s2048/README.md
new file mode 100644
index 0000000000000000000000000000000000000000..d4576fe074287232d3836bf69c21d3f2593290d9
--- /dev/null
+++ b/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa_shuffled-streaming_baseline_tmppos_lora_agg_fc4_s2048/README.md
@@ -0,0 +1,9 @@
+---
+library_name: peft
+---
+## Training procedure
+
+### Framework versions
+
+
+- PEFT 0.4.0
diff --git a/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa_shuffled-streaming_baseline_tmppos_lora_agg_fc4_s2048/adapter_config.json b/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa_shuffled-streaming_baseline_tmppos_lora_agg_fc4_s2048/adapter_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..2dbc82830c03141f1757ac3cc99367792fe4b0b8
--- /dev/null
+++ b/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa_shuffled-streaming_baseline_tmppos_lora_agg_fc4_s2048/adapter_config.json
@@ -0,0 +1,26 @@
+{
+  "auto_mapping": null,
+  "base_model_name_or_path": "/data/wiedmann/hub/models--lmms-lab--llava-onevision-qwen2-7b-ov/snapshots/0b07bf7565e244cf4f39982249eafe8cd799d6dd",
+  "bias": "none",
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "lora_alpha": 256,
+  "lora_dropout": 0.05,
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "r": 128,
+  "revision": null,
+  "target_modules": [
+    "o_proj",
+    "q_proj",
+    "up_proj",
+    "gate_proj",
+    "k_proj",
+    "v_proj",
+    "down_proj"
+  ],
+  "task_type": "CAUSAL_LM"
+}
\ No newline at end of file
diff --git a/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa_shuffled-streaming_baseline_tmppos_lora_agg_fc4_s2048/adapter_model.bin b/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa_shuffled-streaming_baseline_tmppos_lora_agg_fc4_s2048/adapter_model.bin
new file mode 100644
index 0000000000000000000000000000000000000000..e2fb386c6c0b052641a0143a4941ff341b646cd1
--- /dev/null
+++ b/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa_shuffled-streaming_baseline_tmppos_lora_agg_fc4_s2048/adapter_model.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b714e05bef837c53d7713a34bd54c02b4bdb2dc6867154d7fa4aa88bce735902
+size 692127130
diff --git a/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa_shuffled-streaming_baseline_tmppos_lora_agg_fc4_s2048/config.json b/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa_shuffled-streaming_baseline_tmppos_lora_agg_fc4_s2048/config.json
new file mode 100644
index 0000000000000000000000000000000000000000..a48d68518cce4598c62f52102a6c0a9e1adc9faa
--- /dev/null
+++ b/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa_shuffled-streaming_baseline_tmppos_lora_agg_fc4_s2048/config.json
@@ -0,0 +1,68 @@
+{
+  "_name_or_path": "/data/wiedmann/hub/models--lmms-lab--llava-onevision-qwen2-7b-ov/snapshots/0b07bf7565e244cf4f39982249eafe8cd799d6dd",
+  "add_faster_video": false,
+  "add_time_instruction": false,
+  "architectures": [
+    "LlavaQwenForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 151643,
+  "eos_token_id": 151645,
+  "faster_token_stride": 10,
+  "force_sample": false,
+  "hidden_act": "silu",
+  "hidden_size": 3584,
+  "image_aspect_ratio": "square",
+  "image_crop_resolution": null,
+  "image_grid_pinpoints": null,
+  "image_split_resolution": null,
+  "image_token_index": 151646,
+  "initializer_range": 0.02,
+  "intermediate_size": 18944,
+  "max_position_embeddings": 32768,
+  "max_window_layers": 28,
+  "mm_hidden_size": 1152,
+  "mm_newline_position": "no_token",
+  "mm_patch_merge_type": "spatial_unpad",
+  "mm_projector_lr": null,
+  "mm_projector_type": "mlp2x_gelu",
+  "mm_resampler_type": "streaming_agg",
+  "mm_spatial_pool_mode": "bilinear",
+  "mm_spatial_pool_stride": null,
+  "mm_streaming_frames_per_chunk": 4,
+  "mm_streaming_input_dim": 1152,
+  "mm_streaming_mlp_ratio": 4.0,
+  "mm_streaming_num_heads": 8,
+  "mm_streaming_num_layers": 4,
+  "mm_streaming_num_state_tokens": 2048,
+  "mm_streaming_patches_per_frame": 729,
+  "mm_streaming_state_dim": 1152,
+  "mm_streaming_vision_chunk_size": 8,
+  "mm_tunable_parts": "mm_mlp_adapter,mm_vision_resampler",
+  "mm_use_im_patch_token": false,
+  "mm_use_im_start_end": false,
+  "mm_vision_select_feature": "patch",
+  "mm_vision_select_layer": -2,
+  "mm_vision_tower": "google/siglip-so400m-patch14-384",
+  "mm_vision_tower_lr": null,
+  "model_type": "llava_qwen",
+  "num_attention_heads": 28,
+  "num_hidden_layers": 28,
+  "num_key_value_heads": 4,
+  "pos_skipping_range": 4096,
+  "rms_norm_eps": 1e-06,
+  "rope_scaling": null,
+  "rope_theta": 1000000.0,
+  "sliding_window": 131072,
+  "tie_word_embeddings": false,
+  "tokenizer_model_max_length": 8192,
+  "tokenizer_padding_side": "right",
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.40.0.dev0",
+  "use_cache": true,
+  "use_mm_proj": true,
+  "use_pos_skipping": false,
+  "use_sliding_window": false,
+  "vision_tower_pretrained": null,
+  "vocab_size": 152064
+}
diff --git a/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa_shuffled-streaming_baseline_tmppos_lora_agg_fc4_s2048/generation_config.json b/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa_shuffled-streaming_baseline_tmppos_lora_agg_fc4_s2048/generation_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..19a297221acb87418d4388a3decef2282c6d7316
--- /dev/null
+++ b/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa_shuffled-streaming_baseline_tmppos_lora_agg_fc4_s2048/generation_config.json
@@ -0,0 +1,14 @@
+{
+  "bos_token_id": 151643,
+  "do_sample": true,
+  "eos_token_id": [
+    151645,
+    151643
+  ],
+  "pad_token_id": 151643,
+  "repetition_penalty": 1.05,
+  "temperature": 0.7,
+  "top_k": 20,
+  "top_p": 0.8,
+  "transformers_version": "4.40.0.dev0"
+}
diff --git a/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa_shuffled-streaming_baseline_tmppos_lora_agg_fc4_s2048/non_lora_trainables.bin b/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa_shuffled-streaming_baseline_tmppos_lora_agg_fc4_s2048/non_lora_trainables.bin
new file mode 100644
index 0000000000000000000000000000000000000000..12e52dee281380ea258794f08d52063b12ab90cc
--- /dev/null
+++ b/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa_shuffled-streaming_baseline_tmppos_lora_agg_fc4_s2048/non_lora_trainables.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:718d2cfa2c338b8ddd887997515ab051bcfe3edac6b659caccfb936b0e00504f
+size 211441788
diff --git a/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa_shuffled-streaming_baseline_tmppos_lora_agg_fc4_s2048/trainer_state.json b/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa_shuffled-streaming_baseline_tmppos_lora_agg_fc4_s2048/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..1a6d2e0804929d01aa8759d36a82bae11f188ac7
--- /dev/null
+++ b/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa_shuffled-streaming_baseline_tmppos_lora_agg_fc4_s2048/trainer_state.json
@@ -0,0 +1,7495 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9996446760630108,
+  "eval_steps": 100,
+  "global_step": 1055,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 30.804296493530273,
+      "learning_rate": 3.125e-06,
+      "loss": 1.2065,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 23.284700393676758,
+      "learning_rate": 6.25e-06,
+      "loss": 1.0983,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 27.227680206298828,
+      "learning_rate": 9.375000000000001e-06,
+      "loss": 1.1279,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 11.16970443725586,
+      "learning_rate": 1.25e-05,
+      "loss": 1.0219,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.86640453338623,
+      "learning_rate": 1.5625e-05,
+      "loss": 0.7726,
+      "step": 5
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.84248685836792,
+      "learning_rate": 1.8750000000000002e-05,
+      "loss": 0.6763,
+      "step": 6
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.7346138954162598,
+      "learning_rate": 2.1875e-05,
+      "loss": 0.5752,
+      "step": 7
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9248408079147339,
+      "learning_rate": 2.5e-05,
+      "loss": 0.4705,
+      "step": 8
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0167689323425293,
+      "learning_rate": 2.8125000000000003e-05,
+      "loss": 0.3754,
+      "step": 9
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8291449546813965,
+      "learning_rate": 3.125e-05,
+      "loss": 0.3479,
+      "step": 10
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.909376859664917,
+      "learning_rate": 3.4375e-05,
+      "loss": 0.3093,
+      "step": 11
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6860577464103699,
+      "learning_rate": 3.7500000000000003e-05,
+      "loss": 0.2743,
+      "step": 12
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7518047094345093,
+      "learning_rate": 4.0625000000000005e-05,
+      "loss": 0.3408,
+      "step": 13
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7100327014923096,
+      "learning_rate": 4.375e-05,
+      "loss": 0.3462,
+      "step": 14
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8330450654029846,
+      "learning_rate": 4.6875e-05,
+      "loss": 0.5261,
+      "step": 15
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6445046067237854,
+      "learning_rate": 5e-05,
+      "loss": 0.3798,
+      "step": 16
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.773724377155304,
+      "learning_rate": 5.3125000000000004e-05,
+      "loss": 0.4469,
+      "step": 17
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6619732975959778,
+      "learning_rate": 5.6250000000000005e-05,
+      "loss": 0.2812,
+      "step": 18
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5613172054290771,
+      "learning_rate": 5.9375e-05,
+      "loss": 0.3588,
+      "step": 19
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5960156917572021,
+      "learning_rate": 6.25e-05,
+      "loss": 0.2513,
+      "step": 20
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.585588812828064,
+      "learning_rate": 6.562500000000001e-05,
+      "loss": 0.3698,
+      "step": 21
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5156756043434143,
+      "learning_rate": 6.875e-05,
+      "loss": 0.175,
+      "step": 22
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.49335452914237976,
+      "learning_rate": 7.1875e-05,
+      "loss": 0.299,
+      "step": 23
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7429764270782471,
+      "learning_rate": 7.500000000000001e-05,
+      "loss": 0.4715,
+      "step": 24
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5946854948997498,
+      "learning_rate": 7.8125e-05,
+      "loss": 0.2294,
+      "step": 25
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6120976209640503,
+      "learning_rate": 8.125000000000001e-05,
+      "loss": 0.3846,
+      "step": 26
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6154862642288208,
+      "learning_rate": 8.4375e-05,
+      "loss": 0.365,
+      "step": 27
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5157431364059448,
+      "learning_rate": 8.75e-05,
+      "loss": 0.3275,
+      "step": 28
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5386487245559692,
+      "learning_rate": 9.062500000000001e-05,
+      "loss": 0.3336,
+      "step": 29
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.526266872882843,
+      "learning_rate": 9.375e-05,
+      "loss": 0.2939,
+      "step": 30
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5345079302787781,
+      "learning_rate": 9.687500000000001e-05,
+      "loss": 0.3086,
+      "step": 31
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6255465745925903,
+      "learning_rate": 0.0001,
+      "loss": 0.332,
+      "step": 32
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6006800532341003,
+      "learning_rate": 9.999976423021617e-05,
+      "loss": 0.3782,
+      "step": 33
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5360583066940308,
+      "learning_rate": 9.999905692308813e-05,
+      "loss": 0.3754,
+      "step": 34
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6360620260238647,
+      "learning_rate": 9.999787808528638e-05,
+      "loss": 0.4859,
+      "step": 35
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.44070759415626526,
+      "learning_rate": 9.999622772792828e-05,
+      "loss": 0.2861,
+      "step": 36
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.45899343490600586,
+      "learning_rate": 9.9994105866578e-05,
+      "loss": 0.3691,
+      "step": 37
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5495193004608154,
+      "learning_rate": 9.999151252124639e-05,
+      "loss": 0.37,
+      "step": 38
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.452613890171051,
+      "learning_rate": 9.998844771639073e-05,
+      "loss": 0.2928,
+      "step": 39
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3938613831996918,
+      "learning_rate": 9.998491148091457e-05,
+      "loss": 0.3118,
+      "step": 40
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5872858166694641,
+      "learning_rate": 9.99809038481674e-05,
+      "loss": 0.4285,
+      "step": 41
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4832764267921448,
+      "learning_rate": 9.997642485594436e-05,
+      "loss": 0.3354,
+      "step": 42
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.43297770619392395,
+      "learning_rate": 9.99714745464859e-05,
+      "loss": 0.3512,
+      "step": 43
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.454502671957016,
+      "learning_rate": 9.996605296647737e-05,
+      "loss": 0.2908,
+      "step": 44
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.529533863067627,
+      "learning_rate": 9.996016016704854e-05,
+      "loss": 0.5052,
+      "step": 45
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.41448667645454407,
+      "learning_rate": 9.995379620377318e-05,
+      "loss": 0.2739,
+      "step": 46
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.38955628871917725,
+      "learning_rate": 9.99469611366685e-05,
+      "loss": 0.2549,
+      "step": 47
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.45690223574638367,
+      "learning_rate": 9.993965503019457e-05,
+      "loss": 0.4029,
+      "step": 48
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.41869643330574036,
+      "learning_rate": 9.99318779532538e-05,
+      "loss": 0.3377,
+      "step": 49
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.47526320815086365,
+      "learning_rate": 9.992362997919016e-05,
+      "loss": 0.3284,
+      "step": 50
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4810812473297119,
+      "learning_rate": 9.991491118578856e-05,
+      "loss": 0.3745,
+      "step": 51
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4266604781150818,
+      "learning_rate": 9.990572165527413e-05,
+      "loss": 0.3366,
+      "step": 52
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5009778141975403,
+      "learning_rate": 9.98960614743114e-05,
+      "loss": 0.4346,
+      "step": 53
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5063421726226807,
+      "learning_rate": 9.988593073400354e-05,
+      "loss": 0.4424,
+      "step": 54
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4510231018066406,
+      "learning_rate": 9.987532952989145e-05,
+      "loss": 0.3745,
+      "step": 55
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4165729582309723,
+      "learning_rate": 9.986425796195287e-05,
+      "loss": 0.2746,
+      "step": 56
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4778216779232025,
+      "learning_rate": 9.985271613460144e-05,
+      "loss": 0.4199,
+      "step": 57
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.38626334071159363,
+      "learning_rate": 9.984070415668574e-05,
+      "loss": 0.2438,
+      "step": 58
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.48320698738098145,
+      "learning_rate": 9.982822214148821e-05,
+      "loss": 0.4163,
+      "step": 59
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.43068376183509827,
+      "learning_rate": 9.981527020672413e-05,
+      "loss": 0.3129,
+      "step": 60
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.47523853182792664,
+      "learning_rate": 9.980184847454052e-05,
+      "loss": 0.3959,
+      "step": 61
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.43448108434677124,
+      "learning_rate": 9.978795707151492e-05,
+      "loss": 0.3093,
+      "step": 62
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.40242329239845276,
+      "learning_rate": 9.977359612865423e-05,
+      "loss": 0.3048,
+      "step": 63
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.44845131039619446,
+      "learning_rate": 9.975876578139355e-05,
+      "loss": 0.3456,
+      "step": 64
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4786364436149597,
+      "learning_rate": 9.974346616959476e-05,
+      "loss": 0.4488,
+      "step": 65
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4046257734298706,
+      "learning_rate": 9.972769743754531e-05,
+      "loss": 0.2914,
+      "step": 66
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3762134909629822,
+      "learning_rate": 9.971145973395684e-05,
+      "loss": 0.2452,
+      "step": 67
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3934119641780853,
+      "learning_rate": 9.969475321196373e-05,
+      "loss": 0.2175,
+      "step": 68
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.428158164024353,
+      "learning_rate": 9.96775780291217e-05,
+      "loss": 0.3011,
+      "step": 69
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4342409074306488,
+      "learning_rate": 9.965993434740633e-05,
+      "loss": 0.2891,
+      "step": 70
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5298842787742615,
+      "learning_rate": 9.96418223332115e-05,
+      "loss": 0.3521,
+      "step": 71
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.42322099208831787,
+      "learning_rate": 9.962324215734782e-05,
+      "loss": 0.2079,
+      "step": 72
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5400073528289795,
+      "learning_rate": 9.960419399504106e-05,
+      "loss": 0.3937,
+      "step": 73
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4282865822315216,
+      "learning_rate": 9.958467802593046e-05,
+      "loss": 0.2715,
+      "step": 74
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4480207860469818,
+      "learning_rate": 9.956469443406706e-05,
+      "loss": 0.3131,
+      "step": 75
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3894388973712921,
+      "learning_rate": 9.954424340791196e-05,
+      "loss": 0.2954,
+      "step": 76
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.49804937839508057,
+      "learning_rate": 9.952332514033447e-05,
+      "loss": 0.351,
+      "step": 77
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.41495615243911743,
+      "learning_rate": 9.950193982861047e-05,
+      "loss": 0.3886,
+      "step": 78
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4607909619808197,
+      "learning_rate": 9.948008767442034e-05,
+      "loss": 0.3415,
+      "step": 79
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3959217965602875,
+      "learning_rate": 9.94577688838472e-05,
+      "loss": 0.2881,
+      "step": 80
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.41608935594558716,
+      "learning_rate": 9.943498366737487e-05,
+      "loss": 0.3601,
+      "step": 81
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4240490794181824,
+      "learning_rate": 9.941173223988602e-05,
+      "loss": 0.3083,
+      "step": 82
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4540845453739166,
+      "learning_rate": 9.938801482065998e-05,
+      "loss": 0.3354,
+      "step": 83
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4498790502548218,
+      "learning_rate": 9.93638316333708e-05,
+      "loss": 0.3859,
+      "step": 84
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.42697852849960327,
+      "learning_rate": 9.933918290608509e-05,
+      "loss": 0.323,
+      "step": 85
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.43311402201652527,
+      "learning_rate": 9.93140688712598e-05,
+      "loss": 0.345,
+      "step": 86
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4836309850215912,
+      "learning_rate": 9.928848976574019e-05,
+      "loss": 0.4193,
+      "step": 87
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.400007963180542,
+      "learning_rate": 9.926244583075748e-05,
+      "loss": 0.3276,
+      "step": 88
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4656391441822052,
+      "learning_rate": 9.923593731192655e-05,
+      "loss": 0.3709,
+      "step": 89
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3918564021587372,
+      "learning_rate": 9.920896445924373e-05,
+      "loss": 0.2899,
+      "step": 90
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.37206435203552246,
+      "learning_rate": 9.918152752708438e-05,
+      "loss": 0.1981,
+      "step": 91
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4430893659591675,
+      "learning_rate": 9.915362677420044e-05,
+      "loss": 0.3654,
+      "step": 92
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.44486117362976074,
+      "learning_rate": 9.912526246371814e-05,
+      "loss": 0.4276,
+      "step": 93
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.47260475158691406,
+      "learning_rate": 9.909643486313533e-05,
+      "loss": 0.3608,
+      "step": 94
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4907275438308716,
+      "learning_rate": 9.906714424431913e-05,
+      "loss": 0.3866,
+      "step": 95
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.46724241971969604,
+      "learning_rate": 9.903739088350325e-05,
+      "loss": 0.2709,
+      "step": 96
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.47604647278785706,
+      "learning_rate": 9.900717506128541e-05,
+      "loss": 0.3038,
+      "step": 97
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4240775406360626,
+      "learning_rate": 9.897649706262473e-05,
+      "loss": 0.3064,
+      "step": 98
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5179862380027771,
+      "learning_rate": 9.894535717683902e-05,
+      "loss": 0.3739,
+      "step": 99
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4301639795303345,
+      "learning_rate": 9.891375569760204e-05,
+      "loss": 0.342,
+      "step": 100
+    },
+    {
+      "epoch": 0.09,
+      "eval_loss": 0.3124053478240967,
+      "eval_runtime": 371.4195,
+      "eval_samples_per_second": 0.918,
+      "eval_steps_per_second": 0.232,
+      "step": 100
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.46046969294548035,
+      "learning_rate": 9.888169292294076e-05,
+      "loss": 0.2961,
+      "step": 101
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.45049476623535156,
+      "learning_rate": 9.884916915523251e-05,
+      "loss": 0.2909,
+      "step": 102
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.521728515625,
+      "learning_rate": 9.881618470120215e-05,
+      "loss": 0.3543,
+      "step": 103
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.47198793292045593,
+      "learning_rate": 9.87827398719192e-05,
+      "loss": 0.3814,
+      "step": 104
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.46089521050453186,
+      "learning_rate": 9.874883498279484e-05,
+      "loss": 0.2694,
+      "step": 105
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4399769902229309,
+      "learning_rate": 9.871447035357903e-05,
+      "loss": 0.3216,
+      "step": 106
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4072099030017853,
+      "learning_rate": 9.867964630835743e-05,
+      "loss": 0.2752,
+      "step": 107
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3770722448825836,
+      "learning_rate": 9.864436317554831e-05,
+      "loss": 0.2752,
+      "step": 108
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4324779808521271,
+      "learning_rate": 9.860862128789953e-05,
+      "loss": 0.285,
+      "step": 109
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4228529632091522,
+      "learning_rate": 9.857242098248542e-05,
+      "loss": 0.2292,
+      "step": 110
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.43217021226882935,
+      "learning_rate": 9.853576260070347e-05,
+      "loss": 0.3298,
+      "step": 111
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.39829280972480774,
+      "learning_rate": 9.849864648827125e-05,
+      "loss": 0.2835,
+      "step": 112
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.46612149477005005,
+      "learning_rate": 9.846107299522304e-05,
+      "loss": 0.3774,
+      "step": 113
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5025432109832764,
+      "learning_rate": 9.842304247590667e-05,
+      "loss": 0.4244,
+      "step": 114
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.407463401556015,
+      "learning_rate": 9.838455528897998e-05,
+      "loss": 0.2726,
+      "step": 115
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.44039663672447205,
+      "learning_rate": 9.834561179740762e-05,
+      "loss": 0.2931,
+      "step": 116
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4455850124359131,
+      "learning_rate": 9.830621236845754e-05,
+      "loss": 0.3054,
+      "step": 117
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4223516881465912,
+      "learning_rate": 9.826635737369751e-05,
+      "loss": 0.3173,
+      "step": 118
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.43214577436447144,
+      "learning_rate": 9.82260471889917e-05,
+      "loss": 0.2673,
+      "step": 119
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4163484573364258,
+      "learning_rate": 9.818528219449705e-05,
+      "loss": 0.2587,
+      "step": 120
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.39266541600227356,
+      "learning_rate": 9.814406277465968e-05,
+      "loss": 0.2547,
+      "step": 121
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.45521286129951477,
+      "learning_rate": 9.810238931821138e-05,
+      "loss": 0.3337,
+      "step": 122
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4831443727016449,
+      "learning_rate": 9.806026221816581e-05,
+      "loss": 0.4037,
+      "step": 123
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.42679184675216675,
+      "learning_rate": 9.801768187181486e-05,
+      "loss": 0.3684,
+      "step": 124
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4037214517593384,
+      "learning_rate": 9.797464868072488e-05,
+      "loss": 0.2681,
+      "step": 125
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.444090336561203,
+      "learning_rate": 9.793116305073292e-05,
+      "loss": 0.3424,
+      "step": 126
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.45706042647361755,
+      "learning_rate": 9.788722539194291e-05,
+      "loss": 0.3811,
+      "step": 127
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4240741431713104,
+      "learning_rate": 9.784283611872169e-05,
+      "loss": 0.2971,
+      "step": 128
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4279850721359253,
+      "learning_rate": 9.77979956496953e-05,
+      "loss": 0.35,
+      "step": 129
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.49542683362960815,
+      "learning_rate": 9.77527044077448e-05,
+      "loss": 0.4575,
+      "step": 130
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4722813069820404,
+      "learning_rate": 9.770696282000244e-05,
+      "loss": 0.3603,
+      "step": 131
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4850144386291504,
+      "learning_rate": 9.766077131784763e-05,
+      "loss": 0.4054,
+      "step": 132
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.41566139459609985,
+      "learning_rate": 9.761413033690276e-05,
+      "loss": 0.3093,
+      "step": 133
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4394778907299042,
+      "learning_rate": 9.756704031702918e-05,
+      "loss": 0.3784,
+      "step": 134
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.40402501821517944,
+      "learning_rate": 9.751950170232309e-05,
+      "loss": 0.2863,
+      "step": 135
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4203854203224182,
+      "learning_rate": 9.74715149411112e-05,
+      "loss": 0.3284,
+      "step": 136
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.45846274495124817,
+      "learning_rate": 9.742308048594665e-05,
+      "loss": 0.3771,
+      "step": 137
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.2881685793399811,
+      "learning_rate": 9.73741987936047e-05,
+      "loss": 0.1712,
+      "step": 138
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.35943669080734253,
+      "learning_rate": 9.732487032507837e-05,
+      "loss": 0.2327,
+      "step": 139
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.48380306363105774,
+      "learning_rate": 9.727509554557417e-05,
+      "loss": 0.3955,
+      "step": 140
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.40915006399154663,
+      "learning_rate": 9.722487492450764e-05,
+      "loss": 0.2794,
+      "step": 141
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.46555307507514954,
+      "learning_rate": 9.717420893549902e-05,
+      "loss": 0.3706,
+      "step": 142
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.48473846912384033,
+      "learning_rate": 9.712309805636863e-05,
+      "loss": 0.3105,
+      "step": 143
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.49143391847610474,
+      "learning_rate": 9.707154276913256e-05,
+      "loss": 0.3829,
+      "step": 144
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.477644145488739,
+      "learning_rate": 9.701954355999792e-05,
+      "loss": 0.3629,
+      "step": 145
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.49221259355545044,
+      "learning_rate": 9.696710091935841e-05,
+      "loss": 0.3374,
+      "step": 146
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3983844816684723,
+      "learning_rate": 9.691421534178966e-05,
+      "loss": 0.2839,
+      "step": 147
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4182831645011902,
+      "learning_rate": 9.686088732604449e-05,
+      "loss": 0.305,
+      "step": 148
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.343465119600296,
+      "learning_rate": 9.68071173750483e-05,
+      "loss": 0.2242,
+      "step": 149
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4513535499572754,
+      "learning_rate": 9.675290599589429e-05,
+      "loss": 0.3739,
+      "step": 150
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4577466547489166,
+      "learning_rate": 9.669825369983865e-05,
+      "loss": 0.2852,
+      "step": 151
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.41453319787979126,
+      "learning_rate": 9.664316100229578e-05,
+      "loss": 0.3369,
+      "step": 152
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4898837208747864,
+      "learning_rate": 9.658762842283343e-05,
+      "loss": 0.3611,
+      "step": 153
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.35460522770881653,
+      "learning_rate": 9.653165648516776e-05,
+      "loss": 0.2565,
+      "step": 154
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.519338071346283,
+      "learning_rate": 9.647524571715843e-05,
+      "loss": 0.3722,
+      "step": 155
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.39072704315185547,
+      "learning_rate": 9.641839665080363e-05,
+      "loss": 0.3016,
+      "step": 156
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.40939077734947205,
+      "learning_rate": 9.636110982223505e-05,
+      "loss": 0.2837,
+      "step": 157
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.41712358593940735,
+      "learning_rate": 9.630338577171282e-05,
+      "loss": 0.3159,
+      "step": 158
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3813435435295105,
+      "learning_rate": 9.624522504362039e-05,
+      "loss": 0.2956,
+      "step": 159
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4338795840740204,
+      "learning_rate": 9.618662818645948e-05,
+      "loss": 0.3289,
+      "step": 160
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3376855254173279,
+      "learning_rate": 9.612759575284483e-05,
+      "loss": 0.2108,
+      "step": 161
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.47196638584136963,
+      "learning_rate": 9.606812829949896e-05,
+      "loss": 0.3092,
+      "step": 162
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4025622010231018,
+      "learning_rate": 9.600822638724705e-05,
+      "loss": 0.2853,
+      "step": 163
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.39661628007888794,
+      "learning_rate": 9.594789058101153e-05,
+      "loss": 0.2593,
+      "step": 164
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4002563953399658,
+      "learning_rate": 9.588712144980681e-05,
+      "loss": 0.3207,
+      "step": 165
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.44630905985832214,
+      "learning_rate": 9.582591956673387e-05,
+      "loss": 0.3185,
+      "step": 166
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4276568293571472,
+      "learning_rate": 9.576428550897489e-05,
+      "loss": 0.2568,
+      "step": 167
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.46223482489585876,
+      "learning_rate": 9.570221985778785e-05,
+      "loss": 0.3475,
+      "step": 168
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.42892640829086304,
+      "learning_rate": 9.563972319850092e-05,
+      "loss": 0.3124,
+      "step": 169
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.29332005977630615,
+      "learning_rate": 9.557679612050708e-05,
+      "loss": 0.2059,
+      "step": 170
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.42561015486717224,
+      "learning_rate": 9.551343921725843e-05,
+      "loss": 0.3656,
+      "step": 171
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5277482271194458,
+      "learning_rate": 9.544965308626074e-05,
+      "loss": 0.4538,
+      "step": 172
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.35639628767967224,
+      "learning_rate": 9.538543832906773e-05,
+      "loss": 0.2571,
+      "step": 173
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4271547198295593,
+      "learning_rate": 9.53207955512753e-05,
+      "loss": 0.3118,
+      "step": 174
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.46997734904289246,
+      "learning_rate": 9.525572536251607e-05,
+      "loss": 0.4165,
+      "step": 175
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.43963125348091125,
+      "learning_rate": 9.519022837645337e-05,
+      "loss": 0.3364,
+      "step": 176
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4160256087779999,
+      "learning_rate": 9.512430521077565e-05,
+      "loss": 0.2851,
+      "step": 177
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4184534251689911,
+      "learning_rate": 9.505795648719048e-05,
+      "loss": 0.3379,
+      "step": 178
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4775907099246979,
+      "learning_rate": 9.499118283141887e-05,
+      "loss": 0.3706,
+      "step": 179
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3853232264518738,
+      "learning_rate": 9.492398487318921e-05,
+      "loss": 0.3175,
+      "step": 180
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.37623131275177,
+      "learning_rate": 9.485636324623146e-05,
+      "loss": 0.2571,
+      "step": 181
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.41286978125572205,
+      "learning_rate": 9.478831858827104e-05,
+      "loss": 0.3544,
+      "step": 182
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4084218442440033,
+      "learning_rate": 9.471985154102292e-05,
+      "loss": 0.3279,
+      "step": 183
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5742401480674744,
+      "learning_rate": 9.465096275018556e-05,
+      "loss": 0.4284,
+      "step": 184
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.48429086804389954,
+      "learning_rate": 9.458165286543476e-05,
+      "loss": 0.3218,
+      "step": 185
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.48525944352149963,
+      "learning_rate": 9.451192254041758e-05,
+      "loss": 0.3434,
+      "step": 186
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.49932390451431274,
+      "learning_rate": 9.444177243274618e-05,
+      "loss": 0.4464,
+      "step": 187
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.45250341296195984,
+      "learning_rate": 9.437120320399158e-05,
+      "loss": 0.3498,
+      "step": 188
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3924804925918579,
+      "learning_rate": 9.430021551967744e-05,
+      "loss": 0.2776,
+      "step": 189
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3793688714504242,
+      "learning_rate": 9.422881004927381e-05,
+      "loss": 0.2747,
+      "step": 190
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.41701602935791016,
+      "learning_rate": 9.415698746619079e-05,
+      "loss": 0.3365,
+      "step": 191
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4095969498157501,
+      "learning_rate": 9.408474844777217e-05,
+      "loss": 0.2589,
+      "step": 192
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4828776717185974,
+      "learning_rate": 9.401209367528908e-05,
+      "loss": 0.4754,
+      "step": 193
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.41487669944763184,
+      "learning_rate": 9.393902383393347e-05,
+      "loss": 0.3103,
+      "step": 194
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5677187442779541,
+      "learning_rate": 9.386553961281179e-05,
+      "loss": 0.4281,
+      "step": 195
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4494563639163971,
+      "learning_rate": 9.379164170493844e-05,
+      "loss": 0.3802,
+      "step": 196
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4396170973777771,
+      "learning_rate": 9.371733080722911e-05,
+      "loss": 0.3371,
+      "step": 197
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3405661880970001,
+      "learning_rate": 9.36426076204944e-05,
+      "loss": 0.2232,
+      "step": 198
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.41570788621902466,
+      "learning_rate": 9.35674728494331e-05,
+      "loss": 0.3509,
+      "step": 199
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3930405080318451,
+      "learning_rate": 9.349192720262555e-05,
+      "loss": 0.3006,
+      "step": 200
+    },
+    {
+      "epoch": 0.19,
+      "eval_loss": 0.3084467351436615,
+      "eval_runtime": 388.9171,
+      "eval_samples_per_second": 0.877,
+      "eval_steps_per_second": 0.221,
+      "step": 200
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4046098291873932,
+      "learning_rate": 9.341597139252698e-05,
+      "loss": 0.3208,
+      "step": 201
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.43528658151626587,
+      "learning_rate": 9.333960613546079e-05,
+      "loss": 0.3081,
+      "step": 202
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.36973515152931213,
+      "learning_rate": 9.326283215161178e-05,
+      "loss": 0.2638,
+      "step": 203
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3538803458213806,
+      "learning_rate": 9.31856501650194e-05,
+      "loss": 0.1659,
+      "step": 204
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.402567058801651,
+      "learning_rate": 9.310806090357082e-05,
+      "loss": 0.357,
+      "step": 205
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.33732229471206665,
+      "learning_rate": 9.303006509899418e-05,
+      "loss": 0.1823,
+      "step": 206
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4957587420940399,
+      "learning_rate": 9.29516634868517e-05,
+      "loss": 0.3466,
+      "step": 207
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5121288895606995,
+      "learning_rate": 9.287285680653254e-05,
+      "loss": 0.3716,
+      "step": 208
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4042970538139343,
+      "learning_rate": 9.279364580124614e-05,
+      "loss": 0.3038,
+      "step": 209
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4011918902397156,
+      "learning_rate": 9.27140312180149e-05,
+      "loss": 0.2283,
+      "step": 210
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5329661965370178,
+      "learning_rate": 9.263401380766739e-05,
+      "loss": 0.4141,
+      "step": 211
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.47637537121772766,
+      "learning_rate": 9.255359432483105e-05,
+      "loss": 0.349,
+      "step": 212
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4198983609676361,
+      "learning_rate": 9.247277352792533e-05,
+      "loss": 0.2461,
+      "step": 213
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.47595182061195374,
+      "learning_rate": 9.239155217915422e-05,
+      "loss": 0.3589,
+      "step": 214
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3810179829597473,
+      "learning_rate": 9.230993104449939e-05,
+      "loss": 0.2334,
+      "step": 215
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.38976696133613586,
+      "learning_rate": 9.222791089371265e-05,
+      "loss": 0.295,
+      "step": 216
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.38406479358673096,
+      "learning_rate": 9.214549250030899e-05,
+      "loss": 0.2749,
+      "step": 217
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4411284029483795,
+      "learning_rate": 9.206267664155907e-05,
+      "loss": 0.3951,
+      "step": 218
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3465658724308014,
+      "learning_rate": 9.197946409848194e-05,
+      "loss": 0.2309,
+      "step": 219
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.367019921541214,
+      "learning_rate": 9.189585565583778e-05,
+      "loss": 0.2964,
+      "step": 220
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.37305861711502075,
+      "learning_rate": 9.181185210212034e-05,
+      "loss": 0.323,
+      "step": 221
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.37553930282592773,
+      "learning_rate": 9.172745422954961e-05,
+      "loss": 0.2614,
+      "step": 222
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.38929733633995056,
+      "learning_rate": 9.164266283406433e-05,
+      "loss": 0.3389,
+      "step": 223
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3998091518878937,
+      "learning_rate": 9.155747871531444e-05,
+      "loss": 0.322,
+      "step": 224
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4623723030090332,
+      "learning_rate": 9.14719026766536e-05,
+      "loss": 0.366,
+      "step": 225
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.41633743047714233,
+      "learning_rate": 9.13859355251316e-05,
+      "loss": 0.2965,
+      "step": 226
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3783974349498749,
+      "learning_rate": 9.129957807148666e-05,
+      "loss": 0.2263,
+      "step": 227
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4868335425853729,
+      "learning_rate": 9.121283113013795e-05,
+      "loss": 0.4145,
+      "step": 228
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3790939748287201,
+      "learning_rate": 9.112569551917774e-05,
+      "loss": 0.2589,
+      "step": 229
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.48874813318252563,
+      "learning_rate": 9.103817206036382e-05,
+      "loss": 0.3771,
+      "step": 230
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.47203895449638367,
+      "learning_rate": 9.095026157911166e-05,
+      "loss": 0.3425,
+      "step": 231
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4333976209163666,
+      "learning_rate": 9.086196490448668e-05,
+      "loss": 0.2626,
+      "step": 232
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.46086961030960083,
+      "learning_rate": 9.077328286919638e-05,
+      "loss": 0.3654,
+      "step": 233
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4530964195728302,
+      "learning_rate": 9.068421630958254e-05,
+      "loss": 0.3772,
+      "step": 234
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.34641772508621216,
+      "learning_rate": 9.059476606561328e-05,
+      "loss": 0.2236,
+      "step": 235
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4184739589691162,
+      "learning_rate": 9.050493298087523e-05,
+      "loss": 0.3458,
+      "step": 236
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.39227980375289917,
+      "learning_rate": 9.041471790256542e-05,
+      "loss": 0.3097,
+      "step": 237
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.39379796385765076,
+      "learning_rate": 9.032412168148345e-05,
+      "loss": 0.313,
+      "step": 238
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4767160415649414,
+      "learning_rate": 9.02331451720234e-05,
+      "loss": 0.3428,
+      "step": 239
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3696959316730499,
+      "learning_rate": 9.014178923216572e-05,
+      "loss": 0.2406,
+      "step": 240
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3842311501502991,
+      "learning_rate": 9.005005472346924e-05,
+      "loss": 0.253,
+      "step": 241
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.44675490260124207,
+      "learning_rate": 8.995794251106295e-05,
+      "loss": 0.3911,
+      "step": 242
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4012981057167053,
+      "learning_rate": 8.986545346363792e-05,
+      "loss": 0.3122,
+      "step": 243
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4179876744747162,
+      "learning_rate": 8.977258845343905e-05,
+      "loss": 0.3407,
+      "step": 244
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3772086203098297,
+      "learning_rate": 8.967934835625689e-05,
+      "loss": 0.2932,
+      "step": 245
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.46868202090263367,
+      "learning_rate": 8.958573405141932e-05,
+      "loss": 0.3366,
+      "step": 246
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4833935499191284,
+      "learning_rate": 8.949174642178332e-05,
+      "loss": 0.3927,
+      "step": 247
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4582580626010895,
+      "learning_rate": 8.939738635372664e-05,
+      "loss": 0.3322,
+      "step": 248
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4749627113342285,
+      "learning_rate": 8.930265473713938e-05,
+      "loss": 0.3708,
+      "step": 249
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.45011281967163086,
+      "learning_rate": 8.920755246541563e-05,
+      "loss": 0.3485,
+      "step": 250
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.40162166953086853,
+      "learning_rate": 8.911208043544513e-05,
+      "loss": 0.3136,
+      "step": 251
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3785412907600403,
+      "learning_rate": 8.90162395476046e-05,
+      "loss": 0.2934,
+      "step": 252
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.521395206451416,
+      "learning_rate": 8.89200307057495e-05,
+      "loss": 0.4588,
+      "step": 253
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.44272148609161377,
+      "learning_rate": 8.882345481720533e-05,
+      "loss": 0.3945,
+      "step": 254
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4032283127307892,
+      "learning_rate": 8.872651279275916e-05,
+      "loss": 0.3241,
+      "step": 255
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.48827365040779114,
+      "learning_rate": 8.862920554665098e-05,
+      "loss": 0.3371,
+      "step": 256
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4076429009437561,
+      "learning_rate": 8.853153399656513e-05,
+      "loss": 0.3072,
+      "step": 257
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.45559951663017273,
+      "learning_rate": 8.843349906362163e-05,
+      "loss": 0.3794,
+      "step": 258
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4802367687225342,
+      "learning_rate": 8.833510167236747e-05,
+      "loss": 0.3999,
+      "step": 259
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4042610228061676,
+      "learning_rate": 8.823634275076791e-05,
+      "loss": 0.3253,
+      "step": 260
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4060603678226471,
+      "learning_rate": 8.813722323019773e-05,
+      "loss": 0.3739,
+      "step": 261
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.37076127529144287,
+      "learning_rate": 8.803774404543247e-05,
+      "loss": 0.257,
+      "step": 262
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3881576657295227,
+      "learning_rate": 8.793790613463955e-05,
+      "loss": 0.3287,
+      "step": 263
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.31290051341056824,
+      "learning_rate": 8.783771043936948e-05,
+      "loss": 0.2226,
+      "step": 264
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.40022042393684387,
+      "learning_rate": 8.773715790454695e-05,
+      "loss": 0.2917,
+      "step": 265
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3606204688549042,
+      "learning_rate": 8.763624947846195e-05,
+      "loss": 0.224,
+      "step": 266
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.48021483421325684,
+      "learning_rate": 8.753498611276079e-05,
+      "loss": 0.3596,
+      "step": 267
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.47170498967170715,
+      "learning_rate": 8.743336876243712e-05,
+      "loss": 0.4077,
+      "step": 268
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5183234810829163,
+      "learning_rate": 8.7331398385823e-05,
+      "loss": 0.4477,
+      "step": 269
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.47882112860679626,
+      "learning_rate": 8.722907594457974e-05,
+      "loss": 0.3859,
+      "step": 270
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4458145797252655,
+      "learning_rate": 8.712640240368899e-05,
+      "loss": 0.3881,
+      "step": 271
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.47427937388420105,
+      "learning_rate": 8.702337873144343e-05,
+      "loss": 0.3089,
+      "step": 272
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3881189525127411,
+      "learning_rate": 8.692000589943785e-05,
+      "loss": 0.2669,
+      "step": 273
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.47891557216644287,
+      "learning_rate": 8.681628488255986e-05,
+      "loss": 0.3953,
+      "step": 274
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3770958483219147,
+      "learning_rate": 8.671221665898073e-05,
+      "loss": 0.2777,
+      "step": 275
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.43087834119796753,
+      "learning_rate": 8.660780221014616e-05,
+      "loss": 0.326,
+      "step": 276
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4207721948623657,
+      "learning_rate": 8.650304252076703e-05,
+      "loss": 0.3052,
+      "step": 277
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3829022943973541,
+      "learning_rate": 8.639793857881011e-05,
+      "loss": 0.2636,
+      "step": 278
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5186901688575745,
+      "learning_rate": 8.629249137548874e-05,
+      "loss": 0.3903,
+      "step": 279
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4041597843170166,
+      "learning_rate": 8.618670190525352e-05,
+      "loss": 0.2611,
+      "step": 280
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.41534358263015747,
+      "learning_rate": 8.608057116578283e-05,
+      "loss": 0.2872,
+      "step": 281
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.47262248396873474,
+      "learning_rate": 8.597410015797358e-05,
+      "loss": 0.4081,
+      "step": 282
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3860476315021515,
+      "learning_rate": 8.586728988593158e-05,
+      "loss": 0.2329,
+      "step": 283
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.48730090260505676,
+      "learning_rate": 8.576014135696226e-05,
+      "loss": 0.3219,
+      "step": 284
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4236432611942291,
+      "learning_rate": 8.565265558156101e-05,
+      "loss": 0.2836,
+      "step": 285
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5237382054328918,
+      "learning_rate": 8.554483357340379e-05,
+      "loss": 0.353,
+      "step": 286
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.37562325596809387,
+      "learning_rate": 8.543667634933742e-05,
+      "loss": 0.2464,
+      "step": 287
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.40596145391464233,
+      "learning_rate": 8.532818492937014e-05,
+      "loss": 0.2966,
+      "step": 288
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.42180100083351135,
+      "learning_rate": 8.521936033666187e-05,
+      "loss": 0.2999,
+      "step": 289
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4076705873012543,
+      "learning_rate": 8.511020359751466e-05,
+      "loss": 0.3086,
+      "step": 290
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3389161229133606,
+      "learning_rate": 8.500071574136295e-05,
+      "loss": 0.2131,
+      "step": 291
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4136177897453308,
+      "learning_rate": 8.489089780076386e-05,
+      "loss": 0.2865,
+      "step": 292
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.431543231010437,
+      "learning_rate": 8.478075081138745e-05,
+      "loss": 0.2909,
+      "step": 293
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5177910327911377,
+      "learning_rate": 8.467027581200702e-05,
+      "loss": 0.4105,
+      "step": 294
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.46816176176071167,
+      "learning_rate": 8.455947384448924e-05,
+      "loss": 0.3566,
+      "step": 295
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.35551542043685913,
+      "learning_rate": 8.444834595378434e-05,
+      "loss": 0.2521,
+      "step": 296
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.41277021169662476,
+      "learning_rate": 8.433689318791626e-05,
+      "loss": 0.3427,
+      "step": 297
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.40482524037361145,
+      "learning_rate": 8.422511659797279e-05,
+      "loss": 0.2991,
+      "step": 298
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.41236555576324463,
+      "learning_rate": 8.411301723809563e-05,
+      "loss": 0.2905,
+      "step": 299
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.39597707986831665,
+      "learning_rate": 8.400059616547047e-05,
+      "loss": 0.3251,
+      "step": 300
+    },
+    {
+      "epoch": 0.28,
+      "eval_loss": 0.3053378164768219,
+      "eval_runtime": 376.772,
+      "eval_samples_per_second": 0.905,
+      "eval_steps_per_second": 0.228,
+      "step": 300
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4647742807865143,
+      "learning_rate": 8.388785444031695e-05,
+      "loss": 0.395,
+      "step": 301
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4007340967655182,
+      "learning_rate": 8.377479312587879e-05,
+      "loss": 0.2737,
+      "step": 302
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.417574942111969,
+      "learning_rate": 8.366141328841366e-05,
+      "loss": 0.3351,
+      "step": 303
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.347423791885376,
+      "learning_rate": 8.354771599718312e-05,
+      "loss": 0.2653,
+      "step": 304
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.433025062084198,
+      "learning_rate": 8.343370232444261e-05,
+      "loss": 0.3532,
+      "step": 305
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3730097711086273,
+      "learning_rate": 8.331937334543132e-05,
+      "loss": 0.2968,
+      "step": 306
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.48292431235313416,
+      "learning_rate": 8.320473013836196e-05,
+      "loss": 0.4295,
+      "step": 307
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.439970463514328,
+      "learning_rate": 8.308977378441072e-05,
+      "loss": 0.4032,
+      "step": 308
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3936116397380829,
+      "learning_rate": 8.297450536770697e-05,
+      "loss": 0.3051,
+      "step": 309
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.41220739483833313,
+      "learning_rate": 8.28589259753231e-05,
+      "loss": 0.292,
+      "step": 310
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.41591155529022217,
+      "learning_rate": 8.274303669726426e-05,
+      "loss": 0.3345,
+      "step": 311
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.39442381262779236,
+      "learning_rate": 8.262683862645804e-05,
+      "loss": 0.3164,
+      "step": 312
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4502621293067932,
+      "learning_rate": 8.251033285874421e-05,
+      "loss": 0.3558,
+      "step": 313
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.42802926898002625,
+      "learning_rate": 8.239352049286435e-05,
+      "loss": 0.3198,
+      "step": 314
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4209524095058441,
+      "learning_rate": 8.22764026304515e-05,
+      "loss": 0.3313,
+      "step": 315
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.43702828884124756,
+      "learning_rate": 8.21589803760198e-05,
+      "loss": 0.3498,
+      "step": 316
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3823443055152893,
+      "learning_rate": 8.204125483695403e-05,
+      "loss": 0.3173,
+      "step": 317
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.37982624769210815,
+      "learning_rate": 8.192322712349917e-05,
+      "loss": 0.2638,
+      "step": 318
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.41392555832862854,
+      "learning_rate": 8.180489834875e-05,
+      "loss": 0.3171,
+      "step": 319
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4302511215209961,
+      "learning_rate": 8.168626962864045e-05,
+      "loss": 0.2687,
+      "step": 320
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.47123268246650696,
+      "learning_rate": 8.156734208193327e-05,
+      "loss": 0.3131,
+      "step": 321
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4142516851425171,
+      "learning_rate": 8.144811683020932e-05,
+      "loss": 0.2877,
+      "step": 322
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.43524178862571716,
+      "learning_rate": 8.132859499785707e-05,
+      "loss": 0.3487,
+      "step": 323
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4116482436656952,
+      "learning_rate": 8.120877771206201e-05,
+      "loss": 0.3156,
+      "step": 324
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.49788716435432434,
+      "learning_rate": 8.108866610279593e-05,
+      "loss": 0.3947,
+      "step": 325
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4383057951927185,
+      "learning_rate": 8.096826130280639e-05,
+      "loss": 0.3066,
+      "step": 326
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4812861382961273,
+      "learning_rate": 8.08475644476059e-05,
+      "loss": 0.3646,
+      "step": 327
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.42143720388412476,
+      "learning_rate": 8.072657667546136e-05,
+      "loss": 0.3719,
+      "step": 328
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.43385979533195496,
+      "learning_rate": 8.060529912738315e-05,
+      "loss": 0.3691,
+      "step": 329
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5112125277519226,
+      "learning_rate": 8.048373294711455e-05,
+      "loss": 0.3336,
+      "step": 330
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.742655634880066,
+      "learning_rate": 8.036187928112087e-05,
+      "loss": 0.3463,
+      "step": 331
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.45481520891189575,
+      "learning_rate": 8.023973927857857e-05,
+      "loss": 0.3956,
+      "step": 332
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.529845654964447,
+      "learning_rate": 8.011731409136454e-05,
+      "loss": 0.3941,
+      "step": 333
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4360586106777191,
+      "learning_rate": 7.99946048740452e-05,
+      "loss": 0.2524,
+      "step": 334
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5153516530990601,
+      "learning_rate": 7.987161278386554e-05,
+      "loss": 0.365,
+      "step": 335
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4026721715927124,
+      "learning_rate": 7.974833898073832e-05,
+      "loss": 0.3111,
+      "step": 336
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.45282304286956787,
+      "learning_rate": 7.962478462723307e-05,
+      "loss": 0.3708,
+      "step": 337
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4389725923538208,
+      "learning_rate": 7.950095088856508e-05,
+      "loss": 0.3558,
+      "step": 338
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.34882915019989014,
+      "learning_rate": 7.937683893258454e-05,
+      "loss": 0.2254,
+      "step": 339
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4819268584251404,
+      "learning_rate": 7.925244992976538e-05,
+      "loss": 0.4112,
+      "step": 340
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3798743486404419,
+      "learning_rate": 7.912778505319436e-05,
+      "loss": 0.3277,
+      "step": 341
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.32814425230026245,
+      "learning_rate": 7.900284547855991e-05,
+      "loss": 0.1931,
+      "step": 342
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.49891549348831177,
+      "learning_rate": 7.88776323841411e-05,
+      "loss": 0.368,
+      "step": 343
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5247824192047119,
+      "learning_rate": 7.875214695079647e-05,
+      "loss": 0.3932,
+      "step": 344
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.33255767822265625,
+      "learning_rate": 7.862639036195298e-05,
+      "loss": 0.2609,
+      "step": 345
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.48547419905662537,
+      "learning_rate": 7.850036380359478e-05,
+      "loss": 0.3502,
+      "step": 346
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4334273338317871,
+      "learning_rate": 7.837406846425204e-05,
+      "loss": 0.3819,
+      "step": 347
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4432954490184784,
+      "learning_rate": 7.824750553498976e-05,
+      "loss": 0.394,
+      "step": 348
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.48492997884750366,
+      "learning_rate": 7.812067620939652e-05,
+      "loss": 0.3532,
+      "step": 349
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3824009895324707,
+      "learning_rate": 7.799358168357323e-05,
+      "loss": 0.2875,
+      "step": 350
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5168920159339905,
+      "learning_rate": 7.786622315612183e-05,
+      "loss": 0.3885,
+      "step": 351
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3767426013946533,
+      "learning_rate": 7.773860182813403e-05,
+      "loss": 0.2681,
+      "step": 352
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.45943331718444824,
+      "learning_rate": 7.761071890317994e-05,
+      "loss": 0.341,
+      "step": 353
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.40229344367980957,
+      "learning_rate": 7.748257558729677e-05,
+      "loss": 0.3217,
+      "step": 354
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.43115130066871643,
+      "learning_rate": 7.735417308897736e-05,
+      "loss": 0.3699,
+      "step": 355
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4053991138935089,
+      "learning_rate": 7.72255126191589e-05,
+      "loss": 0.3191,
+      "step": 356
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.46663257479667664,
+      "learning_rate": 7.709659539121144e-05,
+      "loss": 0.3258,
+      "step": 357
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.39004477858543396,
+      "learning_rate": 7.696742262092644e-05,
+      "loss": 0.2996,
+      "step": 358
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3414752185344696,
+      "learning_rate": 7.683799552650534e-05,
+      "loss": 0.2297,
+      "step": 359
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3376968502998352,
+      "learning_rate": 7.67083153285481e-05,
+      "loss": 0.2442,
+      "step": 360
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.37360575795173645,
+      "learning_rate": 7.657838325004159e-05,
+      "loss": 0.2628,
+      "step": 361
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4016753137111664,
+      "learning_rate": 7.644820051634812e-05,
+      "loss": 0.3307,
+      "step": 362
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4536408483982086,
+      "learning_rate": 7.63177683551939e-05,
+      "loss": 0.2852,
+      "step": 363
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.42366743087768555,
+      "learning_rate": 7.618708799665744e-05,
+      "loss": 0.2858,
+      "step": 364
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4187493920326233,
+      "learning_rate": 7.605616067315793e-05,
+      "loss": 0.3561,
+      "step": 365
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.37502115964889526,
+      "learning_rate": 7.592498761944362e-05,
+      "loss": 0.2209,
+      "step": 366
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.30580833554267883,
+      "learning_rate": 7.579357007258022e-05,
+      "loss": 0.1708,
+      "step": 367
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4009668529033661,
+      "learning_rate": 7.56619092719392e-05,
+      "loss": 0.2685,
+      "step": 368
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.37424060702323914,
+      "learning_rate": 7.553000645918611e-05,
+      "loss": 0.2034,
+      "step": 369
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.33520057797431946,
+      "learning_rate": 7.539786287826885e-05,
+      "loss": 0.2141,
+      "step": 370
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3602328300476074,
+      "learning_rate": 7.526547977540592e-05,
+      "loss": 0.2214,
+      "step": 371
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3910132944583893,
+      "learning_rate": 7.51328583990748e-05,
+      "loss": 0.2584,
+      "step": 372
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.43052783608436584,
+      "learning_rate": 7.500000000000001e-05,
+      "loss": 0.2919,
+      "step": 373
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3827561140060425,
+      "learning_rate": 7.486690583114136e-05,
+      "loss": 0.2129,
+      "step": 374
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.48869532346725464,
+      "learning_rate": 7.473357714768222e-05,
+      "loss": 0.3665,
+      "step": 375
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5260458588600159,
+      "learning_rate": 7.460001520701755e-05,
+      "loss": 0.45,
+      "step": 376
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.49762579798698425,
+      "learning_rate": 7.446622126874218e-05,
+      "loss": 0.3492,
+      "step": 377
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.49906134605407715,
+      "learning_rate": 7.433219659463881e-05,
+      "loss": 0.3686,
+      "step": 378
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.43443796038627625,
+      "learning_rate": 7.41979424486662e-05,
+      "loss": 0.3532,
+      "step": 379
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4305638372898102,
+      "learning_rate": 7.406346009694713e-05,
+      "loss": 0.2751,
+      "step": 380
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3603278696537018,
+      "learning_rate": 7.392875080775665e-05,
+      "loss": 0.2739,
+      "step": 381
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.44212138652801514,
+      "learning_rate": 7.379381585150996e-05,
+      "loss": 0.3136,
+      "step": 382
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3916763961315155,
+      "learning_rate": 7.365865650075046e-05,
+      "loss": 0.287,
+      "step": 383
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3192064166069031,
+      "learning_rate": 7.35232740301378e-05,
+      "loss": 0.2272,
+      "step": 384
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.38105249404907227,
+      "learning_rate": 7.338766971643579e-05,
+      "loss": 0.3102,
+      "step": 385
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4224295914173126,
+      "learning_rate": 7.325184483850042e-05,
+      "loss": 0.3151,
+      "step": 386
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.361806184053421,
+      "learning_rate": 7.311580067726783e-05,
+      "loss": 0.2248,
+      "step": 387
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3746489882469177,
+      "learning_rate": 7.297953851574206e-05,
+      "loss": 0.2809,
+      "step": 388
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.48226237297058105,
+      "learning_rate": 7.284305963898314e-05,
+      "loss": 0.4617,
+      "step": 389
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.42972224950790405,
+      "learning_rate": 7.27063653340949e-05,
+      "loss": 0.3475,
+      "step": 390
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.42233508825302124,
+      "learning_rate": 7.25694568902128e-05,
+      "loss": 0.3409,
+      "step": 391
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2803710401058197,
+      "learning_rate": 7.243233559849179e-05,
+      "loss": 0.2025,
+      "step": 392
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.34155765175819397,
+      "learning_rate": 7.229500275209418e-05,
+      "loss": 0.2606,
+      "step": 393
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.38971152901649475,
+      "learning_rate": 7.215745964617737e-05,
+      "loss": 0.2833,
+      "step": 394
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.37858226895332336,
+      "learning_rate": 7.201970757788172e-05,
+      "loss": 0.3383,
+      "step": 395
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3657345175743103,
+      "learning_rate": 7.188174784631824e-05,
+      "loss": 0.2336,
+      "step": 396
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3923492729663849,
+      "learning_rate": 7.174358175255636e-05,
+      "loss": 0.2972,
+      "step": 397
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3969753086566925,
+      "learning_rate": 7.160521059961169e-05,
+      "loss": 0.2849,
+      "step": 398
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.2949353754520416,
+      "learning_rate": 7.146663569243371e-05,
+      "loss": 0.2215,
+      "step": 399
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.45280420780181885,
+      "learning_rate": 7.132785833789344e-05,
+      "loss": 0.3974,
+      "step": 400
+    },
+    {
+      "epoch": 0.38,
+      "eval_loss": 0.30068090558052063,
+      "eval_runtime": 369.9676,
+      "eval_samples_per_second": 0.922,
+      "eval_steps_per_second": 0.232,
+      "step": 400
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.36197903752326965,
+      "learning_rate": 7.118887984477116e-05,
+      "loss": 0.2835,
+      "step": 401
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.34623628854751587,
+      "learning_rate": 7.104970152374405e-05,
+      "loss": 0.3069,
+      "step": 402
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3760015666484833,
+      "learning_rate": 7.091032468737381e-05,
+      "loss": 0.2888,
+      "step": 403
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3890538513660431,
+      "learning_rate": 7.077075065009433e-05,
+      "loss": 0.299,
+      "step": 404
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.36255285143852234,
+      "learning_rate": 7.063098072819919e-05,
+      "loss": 0.3058,
+      "step": 405
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3318592309951782,
+      "learning_rate": 7.049101623982937e-05,
+      "loss": 0.2326,
+      "step": 406
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.36836788058280945,
+      "learning_rate": 7.03508585049608e-05,
+      "loss": 0.2571,
+      "step": 407
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4255884885787964,
+      "learning_rate": 7.021050884539177e-05,
+      "loss": 0.3345,
+      "step": 408
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.36213773488998413,
+      "learning_rate": 7.006996858473069e-05,
+      "loss": 0.2725,
+      "step": 409
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3614218235015869,
+      "learning_rate": 6.99292390483834e-05,
+      "loss": 0.2409,
+      "step": 410
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.34119337797164917,
+      "learning_rate": 6.97883215635408e-05,
+      "loss": 0.2615,
+      "step": 411
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4752042293548584,
+      "learning_rate": 6.964721745916631e-05,
+      "loss": 0.3332,
+      "step": 412
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3704507052898407,
+      "learning_rate": 6.950592806598327e-05,
+      "loss": 0.2744,
+      "step": 413
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3751181662082672,
+      "learning_rate": 6.936445471646249e-05,
+      "loss": 0.2733,
+      "step": 414
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.38812124729156494,
+      "learning_rate": 6.922279874480959e-05,
+      "loss": 0.3402,
+      "step": 415
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3884803056716919,
+      "learning_rate": 6.908096148695251e-05,
+      "loss": 0.3195,
+      "step": 416
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3704192042350769,
+      "learning_rate": 6.89389442805288e-05,
+      "loss": 0.2742,
+      "step": 417
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4019993543624878,
+      "learning_rate": 6.879674846487314e-05,
+      "loss": 0.3018,
+      "step": 418
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.42715343832969666,
+      "learning_rate": 6.865437538100457e-05,
+      "loss": 0.3402,
+      "step": 419
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.33204010128974915,
+      "learning_rate": 6.851182637161396e-05,
+      "loss": 0.2293,
+      "step": 420
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3886491060256958,
+      "learning_rate": 6.836910278105124e-05,
+      "loss": 0.2975,
+      "step": 421
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.429673433303833,
+      "learning_rate": 6.822620595531286e-05,
+      "loss": 0.3454,
+      "step": 422
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.392663836479187,
+      "learning_rate": 6.808313724202894e-05,
+      "loss": 0.2883,
+      "step": 423
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3442196547985077,
+      "learning_rate": 6.793989799045066e-05,
+      "loss": 0.232,
+      "step": 424
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.43839895725250244,
+      "learning_rate": 6.779648955143754e-05,
+      "loss": 0.3182,
+      "step": 425
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3995040953159332,
+      "learning_rate": 6.765291327744463e-05,
+      "loss": 0.3276,
+      "step": 426
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.36061960458755493,
+      "learning_rate": 6.750917052250982e-05,
+      "loss": 0.2887,
+      "step": 427
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3944300413131714,
+      "learning_rate": 6.736526264224101e-05,
+      "loss": 0.3062,
+      "step": 428
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.42164069414138794,
+      "learning_rate": 6.722119099380345e-05,
+      "loss": 0.3433,
+      "step": 429
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.33990684151649475,
+      "learning_rate": 6.707695693590676e-05,
+      "loss": 0.2064,
+      "step": 430
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3268337547779083,
+      "learning_rate": 6.693256182879225e-05,
+      "loss": 0.2728,
+      "step": 431
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4086841940879822,
+      "learning_rate": 6.678800703422003e-05,
+      "loss": 0.3605,
+      "step": 432
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.465461403131485,
+      "learning_rate": 6.664329391545626e-05,
+      "loss": 0.3917,
+      "step": 433
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4519641697406769,
+      "learning_rate": 6.64984238372601e-05,
+      "loss": 0.4305,
+      "step": 434
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.36165934801101685,
+      "learning_rate": 6.635339816587109e-05,
+      "loss": 0.2957,
+      "step": 435
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.49502721428871155,
+      "learning_rate": 6.620821826899606e-05,
+      "loss": 0.3882,
+      "step": 436
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4157223105430603,
+      "learning_rate": 6.606288551579629e-05,
+      "loss": 0.3393,
+      "step": 437
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.42530369758605957,
+      "learning_rate": 6.59174012768747e-05,
+      "loss": 0.3835,
+      "step": 438
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3848186731338501,
+      "learning_rate": 6.577176692426279e-05,
+      "loss": 0.3403,
+      "step": 439
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4040294289588928,
+      "learning_rate": 6.562598383140772e-05,
+      "loss": 0.3292,
+      "step": 440
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.38477933406829834,
+      "learning_rate": 6.548005337315944e-05,
+      "loss": 0.2976,
+      "step": 441
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3160758316516876,
+      "learning_rate": 6.533397692575766e-05,
+      "loss": 0.1825,
+      "step": 442
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5142346620559692,
+      "learning_rate": 6.518775586681887e-05,
+      "loss": 0.4348,
+      "step": 443
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3608967363834381,
+      "learning_rate": 6.504139157532337e-05,
+      "loss": 0.2784,
+      "step": 444
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.33182722330093384,
+      "learning_rate": 6.489488543160224e-05,
+      "loss": 0.2209,
+      "step": 445
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3692116141319275,
+      "learning_rate": 6.47482388173244e-05,
+      "loss": 0.2427,
+      "step": 446
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4166927635669708,
+      "learning_rate": 6.460145311548341e-05,
+      "loss": 0.368,
+      "step": 447
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4535463750362396,
+      "learning_rate": 6.445452971038464e-05,
+      "loss": 0.3308,
+      "step": 448
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4454578459262848,
+      "learning_rate": 6.430746998763204e-05,
+      "loss": 0.3745,
+      "step": 449
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.41668403148651123,
+      "learning_rate": 6.416027533411519e-05,
+      "loss": 0.3766,
+      "step": 450
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5912679433822632,
+      "learning_rate": 6.401294713799618e-05,
+      "loss": 0.4387,
+      "step": 451
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3945387899875641,
+      "learning_rate": 6.386548678869644e-05,
+      "loss": 0.3043,
+      "step": 452
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3480340242385864,
+      "learning_rate": 6.37178956768838e-05,
+      "loss": 0.2651,
+      "step": 453
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.39592987298965454,
+      "learning_rate": 6.35701751944592e-05,
+      "loss": 0.311,
+      "step": 454
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.40607360005378723,
+      "learning_rate": 6.342232673454371e-05,
+      "loss": 0.3391,
+      "step": 455
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.32058456540107727,
+      "learning_rate": 6.32743516914653e-05,
+      "loss": 0.2055,
+      "step": 456
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4424896240234375,
+      "learning_rate": 6.312625146074575e-05,
+      "loss": 0.348,
+      "step": 457
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.36645376682281494,
+      "learning_rate": 6.297802743908741e-05,
+      "loss": 0.2687,
+      "step": 458
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.379740834236145,
+      "learning_rate": 6.282968102436009e-05,
+      "loss": 0.3027,
+      "step": 459
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3414551317691803,
+      "learning_rate": 6.268121361558792e-05,
+      "loss": 0.2337,
+      "step": 460
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3833504915237427,
+      "learning_rate": 6.253262661293604e-05,
+      "loss": 0.3471,
+      "step": 461
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.37215739488601685,
+      "learning_rate": 6.238392141769743e-05,
+      "loss": 0.2434,
+      "step": 462
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4152020812034607,
+      "learning_rate": 6.22350994322798e-05,
+      "loss": 0.3909,
+      "step": 463
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3794286847114563,
+      "learning_rate": 6.208616206019224e-05,
+      "loss": 0.2629,
+      "step": 464
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3543379604816437,
+      "learning_rate": 6.193711070603202e-05,
+      "loss": 0.252,
+      "step": 465
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.36758437752723694,
+      "learning_rate": 6.178794677547137e-05,
+      "loss": 0.273,
+      "step": 466
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.41124922037124634,
+      "learning_rate": 6.163867167524419e-05,
+      "loss": 0.3192,
+      "step": 467
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.42042025923728943,
+      "learning_rate": 6.14892868131328e-05,
+      "loss": 0.2983,
+      "step": 468
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3670899569988251,
+      "learning_rate": 6.133979359795468e-05,
+      "loss": 0.2665,
+      "step": 469
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.39666733145713806,
+      "learning_rate": 6.119019343954914e-05,
+      "loss": 0.2879,
+      "step": 470
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.37360242009162903,
+      "learning_rate": 6.104048774876406e-05,
+      "loss": 0.2979,
+      "step": 471
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4077526926994324,
+      "learning_rate": 6.0890677937442574e-05,
+      "loss": 0.3643,
+      "step": 472
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5177510976791382,
+      "learning_rate": 6.074076541840977e-05,
+      "loss": 0.478,
+      "step": 473
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3729604482650757,
+      "learning_rate": 6.059075160545933e-05,
+      "loss": 0.2348,
+      "step": 474
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3867332935333252,
+      "learning_rate": 6.0440637913340224e-05,
+      "loss": 0.2397,
+      "step": 475
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3798309564590454,
+      "learning_rate": 6.029042575774334e-05,
+      "loss": 0.2985,
+      "step": 476
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4377082884311676,
+      "learning_rate": 6.0140116555288204e-05,
+      "loss": 0.3347,
+      "step": 477
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3695531189441681,
+      "learning_rate": 5.998971172350953e-05,
+      "loss": 0.3133,
+      "step": 478
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3421245217323303,
+      "learning_rate": 5.9839212680843925e-05,
+      "loss": 0.2555,
+      "step": 479
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.44313791394233704,
+      "learning_rate": 5.968862084661643e-05,
+      "loss": 0.3724,
+      "step": 480
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.36526092886924744,
+      "learning_rate": 5.953793764102722e-05,
+      "loss": 0.2995,
+      "step": 481
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3772035539150238,
+      "learning_rate": 5.938716448513818e-05,
+      "loss": 0.2864,
+      "step": 482
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.49520444869995117,
+      "learning_rate": 5.923630280085948e-05,
+      "loss": 0.4505,
+      "step": 483
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4190358519554138,
+      "learning_rate": 5.908535401093618e-05,
+      "loss": 0.3696,
+      "step": 484
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.357243150472641,
+      "learning_rate": 5.8934319538934824e-05,
+      "loss": 0.2911,
+      "step": 485
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3449594974517822,
+      "learning_rate": 5.878320080923001e-05,
+      "loss": 0.2866,
+      "step": 486
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4797684848308563,
+      "learning_rate": 5.863199924699095e-05,
+      "loss": 0.3019,
+      "step": 487
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4785318970680237,
+      "learning_rate": 5.848071627816803e-05,
+      "loss": 0.4253,
+      "step": 488
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.42927151918411255,
+      "learning_rate": 5.8329353329479366e-05,
+      "loss": 0.3504,
+      "step": 489
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3368590772151947,
+      "learning_rate": 5.8177911828397336e-05,
+      "loss": 0.2499,
+      "step": 490
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4111219048500061,
+      "learning_rate": 5.802639320313514e-05,
+      "loss": 0.3663,
+      "step": 491
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.41821691393852234,
+      "learning_rate": 5.787479888263333e-05,
+      "loss": 0.3296,
+      "step": 492
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.40745142102241516,
+      "learning_rate": 5.77231302965463e-05,
+      "loss": 0.2949,
+      "step": 493
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.39822155237197876,
+      "learning_rate": 5.757138887522884e-05,
+      "loss": 0.2893,
+      "step": 494
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.34316396713256836,
+      "learning_rate": 5.7419576049722646e-05,
+      "loss": 0.2451,
+      "step": 495
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3645155727863312,
+      "learning_rate": 5.7267693251742794e-05,
+      "loss": 0.2859,
+      "step": 496
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4316769540309906,
+      "learning_rate": 5.7115741913664264e-05,
+      "loss": 0.2947,
+      "step": 497
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4154459834098816,
+      "learning_rate": 5.696372346850841e-05,
+      "loss": 0.3844,
+      "step": 498
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.40410280227661133,
+      "learning_rate": 5.68116393499295e-05,
+      "loss": 0.3262,
+      "step": 499
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.352742999792099,
+      "learning_rate": 5.6659490992201094e-05,
+      "loss": 0.258,
+      "step": 500
+    },
+    {
+      "epoch": 0.47,
+      "eval_loss": 0.2935165762901306,
+      "eval_runtime": 370.7332,
+      "eval_samples_per_second": 0.92,
+      "eval_steps_per_second": 0.232,
+      "step": 500
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4673909544944763,
+      "learning_rate": 5.650727983020262e-05,
+      "loss": 0.4462,
+      "step": 501
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3418557643890381,
+      "learning_rate": 5.635500729940577e-05,
+      "loss": 0.2146,
+      "step": 502
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4534144997596741,
+      "learning_rate": 5.620267483586105e-05,
+      "loss": 0.3785,
+      "step": 503
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.42655184864997864,
+      "learning_rate": 5.6050283876184116e-05,
+      "loss": 0.3626,
+      "step": 504
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4821220636367798,
+      "learning_rate": 5.5897835857542317e-05,
+      "loss": 0.3625,
+      "step": 505
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.40257149934768677,
+      "learning_rate": 5.574533221764109e-05,
+      "loss": 0.2671,
+      "step": 506
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3685300052165985,
+      "learning_rate": 5.5592774394710465e-05,
+      "loss": 0.2824,
+      "step": 507
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4492349624633789,
+      "learning_rate": 5.5440163827491454e-05,
+      "loss": 0.2858,
+      "step": 508
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.38566920161247253,
+      "learning_rate": 5.5287501955222444e-05,
+      "loss": 0.2672,
+      "step": 509
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3556060194969177,
+      "learning_rate": 5.513479021762572e-05,
+      "loss": 0.2846,
+      "step": 510
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.37076959013938904,
+      "learning_rate": 5.4982030054893785e-05,
+      "loss": 0.2247,
+      "step": 511
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4032704830169678,
+      "learning_rate": 5.482922290767589e-05,
+      "loss": 0.2576,
+      "step": 512
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.38813984394073486,
+      "learning_rate": 5.467637021706438e-05,
+      "loss": 0.2866,
+      "step": 513
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5001494884490967,
+      "learning_rate": 5.452347342458104e-05,
+      "loss": 0.3794,
+      "step": 514
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.40518251061439514,
+      "learning_rate": 5.437053397216364e-05,
+      "loss": 0.312,
+      "step": 515
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3968459367752075,
+      "learning_rate": 5.4217553302152237e-05,
+      "loss": 0.3138,
+      "step": 516
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4698263108730316,
+      "learning_rate": 5.406453285727564e-05,
+      "loss": 0.3502,
+      "step": 517
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3987574279308319,
+      "learning_rate": 5.39114740806377e-05,
+      "loss": 0.2774,
+      "step": 518
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3404723107814789,
+      "learning_rate": 5.3758378415703825e-05,
+      "loss": 0.2421,
+      "step": 519
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.47483158111572266,
+      "learning_rate": 5.360524730628728e-05,
+      "loss": 0.4053,
+      "step": 520
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4022204279899597,
+      "learning_rate": 5.345208219653561e-05,
+      "loss": 0.2897,
+      "step": 521
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4006354808807373,
+      "learning_rate": 5.3298884530917014e-05,
+      "loss": 0.3432,
+      "step": 522
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3592531979084015,
+      "learning_rate": 5.3145655754206714e-05,
+      "loss": 0.257,
+      "step": 523
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3259998857975006,
+      "learning_rate": 5.2992397311473316e-05,
+      "loss": 0.2084,
+      "step": 524
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4196920394897461,
+      "learning_rate": 5.2839110648065214e-05,
+      "loss": 0.321,
+      "step": 525
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4363613426685333,
+      "learning_rate": 5.2685797209596976e-05,
+      "loss": 0.3242,
+      "step": 526
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4313899874687195,
+      "learning_rate": 5.2532458441935636e-05,
+      "loss": 0.3146,
+      "step": 527
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.44467124342918396,
+      "learning_rate": 5.2379095791187124e-05,
+      "loss": 0.3447,
+      "step": 528
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3862348794937134,
+      "learning_rate": 5.222571070368257e-05,
+      "loss": 0.2889,
+      "step": 529
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.385376513004303,
+      "learning_rate": 5.2072304625964785e-05,
+      "loss": 0.269,
+      "step": 530
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3828996419906616,
+      "learning_rate": 5.191887900477443e-05,
+      "loss": 0.3413,
+      "step": 531
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3872687816619873,
+      "learning_rate": 5.176543528703657e-05,
+      "loss": 0.316,
+      "step": 532
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.37200242280960083,
+      "learning_rate": 5.161197491984684e-05,
+      "loss": 0.2868,
+      "step": 533
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.39529964327812195,
+      "learning_rate": 5.1458499350458e-05,
+      "loss": 0.2231,
+      "step": 534
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.44413501024246216,
+      "learning_rate": 5.130501002626609e-05,
+      "loss": 0.4653,
+      "step": 535
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4013352692127228,
+      "learning_rate": 5.11515083947969e-05,
+      "loss": 0.284,
+      "step": 536
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.39808186888694763,
+      "learning_rate": 5.099799590369231e-05,
+      "loss": 0.3255,
+      "step": 537
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4179256856441498,
+      "learning_rate": 5.084447400069655e-05,
+      "loss": 0.3378,
+      "step": 538
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3801272213459015,
+      "learning_rate": 5.069094413364272e-05,
+      "loss": 0.2895,
+      "step": 539
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4310075640678406,
+      "learning_rate": 5.053740775043892e-05,
+      "loss": 0.3423,
+      "step": 540
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.47626036405563354,
+      "learning_rate": 5.038386629905475e-05,
+      "loss": 0.3537,
+      "step": 541
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.444980651140213,
+      "learning_rate": 5.023032122750759e-05,
+      "loss": 0.339,
+      "step": 542
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.363412082195282,
+      "learning_rate": 5.007677398384901e-05,
+      "loss": 0.2565,
+      "step": 543
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4025689363479614,
+      "learning_rate": 4.9923226016151e-05,
+      "loss": 0.3197,
+      "step": 544
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.39782440662384033,
+      "learning_rate": 4.976967877249242e-05,
+      "loss": 0.3417,
+      "step": 545
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.29611390829086304,
+      "learning_rate": 4.961613370094527e-05,
+      "loss": 0.2382,
+      "step": 546
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.41448062658309937,
+      "learning_rate": 4.9462592249561095e-05,
+      "loss": 0.3523,
+      "step": 547
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3686479330062866,
+      "learning_rate": 4.9309055866357295e-05,
+      "loss": 0.2325,
+      "step": 548
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.40558409690856934,
+      "learning_rate": 4.915552599930345e-05,
+      "loss": 0.3463,
+      "step": 549
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.40102720260620117,
+      "learning_rate": 4.900200409630771e-05,
+      "loss": 0.2792,
+      "step": 550
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.344117134809494,
+      "learning_rate": 4.884849160520311e-05,
+      "loss": 0.2309,
+      "step": 551
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4005812108516693,
+      "learning_rate": 4.869498997373393e-05,
+      "loss": 0.2918,
+      "step": 552
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4510882496833801,
+      "learning_rate": 4.8541500649542014e-05,
+      "loss": 0.3854,
+      "step": 553
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4111492931842804,
+      "learning_rate": 4.838802508015316e-05,
+      "loss": 0.3009,
+      "step": 554
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.41969767212867737,
+      "learning_rate": 4.823456471296344e-05,
+      "loss": 0.2923,
+      "step": 555
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3440259099006653,
+      "learning_rate": 4.8081120995225574e-05,
+      "loss": 0.2677,
+      "step": 556
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.44143709540367126,
+      "learning_rate": 4.792769537403523e-05,
+      "loss": 0.4032,
+      "step": 557
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4543381631374359,
+      "learning_rate": 4.777428929631743e-05,
+      "loss": 0.3774,
+      "step": 558
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.44187963008880615,
+      "learning_rate": 4.762090420881289e-05,
+      "loss": 0.3666,
+      "step": 559
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4897536635398865,
+      "learning_rate": 4.746754155806437e-05,
+      "loss": 0.4359,
+      "step": 560
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3388974368572235,
+      "learning_rate": 4.731420279040303e-05,
+      "loss": 0.293,
+      "step": 561
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.34526583552360535,
+      "learning_rate": 4.716088935193479e-05,
+      "loss": 0.1805,
+      "step": 562
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.43450701236724854,
+      "learning_rate": 4.7007602688526695e-05,
+      "loss": 0.3381,
+      "step": 563
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3250710070133209,
+      "learning_rate": 4.6854344245793305e-05,
+      "loss": 0.2123,
+      "step": 564
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3698304295539856,
+      "learning_rate": 4.6701115469083e-05,
+      "loss": 0.2623,
+      "step": 565
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3884919285774231,
+      "learning_rate": 4.65479178034644e-05,
+      "loss": 0.1959,
+      "step": 566
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3933667540550232,
+      "learning_rate": 4.6394752693712734e-05,
+      "loss": 0.2888,
+      "step": 567
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4416023790836334,
+      "learning_rate": 4.624162158429619e-05,
+      "loss": 0.2658,
+      "step": 568
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4641416668891907,
+      "learning_rate": 4.608852591936231e-05,
+      "loss": 0.3646,
+      "step": 569
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4375886619091034,
+      "learning_rate": 4.593546714272438e-05,
+      "loss": 0.3381,
+      "step": 570
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4571078419685364,
+      "learning_rate": 4.578244669784777e-05,
+      "loss": 0.3569,
+      "step": 571
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3207874000072479,
+      "learning_rate": 4.562946602783636e-05,
+      "loss": 0.1999,
+      "step": 572
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.49878060817718506,
+      "learning_rate": 4.547652657541897e-05,
+      "loss": 0.4471,
+      "step": 573
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.41017472743988037,
+      "learning_rate": 4.532362978293564e-05,
+      "loss": 0.3663,
+      "step": 574
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.35150402784347534,
+      "learning_rate": 4.517077709232411e-05,
+      "loss": 0.2319,
+      "step": 575
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.49029800295829773,
+      "learning_rate": 4.501796994510622e-05,
+      "loss": 0.3865,
+      "step": 576
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2655192017555237,
+      "learning_rate": 4.4865209782374304e-05,
+      "loss": 0.1876,
+      "step": 577
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3785124719142914,
+      "learning_rate": 4.471249804477758e-05,
+      "loss": 0.2123,
+      "step": 578
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.37880706787109375,
+      "learning_rate": 4.4559836172508565e-05,
+      "loss": 0.3157,
+      "step": 579
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9027198553085327,
+      "learning_rate": 4.4407225605289547e-05,
+      "loss": 0.2307,
+      "step": 580
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.31004229187965393,
+      "learning_rate": 4.4254667782358924e-05,
+      "loss": 0.247,
+      "step": 581
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.42295175790786743,
+      "learning_rate": 4.410216414245771e-05,
+      "loss": 0.3613,
+      "step": 582
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.37426963448524475,
+      "learning_rate": 4.394971612381591e-05,
+      "loss": 0.2915,
+      "step": 583
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4083792269229889,
+      "learning_rate": 4.379732516413897e-05,
+      "loss": 0.2728,
+      "step": 584
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.39929839968681335,
+      "learning_rate": 4.364499270059424e-05,
+      "loss": 0.2983,
+      "step": 585
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.39979851245880127,
+      "learning_rate": 4.34927201697974e-05,
+      "loss": 0.3143,
+      "step": 586
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.38498011231422424,
+      "learning_rate": 4.334050900779893e-05,
+      "loss": 0.2758,
+      "step": 587
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4555414021015167,
+      "learning_rate": 4.3188360650070525e-05,
+      "loss": 0.3462,
+      "step": 588
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.35760870575904846,
+      "learning_rate": 4.303627653149159e-05,
+      "loss": 0.22,
+      "step": 589
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.36845043301582336,
+      "learning_rate": 4.288425808633575e-05,
+      "loss": 0.2353,
+      "step": 590
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3949494957923889,
+      "learning_rate": 4.2732306748257224e-05,
+      "loss": 0.2841,
+      "step": 591
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3524051606655121,
+      "learning_rate": 4.258042395027738e-05,
+      "loss": 0.2303,
+      "step": 592
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4188311696052551,
+      "learning_rate": 4.2428611124771184e-05,
+      "loss": 0.2779,
+      "step": 593
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3856961131095886,
+      "learning_rate": 4.2276869703453724e-05,
+      "loss": 0.2968,
+      "step": 594
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.39202168583869934,
+      "learning_rate": 4.21252011173667e-05,
+      "loss": 0.2622,
+      "step": 595
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.40051335096359253,
+      "learning_rate": 4.1973606796864884e-05,
+      "loss": 0.2712,
+      "step": 596
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.43389976024627686,
+      "learning_rate": 4.182208817160269e-05,
+      "loss": 0.3661,
+      "step": 597
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3527527153491974,
+      "learning_rate": 4.167064667052065e-05,
+      "loss": 0.2428,
+      "step": 598
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4219164550304413,
+      "learning_rate": 4.151928372183198e-05,
+      "loss": 0.3061,
+      "step": 599
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4185023605823517,
+      "learning_rate": 4.136800075300906e-05,
+      "loss": 0.3067,
+      "step": 600
+    },
+    {
+      "epoch": 0.57,
+      "eval_loss": 0.28891077637672424,
+      "eval_runtime": 361.7732,
+      "eval_samples_per_second": 0.943,
+      "eval_steps_per_second": 0.238,
+      "step": 600
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.47042807936668396,
+      "learning_rate": 4.1216799190770006e-05,
+      "loss": 0.4265,
+      "step": 601
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3343164026737213,
+      "learning_rate": 4.1065680461065194e-05,
+      "loss": 0.2385,
+      "step": 602
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3730446994304657,
+      "learning_rate": 4.0914645989063845e-05,
+      "loss": 0.2809,
+      "step": 603
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4188142716884613,
+      "learning_rate": 4.076369719914055e-05,
+      "loss": 0.2773,
+      "step": 604
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5032288432121277,
+      "learning_rate": 4.061283551486185e-05,
+      "loss": 0.4581,
+      "step": 605
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4637194573879242,
+      "learning_rate": 4.0462062358972796e-05,
+      "loss": 0.4043,
+      "step": 606
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4894029200077057,
+      "learning_rate": 4.03113791533836e-05,
+      "loss": 0.4303,
+      "step": 607
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4027673006057739,
+      "learning_rate": 4.016078731915608e-05,
+      "loss": 0.2806,
+      "step": 608
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4372347295284271,
+      "learning_rate": 4.001028827649046e-05,
+      "loss": 0.3118,
+      "step": 609
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3647504448890686,
+      "learning_rate": 3.9859883444711794e-05,
+      "loss": 0.2655,
+      "step": 610
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3525468111038208,
+      "learning_rate": 3.970957424225666e-05,
+      "loss": 0.2481,
+      "step": 611
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3449767231941223,
+      "learning_rate": 3.955936208665979e-05,
+      "loss": 0.2366,
+      "step": 612
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3690239191055298,
+      "learning_rate": 3.940924839454066e-05,
+      "loss": 0.2782,
+      "step": 613
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.41123005747795105,
+      "learning_rate": 3.925923458159023e-05,
+      "loss": 0.3812,
+      "step": 614
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.36917823553085327,
+      "learning_rate": 3.9109322062557424e-05,
+      "loss": 0.2802,
+      "step": 615
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.28429797291755676,
+      "learning_rate": 3.895951225123595e-05,
+      "loss": 0.1897,
+      "step": 616
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.37444809079170227,
+      "learning_rate": 3.880980656045087e-05,
+      "loss": 0.3077,
+      "step": 617
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3328472971916199,
+      "learning_rate": 3.866020640204533e-05,
+      "loss": 0.2611,
+      "step": 618
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3674910366535187,
+      "learning_rate": 3.851071318686721e-05,
+      "loss": 0.2602,
+      "step": 619
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3362487554550171,
+      "learning_rate": 3.8361328324755825e-05,
+      "loss": 0.2294,
+      "step": 620
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4096759259700775,
+      "learning_rate": 3.821205322452863e-05,
+      "loss": 0.3413,
+      "step": 621
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4167114198207855,
+      "learning_rate": 3.8062889293967976e-05,
+      "loss": 0.4133,
+      "step": 622
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.40718039870262146,
+      "learning_rate": 3.791383793980776e-05,
+      "loss": 0.3487,
+      "step": 623
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.29639190435409546,
+      "learning_rate": 3.7764900567720196e-05,
+      "loss": 0.2037,
+      "step": 624
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.41772112250328064,
+      "learning_rate": 3.761607858230257e-05,
+      "loss": 0.3954,
+      "step": 625
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3576328158378601,
+      "learning_rate": 3.746737338706397e-05,
+      "loss": 0.2496,
+      "step": 626
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4080166220664978,
+      "learning_rate": 3.7318786384412074e-05,
+      "loss": 0.3798,
+      "step": 627
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.38553595542907715,
+      "learning_rate": 3.71703189756399e-05,
+      "loss": 0.32,
+      "step": 628
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.36167973279953003,
+      "learning_rate": 3.70219725609126e-05,
+      "loss": 0.2788,
+      "step": 629
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3779031038284302,
+      "learning_rate": 3.6873748539254246e-05,
+      "loss": 0.2659,
+      "step": 630
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.42400118708610535,
+      "learning_rate": 3.6725648308534696e-05,
+      "loss": 0.3887,
+      "step": 631
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3923243582248688,
+      "learning_rate": 3.65776732654563e-05,
+      "loss": 0.3409,
+      "step": 632
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.39692896604537964,
+      "learning_rate": 3.6429824805540815e-05,
+      "loss": 0.3548,
+      "step": 633
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3133631944656372,
+      "learning_rate": 3.628210432311621e-05,
+      "loss": 0.2056,
+      "step": 634
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4378962814807892,
+      "learning_rate": 3.6134513211303556e-05,
+      "loss": 0.3752,
+      "step": 635
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4492775499820709,
+      "learning_rate": 3.598705286200382e-05,
+      "loss": 0.2888,
+      "step": 636
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.41877108812332153,
+      "learning_rate": 3.58397246658848e-05,
+      "loss": 0.3696,
+      "step": 637
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4388618767261505,
+      "learning_rate": 3.5692530012367955e-05,
+      "loss": 0.3458,
+      "step": 638
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.39103490114212036,
+      "learning_rate": 3.554547028961537e-05,
+      "loss": 0.2919,
+      "step": 639
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.35321909189224243,
+      "learning_rate": 3.5398546884516604e-05,
+      "loss": 0.2159,
+      "step": 640
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.40863534808158875,
+      "learning_rate": 3.5251761182675625e-05,
+      "loss": 0.3496,
+      "step": 641
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.42295730113983154,
+      "learning_rate": 3.510511456839777e-05,
+      "loss": 0.3141,
+      "step": 642
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4626457393169403,
+      "learning_rate": 3.495860842467664e-05,
+      "loss": 0.3515,
+      "step": 643
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.38017332553863525,
+      "learning_rate": 3.481224413318114e-05,
+      "loss": 0.3137,
+      "step": 644
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4441210925579071,
+      "learning_rate": 3.4666023074242356e-05,
+      "loss": 0.3372,
+      "step": 645
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4361349046230316,
+      "learning_rate": 3.4519946626840573e-05,
+      "loss": 0.3249,
+      "step": 646
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4044022560119629,
+      "learning_rate": 3.4374016168592296e-05,
+      "loss": 0.306,
+      "step": 647
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.42988958954811096,
+      "learning_rate": 3.422823307573722e-05,
+      "loss": 0.3451,
+      "step": 648
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4294050931930542,
+      "learning_rate": 3.40825987231253e-05,
+      "loss": 0.3824,
+      "step": 649
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.4360145926475525,
+      "learning_rate": 3.393711448420372e-05,
+      "loss": 0.3011,
+      "step": 650
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3989863693714142,
+      "learning_rate": 3.379178173100396e-05,
+      "loss": 0.2847,
+      "step": 651
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3406922519207001,
+      "learning_rate": 3.364660183412892e-05,
+      "loss": 0.1905,
+      "step": 652
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.45024675130844116,
+      "learning_rate": 3.3501576162739904e-05,
+      "loss": 0.2829,
+      "step": 653
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.38887667655944824,
+      "learning_rate": 3.335670608454376e-05,
+      "loss": 0.3417,
+      "step": 654
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3622775673866272,
+      "learning_rate": 3.321199296577998e-05,
+      "loss": 0.2042,
+      "step": 655
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.41765519976615906,
+      "learning_rate": 3.3067438171207766e-05,
+      "loss": 0.3167,
+      "step": 656
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.4232788681983948,
+      "learning_rate": 3.292304306409325e-05,
+      "loss": 0.3083,
+      "step": 657
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.39359962940216064,
+      "learning_rate": 3.2778809006196564e-05,
+      "loss": 0.2991,
+      "step": 658
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.33065277338027954,
+      "learning_rate": 3.263473735775899e-05,
+      "loss": 0.1897,
+      "step": 659
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3721403181552887,
+      "learning_rate": 3.249082947749019e-05,
+      "loss": 0.2422,
+      "step": 660
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.40056538581848145,
+      "learning_rate": 3.234708672255538e-05,
+      "loss": 0.2884,
+      "step": 661
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.35371360182762146,
+      "learning_rate": 3.220351044856247e-05,
+      "loss": 0.2541,
+      "step": 662
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3902556896209717,
+      "learning_rate": 3.206010200954935e-05,
+      "loss": 0.2688,
+      "step": 663
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4552339017391205,
+      "learning_rate": 3.191686275797107e-05,
+      "loss": 0.3737,
+      "step": 664
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3884057104587555,
+      "learning_rate": 3.177379404468715e-05,
+      "loss": 0.2839,
+      "step": 665
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3160427510738373,
+      "learning_rate": 3.163089721894876e-05,
+      "loss": 0.1856,
+      "step": 666
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4080512225627899,
+      "learning_rate": 3.148817362838606e-05,
+      "loss": 0.317,
+      "step": 667
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4197518229484558,
+      "learning_rate": 3.134562461899545e-05,
+      "loss": 0.3152,
+      "step": 668
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3788788914680481,
+      "learning_rate": 3.120325153512687e-05,
+      "loss": 0.2579,
+      "step": 669
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.40599849820137024,
+      "learning_rate": 3.10610557194712e-05,
+      "loss": 0.2831,
+      "step": 670
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.44576358795166016,
+      "learning_rate": 3.091903851304751e-05,
+      "loss": 0.4155,
+      "step": 671
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.34686872363090515,
+      "learning_rate": 3.077720125519042e-05,
+      "loss": 0.2532,
+      "step": 672
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4478817880153656,
+      "learning_rate": 3.063554528353752e-05,
+      "loss": 0.3305,
+      "step": 673
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.37755653262138367,
+      "learning_rate": 3.0494071934016737e-05,
+      "loss": 0.2545,
+      "step": 674
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.40632200241088867,
+      "learning_rate": 3.03527825408337e-05,
+      "loss": 0.339,
+      "step": 675
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3476860225200653,
+      "learning_rate": 3.0211678436459212e-05,
+      "loss": 0.2573,
+      "step": 676
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.40307381749153137,
+      "learning_rate": 3.0070760951616618e-05,
+      "loss": 0.3119,
+      "step": 677
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.41007760167121887,
+      "learning_rate": 2.993003141526933e-05,
+      "loss": 0.3249,
+      "step": 678
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3331148326396942,
+      "learning_rate": 2.978949115460824e-05,
+      "loss": 0.2164,
+      "step": 679
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4161522686481476,
+      "learning_rate": 2.9649141495039223e-05,
+      "loss": 0.2686,
+      "step": 680
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.38246211409568787,
+      "learning_rate": 2.950898376017064e-05,
+      "loss": 0.2848,
+      "step": 681
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3497626483440399,
+      "learning_rate": 2.9369019271800823e-05,
+      "loss": 0.2332,
+      "step": 682
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.41974884271621704,
+      "learning_rate": 2.9229249349905684e-05,
+      "loss": 0.3509,
+      "step": 683
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3688376545906067,
+      "learning_rate": 2.908967531262618e-05,
+      "loss": 0.2131,
+      "step": 684
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4084636867046356,
+      "learning_rate": 2.895029847625595e-05,
+      "loss": 0.2962,
+      "step": 685
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4998319149017334,
+      "learning_rate": 2.8811120155228844e-05,
+      "loss": 0.431,
+      "step": 686
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.42586615681648254,
+      "learning_rate": 2.8672141662106578e-05,
+      "loss": 0.3273,
+      "step": 687
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4223412275314331,
+      "learning_rate": 2.853336430756631e-05,
+      "loss": 0.345,
+      "step": 688
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4866025149822235,
+      "learning_rate": 2.8394789400388328e-05,
+      "loss": 0.3754,
+      "step": 689
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.46881648898124695,
+      "learning_rate": 2.8256418247443662e-05,
+      "loss": 0.3044,
+      "step": 690
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3632194697856903,
+      "learning_rate": 2.811825215368179e-05,
+      "loss": 0.2588,
+      "step": 691
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.44844743609428406,
+      "learning_rate": 2.798029242211828e-05,
+      "loss": 0.3665,
+      "step": 692
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.40287724137306213,
+      "learning_rate": 2.7842540353822632e-05,
+      "loss": 0.2894,
+      "step": 693
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3673706352710724,
+      "learning_rate": 2.770499724790584e-05,
+      "loss": 0.263,
+      "step": 694
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.41256505250930786,
+      "learning_rate": 2.7567664401508225e-05,
+      "loss": 0.3289,
+      "step": 695
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.45802727341651917,
+      "learning_rate": 2.7430543109787222e-05,
+      "loss": 0.3818,
+      "step": 696
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3631078600883484,
+      "learning_rate": 2.729363466590511e-05,
+      "loss": 0.2179,
+      "step": 697
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.37977662682533264,
+      "learning_rate": 2.7156940361016864e-05,
+      "loss": 0.2731,
+      "step": 698
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3976805508136749,
+      "learning_rate": 2.702046148425795e-05,
+      "loss": 0.2668,
+      "step": 699
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.46016037464141846,
+      "learning_rate": 2.688419932273219e-05,
+      "loss": 0.3662,
+      "step": 700
+    },
+    {
+      "epoch": 0.66,
+      "eval_loss": 0.28551357984542847,
+      "eval_runtime": 366.1246,
+      "eval_samples_per_second": 0.931,
+      "eval_steps_per_second": 0.235,
+      "step": 700
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.42616188526153564,
+      "learning_rate": 2.6748155161499567e-05,
+      "loss": 0.3419,
+      "step": 701
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.33724236488342285,
+      "learning_rate": 2.6612330283564223e-05,
+      "loss": 0.205,
+      "step": 702
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.362380713224411,
+      "learning_rate": 2.6476725969862227e-05,
+      "loss": 0.1811,
+      "step": 703
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.34569138288497925,
+      "learning_rate": 2.634134349924956e-05,
+      "loss": 0.2346,
+      "step": 704
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.425965279340744,
+      "learning_rate": 2.6206184148490066e-05,
+      "loss": 0.3097,
+      "step": 705
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.40935182571411133,
+      "learning_rate": 2.6071249192243363e-05,
+      "loss": 0.3109,
+      "step": 706
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.4438038468360901,
+      "learning_rate": 2.5936539903052892e-05,
+      "loss": 0.3451,
+      "step": 707
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.4346877634525299,
+      "learning_rate": 2.580205755133384e-05,
+      "loss": 0.3923,
+      "step": 708
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.44731929898262024,
+      "learning_rate": 2.5667803405361213e-05,
+      "loss": 0.3944,
+      "step": 709
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.4447585940361023,
+      "learning_rate": 2.5533778731257824e-05,
+      "loss": 0.3867,
+      "step": 710
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.38740718364715576,
+      "learning_rate": 2.5399984792982457e-05,
+      "loss": 0.2711,
+      "step": 711
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.38787004351615906,
+      "learning_rate": 2.5266422852317795e-05,
+      "loss": 0.2978,
+      "step": 712
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4358702600002289,
+      "learning_rate": 2.513309416885865e-05,
+      "loss": 0.3553,
+      "step": 713
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4042569100856781,
+      "learning_rate": 2.500000000000001e-05,
+      "loss": 0.2557,
+      "step": 714
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.409673810005188,
+      "learning_rate": 2.4867141600925214e-05,
+      "loss": 0.2907,
+      "step": 715
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4438133239746094,
+      "learning_rate": 2.4734520224594093e-05,
+      "loss": 0.3444,
+      "step": 716
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4256063401699066,
+      "learning_rate": 2.4602137121731195e-05,
+      "loss": 0.3204,
+      "step": 717
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.39291906356811523,
+      "learning_rate": 2.44699935408139e-05,
+      "loss": 0.3008,
+      "step": 718
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2770516872406006,
+      "learning_rate": 2.433809072806081e-05,
+      "loss": 0.1442,
+      "step": 719
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.40489858388900757,
+      "learning_rate": 2.4206429927419794e-05,
+      "loss": 0.2674,
+      "step": 720
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4332938492298126,
+      "learning_rate": 2.40750123805564e-05,
+      "loss": 0.2637,
+      "step": 721
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.422156423330307,
+      "learning_rate": 2.3943839326842092e-05,
+      "loss": 0.3095,
+      "step": 722
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3582759499549866,
+      "learning_rate": 2.3812912003342568e-05,
+      "loss": 0.2502,
+      "step": 723
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4367879331111908,
+      "learning_rate": 2.3682231644806108e-05,
+      "loss": 0.3335,
+      "step": 724
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3726557493209839,
+      "learning_rate": 2.3551799483651894e-05,
+      "loss": 0.2839,
+      "step": 725
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.49943849444389343,
+      "learning_rate": 2.342161674995843e-05,
+      "loss": 0.3748,
+      "step": 726
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.39246371388435364,
+      "learning_rate": 2.3291684671451903e-05,
+      "loss": 0.3061,
+      "step": 727
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.38669201731681824,
+      "learning_rate": 2.3162004473494657e-05,
+      "loss": 0.2495,
+      "step": 728
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.37277933955192566,
+      "learning_rate": 2.3032577379073577e-05,
+      "loss": 0.305,
+      "step": 729
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.373652845621109,
+      "learning_rate": 2.2903404608788582e-05,
+      "loss": 0.242,
+      "step": 730
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4035954177379608,
+      "learning_rate": 2.2774487380841115e-05,
+      "loss": 0.3202,
+      "step": 731
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.41326603293418884,
+      "learning_rate": 2.2645826911022655e-05,
+      "loss": 0.3054,
+      "step": 732
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.40127384662628174,
+      "learning_rate": 2.2517424412703252e-05,
+      "loss": 0.2723,
+      "step": 733
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.34013038873672485,
+      "learning_rate": 2.2389281096820075e-05,
+      "loss": 0.2478,
+      "step": 734
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.41174599528312683,
+      "learning_rate": 2.2261398171865978e-05,
+      "loss": 0.3147,
+      "step": 735
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3763292133808136,
+      "learning_rate": 2.2133776843878186e-05,
+      "loss": 0.2602,
+      "step": 736
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4248455762863159,
+      "learning_rate": 2.2006418316426775e-05,
+      "loss": 0.2847,
+      "step": 737
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3882761597633362,
+      "learning_rate": 2.187932379060348e-05,
+      "loss": 0.3172,
+      "step": 738
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3807338774204254,
+      "learning_rate": 2.1752494465010242e-05,
+      "loss": 0.2203,
+      "step": 739
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.41011834144592285,
+      "learning_rate": 2.1625931535747964e-05,
+      "loss": 0.325,
+      "step": 740
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4191513657569885,
+      "learning_rate": 2.1499636196405228e-05,
+      "loss": 0.3497,
+      "step": 741
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.41447946429252625,
+      "learning_rate": 2.1373609638047032e-05,
+      "loss": 0.2694,
+      "step": 742
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.40466463565826416,
+      "learning_rate": 2.1247853049203543e-05,
+      "loss": 0.2593,
+      "step": 743
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4141696095466614,
+      "learning_rate": 2.112236761585892e-05,
+      "loss": 0.3247,
+      "step": 744
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3989594876766205,
+      "learning_rate": 2.09971545214401e-05,
+      "loss": 0.3054,
+      "step": 745
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4133450984954834,
+      "learning_rate": 2.087221494680563e-05,
+      "loss": 0.3319,
+      "step": 746
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3053062856197357,
+      "learning_rate": 2.074755007023461e-05,
+      "loss": 0.2223,
+      "step": 747
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.28034481406211853,
+      "learning_rate": 2.0623161067415463e-05,
+      "loss": 0.1555,
+      "step": 748
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4164118468761444,
+      "learning_rate": 2.049904911143492e-05,
+      "loss": 0.3041,
+      "step": 749
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3929029703140259,
+      "learning_rate": 2.0375215372766944e-05,
+      "loss": 0.3018,
+      "step": 750
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4244453012943268,
+      "learning_rate": 2.025166101926168e-05,
+      "loss": 0.4086,
+      "step": 751
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.33513087034225464,
+      "learning_rate": 2.012838721613447e-05,
+      "loss": 0.2147,
+      "step": 752
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.432666540145874,
+      "learning_rate": 2.0005395125954812e-05,
+      "loss": 0.3837,
+      "step": 753
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3546793460845947,
+      "learning_rate": 1.9882685908635462e-05,
+      "loss": 0.2649,
+      "step": 754
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.4040340781211853,
+      "learning_rate": 1.9760260721421426e-05,
+      "loss": 0.3134,
+      "step": 755
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.44750577211380005,
+      "learning_rate": 1.963812071887913e-05,
+      "loss": 0.3559,
+      "step": 756
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.34949734807014465,
+      "learning_rate": 1.951626705288544e-05,
+      "loss": 0.2943,
+      "step": 757
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3807576298713684,
+      "learning_rate": 1.9394700872616855e-05,
+      "loss": 0.2297,
+      "step": 758
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.4196769595146179,
+      "learning_rate": 1.9273423324538658e-05,
+      "loss": 0.3513,
+      "step": 759
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3704902231693268,
+      "learning_rate": 1.9152435552394104e-05,
+      "loss": 0.2569,
+      "step": 760
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.32672318816185,
+      "learning_rate": 1.903173869719362e-05,
+      "loss": 0.2352,
+      "step": 761
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.345194011926651,
+      "learning_rate": 1.891133389720407e-05,
+      "loss": 0.2641,
+      "step": 762
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.4048096835613251,
+      "learning_rate": 1.8791222287937982e-05,
+      "loss": 0.2948,
+      "step": 763
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.36946576833724976,
+      "learning_rate": 1.8671405002142918e-05,
+      "loss": 0.2121,
+      "step": 764
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.38662201166152954,
+      "learning_rate": 1.855188316979068e-05,
+      "loss": 0.3062,
+      "step": 765
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.44647669792175293,
+      "learning_rate": 1.8432657918066732e-05,
+      "loss": 0.3065,
+      "step": 766
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.4243049919605255,
+      "learning_rate": 1.831373037135955e-05,
+      "loss": 0.3693,
+      "step": 767
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.35591912269592285,
+      "learning_rate": 1.8195101651250017e-05,
+      "loss": 0.2103,
+      "step": 768
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.42392200231552124,
+      "learning_rate": 1.807677287650083e-05,
+      "loss": 0.2975,
+      "step": 769
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3883664011955261,
+      "learning_rate": 1.7958745163045986e-05,
+      "loss": 0.2916,
+      "step": 770
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.47300446033477783,
+      "learning_rate": 1.7841019623980217e-05,
+      "loss": 0.3095,
+      "step": 771
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.34601467847824097,
+      "learning_rate": 1.77235973695485e-05,
+      "loss": 0.2127,
+      "step": 772
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.39340120553970337,
+      "learning_rate": 1.760647950713566e-05,
+      "loss": 0.3155,
+      "step": 773
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.37569355964660645,
+      "learning_rate": 1.74896671412558e-05,
+      "loss": 0.254,
+      "step": 774
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.4376790225505829,
+      "learning_rate": 1.737316137354197e-05,
+      "loss": 0.3762,
+      "step": 775
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3843180239200592,
+      "learning_rate": 1.725696330273575e-05,
+      "loss": 0.3154,
+      "step": 776
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3781803846359253,
+      "learning_rate": 1.7141074024676912e-05,
+      "loss": 0.2815,
+      "step": 777
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.34335994720458984,
+      "learning_rate": 1.7025494632293048e-05,
+      "loss": 0.1914,
+      "step": 778
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.41264262795448303,
+      "learning_rate": 1.6910226215589303e-05,
+      "loss": 0.3454,
+      "step": 779
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.42045897245407104,
+      "learning_rate": 1.6795269861638042e-05,
+      "loss": 0.3402,
+      "step": 780
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.38134869933128357,
+      "learning_rate": 1.6680626654568686e-05,
+      "loss": 0.2552,
+      "step": 781
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.45176658034324646,
+      "learning_rate": 1.656629767555739e-05,
+      "loss": 0.3775,
+      "step": 782
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.4160310924053192,
+      "learning_rate": 1.645228400281689e-05,
+      "loss": 0.2807,
+      "step": 783
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.37641626596450806,
+      "learning_rate": 1.6338586711586356e-05,
+      "loss": 0.2435,
+      "step": 784
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3558084964752197,
+      "learning_rate": 1.6225206874121218e-05,
+      "loss": 0.3089,
+      "step": 785
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3629785478115082,
+      "learning_rate": 1.6112145559683055e-05,
+      "loss": 0.2732,
+      "step": 786
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4224972426891327,
+      "learning_rate": 1.5999403834529548e-05,
+      "loss": 0.3871,
+      "step": 787
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.362154096364975,
+      "learning_rate": 1.5886982761904377e-05,
+      "loss": 0.2531,
+      "step": 788
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4315077066421509,
+      "learning_rate": 1.5774883402027208e-05,
+      "loss": 0.3197,
+      "step": 789
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4491424858570099,
+      "learning_rate": 1.5663106812083743e-05,
+      "loss": 0.3072,
+      "step": 790
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3408665657043457,
+      "learning_rate": 1.555165404621567e-05,
+      "loss": 0.2549,
+      "step": 791
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4255100190639496,
+      "learning_rate": 1.5440526155510763e-05,
+      "loss": 0.3725,
+      "step": 792
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4389415681362152,
+      "learning_rate": 1.532972418799298e-05,
+      "loss": 0.3652,
+      "step": 793
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.42644500732421875,
+      "learning_rate": 1.5219249188612556e-05,
+      "loss": 0.2914,
+      "step": 794
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4122108519077301,
+      "learning_rate": 1.5109102199236152e-05,
+      "loss": 0.326,
+      "step": 795
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4755607843399048,
+      "learning_rate": 1.4999284258637053e-05,
+      "loss": 0.3403,
+      "step": 796
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.40237653255462646,
+      "learning_rate": 1.488979640248534e-05,
+      "loss": 0.2904,
+      "step": 797
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.34840846061706543,
+      "learning_rate": 1.4780639663338125e-05,
+      "loss": 0.2531,
+      "step": 798
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.4074588716030121,
+      "learning_rate": 1.4671815070629868e-05,
+      "loss": 0.2274,
+      "step": 799
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3214362859725952,
+      "learning_rate": 1.4563323650662586e-05,
+      "loss": 0.1942,
+      "step": 800
+    },
+    {
+      "epoch": 0.76,
+      "eval_loss": 0.2829459607601166,
+      "eval_runtime": 363.763,
+      "eval_samples_per_second": 0.937,
+      "eval_steps_per_second": 0.236,
+      "step": 800
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3532373309135437,
+      "learning_rate": 1.4455166426596223e-05,
+      "loss": 0.2042,
+      "step": 801
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.48360303044319153,
+      "learning_rate": 1.434734441843899e-05,
+      "loss": 0.4728,
+      "step": 802
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3731449246406555,
+      "learning_rate": 1.4239858643037751e-05,
+      "loss": 0.2899,
+      "step": 803
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3608839511871338,
+      "learning_rate": 1.4132710114068426e-05,
+      "loss": 0.2601,
+      "step": 804
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3535532057285309,
+      "learning_rate": 1.4025899842026442e-05,
+      "loss": 0.1733,
+      "step": 805
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3848244845867157,
+      "learning_rate": 1.3919428834217163e-05,
+      "loss": 0.2156,
+      "step": 806
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.36804237961769104,
+      "learning_rate": 1.3813298094746491e-05,
+      "loss": 0.2363,
+      "step": 807
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.34813761711120605,
+      "learning_rate": 1.3707508624511262e-05,
+      "loss": 0.2022,
+      "step": 808
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.4521434009075165,
+      "learning_rate": 1.36020614211899e-05,
+      "loss": 0.3591,
+      "step": 809
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.360469251871109,
+      "learning_rate": 1.349695747923298e-05,
+      "loss": 0.2504,
+      "step": 810
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.35620465874671936,
+      "learning_rate": 1.339219778985385e-05,
+      "loss": 0.2724,
+      "step": 811
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.4041351079940796,
+      "learning_rate": 1.3287783341019278e-05,
+      "loss": 0.2839,
+      "step": 812
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.4110220670700073,
+      "learning_rate": 1.3183715117440142e-05,
+      "loss": 0.3289,
+      "step": 813
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.4474922716617584,
+      "learning_rate": 1.3079994100562159e-05,
+      "loss": 0.3108,
+      "step": 814
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.40211912989616394,
+      "learning_rate": 1.2976621268556571e-05,
+      "loss": 0.3196,
+      "step": 815
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.37706807255744934,
+      "learning_rate": 1.2873597596311027e-05,
+      "loss": 0.2776,
+      "step": 816
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.45289844274520874,
+      "learning_rate": 1.2770924055420258e-05,
+      "loss": 0.297,
+      "step": 817
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.4111120104789734,
+      "learning_rate": 1.2668601614177017e-05,
+      "loss": 0.3389,
+      "step": 818
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.38809487223625183,
+      "learning_rate": 1.2566631237562893e-05,
+      "loss": 0.248,
+      "step": 819
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.37415727972984314,
+      "learning_rate": 1.246501388723923e-05,
+      "loss": 0.2548,
+      "step": 820
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.38985398411750793,
+      "learning_rate": 1.2363750521538064e-05,
+      "loss": 0.2638,
+      "step": 821
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.41149115562438965,
+      "learning_rate": 1.2262842095453065e-05,
+      "loss": 0.3716,
+      "step": 822
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.4018831253051758,
+      "learning_rate": 1.2162289560630524e-05,
+      "loss": 0.2886,
+      "step": 823
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.43984273076057434,
+      "learning_rate": 1.2062093865360458e-05,
+      "loss": 0.3474,
+      "step": 824
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3629956841468811,
+      "learning_rate": 1.1962255954567537e-05,
+      "loss": 0.2357,
+      "step": 825
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.372852623462677,
+      "learning_rate": 1.1862776769802276e-05,
+      "loss": 0.2672,
+      "step": 826
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.4188729524612427,
+      "learning_rate": 1.1763657249232107e-05,
+      "loss": 0.3281,
+      "step": 827
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.33174970746040344,
+      "learning_rate": 1.1664898327632551e-05,
+      "loss": 0.237,
+      "step": 828
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.4201258718967438,
+      "learning_rate": 1.1566500936378388e-05,
+      "loss": 0.337,
+      "step": 829
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.49972257018089294,
+      "learning_rate": 1.146846600343488e-05,
+      "loss": 0.3505,
+      "step": 830
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.458884060382843,
+      "learning_rate": 1.1370794453349037e-05,
+      "loss": 0.4045,
+      "step": 831
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5413430333137512,
+      "learning_rate": 1.1273487207240847e-05,
+      "loss": 0.2857,
+      "step": 832
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5663003921508789,
+      "learning_rate": 1.1176545182794674e-05,
+      "loss": 0.1916,
+      "step": 833
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3781757354736328,
+      "learning_rate": 1.1079969294250514e-05,
+      "loss": 0.2698,
+      "step": 834
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.4103042781352997,
+      "learning_rate": 1.0983760452395413e-05,
+      "loss": 0.3254,
+      "step": 835
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.4255097210407257,
+      "learning_rate": 1.0887919564554894e-05,
+      "loss": 0.3273,
+      "step": 836
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.4411180913448334,
+      "learning_rate": 1.079244753458437e-05,
+      "loss": 0.3329,
+      "step": 837
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.38623252511024475,
+      "learning_rate": 1.0697345262860636e-05,
+      "loss": 0.3241,
+      "step": 838
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3975672721862793,
+      "learning_rate": 1.0602613646273374e-05,
+      "loss": 0.359,
+      "step": 839
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4542789161205292,
+      "learning_rate": 1.0508253578216693e-05,
+      "loss": 0.3751,
+      "step": 840
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.44021835923194885,
+      "learning_rate": 1.0414265948580692e-05,
+      "loss": 0.3188,
+      "step": 841
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.42216500639915466,
+      "learning_rate": 1.032065164374313e-05,
+      "loss": 0.2518,
+      "step": 842
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3464965224266052,
+      "learning_rate": 1.0227411546560962e-05,
+      "loss": 0.2523,
+      "step": 843
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.387981653213501,
+      "learning_rate": 1.01345465363621e-05,
+      "loss": 0.293,
+      "step": 844
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.426349937915802,
+      "learning_rate": 1.0042057488937067e-05,
+      "loss": 0.2949,
+      "step": 845
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4316929876804352,
+      "learning_rate": 9.949945276530781e-06,
+      "loss": 0.391,
+      "step": 846
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.32711362838745117,
+      "learning_rate": 9.858210767834292e-06,
+      "loss": 0.2456,
+      "step": 847
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4321954846382141,
+      "learning_rate": 9.766854827976617e-06,
+      "loss": 0.2865,
+      "step": 848
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3895399272441864,
+      "learning_rate": 9.675878318516546e-06,
+      "loss": 0.2167,
+      "step": 849
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5271614193916321,
+      "learning_rate": 9.58528209743459e-06,
+      "loss": 0.3951,
+      "step": 850
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3950306475162506,
+      "learning_rate": 9.495067019124792e-06,
+      "loss": 0.3125,
+      "step": 851
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.4167860448360443,
+      "learning_rate": 9.405233934386726e-06,
+      "loss": 0.2796,
+      "step": 852
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3829375207424164,
+      "learning_rate": 9.315783690417479e-06,
+      "loss": 0.2468,
+      "step": 853
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.39671874046325684,
+      "learning_rate": 9.226717130803636e-06,
+      "loss": 0.2592,
+      "step": 854
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.38071882724761963,
+      "learning_rate": 9.138035095513336e-06,
+      "loss": 0.3262,
+      "step": 855
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.32630348205566406,
+      "learning_rate": 9.049738420888349e-06,
+      "loss": 0.2522,
+      "step": 856
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.394283652305603,
+      "learning_rate": 8.961827939636196e-06,
+      "loss": 0.2685,
+      "step": 857
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3125092089176178,
+      "learning_rate": 8.87430448082227e-06,
+      "loss": 0.1336,
+      "step": 858
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3436705768108368,
+      "learning_rate": 8.787168869862066e-06,
+      "loss": 0.2223,
+      "step": 859
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.41198545694351196,
+      "learning_rate": 8.700421928513352e-06,
+      "loss": 0.3889,
+      "step": 860
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.40527185797691345,
+      "learning_rate": 8.614064474868421e-06,
+      "loss": 0.2744,
+      "step": 861
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.4188045561313629,
+      "learning_rate": 8.528097323346407e-06,
+      "loss": 0.3281,
+      "step": 862
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3231567442417145,
+      "learning_rate": 8.442521284685573e-06,
+      "loss": 0.1885,
+      "step": 863
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3571144640445709,
+      "learning_rate": 8.357337165935675e-06,
+      "loss": 0.2651,
+      "step": 864
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.44496673345565796,
+      "learning_rate": 8.27254577045039e-06,
+      "loss": 0.3386,
+      "step": 865
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.37217986583709717,
+      "learning_rate": 8.188147897879667e-06,
+      "loss": 0.2207,
+      "step": 866
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3924148678779602,
+      "learning_rate": 8.104144344162229e-06,
+      "loss": 0.2454,
+      "step": 867
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.35032355785369873,
+      "learning_rate": 8.02053590151805e-06,
+      "loss": 0.2578,
+      "step": 868
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.4049111008644104,
+      "learning_rate": 7.937323358440935e-06,
+      "loss": 0.262,
+      "step": 869
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.31789684295654297,
+      "learning_rate": 7.854507499691005e-06,
+      "loss": 0.156,
+      "step": 870
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.4001528024673462,
+      "learning_rate": 7.772089106287344e-06,
+      "loss": 0.314,
+      "step": 871
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.42885470390319824,
+      "learning_rate": 7.690068955500624e-06,
+      "loss": 0.3621,
+      "step": 872
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.37726956605911255,
+      "learning_rate": 7.608447820845771e-06,
+      "loss": 0.2335,
+      "step": 873
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.37034162878990173,
+      "learning_rate": 7.527226472074678e-06,
+      "loss": 0.2606,
+      "step": 874
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.40577173233032227,
+      "learning_rate": 7.446405675168938e-06,
+      "loss": 0.3217,
+      "step": 875
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3697319030761719,
+      "learning_rate": 7.365986192332625e-06,
+      "loss": 0.1921,
+      "step": 876
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.36988091468811035,
+      "learning_rate": 7.285968781985092e-06,
+      "loss": 0.2452,
+      "step": 877
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3300369679927826,
+      "learning_rate": 7.206354198753862e-06,
+      "loss": 0.1979,
+      "step": 878
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.4411529004573822,
+      "learning_rate": 7.127143193467445e-06,
+      "loss": 0.3513,
+      "step": 879
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5053805112838745,
+      "learning_rate": 7.048336513148307e-06,
+      "loss": 0.3615,
+      "step": 880
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5213754773139954,
+      "learning_rate": 6.969934901005809e-06,
+      "loss": 0.4167,
+      "step": 881
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3984251916408539,
+      "learning_rate": 6.8919390964291895e-06,
+      "loss": 0.3323,
+      "step": 882
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3953109383583069,
+      "learning_rate": 6.814349834980621e-06,
+      "loss": 0.2935,
+      "step": 883
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.453165203332901,
+      "learning_rate": 6.7371678483882264e-06,
+      "loss": 0.4064,
+      "step": 884
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3721240162849426,
+      "learning_rate": 6.660393864539222e-06,
+      "loss": 0.2607,
+      "step": 885
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3390076160430908,
+      "learning_rate": 6.58402860747302e-06,
+      "loss": 0.2042,
+      "step": 886
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3938121795654297,
+      "learning_rate": 6.508072797374454e-06,
+      "loss": 0.3204,
+      "step": 887
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4577292799949646,
+      "learning_rate": 6.4325271505669025e-06,
+      "loss": 0.3225,
+      "step": 888
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.418714314699173,
+      "learning_rate": 6.3573923795056e-06,
+      "loss": 0.3154,
+      "step": 889
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.41570523381233215,
+      "learning_rate": 6.282669192770896e-06,
+      "loss": 0.333,
+      "step": 890
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.41950684785842896,
+      "learning_rate": 6.208358295061572e-06,
+      "loss": 0.2944,
+      "step": 891
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3647615611553192,
+      "learning_rate": 6.134460387188206e-06,
+      "loss": 0.2522,
+      "step": 892
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.387281209230423,
+      "learning_rate": 6.060976166066546e-06,
+      "loss": 0.2683,
+      "step": 893
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.44456931948661804,
+      "learning_rate": 5.9879063247109405e-06,
+      "loss": 0.3275,
+      "step": 894
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.40326932072639465,
+      "learning_rate": 5.915251552227829e-06,
+      "loss": 0.2797,
+      "step": 895
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.4886815845966339,
+      "learning_rate": 5.8430125338092115e-06,
+      "loss": 0.3809,
+      "step": 896
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.36966973543167114,
+      "learning_rate": 5.771189950726191e-06,
+      "loss": 0.2206,
+      "step": 897
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.4477539658546448,
+      "learning_rate": 5.699784480322568e-06,
+      "loss": 0.3625,
+      "step": 898
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.4014408588409424,
+      "learning_rate": 5.628796796008434e-06,
+      "loss": 0.3155,
+      "step": 899
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3876466453075409,
+      "learning_rate": 5.558227567253832e-06,
+      "loss": 0.2865,
+      "step": 900
+    },
+    {
+      "epoch": 0.85,
+      "eval_loss": 0.2800748646259308,
+      "eval_runtime": 363.8216,
+      "eval_samples_per_second": 0.937,
+      "eval_steps_per_second": 0.236,
+      "step": 900
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.44938063621520996,
+      "learning_rate": 5.488077459582425e-06,
+      "loss": 0.3348,
+      "step": 901
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.37703728675842285,
+      "learning_rate": 5.418347134565249e-06,
+      "loss": 0.2928,
+      "step": 902
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.37471407651901245,
+      "learning_rate": 5.349037249814442e-06,
+      "loss": 0.2935,
+      "step": 903
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.423926442861557,
+      "learning_rate": 5.28014845897708e-06,
+      "loss": 0.2895,
+      "step": 904
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.45423102378845215,
+      "learning_rate": 5.211681411728969e-06,
+      "loss": 0.3126,
+      "step": 905
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.37803101539611816,
+      "learning_rate": 5.143636753768549e-06,
+      "loss": 0.2372,
+      "step": 906
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3723331689834595,
+      "learning_rate": 5.076015126810784e-06,
+      "loss": 0.2958,
+      "step": 907
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3997434377670288,
+      "learning_rate": 5.008817168581137e-06,
+      "loss": 0.2758,
+      "step": 908
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.41803666949272156,
+      "learning_rate": 4.94204351280953e-06,
+      "loss": 0.3124,
+      "step": 909
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.42922714352607727,
+      "learning_rate": 4.8756947892243725e-06,
+      "loss": 0.3163,
+      "step": 910
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.37815365195274353,
+      "learning_rate": 4.809771623546627e-06,
+      "loss": 0.217,
+      "step": 911
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.333159863948822,
+      "learning_rate": 4.744274637483936e-06,
+      "loss": 0.2038,
+      "step": 912
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3474457561969757,
+      "learning_rate": 4.6792044487247e-06,
+      "loss": 0.2852,
+      "step": 913
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3912842273712158,
+      "learning_rate": 4.614561670932288e-06,
+      "loss": 0.2713,
+      "step": 914
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3732365369796753,
+      "learning_rate": 4.550346913739256e-06,
+      "loss": 0.2287,
+      "step": 915
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.37471485137939453,
+      "learning_rate": 4.486560782741578e-06,
+      "loss": 0.2219,
+      "step": 916
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.33144327998161316,
+      "learning_rate": 4.423203879492943e-06,
+      "loss": 0.214,
+      "step": 917
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4433518648147583,
+      "learning_rate": 4.360276801499086e-06,
+      "loss": 0.3451,
+      "step": 918
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4415975511074066,
+      "learning_rate": 4.29778014221216e-06,
+      "loss": 0.2725,
+      "step": 919
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4309133291244507,
+      "learning_rate": 4.2357144910251e-06,
+      "loss": 0.2994,
+      "step": 920
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4637148380279541,
+      "learning_rate": 4.174080433266136e-06,
+      "loss": 0.3382,
+      "step": 921
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4580986499786377,
+      "learning_rate": 4.112878550193194e-06,
+      "loss": 0.3218,
+      "step": 922
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3549492061138153,
+      "learning_rate": 4.05210941898847e-06,
+      "loss": 0.2179,
+      "step": 923
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3796776533126831,
+      "learning_rate": 3.991773612752952e-06,
+      "loss": 0.2363,
+      "step": 924
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.4128180146217346,
+      "learning_rate": 3.9318717005010495e-06,
+      "loss": 0.3129,
+      "step": 925
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.386189341545105,
+      "learning_rate": 3.872404247155193e-06,
+      "loss": 0.3162,
+      "step": 926
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.4205312728881836,
+      "learning_rate": 3.8133718135405285e-06,
+      "loss": 0.3472,
+      "step": 927
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3960677981376648,
+      "learning_rate": 3.754774956379614e-06,
+      "loss": 0.2843,
+      "step": 928
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.4133547842502594,
+      "learning_rate": 3.696614228287187e-06,
+      "loss": 0.3069,
+      "step": 929
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.4363679587841034,
+      "learning_rate": 3.6388901777649477e-06,
+      "loss": 0.3128,
+      "step": 930
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.4414587914943695,
+      "learning_rate": 3.581603349196372e-06,
+      "loss": 0.2882,
+      "step": 931
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.46069326996803284,
+      "learning_rate": 3.5247542828415747e-06,
+      "loss": 0.3863,
+      "step": 932
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.35465437173843384,
+      "learning_rate": 3.468343514832251e-06,
+      "loss": 0.229,
+      "step": 933
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3709930181503296,
+      "learning_rate": 3.4123715771665786e-06,
+      "loss": 0.2067,
+      "step": 934
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.39402368664741516,
+      "learning_rate": 3.356838997704226e-06,
+      "loss": 0.3227,
+      "step": 935
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3811134397983551,
+      "learning_rate": 3.3017463001613626e-06,
+      "loss": 0.2466,
+      "step": 936
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.34305959939956665,
+      "learning_rate": 3.2470940041057106e-06,
+      "loss": 0.2065,
+      "step": 937
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.37376561760902405,
+      "learning_rate": 3.1928826249516987e-06,
+      "loss": 0.2088,
+      "step": 938
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.42914170026779175,
+      "learning_rate": 3.139112673955513e-06,
+      "loss": 0.2844,
+      "step": 939
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.39991962909698486,
+      "learning_rate": 3.08578465821035e-06,
+      "loss": 0.3008,
+      "step": 940
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3750627338886261,
+      "learning_rate": 3.0328990806415934e-06,
+      "loss": 0.2889,
+      "step": 941
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.47464945912361145,
+      "learning_rate": 2.9804564400020994e-06,
+      "loss": 0.3441,
+      "step": 942
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.41791674494743347,
+      "learning_rate": 2.9284572308674596e-06,
+      "loss": 0.3341,
+      "step": 943
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.40713489055633545,
+      "learning_rate": 2.8769019436313715e-06,
+      "loss": 0.3142,
+      "step": 944
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.36073338985443115,
+      "learning_rate": 2.825791064500993e-06,
+      "loss": 0.28,
+      "step": 945
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4595106840133667,
+      "learning_rate": 2.7751250754923573e-06,
+      "loss": 0.3045,
+      "step": 946
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.39888688921928406,
+      "learning_rate": 2.7249044544258363e-06,
+      "loss": 0.3151,
+      "step": 947
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.44055163860321045,
+      "learning_rate": 2.675129674921639e-06,
+      "loss": 0.4007,
+      "step": 948
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.506105899810791,
+      "learning_rate": 2.625801206395312e-06,
+      "loss": 0.3858,
+      "step": 949
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3843456208705902,
+      "learning_rate": 2.576919514053355e-06,
+      "loss": 0.2618,
+      "step": 950
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4758426249027252,
+      "learning_rate": 2.528485058888813e-06,
+      "loss": 0.3758,
+      "step": 951
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.40591248869895935,
+      "learning_rate": 2.48049829767692e-06,
+      "loss": 0.327,
+      "step": 952
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4928969442844391,
+      "learning_rate": 2.4329596829708144e-06,
+      "loss": 0.4,
+      "step": 953
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3787176012992859,
+      "learning_rate": 2.385869663097251e-06,
+      "loss": 0.2855,
+      "step": 954
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4510795772075653,
+      "learning_rate": 2.3392286821523724e-06,
+      "loss": 0.3499,
+      "step": 955
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3751595616340637,
+      "learning_rate": 2.2930371799975594e-06,
+      "loss": 0.2166,
+      "step": 956
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3744955062866211,
+      "learning_rate": 2.2472955922552163e-06,
+      "loss": 0.2498,
+      "step": 957
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.42327001690864563,
+      "learning_rate": 2.202004350304715e-06,
+      "loss": 0.3506,
+      "step": 958
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.4099956452846527,
+      "learning_rate": 2.157163881278312e-06,
+      "loss": 0.3139,
+      "step": 959
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3939151167869568,
+      "learning_rate": 2.112774608057111e-06,
+      "loss": 0.2937,
+      "step": 960
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.4438505172729492,
+      "learning_rate": 2.068836949267089e-06,
+      "loss": 0.3383,
+      "step": 961
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.4137350022792816,
+      "learning_rate": 2.0253513192751373e-06,
+      "loss": 0.276,
+      "step": 962
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.39033421874046326,
+      "learning_rate": 1.982318128185151e-06,
+      "loss": 0.2942,
+      "step": 963
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.42454180121421814,
+      "learning_rate": 1.9397377818341944e-06,
+      "loss": 0.3121,
+      "step": 964
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.44790568947792053,
+      "learning_rate": 1.8976106817886196e-06,
+      "loss": 0.3663,
+      "step": 965
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4456357955932617,
+      "learning_rate": 1.855937225340315e-06,
+      "loss": 0.3028,
+      "step": 966
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3860545754432678,
+      "learning_rate": 1.8147178055029579e-06,
+      "loss": 0.269,
+      "step": 967
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.33939704298973083,
+      "learning_rate": 1.7739528110083004e-06,
+      "loss": 0.2498,
+      "step": 968
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.41515877842903137,
+      "learning_rate": 1.7336426263024897e-06,
+      "loss": 0.3072,
+      "step": 969
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3300078809261322,
+      "learning_rate": 1.6937876315424705e-06,
+      "loss": 0.1976,
+      "step": 970
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3762941360473633,
+      "learning_rate": 1.6543882025923886e-06,
+      "loss": 0.2688,
+      "step": 971
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4427281320095062,
+      "learning_rate": 1.6154447110200255e-06,
+      "loss": 0.3461,
+      "step": 972
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3941108286380768,
+      "learning_rate": 1.5769575240933422e-06,
+      "loss": 0.2212,
+      "step": 973
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5374137163162231,
+      "learning_rate": 1.5389270047769578e-06,
+      "loss": 0.4852,
+      "step": 974
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.39923667907714844,
+      "learning_rate": 1.5013535117287647e-06,
+      "loss": 0.2986,
+      "step": 975
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.34776413440704346,
+      "learning_rate": 1.4642373992965364e-06,
+      "loss": 0.1985,
+      "step": 976
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3739027678966522,
+      "learning_rate": 1.4275790175145908e-06,
+      "loss": 0.226,
+      "step": 977
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.36304882168769836,
+      "learning_rate": 1.3913787121004719e-06,
+      "loss": 0.2485,
+      "step": 978
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.4300604462623596,
+      "learning_rate": 1.3556368244517114e-06,
+      "loss": 0.3087,
+      "step": 979
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3972741663455963,
+      "learning_rate": 1.3203536916425841e-06,
+      "loss": 0.2795,
+      "step": 980
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.40066105127334595,
+      "learning_rate": 1.2855296464209688e-06,
+      "loss": 0.3039,
+      "step": 981
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.4563113749027252,
+      "learning_rate": 1.2511650172051637e-06,
+      "loss": 0.4195,
+      "step": 982
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3806251883506775,
+      "learning_rate": 1.217260128080816e-06,
+      "loss": 0.2886,
+      "step": 983
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.38023003935813904,
+      "learning_rate": 1.1838152987978578e-06,
+      "loss": 0.2814,
+      "step": 984
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.4246285855770111,
+      "learning_rate": 1.1508308447674975e-06,
+      "loss": 0.2958,
+      "step": 985
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.395789235830307,
+      "learning_rate": 1.1183070770592441e-06,
+      "loss": 0.2894,
+      "step": 986
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3676890432834625,
+      "learning_rate": 1.0862443023979651e-06,
+      "loss": 0.2567,
+      "step": 987
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3309171795845032,
+      "learning_rate": 1.0546428231609895e-06,
+      "loss": 0.2288,
+      "step": 988
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5018779635429382,
+      "learning_rate": 1.0235029373752758e-06,
+      "loss": 0.3949,
+      "step": 989
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.392005980014801,
+      "learning_rate": 9.928249387145982e-07,
+      "loss": 0.2627,
+      "step": 990
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.37266385555267334,
+      "learning_rate": 9.6260911649676e-07,
+      "loss": 0.2115,
+      "step": 991
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.35400834679603577,
+      "learning_rate": 9.32855755680867e-07,
+      "loss": 0.2217,
+      "step": 992
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3872481882572174,
+      "learning_rate": 9.035651368646648e-07,
+      "loss": 0.3084,
+      "step": 993
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.497378408908844,
+      "learning_rate": 8.747375362818666e-07,
+      "loss": 0.4101,
+      "step": 994
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.43928062915802,
+      "learning_rate": 8.463732257995571e-07,
+      "loss": 0.3203,
+      "step": 995
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.31176984310150146,
+      "learning_rate": 8.184724729156379e-07,
+      "loss": 0.2082,
+      "step": 996
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.4319514334201813,
+      "learning_rate": 7.910355407562742e-07,
+      "loss": 0.2952,
+      "step": 997
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.4082161486148834,
+      "learning_rate": 7.640626880734581e-07,
+      "loss": 0.2191,
+      "step": 998
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3197343945503235,
+      "learning_rate": 7.375541692425326e-07,
+      "loss": 0.2167,
+      "step": 999
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.433448851108551,
+      "learning_rate": 7.115102342598101e-07,
+      "loss": 0.3335,
+      "step": 1000
+    },
+    {
+      "epoch": 0.95,
+      "eval_loss": 0.2786545753479004,
+      "eval_runtime": 360.7934,
+      "eval_samples_per_second": 0.945,
+      "eval_steps_per_second": 0.238,
+      "step": 1000
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.4050236642360687,
+      "learning_rate": 6.859311287402081e-07,
+      "loss": 0.2476,
+      "step": 1001
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.32928723096847534,
+      "learning_rate": 6.608170939149283e-07,
+      "loss": 0.1961,
+      "step": 1002
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.4043203592300415,
+      "learning_rate": 6.361683666291973e-07,
+      "loss": 0.3263,
+      "step": 1003
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.47581639885902405,
+      "learning_rate": 6.119851793400189e-07,
+      "loss": 0.3081,
+      "step": 1004
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.4158116281032562,
+      "learning_rate": 5.882677601139919e-07,
+      "loss": 0.2632,
+      "step": 1005
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.41737163066864014,
+      "learning_rate": 5.650163326251345e-07,
+      "loss": 0.2575,
+      "step": 1006
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.485889732837677,
+      "learning_rate": 5.422311161528193e-07,
+      "loss": 0.3773,
+      "step": 1007
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4189164936542511,
+      "learning_rate": 5.199123255796634e-07,
+      "loss": 0.2623,
+      "step": 1008
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.31727853417396545,
+      "learning_rate": 4.980601713895305e-07,
+      "loss": 0.172,
+      "step": 1009
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4250466823577881,
+      "learning_rate": 4.7667485966552685e-07,
+      "loss": 0.2423,
+      "step": 1010
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3881832957267761,
+      "learning_rate": 4.55756592088058e-07,
+      "loss": 0.3309,
+      "step": 1011
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4162770211696625,
+      "learning_rate": 4.353055659329419e-07,
+      "loss": 0.3784,
+      "step": 1012
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4592972695827484,
+      "learning_rate": 4.153219740695435e-07,
+      "loss": 0.3823,
+      "step": 1013
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4741443991661072,
+      "learning_rate": 3.958060049589485e-07,
+      "loss": 0.3563,
+      "step": 1014
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3570055365562439,
+      "learning_rate": 3.767578426521923e-07,
+      "loss": 0.2459,
+      "step": 1015
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3500770032405853,
+      "learning_rate": 3.5817766678850615e-07,
+      "loss": 0.2512,
+      "step": 1016
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.351227730512619,
+      "learning_rate": 3.4006565259366295e-07,
+      "loss": 0.2494,
+      "step": 1017
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3954888880252838,
+      "learning_rate": 3.224219708782894e-07,
+      "loss": 0.2872,
+      "step": 1018
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.45764344930648804,
+      "learning_rate": 3.052467880362675e-07,
+      "loss": 0.3956,
+      "step": 1019
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.394677996635437,
+      "learning_rate": 2.88540266043158e-07,
+      "loss": 0.3036,
+      "step": 1020
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.41744089126586914,
+      "learning_rate": 2.723025624546849e-07,
+      "loss": 0.3311,
+      "step": 1021
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.37757179141044617,
+      "learning_rate": 2.5653383040524227e-07,
+      "loss": 0.311,
+      "step": 1022
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4209496080875397,
+      "learning_rate": 2.412342186064565e-07,
+      "loss": 0.2342,
+      "step": 1023
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.41492438316345215,
+      "learning_rate": 2.2640387134577058e-07,
+      "loss": 0.3602,
+      "step": 1024
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3416643738746643,
+      "learning_rate": 2.1204292848509555e-07,
+      "loss": 0.2367,
+      "step": 1025
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4362007975578308,
+      "learning_rate": 1.98151525459489e-07,
+      "loss": 0.2689,
+      "step": 1026
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.341298907995224,
+      "learning_rate": 1.8472979327587292e-07,
+      "loss": 0.21,
+      "step": 1027
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4322044551372528,
+      "learning_rate": 1.717778585118013e-07,
+      "loss": 0.3413,
+      "step": 1028
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.39564454555511475,
+      "learning_rate": 1.5929584331427216e-07,
+      "loss": 0.3038,
+      "step": 1029
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5081477761268616,
+      "learning_rate": 1.4728386539856754e-07,
+      "loss": 0.3507,
+      "step": 1030
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.44406288862228394,
+      "learning_rate": 1.3574203804713747e-07,
+      "loss": 0.4077,
+      "step": 1031
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.3900183439254761,
+      "learning_rate": 1.246704701085566e-07,
+      "loss": 0.2743,
+      "step": 1032
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.3891497254371643,
+      "learning_rate": 1.1406926599646373e-07,
+      "loss": 0.2778,
+      "step": 1033
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.3727441430091858,
+      "learning_rate": 1.0393852568860719e-07,
+      "loss": 0.2349,
+      "step": 1034
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.28589460253715515,
+      "learning_rate": 9.427834472588992e-08,
+      "loss": 0.145,
+      "step": 1035
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.41049668192863464,
+      "learning_rate": 8.508881421145365e-08,
+      "loss": 0.2638,
+      "step": 1036
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.3703855574131012,
+      "learning_rate": 7.637002080985168e-08,
+      "loss": 0.2358,
+      "step": 1037
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.41987836360931396,
+      "learning_rate": 6.812204674619959e-08,
+      "loss": 0.3283,
+      "step": 1038
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.3911339342594147,
+      "learning_rate": 6.034496980542037e-08,
+      "loss": 0.2811,
+      "step": 1039
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3822670578956604,
+      "learning_rate": 5.303886333151153e-08,
+      "loss": 0.2701,
+      "step": 1040
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3765828311443329,
+      "learning_rate": 4.620379622682358e-08,
+      "loss": 0.2665,
+      "step": 1041
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.362557590007782,
+      "learning_rate": 3.9839832951465986e-08,
+      "loss": 0.2462,
+      "step": 1042
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3730635344982147,
+      "learning_rate": 3.394703352263551e-08,
+      "loss": 0.2676,
+      "step": 1043
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.4196093678474426,
+      "learning_rate": 2.8525453514099965e-08,
+      "loss": 0.3338,
+      "step": 1044
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.359421968460083,
+      "learning_rate": 2.3575144055643094e-08,
+      "loss": 0.1959,
+      "step": 1045
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3653644323348999,
+      "learning_rate": 1.9096151832609375e-08,
+      "loss": 0.2419,
+      "step": 1046
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.38383719325065613,
+      "learning_rate": 1.5088519085437734e-08,
+      "loss": 0.2995,
+      "step": 1047
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.36653193831443787,
+      "learning_rate": 1.1552283609272962e-08,
+      "loss": 0.2261,
+      "step": 1048
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.38996705412864685,
+      "learning_rate": 8.487478753615997e-09,
+      "loss": 0.3316,
+      "step": 1049
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.4443221092224121,
+      "learning_rate": 5.894133422001957e-09,
+      "loss": 0.3662,
+      "step": 1050
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.33084529638290405,
+      "learning_rate": 3.772272071722593e-09,
+      "loss": 0.2281,
+      "step": 1051
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.3955950140953064,
+      "learning_rate": 2.1219147136264382e-09,
+      "loss": 0.2675,
+      "step": 1052
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.4168340265750885,
+      "learning_rate": 9.43076911874563e-10,
+      "loss": 0.3497,
+      "step": 1053
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.43389514088630676,
+      "learning_rate": 2.3576978384065584e-10,
+      "loss": 0.3197,
+      "step": 1054
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.36339259147644043,
+      "learning_rate": 0.0,
+      "loss": 0.2589,
+      "step": 1055
+    },
+    {
+      "epoch": 1.0,
+      "step": 1055,
+      "total_flos": 1.1584683941193318e+17,
+      "train_loss": 0.31111926025406444,
+      "train_runtime": 56574.5112,
+      "train_samples_per_second": 0.597,
+      "train_steps_per_second": 0.019
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 1055,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 100,
+  "total_flos": 1.1584683941193318e+17,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}