diff --git a/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa-shuffled-streaming_baseline_lora_agg_fc4_s2048/README.md b/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa-shuffled-streaming_baseline_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_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_lora_agg_fc4_s2048/adapter_config.json b/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa-shuffled-streaming_baseline_lora_agg_fc4_s2048/adapter_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..1e995897cdb0a63636ae8a99526dc8cf7918bcb4
--- /dev/null
+++ b/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa-shuffled-streaming_baseline_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": [
+    "gate_proj",
+    "q_proj",
+    "v_proj",
+    "k_proj",
+    "o_proj",
+    "down_proj",
+    "up_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_lora_agg_fc4_s2048/adapter_model.bin b/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa-shuffled-streaming_baseline_lora_agg_fc4_s2048/adapter_model.bin
new file mode 100644
index 0000000000000000000000000000000000000000..424636304627e963e495c3ffb43e2389979644d6
--- /dev/null
+++ b/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa-shuffled-streaming_baseline_lora_agg_fc4_s2048/adapter_model.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:84942ecf9314974ae7a42aec32b6c3ab7e894c3b7f0e4ea74092601c3ee0aa00
+size 692127130
diff --git a/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa-shuffled-streaming_baseline_lora_agg_fc4_s2048/config.json b/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa-shuffled-streaming_baseline_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_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_lora_agg_fc4_s2048/generation_config.json b/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa-shuffled-streaming_baseline_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_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_lora_agg_fc4_s2048/non_lora_trainables.bin b/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa-shuffled-streaming_baseline_lora_agg_fc4_s2048/non_lora_trainables.bin
new file mode 100644
index 0000000000000000000000000000000000000000..ad97cedd06c03568a18ea3462dfbd9ead4248bdf
--- /dev/null
+++ b/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa-shuffled-streaming_baseline_lora_agg_fc4_s2048/non_lora_trainables.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:68da85e53e1131e209320a5d0e2581c67fc71d5138125d7db85780c95597d458
+size 211441788
diff --git a/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa-shuffled-streaming_baseline_lora_agg_fc4_s2048/trainer_state.json b/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa-shuffled-streaming_baseline_lora_agg_fc4_s2048/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..448c18e64ac4bf93efa08a11f2b503e46439e1da
--- /dev/null
+++ b/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa-shuffled-streaming_baseline_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": 18.313032150268555,
+      "learning_rate": 3.125e-06,
+      "loss": 1.0657,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 13.580374717712402,
+      "learning_rate": 6.25e-06,
+      "loss": 0.9818,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 10.899441719055176,
+      "learning_rate": 9.375000000000001e-06,
+      "loss": 1.0724,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.318131446838379,
+      "learning_rate": 1.25e-05,
+      "loss": 0.9968,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.730851173400879,
+      "learning_rate": 1.5625e-05,
+      "loss": 0.6971,
+      "step": 5
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.394482135772705,
+      "learning_rate": 1.8750000000000002e-05,
+      "loss": 0.6248,
+      "step": 6
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.010072946548462,
+      "learning_rate": 2.1875e-05,
+      "loss": 0.5495,
+      "step": 7
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6598902940750122,
+      "learning_rate": 2.5e-05,
+      "loss": 0.4581,
+      "step": 8
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.1555105447769165,
+      "learning_rate": 2.8125000000000003e-05,
+      "loss": 0.3745,
+      "step": 9
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9601075053215027,
+      "learning_rate": 3.125e-05,
+      "loss": 0.3504,
+      "step": 10
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8265823721885681,
+      "learning_rate": 3.4375e-05,
+      "loss": 0.3034,
+      "step": 11
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7626316547393799,
+      "learning_rate": 3.7500000000000003e-05,
+      "loss": 0.2733,
+      "step": 12
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7715867161750793,
+      "learning_rate": 4.0625000000000005e-05,
+      "loss": 0.3472,
+      "step": 13
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.758461058139801,
+      "learning_rate": 4.375e-05,
+      "loss": 0.3511,
+      "step": 14
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8174577355384827,
+      "learning_rate": 4.6875e-05,
+      "loss": 0.5306,
+      "step": 15
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6446866393089294,
+      "learning_rate": 5e-05,
+      "loss": 0.3823,
+      "step": 16
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7688436508178711,
+      "learning_rate": 5.3125000000000004e-05,
+      "loss": 0.4384,
+      "step": 17
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7424086928367615,
+      "learning_rate": 5.6250000000000005e-05,
+      "loss": 0.2853,
+      "step": 18
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5257894396781921,
+      "learning_rate": 5.9375e-05,
+      "loss": 0.3546,
+      "step": 19
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5890655517578125,
+      "learning_rate": 6.25e-05,
+      "loss": 0.2512,
+      "step": 20
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5847811698913574,
+      "learning_rate": 6.562500000000001e-05,
+      "loss": 0.3774,
+      "step": 21
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.46664291620254517,
+      "learning_rate": 6.875e-05,
+      "loss": 0.1707,
+      "step": 22
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5199007987976074,
+      "learning_rate": 7.1875e-05,
+      "loss": 0.3012,
+      "step": 23
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7765957117080688,
+      "learning_rate": 7.500000000000001e-05,
+      "loss": 0.4722,
+      "step": 24
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5152903199195862,
+      "learning_rate": 7.8125e-05,
+      "loss": 0.2235,
+      "step": 25
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6177201867103577,
+      "learning_rate": 8.125000000000001e-05,
+      "loss": 0.3814,
+      "step": 26
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6518827676773071,
+      "learning_rate": 8.4375e-05,
+      "loss": 0.3717,
+      "step": 27
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5696550607681274,
+      "learning_rate": 8.75e-05,
+      "loss": 0.3354,
+      "step": 28
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.529187023639679,
+      "learning_rate": 9.062500000000001e-05,
+      "loss": 0.3285,
+      "step": 29
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5582684874534607,
+      "learning_rate": 9.375e-05,
+      "loss": 0.2991,
+      "step": 30
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5471290349960327,
+      "learning_rate": 9.687500000000001e-05,
+      "loss": 0.3064,
+      "step": 31
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6911242604255676,
+      "learning_rate": 0.0001,
+      "loss": 0.3243,
+      "step": 32
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.582001805305481,
+      "learning_rate": 9.999976423021617e-05,
+      "loss": 0.3636,
+      "step": 33
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5498625040054321,
+      "learning_rate": 9.999905692308813e-05,
+      "loss": 0.3758,
+      "step": 34
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6672377586364746,
+      "learning_rate": 9.999787808528638e-05,
+      "loss": 0.4803,
+      "step": 35
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4661501348018646,
+      "learning_rate": 9.999622772792828e-05,
+      "loss": 0.2856,
+      "step": 36
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5137473344802856,
+      "learning_rate": 9.9994105866578e-05,
+      "loss": 0.376,
+      "step": 37
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5769113302230835,
+      "learning_rate": 9.999151252124639e-05,
+      "loss": 0.3687,
+      "step": 38
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4689115285873413,
+      "learning_rate": 9.998844771639073e-05,
+      "loss": 0.2882,
+      "step": 39
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4166828393936157,
+      "learning_rate": 9.998491148091457e-05,
+      "loss": 0.3142,
+      "step": 40
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6350471377372742,
+      "learning_rate": 9.99809038481674e-05,
+      "loss": 0.4258,
+      "step": 41
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5171714425086975,
+      "learning_rate": 9.997642485594436e-05,
+      "loss": 0.337,
+      "step": 42
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4555424749851227,
+      "learning_rate": 9.99714745464859e-05,
+      "loss": 0.3545,
+      "step": 43
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4361424148082733,
+      "learning_rate": 9.996605296647737e-05,
+      "loss": 0.2842,
+      "step": 44
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5546784400939941,
+      "learning_rate": 9.996016016704854e-05,
+      "loss": 0.507,
+      "step": 45
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.41821250319480896,
+      "learning_rate": 9.995379620377318e-05,
+      "loss": 0.2746,
+      "step": 46
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3964790999889374,
+      "learning_rate": 9.99469611366685e-05,
+      "loss": 0.2524,
+      "step": 47
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.44861629605293274,
+      "learning_rate": 9.993965503019457e-05,
+      "loss": 0.3982,
+      "step": 48
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.41635861992836,
+      "learning_rate": 9.99318779532538e-05,
+      "loss": 0.3352,
+      "step": 49
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.48465538024902344,
+      "learning_rate": 9.992362997919016e-05,
+      "loss": 0.3313,
+      "step": 50
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4887374937534332,
+      "learning_rate": 9.991491118578856e-05,
+      "loss": 0.3755,
+      "step": 51
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4448220431804657,
+      "learning_rate": 9.990572165527413e-05,
+      "loss": 0.3402,
+      "step": 52
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.49894165992736816,
+      "learning_rate": 9.98960614743114e-05,
+      "loss": 0.4375,
+      "step": 53
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5184201002120972,
+      "learning_rate": 9.988593073400354e-05,
+      "loss": 0.4377,
+      "step": 54
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.44544926285743713,
+      "learning_rate": 9.987532952989145e-05,
+      "loss": 0.3765,
+      "step": 55
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4068632125854492,
+      "learning_rate": 9.986425796195287e-05,
+      "loss": 0.272,
+      "step": 56
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.47971636056900024,
+      "learning_rate": 9.985271613460144e-05,
+      "loss": 0.4205,
+      "step": 57
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3723823130130768,
+      "learning_rate": 9.984070415668574e-05,
+      "loss": 0.24,
+      "step": 58
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.49853771924972534,
+      "learning_rate": 9.982822214148821e-05,
+      "loss": 0.4158,
+      "step": 59
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.42973271012306213,
+      "learning_rate": 9.981527020672413e-05,
+      "loss": 0.3128,
+      "step": 60
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.48218533396720886,
+      "learning_rate": 9.980184847454052e-05,
+      "loss": 0.4003,
+      "step": 61
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4399939477443695,
+      "learning_rate": 9.978795707151492e-05,
+      "loss": 0.3124,
+      "step": 62
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.40350911021232605,
+      "learning_rate": 9.977359612865423e-05,
+      "loss": 0.3063,
+      "step": 63
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.44324538111686707,
+      "learning_rate": 9.975876578139355e-05,
+      "loss": 0.343,
+      "step": 64
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4793551564216614,
+      "learning_rate": 9.974346616959476e-05,
+      "loss": 0.447,
+      "step": 65
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.40760207176208496,
+      "learning_rate": 9.972769743754531e-05,
+      "loss": 0.293,
+      "step": 66
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.37527546286582947,
+      "learning_rate": 9.971145973395684e-05,
+      "loss": 0.2422,
+      "step": 67
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3915005922317505,
+      "learning_rate": 9.969475321196373e-05,
+      "loss": 0.2127,
+      "step": 68
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4286052882671356,
+      "learning_rate": 9.96775780291217e-05,
+      "loss": 0.2989,
+      "step": 69
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4435606896877289,
+      "learning_rate": 9.965993434740633e-05,
+      "loss": 0.2916,
+      "step": 70
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5245209336280823,
+      "learning_rate": 9.96418223332115e-05,
+      "loss": 0.3567,
+      "step": 71
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.41631606221199036,
+      "learning_rate": 9.962324215734782e-05,
+      "loss": 0.2072,
+      "step": 72
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5242307186126709,
+      "learning_rate": 9.960419399504106e-05,
+      "loss": 0.3906,
+      "step": 73
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4245007038116455,
+      "learning_rate": 9.958467802593046e-05,
+      "loss": 0.2715,
+      "step": 74
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4629935920238495,
+      "learning_rate": 9.956469443406706e-05,
+      "loss": 0.316,
+      "step": 75
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3802966773509979,
+      "learning_rate": 9.954424340791196e-05,
+      "loss": 0.2947,
+      "step": 76
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.495551735162735,
+      "learning_rate": 9.952332514033447e-05,
+      "loss": 0.347,
+      "step": 77
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4242008626461029,
+      "learning_rate": 9.950193982861047e-05,
+      "loss": 0.3883,
+      "step": 78
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4434719681739807,
+      "learning_rate": 9.948008767442034e-05,
+      "loss": 0.3383,
+      "step": 79
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3872564733028412,
+      "learning_rate": 9.94577688838472e-05,
+      "loss": 0.2867,
+      "step": 80
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4217293858528137,
+      "learning_rate": 9.943498366737487e-05,
+      "loss": 0.3575,
+      "step": 81
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.43873757123947144,
+      "learning_rate": 9.941173223988602e-05,
+      "loss": 0.308,
+      "step": 82
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4716368615627289,
+      "learning_rate": 9.938801482065998e-05,
+      "loss": 0.3401,
+      "step": 83
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4649840295314789,
+      "learning_rate": 9.93638316333708e-05,
+      "loss": 0.389,
+      "step": 84
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4310089647769928,
+      "learning_rate": 9.933918290608509e-05,
+      "loss": 0.3216,
+      "step": 85
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.44049933552742004,
+      "learning_rate": 9.93140688712598e-05,
+      "loss": 0.345,
+      "step": 86
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4817194938659668,
+      "learning_rate": 9.928848976574019e-05,
+      "loss": 0.4149,
+      "step": 87
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.400789737701416,
+      "learning_rate": 9.926244583075748e-05,
+      "loss": 0.3222,
+      "step": 88
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4780321717262268,
+      "learning_rate": 9.923593731192655e-05,
+      "loss": 0.37,
+      "step": 89
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.39723774790763855,
+      "learning_rate": 9.920896445924373e-05,
+      "loss": 0.2912,
+      "step": 90
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3887556791305542,
+      "learning_rate": 9.918152752708438e-05,
+      "loss": 0.2001,
+      "step": 91
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4470643103122711,
+      "learning_rate": 9.915362677420044e-05,
+      "loss": 0.3626,
+      "step": 92
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.44813528656959534,
+      "learning_rate": 9.912526246371814e-05,
+      "loss": 0.4236,
+      "step": 93
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.47168847918510437,
+      "learning_rate": 9.909643486313533e-05,
+      "loss": 0.3601,
+      "step": 94
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4922250211238861,
+      "learning_rate": 9.906714424431913e-05,
+      "loss": 0.384,
+      "step": 95
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.48167434334754944,
+      "learning_rate": 9.903739088350325e-05,
+      "loss": 0.2727,
+      "step": 96
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.47917285561561584,
+      "learning_rate": 9.900717506128541e-05,
+      "loss": 0.3053,
+      "step": 97
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.42024195194244385,
+      "learning_rate": 9.897649706262473e-05,
+      "loss": 0.3052,
+      "step": 98
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.517518162727356,
+      "learning_rate": 9.894535717683902e-05,
+      "loss": 0.3758,
+      "step": 99
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4286741018295288,
+      "learning_rate": 9.891375569760204e-05,
+      "loss": 0.3417,
+      "step": 100
+    },
+    {
+      "epoch": 0.09,
+      "eval_loss": 0.31432420015335083,
+      "eval_runtime": 362.7062,
+      "eval_samples_per_second": 0.94,
+      "eval_steps_per_second": 0.237,
+      "step": 100
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.45813727378845215,
+      "learning_rate": 9.888169292294076e-05,
+      "loss": 0.2973,
+      "step": 101
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4520292282104492,
+      "learning_rate": 9.884916915523251e-05,
+      "loss": 0.2907,
+      "step": 102
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5243715643882751,
+      "learning_rate": 9.881618470120215e-05,
+      "loss": 0.3572,
+      "step": 103
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4676945209503174,
+      "learning_rate": 9.87827398719192e-05,
+      "loss": 0.3796,
+      "step": 104
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.46876049041748047,
+      "learning_rate": 9.874883498279484e-05,
+      "loss": 0.2695,
+      "step": 105
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.43690720200538635,
+      "learning_rate": 9.871447035357903e-05,
+      "loss": 0.3192,
+      "step": 106
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.41145750880241394,
+      "learning_rate": 9.867964630835743e-05,
+      "loss": 0.2737,
+      "step": 107
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3777146339416504,
+      "learning_rate": 9.864436317554831e-05,
+      "loss": 0.2753,
+      "step": 108
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4422137141227722,
+      "learning_rate": 9.860862128789953e-05,
+      "loss": 0.2839,
+      "step": 109
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.42014750838279724,
+      "learning_rate": 9.857242098248542e-05,
+      "loss": 0.229,
+      "step": 110
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4446079730987549,
+      "learning_rate": 9.853576260070347e-05,
+      "loss": 0.3326,
+      "step": 111
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.405568927526474,
+      "learning_rate": 9.849864648827125e-05,
+      "loss": 0.2834,
+      "step": 112
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.46995070576667786,
+      "learning_rate": 9.846107299522304e-05,
+      "loss": 0.3771,
+      "step": 113
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5096259117126465,
+      "learning_rate": 9.842304247590667e-05,
+      "loss": 0.4228,
+      "step": 114
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.42030078172683716,
+      "learning_rate": 9.838455528897998e-05,
+      "loss": 0.2722,
+      "step": 115
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4473590552806854,
+      "learning_rate": 9.834561179740762e-05,
+      "loss": 0.2924,
+      "step": 116
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4592130482196808,
+      "learning_rate": 9.830621236845754e-05,
+      "loss": 0.3065,
+      "step": 117
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.42212823033332825,
+      "learning_rate": 9.826635737369751e-05,
+      "loss": 0.3115,
+      "step": 118
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4238523542881012,
+      "learning_rate": 9.82260471889917e-05,
+      "loss": 0.2673,
+      "step": 119
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.41788744926452637,
+      "learning_rate": 9.818528219449705e-05,
+      "loss": 0.2598,
+      "step": 120
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.404971182346344,
+      "learning_rate": 9.814406277465968e-05,
+      "loss": 0.2552,
+      "step": 121
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4492509067058563,
+      "learning_rate": 9.810238931821138e-05,
+      "loss": 0.3353,
+      "step": 122
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.48347416520118713,
+      "learning_rate": 9.806026221816581e-05,
+      "loss": 0.4063,
+      "step": 123
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4379715025424957,
+      "learning_rate": 9.801768187181486e-05,
+      "loss": 0.3709,
+      "step": 124
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.409445196390152,
+      "learning_rate": 9.797464868072488e-05,
+      "loss": 0.2701,
+      "step": 125
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.45940884947776794,
+      "learning_rate": 9.793116305073292e-05,
+      "loss": 0.3433,
+      "step": 126
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4554141163825989,
+      "learning_rate": 9.788722539194291e-05,
+      "loss": 0.3794,
+      "step": 127
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4365048110485077,
+      "learning_rate": 9.784283611872169e-05,
+      "loss": 0.2995,
+      "step": 128
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4406554698944092,
+      "learning_rate": 9.77979956496953e-05,
+      "loss": 0.3568,
+      "step": 129
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4969058632850647,
+      "learning_rate": 9.77527044077448e-05,
+      "loss": 0.4574,
+      "step": 130
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.47635579109191895,
+      "learning_rate": 9.770696282000244e-05,
+      "loss": 0.3573,
+      "step": 131
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.48415517807006836,
+      "learning_rate": 9.766077131784763e-05,
+      "loss": 0.407,
+      "step": 132
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4078788757324219,
+      "learning_rate": 9.761413033690276e-05,
+      "loss": 0.3063,
+      "step": 133
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.44089677929878235,
+      "learning_rate": 9.756704031702918e-05,
+      "loss": 0.3792,
+      "step": 134
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4016633629798889,
+      "learning_rate": 9.751950170232309e-05,
+      "loss": 0.2867,
+      "step": 135
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.41537660360336304,
+      "learning_rate": 9.74715149411112e-05,
+      "loss": 0.3279,
+      "step": 136
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4650765061378479,
+      "learning_rate": 9.742308048594665e-05,
+      "loss": 0.3799,
+      "step": 137
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.2895463705062866,
+      "learning_rate": 9.73741987936047e-05,
+      "loss": 0.1716,
+      "step": 138
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.36121901869773865,
+      "learning_rate": 9.732487032507837e-05,
+      "loss": 0.2301,
+      "step": 139
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.47575443983078003,
+      "learning_rate": 9.727509554557417e-05,
+      "loss": 0.3923,
+      "step": 140
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4208732545375824,
+      "learning_rate": 9.722487492450764e-05,
+      "loss": 0.2809,
+      "step": 141
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4757351279258728,
+      "learning_rate": 9.717420893549902e-05,
+      "loss": 0.3719,
+      "step": 142
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4793698787689209,
+      "learning_rate": 9.712309805636863e-05,
+      "loss": 0.3067,
+      "step": 143
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4953964650630951,
+      "learning_rate": 9.707154276913256e-05,
+      "loss": 0.3812,
+      "step": 144
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4665147066116333,
+      "learning_rate": 9.701954355999792e-05,
+      "loss": 0.3664,
+      "step": 145
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.49213695526123047,
+      "learning_rate": 9.696710091935841e-05,
+      "loss": 0.3376,
+      "step": 146
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.399131715297699,
+      "learning_rate": 9.691421534178966e-05,
+      "loss": 0.2809,
+      "step": 147
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4207635819911957,
+      "learning_rate": 9.686088732604449e-05,
+      "loss": 0.305,
+      "step": 148
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3471522927284241,
+      "learning_rate": 9.68071173750483e-05,
+      "loss": 0.2206,
+      "step": 149
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4532647430896759,
+      "learning_rate": 9.675290599589429e-05,
+      "loss": 0.3718,
+      "step": 150
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.46664243936538696,
+      "learning_rate": 9.669825369983865e-05,
+      "loss": 0.2883,
+      "step": 151
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.41483238339424133,
+      "learning_rate": 9.664316100229578e-05,
+      "loss": 0.3345,
+      "step": 152
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5003163814544678,
+      "learning_rate": 9.658762842283343e-05,
+      "loss": 0.3629,
+      "step": 153
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.35678866505622864,
+      "learning_rate": 9.653165648516776e-05,
+      "loss": 0.2574,
+      "step": 154
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5268349051475525,
+      "learning_rate": 9.647524571715843e-05,
+      "loss": 0.3759,
+      "step": 155
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.40246081352233887,
+      "learning_rate": 9.641839665080363e-05,
+      "loss": 0.3019,
+      "step": 156
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4237998425960541,
+      "learning_rate": 9.636110982223505e-05,
+      "loss": 0.285,
+      "step": 157
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.41048043966293335,
+      "learning_rate": 9.630338577171282e-05,
+      "loss": 0.3175,
+      "step": 158
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.38134410977363586,
+      "learning_rate": 9.624522504362039e-05,
+      "loss": 0.2969,
+      "step": 159
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.46802300214767456,
+      "learning_rate": 9.618662818645948e-05,
+      "loss": 0.3304,
+      "step": 160
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.33313772082328796,
+      "learning_rate": 9.612759575284483e-05,
+      "loss": 0.2122,
+      "step": 161
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4662613272666931,
+      "learning_rate": 9.606812829949896e-05,
+      "loss": 0.309,
+      "step": 162
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.39246639609336853,
+      "learning_rate": 9.600822638724705e-05,
+      "loss": 0.2874,
+      "step": 163
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3868423104286194,
+      "learning_rate": 9.594789058101153e-05,
+      "loss": 0.2578,
+      "step": 164
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4004722833633423,
+      "learning_rate": 9.588712144980681e-05,
+      "loss": 0.3182,
+      "step": 165
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.44403666257858276,
+      "learning_rate": 9.582591956673387e-05,
+      "loss": 0.3193,
+      "step": 166
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.43500831723213196,
+      "learning_rate": 9.576428550897489e-05,
+      "loss": 0.2593,
+      "step": 167
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.44819876551628113,
+      "learning_rate": 9.570221985778785e-05,
+      "loss": 0.3442,
+      "step": 168
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4313012659549713,
+      "learning_rate": 9.563972319850092e-05,
+      "loss": 0.314,
+      "step": 169
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.293268620967865,
+      "learning_rate": 9.557679612050708e-05,
+      "loss": 0.2051,
+      "step": 170
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4209660291671753,
+      "learning_rate": 9.551343921725843e-05,
+      "loss": 0.3652,
+      "step": 171
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5210956335067749,
+      "learning_rate": 9.544965308626074e-05,
+      "loss": 0.459,
+      "step": 172
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3629516065120697,
+      "learning_rate": 9.538543832906773e-05,
+      "loss": 0.2583,
+      "step": 173
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4232120215892792,
+      "learning_rate": 9.53207955512753e-05,
+      "loss": 0.3077,
+      "step": 174
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4677829444408417,
+      "learning_rate": 9.525572536251607e-05,
+      "loss": 0.412,
+      "step": 175
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4338430166244507,
+      "learning_rate": 9.519022837645337e-05,
+      "loss": 0.3353,
+      "step": 176
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4187476336956024,
+      "learning_rate": 9.512430521077565e-05,
+      "loss": 0.2868,
+      "step": 177
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4185602068901062,
+      "learning_rate": 9.505795648719048e-05,
+      "loss": 0.3379,
+      "step": 178
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4814926087856293,
+      "learning_rate": 9.499118283141887e-05,
+      "loss": 0.3762,
+      "step": 179
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.38241589069366455,
+      "learning_rate": 9.492398487318921e-05,
+      "loss": 0.3191,
+      "step": 180
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3785567581653595,
+      "learning_rate": 9.485636324623146e-05,
+      "loss": 0.2583,
+      "step": 181
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.40947380661964417,
+      "learning_rate": 9.478831858827104e-05,
+      "loss": 0.3548,
+      "step": 182
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3985261619091034,
+      "learning_rate": 9.471985154102292e-05,
+      "loss": 0.3258,
+      "step": 183
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5790777802467346,
+      "learning_rate": 9.465096275018556e-05,
+      "loss": 0.4176,
+      "step": 184
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4602338969707489,
+      "learning_rate": 9.458165286543476e-05,
+      "loss": 0.3195,
+      "step": 185
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.47313788533210754,
+      "learning_rate": 9.451192254041758e-05,
+      "loss": 0.3449,
+      "step": 186
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5002240538597107,
+      "learning_rate": 9.444177243274618e-05,
+      "loss": 0.4463,
+      "step": 187
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.45640143752098083,
+      "learning_rate": 9.437120320399158e-05,
+      "loss": 0.3505,
+      "step": 188
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4020766019821167,
+      "learning_rate": 9.430021551967744e-05,
+      "loss": 0.2786,
+      "step": 189
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3839154541492462,
+      "learning_rate": 9.422881004927381e-05,
+      "loss": 0.2762,
+      "step": 190
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.41408756375312805,
+      "learning_rate": 9.415698746619079e-05,
+      "loss": 0.332,
+      "step": 191
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4137730896472931,
+      "learning_rate": 9.408474844777217e-05,
+      "loss": 0.2577,
+      "step": 192
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4891367554664612,
+      "learning_rate": 9.401209367528908e-05,
+      "loss": 0.48,
+      "step": 193
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.42042845487594604,
+      "learning_rate": 9.393902383393347e-05,
+      "loss": 0.3098,
+      "step": 194
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5515792965888977,
+      "learning_rate": 9.386553961281179e-05,
+      "loss": 0.4273,
+      "step": 195
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.44671303033828735,
+      "learning_rate": 9.379164170493844e-05,
+      "loss": 0.3797,
+      "step": 196
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4394177794456482,
+      "learning_rate": 9.371733080722911e-05,
+      "loss": 0.3377,
+      "step": 197
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3395167291164398,
+      "learning_rate": 9.36426076204944e-05,
+      "loss": 0.2225,
+      "step": 198
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.41627058386802673,
+      "learning_rate": 9.35674728494331e-05,
+      "loss": 0.3545,
+      "step": 199
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.39437538385391235,
+      "learning_rate": 9.349192720262555e-05,
+      "loss": 0.3042,
+      "step": 200
+    },
+    {
+      "epoch": 0.19,
+      "eval_loss": 0.306856632232666,
+      "eval_runtime": 357.8476,
+      "eval_samples_per_second": 0.953,
+      "eval_steps_per_second": 0.24,
+      "step": 200
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3970376253128052,
+      "learning_rate": 9.341597139252698e-05,
+      "loss": 0.3203,
+      "step": 201
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4313250184059143,
+      "learning_rate": 9.333960613546079e-05,
+      "loss": 0.3051,
+      "step": 202
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3659449815750122,
+      "learning_rate": 9.326283215161178e-05,
+      "loss": 0.2609,
+      "step": 203
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3503909707069397,
+      "learning_rate": 9.31856501650194e-05,
+      "loss": 0.1664,
+      "step": 204
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4019453823566437,
+      "learning_rate": 9.310806090357082e-05,
+      "loss": 0.3593,
+      "step": 205
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3274018168449402,
+      "learning_rate": 9.303006509899418e-05,
+      "loss": 0.1791,
+      "step": 206
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.47928860783576965,
+      "learning_rate": 9.29516634868517e-05,
+      "loss": 0.3436,
+      "step": 207
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5058536529541016,
+      "learning_rate": 9.287285680653254e-05,
+      "loss": 0.378,
+      "step": 208
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.39776819944381714,
+      "learning_rate": 9.279364580124614e-05,
+      "loss": 0.3029,
+      "step": 209
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3958539366722107,
+      "learning_rate": 9.27140312180149e-05,
+      "loss": 0.2318,
+      "step": 210
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5175220966339111,
+      "learning_rate": 9.263401380766739e-05,
+      "loss": 0.4109,
+      "step": 211
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4764615595340729,
+      "learning_rate": 9.255359432483105e-05,
+      "loss": 0.3554,
+      "step": 212
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.37956544756889343,
+      "learning_rate": 9.247277352792533e-05,
+      "loss": 0.2446,
+      "step": 213
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.470720112323761,
+      "learning_rate": 9.239155217915422e-05,
+      "loss": 0.3601,
+      "step": 214
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.36287134885787964,
+      "learning_rate": 9.230993104449939e-05,
+      "loss": 0.2317,
+      "step": 215
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.38994041085243225,
+      "learning_rate": 9.222791089371265e-05,
+      "loss": 0.2969,
+      "step": 216
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.436535507440567,
+      "learning_rate": 9.214549250030899e-05,
+      "loss": 0.2721,
+      "step": 217
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4360625147819519,
+      "learning_rate": 9.206267664155907e-05,
+      "loss": 0.3915,
+      "step": 218
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.34940338134765625,
+      "learning_rate": 9.197946409848194e-05,
+      "loss": 0.2328,
+      "step": 219
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3634420931339264,
+      "learning_rate": 9.189585565583778e-05,
+      "loss": 0.2944,
+      "step": 220
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3788878619670868,
+      "learning_rate": 9.181185210212034e-05,
+      "loss": 0.3228,
+      "step": 221
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3737380802631378,
+      "learning_rate": 9.172745422954961e-05,
+      "loss": 0.2613,
+      "step": 222
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.39093953371047974,
+      "learning_rate": 9.164266283406433e-05,
+      "loss": 0.3378,
+      "step": 223
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.39257872104644775,
+      "learning_rate": 9.155747871531444e-05,
+      "loss": 0.3236,
+      "step": 224
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4497702717781067,
+      "learning_rate": 9.14719026766536e-05,
+      "loss": 0.3658,
+      "step": 225
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4174399971961975,
+      "learning_rate": 9.13859355251316e-05,
+      "loss": 0.2982,
+      "step": 226
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.38028162717819214,
+      "learning_rate": 9.129957807148666e-05,
+      "loss": 0.2305,
+      "step": 227
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.48720794916152954,
+      "learning_rate": 9.121283113013795e-05,
+      "loss": 0.4153,
+      "step": 228
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3722657561302185,
+      "learning_rate": 9.112569551917774e-05,
+      "loss": 0.2532,
+      "step": 229
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.49003851413726807,
+      "learning_rate": 9.103817206036382e-05,
+      "loss": 0.384,
+      "step": 230
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4677237272262573,
+      "learning_rate": 9.095026157911166e-05,
+      "loss": 0.3413,
+      "step": 231
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.42675694823265076,
+      "learning_rate": 9.086196490448668e-05,
+      "loss": 0.2598,
+      "step": 232
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.45176050066947937,
+      "learning_rate": 9.077328286919638e-05,
+      "loss": 0.3627,
+      "step": 233
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4472631514072418,
+      "learning_rate": 9.068421630958254e-05,
+      "loss": 0.3798,
+      "step": 234
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3411577045917511,
+      "learning_rate": 9.059476606561328e-05,
+      "loss": 0.223,
+      "step": 235
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4092637598514557,
+      "learning_rate": 9.050493298087523e-05,
+      "loss": 0.3479,
+      "step": 236
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.39827343821525574,
+      "learning_rate": 9.041471790256542e-05,
+      "loss": 0.3141,
+      "step": 237
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.392585813999176,
+      "learning_rate": 9.032412168148345e-05,
+      "loss": 0.3175,
+      "step": 238
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4632207453250885,
+      "learning_rate": 9.02331451720234e-05,
+      "loss": 0.3388,
+      "step": 239
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.36682212352752686,
+      "learning_rate": 9.014178923216572e-05,
+      "loss": 0.2389,
+      "step": 240
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3975673317909241,
+      "learning_rate": 9.005005472346924e-05,
+      "loss": 0.2601,
+      "step": 241
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.44503253698349,
+      "learning_rate": 8.995794251106295e-05,
+      "loss": 0.3923,
+      "step": 242
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.40452277660369873,
+      "learning_rate": 8.986545346363792e-05,
+      "loss": 0.311,
+      "step": 243
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4056435525417328,
+      "learning_rate": 8.977258845343905e-05,
+      "loss": 0.3346,
+      "step": 244
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.377971351146698,
+      "learning_rate": 8.967934835625689e-05,
+      "loss": 0.2941,
+      "step": 245
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.47536715865135193,
+      "learning_rate": 8.958573405141932e-05,
+      "loss": 0.3365,
+      "step": 246
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4816979467868805,
+      "learning_rate": 8.949174642178332e-05,
+      "loss": 0.386,
+      "step": 247
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.46703040599823,
+      "learning_rate": 8.939738635372664e-05,
+      "loss": 0.3306,
+      "step": 248
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4739988148212433,
+      "learning_rate": 8.930265473713938e-05,
+      "loss": 0.3695,
+      "step": 249
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4607354402542114,
+      "learning_rate": 8.920755246541563e-05,
+      "loss": 0.3515,
+      "step": 250
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.414195716381073,
+      "learning_rate": 8.911208043544513e-05,
+      "loss": 0.3141,
+      "step": 251
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3896920680999756,
+      "learning_rate": 8.90162395476046e-05,
+      "loss": 0.2951,
+      "step": 252
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5263822078704834,
+      "learning_rate": 8.89200307057495e-05,
+      "loss": 0.462,
+      "step": 253
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4477658271789551,
+      "learning_rate": 8.882345481720533e-05,
+      "loss": 0.3955,
+      "step": 254
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4065832197666168,
+      "learning_rate": 8.872651279275916e-05,
+      "loss": 0.3229,
+      "step": 255
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.48063144087791443,
+      "learning_rate": 8.862920554665098e-05,
+      "loss": 0.3314,
+      "step": 256
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4026496708393097,
+      "learning_rate": 8.853153399656513e-05,
+      "loss": 0.3076,
+      "step": 257
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4567382335662842,
+      "learning_rate": 8.843349906362163e-05,
+      "loss": 0.3804,
+      "step": 258
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4693814516067505,
+      "learning_rate": 8.833510167236747e-05,
+      "loss": 0.4003,
+      "step": 259
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.38870713114738464,
+      "learning_rate": 8.823634275076791e-05,
+      "loss": 0.3228,
+      "step": 260
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.40066248178482056,
+      "learning_rate": 8.813722323019773e-05,
+      "loss": 0.3704,
+      "step": 261
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.359891802072525,
+      "learning_rate": 8.803774404543247e-05,
+      "loss": 0.2549,
+      "step": 262
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.38821840286254883,
+      "learning_rate": 8.793790613463955e-05,
+      "loss": 0.3281,
+      "step": 263
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3141961693763733,
+      "learning_rate": 8.783771043936948e-05,
+      "loss": 0.2215,
+      "step": 264
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.40380561351776123,
+      "learning_rate": 8.773715790454695e-05,
+      "loss": 0.2932,
+      "step": 265
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3581491708755493,
+      "learning_rate": 8.763624947846195e-05,
+      "loss": 0.2205,
+      "step": 266
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4690437316894531,
+      "learning_rate": 8.753498611276079e-05,
+      "loss": 0.3541,
+      "step": 267
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.48447108268737793,
+      "learning_rate": 8.743336876243712e-05,
+      "loss": 0.4139,
+      "step": 268
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5336358547210693,
+      "learning_rate": 8.7331398385823e-05,
+      "loss": 0.4525,
+      "step": 269
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.49868297576904297,
+      "learning_rate": 8.722907594457974e-05,
+      "loss": 0.3918,
+      "step": 270
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.45554542541503906,
+      "learning_rate": 8.712640240368899e-05,
+      "loss": 0.3909,
+      "step": 271
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4768046736717224,
+      "learning_rate": 8.702337873144343e-05,
+      "loss": 0.3129,
+      "step": 272
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3925222158432007,
+      "learning_rate": 8.692000589943785e-05,
+      "loss": 0.2683,
+      "step": 273
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.47496914863586426,
+      "learning_rate": 8.681628488255986e-05,
+      "loss": 0.393,
+      "step": 274
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.38603782653808594,
+      "learning_rate": 8.671221665898073e-05,
+      "loss": 0.279,
+      "step": 275
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.429652601480484,
+      "learning_rate": 8.660780221014616e-05,
+      "loss": 0.3239,
+      "step": 276
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.42030632495880127,
+      "learning_rate": 8.650304252076703e-05,
+      "loss": 0.3056,
+      "step": 277
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.38885384798049927,
+      "learning_rate": 8.639793857881011e-05,
+      "loss": 0.2638,
+      "step": 278
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5263670682907104,
+      "learning_rate": 8.629249137548874e-05,
+      "loss": 0.3861,
+      "step": 279
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.40443241596221924,
+      "learning_rate": 8.618670190525352e-05,
+      "loss": 0.2613,
+      "step": 280
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4165252149105072,
+      "learning_rate": 8.608057116578283e-05,
+      "loss": 0.2894,
+      "step": 281
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4690636992454529,
+      "learning_rate": 8.597410015797358e-05,
+      "loss": 0.4027,
+      "step": 282
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.38352808356285095,
+      "learning_rate": 8.586728988593158e-05,
+      "loss": 0.2295,
+      "step": 283
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4734778106212616,
+      "learning_rate": 8.576014135696226e-05,
+      "loss": 0.3228,
+      "step": 284
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4245835840702057,
+      "learning_rate": 8.565265558156101e-05,
+      "loss": 0.2818,
+      "step": 285
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5121389627456665,
+      "learning_rate": 8.554483357340379e-05,
+      "loss": 0.3555,
+      "step": 286
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.38617703318595886,
+      "learning_rate": 8.543667634933742e-05,
+      "loss": 0.2483,
+      "step": 287
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4135371446609497,
+      "learning_rate": 8.532818492937014e-05,
+      "loss": 0.2959,
+      "step": 288
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.41642940044403076,
+      "learning_rate": 8.521936033666187e-05,
+      "loss": 0.3015,
+      "step": 289
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4078030288219452,
+      "learning_rate": 8.511020359751466e-05,
+      "loss": 0.3106,
+      "step": 290
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.34008803963661194,
+      "learning_rate": 8.500071574136295e-05,
+      "loss": 0.2111,
+      "step": 291
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.41384372115135193,
+      "learning_rate": 8.489089780076386e-05,
+      "loss": 0.2885,
+      "step": 292
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4303152561187744,
+      "learning_rate": 8.478075081138745e-05,
+      "loss": 0.2937,
+      "step": 293
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5157351493835449,
+      "learning_rate": 8.467027581200702e-05,
+      "loss": 0.4072,
+      "step": 294
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.464619904756546,
+      "learning_rate": 8.455947384448924e-05,
+      "loss": 0.3587,
+      "step": 295
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3617192506790161,
+      "learning_rate": 8.444834595378434e-05,
+      "loss": 0.253,
+      "step": 296
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.42125195264816284,
+      "learning_rate": 8.433689318791626e-05,
+      "loss": 0.3451,
+      "step": 297
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4058302640914917,
+      "learning_rate": 8.422511659797279e-05,
+      "loss": 0.3005,
+      "step": 298
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.41211822628974915,
+      "learning_rate": 8.411301723809563e-05,
+      "loss": 0.2944,
+      "step": 299
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3935855031013489,
+      "learning_rate": 8.400059616547047e-05,
+      "loss": 0.3223,
+      "step": 300
+    },
+    {
+      "epoch": 0.28,
+      "eval_loss": 0.3055976927280426,
+      "eval_runtime": 358.0992,
+      "eval_samples_per_second": 0.952,
+      "eval_steps_per_second": 0.24,
+      "step": 300
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4585422873497009,
+      "learning_rate": 8.388785444031695e-05,
+      "loss": 0.3923,
+      "step": 301
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3958037495613098,
+      "learning_rate": 8.377479312587879e-05,
+      "loss": 0.2749,
+      "step": 302
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3999941945075989,
+      "learning_rate": 8.366141328841366e-05,
+      "loss": 0.3311,
+      "step": 303
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.34885841608047485,
+      "learning_rate": 8.354771599718312e-05,
+      "loss": 0.2679,
+      "step": 304
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4293624758720398,
+      "learning_rate": 8.343370232444261e-05,
+      "loss": 0.35,
+      "step": 305
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3836628496646881,
+      "learning_rate": 8.331937334543132e-05,
+      "loss": 0.2999,
+      "step": 306
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.48800304532051086,
+      "learning_rate": 8.320473013836196e-05,
+      "loss": 0.4362,
+      "step": 307
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4378533363342285,
+      "learning_rate": 8.308977378441072e-05,
+      "loss": 0.4021,
+      "step": 308
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.39570435881614685,
+      "learning_rate": 8.297450536770697e-05,
+      "loss": 0.3008,
+      "step": 309
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4032250940799713,
+      "learning_rate": 8.28589259753231e-05,
+      "loss": 0.2949,
+      "step": 310
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.41086429357528687,
+      "learning_rate": 8.274303669726426e-05,
+      "loss": 0.334,
+      "step": 311
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3900216221809387,
+      "learning_rate": 8.262683862645804e-05,
+      "loss": 0.3194,
+      "step": 312
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.44460099935531616,
+      "learning_rate": 8.251033285874421e-05,
+      "loss": 0.3536,
+      "step": 313
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.43662020564079285,
+      "learning_rate": 8.239352049286435e-05,
+      "loss": 0.3249,
+      "step": 314
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4098747968673706,
+      "learning_rate": 8.22764026304515e-05,
+      "loss": 0.3284,
+      "step": 315
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.43044736981391907,
+      "learning_rate": 8.21589803760198e-05,
+      "loss": 0.3484,
+      "step": 316
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3859519362449646,
+      "learning_rate": 8.204125483695403e-05,
+      "loss": 0.3187,
+      "step": 317
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.38338860869407654,
+      "learning_rate": 8.192322712349917e-05,
+      "loss": 0.265,
+      "step": 318
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4141724705696106,
+      "learning_rate": 8.180489834875e-05,
+      "loss": 0.3184,
+      "step": 319
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.43011578917503357,
+      "learning_rate": 8.168626962864045e-05,
+      "loss": 0.2684,
+      "step": 320
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4754847288131714,
+      "learning_rate": 8.156734208193327e-05,
+      "loss": 0.3106,
+      "step": 321
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4177388846874237,
+      "learning_rate": 8.144811683020932e-05,
+      "loss": 0.2894,
+      "step": 322
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.43930912017822266,
+      "learning_rate": 8.132859499785707e-05,
+      "loss": 0.3511,
+      "step": 323
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4168340265750885,
+      "learning_rate": 8.120877771206201e-05,
+      "loss": 0.3147,
+      "step": 324
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5014703273773193,
+      "learning_rate": 8.108866610279593e-05,
+      "loss": 0.3998,
+      "step": 325
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.44383499026298523,
+      "learning_rate": 8.096826130280639e-05,
+      "loss": 0.3103,
+      "step": 326
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.45255860686302185,
+      "learning_rate": 8.08475644476059e-05,
+      "loss": 0.3614,
+      "step": 327
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4339107573032379,
+      "learning_rate": 8.072657667546136e-05,
+      "loss": 0.3712,
+      "step": 328
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.41346800327301025,
+      "learning_rate": 8.060529912738315e-05,
+      "loss": 0.366,
+      "step": 329
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.43535444140434265,
+      "learning_rate": 8.048373294711455e-05,
+      "loss": 0.335,
+      "step": 330
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.452105849981308,
+      "learning_rate": 8.036187928112087e-05,
+      "loss": 0.3278,
+      "step": 331
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.45988729596138,
+      "learning_rate": 8.023973927857857e-05,
+      "loss": 0.3988,
+      "step": 332
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5124867558479309,
+      "learning_rate": 8.011731409136454e-05,
+      "loss": 0.3895,
+      "step": 333
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4300025701522827,
+      "learning_rate": 7.99946048740452e-05,
+      "loss": 0.2483,
+      "step": 334
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.44012773036956787,
+      "learning_rate": 7.987161278386554e-05,
+      "loss": 0.3666,
+      "step": 335
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3999829590320587,
+      "learning_rate": 7.974833898073832e-05,
+      "loss": 0.3089,
+      "step": 336
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4425343871116638,
+      "learning_rate": 7.962478462723307e-05,
+      "loss": 0.3655,
+      "step": 337
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4420604705810547,
+      "learning_rate": 7.950095088856508e-05,
+      "loss": 0.3614,
+      "step": 338
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.34093841910362244,
+      "learning_rate": 7.937683893258454e-05,
+      "loss": 0.2206,
+      "step": 339
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4765128791332245,
+      "learning_rate": 7.925244992976538e-05,
+      "loss": 0.4098,
+      "step": 340
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3809426724910736,
+      "learning_rate": 7.912778505319436e-05,
+      "loss": 0.3279,
+      "step": 341
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.32797250151634216,
+      "learning_rate": 7.900284547855991e-05,
+      "loss": 0.1959,
+      "step": 342
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4985835552215576,
+      "learning_rate": 7.88776323841411e-05,
+      "loss": 0.3687,
+      "step": 343
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5066214799880981,
+      "learning_rate": 7.875214695079647e-05,
+      "loss": 0.3889,
+      "step": 344
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3283845782279968,
+      "learning_rate": 7.862639036195298e-05,
+      "loss": 0.2603,
+      "step": 345
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.49629607796669006,
+      "learning_rate": 7.850036380359478e-05,
+      "loss": 0.3502,
+      "step": 346
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4179321527481079,
+      "learning_rate": 7.837406846425204e-05,
+      "loss": 0.3758,
+      "step": 347
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4417211413383484,
+      "learning_rate": 7.824750553498976e-05,
+      "loss": 0.3933,
+      "step": 348
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.49293336272239685,
+      "learning_rate": 7.812067620939652e-05,
+      "loss": 0.3574,
+      "step": 349
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.38600727915763855,
+      "learning_rate": 7.799358168357323e-05,
+      "loss": 0.2905,
+      "step": 350
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5276869535446167,
+      "learning_rate": 7.786622315612183e-05,
+      "loss": 0.3882,
+      "step": 351
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.36139774322509766,
+      "learning_rate": 7.773860182813403e-05,
+      "loss": 0.2631,
+      "step": 352
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4584254026412964,
+      "learning_rate": 7.761071890317994e-05,
+      "loss": 0.3402,
+      "step": 353
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.38829055428504944,
+      "learning_rate": 7.748257558729677e-05,
+      "loss": 0.315,
+      "step": 354
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.41955825686454773,
+      "learning_rate": 7.735417308897736e-05,
+      "loss": 0.3646,
+      "step": 355
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.37745559215545654,
+      "learning_rate": 7.72255126191589e-05,
+      "loss": 0.3035,
+      "step": 356
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4932405948638916,
+      "learning_rate": 7.709659539121144e-05,
+      "loss": 0.3249,
+      "step": 357
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3724788427352905,
+      "learning_rate": 7.696742262092644e-05,
+      "loss": 0.2962,
+      "step": 358
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3544734716415405,
+      "learning_rate": 7.683799552650534e-05,
+      "loss": 0.2311,
+      "step": 359
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3467218577861786,
+      "learning_rate": 7.67083153285481e-05,
+      "loss": 0.2434,
+      "step": 360
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3818730413913727,
+      "learning_rate": 7.657838325004159e-05,
+      "loss": 0.2621,
+      "step": 361
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4077473282814026,
+      "learning_rate": 7.644820051634812e-05,
+      "loss": 0.3281,
+      "step": 362
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.44180914759635925,
+      "learning_rate": 7.63177683551939e-05,
+      "loss": 0.2871,
+      "step": 363
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4257446825504303,
+      "learning_rate": 7.618708799665744e-05,
+      "loss": 0.2843,
+      "step": 364
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.41894400119781494,
+      "learning_rate": 7.605616067315793e-05,
+      "loss": 0.3566,
+      "step": 365
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3829064667224884,
+      "learning_rate": 7.592498761944362e-05,
+      "loss": 0.2278,
+      "step": 366
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3118450343608856,
+      "learning_rate": 7.579357007258022e-05,
+      "loss": 0.1689,
+      "step": 367
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.411563515663147,
+      "learning_rate": 7.56619092719392e-05,
+      "loss": 0.269,
+      "step": 368
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.39412274956703186,
+      "learning_rate": 7.553000645918611e-05,
+      "loss": 0.1981,
+      "step": 369
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.34632736444473267,
+      "learning_rate": 7.539786287826885e-05,
+      "loss": 0.2169,
+      "step": 370
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.35404646396636963,
+      "learning_rate": 7.526547977540592e-05,
+      "loss": 0.2159,
+      "step": 371
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.394235223531723,
+      "learning_rate": 7.51328583990748e-05,
+      "loss": 0.2549,
+      "step": 372
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4355524480342865,
+      "learning_rate": 7.500000000000001e-05,
+      "loss": 0.2943,
+      "step": 373
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3967321813106537,
+      "learning_rate": 7.486690583114136e-05,
+      "loss": 0.2145,
+      "step": 374
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.49327510595321655,
+      "learning_rate": 7.473357714768222e-05,
+      "loss": 0.3641,
+      "step": 375
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5312338471412659,
+      "learning_rate": 7.460001520701755e-05,
+      "loss": 0.4525,
+      "step": 376
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.48641496896743774,
+      "learning_rate": 7.446622126874218e-05,
+      "loss": 0.3505,
+      "step": 377
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.49293845891952515,
+      "learning_rate": 7.433219659463881e-05,
+      "loss": 0.373,
+      "step": 378
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4420691728591919,
+      "learning_rate": 7.41979424486662e-05,
+      "loss": 0.3583,
+      "step": 379
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4443674385547638,
+      "learning_rate": 7.406346009694713e-05,
+      "loss": 0.2788,
+      "step": 380
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3645838797092438,
+      "learning_rate": 7.392875080775665e-05,
+      "loss": 0.2802,
+      "step": 381
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.44193270802497864,
+      "learning_rate": 7.379381585150996e-05,
+      "loss": 0.3137,
+      "step": 382
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.38843849301338196,
+      "learning_rate": 7.365865650075046e-05,
+      "loss": 0.2877,
+      "step": 383
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.32055163383483887,
+      "learning_rate": 7.35232740301378e-05,
+      "loss": 0.2278,
+      "step": 384
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3780610263347626,
+      "learning_rate": 7.338766971643579e-05,
+      "loss": 0.3099,
+      "step": 385
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.415201872587204,
+      "learning_rate": 7.325184483850042e-05,
+      "loss": 0.3172,
+      "step": 386
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.35397133231163025,
+      "learning_rate": 7.311580067726783e-05,
+      "loss": 0.2253,
+      "step": 387
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.36422204971313477,
+      "learning_rate": 7.297953851574206e-05,
+      "loss": 0.2806,
+      "step": 388
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.47656863927841187,
+      "learning_rate": 7.284305963898314e-05,
+      "loss": 0.456,
+      "step": 389
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4232575297355652,
+      "learning_rate": 7.27063653340949e-05,
+      "loss": 0.3449,
+      "step": 390
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.40987321734428406,
+      "learning_rate": 7.25694568902128e-05,
+      "loss": 0.3401,
+      "step": 391
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2791239321231842,
+      "learning_rate": 7.243233559849179e-05,
+      "loss": 0.2076,
+      "step": 392
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.33301910758018494,
+      "learning_rate": 7.229500275209418e-05,
+      "loss": 0.26,
+      "step": 393
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.38714462518692017,
+      "learning_rate": 7.215745964617737e-05,
+      "loss": 0.2855,
+      "step": 394
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.37874332070350647,
+      "learning_rate": 7.201970757788172e-05,
+      "loss": 0.3419,
+      "step": 395
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3609391450881958,
+      "learning_rate": 7.188174784631824e-05,
+      "loss": 0.2309,
+      "step": 396
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4016916751861572,
+      "learning_rate": 7.174358175255636e-05,
+      "loss": 0.2997,
+      "step": 397
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3981953561306,
+      "learning_rate": 7.160521059961169e-05,
+      "loss": 0.278,
+      "step": 398
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3000105321407318,
+      "learning_rate": 7.146663569243371e-05,
+      "loss": 0.2264,
+      "step": 399
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.45137372612953186,
+      "learning_rate": 7.132785833789344e-05,
+      "loss": 0.3996,
+      "step": 400
+    },
+    {
+      "epoch": 0.38,
+      "eval_loss": 0.30177345871925354,
+      "eval_runtime": 360.7687,
+      "eval_samples_per_second": 0.945,
+      "eval_steps_per_second": 0.238,
+      "step": 400
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3610036373138428,
+      "learning_rate": 7.118887984477116e-05,
+      "loss": 0.2843,
+      "step": 401
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.34985464811325073,
+      "learning_rate": 7.104970152374405e-05,
+      "loss": 0.3053,
+      "step": 402
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3719669580459595,
+      "learning_rate": 7.091032468737381e-05,
+      "loss": 0.2868,
+      "step": 403
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.39235973358154297,
+      "learning_rate": 7.077075065009433e-05,
+      "loss": 0.2991,
+      "step": 404
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.36070242524147034,
+      "learning_rate": 7.063098072819919e-05,
+      "loss": 0.3026,
+      "step": 405
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.33055534958839417,
+      "learning_rate": 7.049101623982937e-05,
+      "loss": 0.2357,
+      "step": 406
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.366161048412323,
+      "learning_rate": 7.03508585049608e-05,
+      "loss": 0.255,
+      "step": 407
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4219996929168701,
+      "learning_rate": 7.021050884539177e-05,
+      "loss": 0.331,
+      "step": 408
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.37790825963020325,
+      "learning_rate": 7.006996858473069e-05,
+      "loss": 0.2791,
+      "step": 409
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3606541156768799,
+      "learning_rate": 6.99292390483834e-05,
+      "loss": 0.2376,
+      "step": 410
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3408856689929962,
+      "learning_rate": 6.97883215635408e-05,
+      "loss": 0.2691,
+      "step": 411
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5256190299987793,
+      "learning_rate": 6.964721745916631e-05,
+      "loss": 0.3391,
+      "step": 412
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.37570720911026,
+      "learning_rate": 6.950592806598327e-05,
+      "loss": 0.274,
+      "step": 413
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.37042880058288574,
+      "learning_rate": 6.936445471646249e-05,
+      "loss": 0.267,
+      "step": 414
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3912036418914795,
+      "learning_rate": 6.922279874480959e-05,
+      "loss": 0.3407,
+      "step": 415
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3941766321659088,
+      "learning_rate": 6.908096148695251e-05,
+      "loss": 0.3241,
+      "step": 416
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.37705329060554504,
+      "learning_rate": 6.89389442805288e-05,
+      "loss": 0.2802,
+      "step": 417
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.40080922842025757,
+      "learning_rate": 6.879674846487314e-05,
+      "loss": 0.3082,
+      "step": 418
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4235745966434479,
+      "learning_rate": 6.865437538100457e-05,
+      "loss": 0.3401,
+      "step": 419
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.34460410475730896,
+      "learning_rate": 6.851182637161396e-05,
+      "loss": 0.2323,
+      "step": 420
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.38471102714538574,
+      "learning_rate": 6.836910278105124e-05,
+      "loss": 0.299,
+      "step": 421
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.42625561356544495,
+      "learning_rate": 6.822620595531286e-05,
+      "loss": 0.3428,
+      "step": 422
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.38997682929039,
+      "learning_rate": 6.808313724202894e-05,
+      "loss": 0.2855,
+      "step": 423
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.338751882314682,
+      "learning_rate": 6.793989799045066e-05,
+      "loss": 0.2293,
+      "step": 424
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.44828277826309204,
+      "learning_rate": 6.779648955143754e-05,
+      "loss": 0.3219,
+      "step": 425
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.40802568197250366,
+      "learning_rate": 6.765291327744463e-05,
+      "loss": 0.3313,
+      "step": 426
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.35456421971321106,
+      "learning_rate": 6.750917052250982e-05,
+      "loss": 0.2901,
+      "step": 427
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3902589678764343,
+      "learning_rate": 6.736526264224101e-05,
+      "loss": 0.3094,
+      "step": 428
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.42279359698295593,
+      "learning_rate": 6.722119099380345e-05,
+      "loss": 0.3441,
+      "step": 429
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3679502010345459,
+      "learning_rate": 6.707695693590676e-05,
+      "loss": 0.2131,
+      "step": 430
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.33404216170310974,
+      "learning_rate": 6.693256182879225e-05,
+      "loss": 0.2803,
+      "step": 431
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.41299375891685486,
+      "learning_rate": 6.678800703422003e-05,
+      "loss": 0.3617,
+      "step": 432
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.46660467982292175,
+      "learning_rate": 6.664329391545626e-05,
+      "loss": 0.394,
+      "step": 433
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.45270633697509766,
+      "learning_rate": 6.64984238372601e-05,
+      "loss": 0.4358,
+      "step": 434
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.36836928129196167,
+      "learning_rate": 6.635339816587109e-05,
+      "loss": 0.3001,
+      "step": 435
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.46978554129600525,
+      "learning_rate": 6.620821826899606e-05,
+      "loss": 0.3839,
+      "step": 436
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4234214127063751,
+      "learning_rate": 6.606288551579629e-05,
+      "loss": 0.3462,
+      "step": 437
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4137416183948517,
+      "learning_rate": 6.59174012768747e-05,
+      "loss": 0.3909,
+      "step": 438
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3862276077270508,
+      "learning_rate": 6.577176692426279e-05,
+      "loss": 0.3432,
+      "step": 439
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.39264771342277527,
+      "learning_rate": 6.562598383140772e-05,
+      "loss": 0.3251,
+      "step": 440
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3825792968273163,
+      "learning_rate": 6.548005337315944e-05,
+      "loss": 0.2952,
+      "step": 441
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3140023946762085,
+      "learning_rate": 6.533397692575766e-05,
+      "loss": 0.1815,
+      "step": 442
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.47646018862724304,
+      "learning_rate": 6.518775586681887e-05,
+      "loss": 0.4319,
+      "step": 443
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3555053472518921,
+      "learning_rate": 6.504139157532337e-05,
+      "loss": 0.2815,
+      "step": 444
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3391650319099426,
+      "learning_rate": 6.489488543160224e-05,
+      "loss": 0.2178,
+      "step": 445
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3642699420452118,
+      "learning_rate": 6.47482388173244e-05,
+      "loss": 0.2441,
+      "step": 446
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.41385436058044434,
+      "learning_rate": 6.460145311548341e-05,
+      "loss": 0.3686,
+      "step": 447
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4535539150238037,
+      "learning_rate": 6.445452971038464e-05,
+      "loss": 0.3247,
+      "step": 448
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4449523389339447,
+      "learning_rate": 6.430746998763204e-05,
+      "loss": 0.3756,
+      "step": 449
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.42415305972099304,
+      "learning_rate": 6.416027533411519e-05,
+      "loss": 0.3782,
+      "step": 450
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6403354406356812,
+      "learning_rate": 6.401294713799618e-05,
+      "loss": 0.4349,
+      "step": 451
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3936477601528168,
+      "learning_rate": 6.386548678869644e-05,
+      "loss": 0.3031,
+      "step": 452
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.34828445315361023,
+      "learning_rate": 6.37178956768838e-05,
+      "loss": 0.2739,
+      "step": 453
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.41428014636039734,
+      "learning_rate": 6.35701751944592e-05,
+      "loss": 0.3132,
+      "step": 454
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4079180657863617,
+      "learning_rate": 6.342232673454371e-05,
+      "loss": 0.3405,
+      "step": 455
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3183709383010864,
+      "learning_rate": 6.32743516914653e-05,
+      "loss": 0.2068,
+      "step": 456
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.44731831550598145,
+      "learning_rate": 6.312625146074575e-05,
+      "loss": 0.3524,
+      "step": 457
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.37294724583625793,
+      "learning_rate": 6.297802743908741e-05,
+      "loss": 0.2706,
+      "step": 458
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.38344302773475647,
+      "learning_rate": 6.282968102436009e-05,
+      "loss": 0.2983,
+      "step": 459
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3487352430820465,
+      "learning_rate": 6.268121361558792e-05,
+      "loss": 0.2346,
+      "step": 460
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3923630118370056,
+      "learning_rate": 6.253262661293604e-05,
+      "loss": 0.3481,
+      "step": 461
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.39114248752593994,
+      "learning_rate": 6.238392141769743e-05,
+      "loss": 0.2472,
+      "step": 462
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.42223629355430603,
+      "learning_rate": 6.22350994322798e-05,
+      "loss": 0.3944,
+      "step": 463
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4153856635093689,
+      "learning_rate": 6.208616206019224e-05,
+      "loss": 0.2665,
+      "step": 464
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.36196765303611755,
+      "learning_rate": 6.193711070603202e-05,
+      "loss": 0.2551,
+      "step": 465
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3687761425971985,
+      "learning_rate": 6.178794677547137e-05,
+      "loss": 0.276,
+      "step": 466
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.42377543449401855,
+      "learning_rate": 6.163867167524419e-05,
+      "loss": 0.3256,
+      "step": 467
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4185924828052521,
+      "learning_rate": 6.14892868131328e-05,
+      "loss": 0.3024,
+      "step": 468
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.36687469482421875,
+      "learning_rate": 6.133979359795468e-05,
+      "loss": 0.2726,
+      "step": 469
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3929303288459778,
+      "learning_rate": 6.119019343954914e-05,
+      "loss": 0.2892,
+      "step": 470
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.37266331911087036,
+      "learning_rate": 6.104048774876406e-05,
+      "loss": 0.3042,
+      "step": 471
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4059527814388275,
+      "learning_rate": 6.0890677937442574e-05,
+      "loss": 0.3647,
+      "step": 472
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.525771975517273,
+      "learning_rate": 6.074076541840977e-05,
+      "loss": 0.48,
+      "step": 473
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3635346591472626,
+      "learning_rate": 6.059075160545933e-05,
+      "loss": 0.2356,
+      "step": 474
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3723987340927124,
+      "learning_rate": 6.0440637913340224e-05,
+      "loss": 0.2371,
+      "step": 475
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3875061869621277,
+      "learning_rate": 6.029042575774334e-05,
+      "loss": 0.302,
+      "step": 476
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.43584808707237244,
+      "learning_rate": 6.0140116555288204e-05,
+      "loss": 0.3386,
+      "step": 477
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3697744309902191,
+      "learning_rate": 5.998971172350953e-05,
+      "loss": 0.319,
+      "step": 478
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3352377712726593,
+      "learning_rate": 5.9839212680843925e-05,
+      "loss": 0.2596,
+      "step": 479
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4484308362007141,
+      "learning_rate": 5.968862084661643e-05,
+      "loss": 0.3744,
+      "step": 480
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.37496909499168396,
+      "learning_rate": 5.953793764102722e-05,
+      "loss": 0.3056,
+      "step": 481
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3706286549568176,
+      "learning_rate": 5.938716448513818e-05,
+      "loss": 0.2799,
+      "step": 482
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5032379627227783,
+      "learning_rate": 5.923630280085948e-05,
+      "loss": 0.448,
+      "step": 483
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4166668951511383,
+      "learning_rate": 5.908535401093618e-05,
+      "loss": 0.3674,
+      "step": 484
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3535380959510803,
+      "learning_rate": 5.8934319538934824e-05,
+      "loss": 0.2952,
+      "step": 485
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.34405651688575745,
+      "learning_rate": 5.878320080923001e-05,
+      "loss": 0.2903,
+      "step": 486
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.47468727827072144,
+      "learning_rate": 5.863199924699095e-05,
+      "loss": 0.302,
+      "step": 487
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4812737703323364,
+      "learning_rate": 5.848071627816803e-05,
+      "loss": 0.4249,
+      "step": 488
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.42257994413375854,
+      "learning_rate": 5.8329353329479366e-05,
+      "loss": 0.3483,
+      "step": 489
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3405510485172272,
+      "learning_rate": 5.8177911828397336e-05,
+      "loss": 0.2508,
+      "step": 490
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.41100460290908813,
+      "learning_rate": 5.802639320313514e-05,
+      "loss": 0.3736,
+      "step": 491
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4147120714187622,
+      "learning_rate": 5.787479888263333e-05,
+      "loss": 0.3248,
+      "step": 492
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.42161324620246887,
+      "learning_rate": 5.77231302965463e-05,
+      "loss": 0.3005,
+      "step": 493
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3975301682949066,
+      "learning_rate": 5.757138887522884e-05,
+      "loss": 0.2928,
+      "step": 494
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.33901986479759216,
+      "learning_rate": 5.7419576049722646e-05,
+      "loss": 0.2455,
+      "step": 495
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3582552969455719,
+      "learning_rate": 5.7267693251742794e-05,
+      "loss": 0.2884,
+      "step": 496
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4242113530635834,
+      "learning_rate": 5.7115741913664264e-05,
+      "loss": 0.3009,
+      "step": 497
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.41590404510498047,
+      "learning_rate": 5.696372346850841e-05,
+      "loss": 0.3887,
+      "step": 498
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.40508371591567993,
+      "learning_rate": 5.68116393499295e-05,
+      "loss": 0.3294,
+      "step": 499
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.34662240743637085,
+      "learning_rate": 5.6659490992201094e-05,
+      "loss": 0.2601,
+      "step": 500
+    },
+    {
+      "epoch": 0.47,
+      "eval_loss": 0.2948228418827057,
+      "eval_runtime": 357.3259,
+      "eval_samples_per_second": 0.954,
+      "eval_steps_per_second": 0.241,
+      "step": 500
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.46786829829216003,
+      "learning_rate": 5.650727983020262e-05,
+      "loss": 0.4535,
+      "step": 501
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3314293622970581,
+      "learning_rate": 5.635500729940577e-05,
+      "loss": 0.2142,
+      "step": 502
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4365943968296051,
+      "learning_rate": 5.620267483586105e-05,
+      "loss": 0.3854,
+      "step": 503
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.42165035009384155,
+      "learning_rate": 5.6050283876184116e-05,
+      "loss": 0.3635,
+      "step": 504
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4694943428039551,
+      "learning_rate": 5.5897835857542317e-05,
+      "loss": 0.3568,
+      "step": 505
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.38838309049606323,
+      "learning_rate": 5.574533221764109e-05,
+      "loss": 0.2734,
+      "step": 506
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3702481985092163,
+      "learning_rate": 5.5592774394710465e-05,
+      "loss": 0.2872,
+      "step": 507
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.43085864186286926,
+      "learning_rate": 5.5440163827491454e-05,
+      "loss": 0.29,
+      "step": 508
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.38785114884376526,
+      "learning_rate": 5.5287501955222444e-05,
+      "loss": 0.2727,
+      "step": 509
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.353714257478714,
+      "learning_rate": 5.513479021762572e-05,
+      "loss": 0.2928,
+      "step": 510
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.35981354117393494,
+      "learning_rate": 5.4982030054893785e-05,
+      "loss": 0.2276,
+      "step": 511
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3932090103626251,
+      "learning_rate": 5.482922290767589e-05,
+      "loss": 0.2645,
+      "step": 512
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3919064998626709,
+      "learning_rate": 5.467637021706438e-05,
+      "loss": 0.2883,
+      "step": 513
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5191507339477539,
+      "learning_rate": 5.452347342458104e-05,
+      "loss": 0.3939,
+      "step": 514
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3995441794395447,
+      "learning_rate": 5.437053397216364e-05,
+      "loss": 0.3091,
+      "step": 515
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3995295763015747,
+      "learning_rate": 5.4217553302152237e-05,
+      "loss": 0.3167,
+      "step": 516
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.46911555528640747,
+      "learning_rate": 5.406453285727564e-05,
+      "loss": 0.3569,
+      "step": 517
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4142760634422302,
+      "learning_rate": 5.39114740806377e-05,
+      "loss": 0.2799,
+      "step": 518
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.34572380781173706,
+      "learning_rate": 5.3758378415703825e-05,
+      "loss": 0.2464,
+      "step": 519
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4616849720478058,
+      "learning_rate": 5.360524730628728e-05,
+      "loss": 0.3943,
+      "step": 520
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.40516605973243713,
+      "learning_rate": 5.345208219653561e-05,
+      "loss": 0.2954,
+      "step": 521
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.405485600233078,
+      "learning_rate": 5.3298884530917014e-05,
+      "loss": 0.345,
+      "step": 522
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3549901843070984,
+      "learning_rate": 5.3145655754206714e-05,
+      "loss": 0.2562,
+      "step": 523
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.31494027376174927,
+      "learning_rate": 5.2992397311473316e-05,
+      "loss": 0.2084,
+      "step": 524
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4172990024089813,
+      "learning_rate": 5.2839110648065214e-05,
+      "loss": 0.3242,
+      "step": 525
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4401204288005829,
+      "learning_rate": 5.2685797209596976e-05,
+      "loss": 0.3262,
+      "step": 526
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4312308132648468,
+      "learning_rate": 5.2532458441935636e-05,
+      "loss": 0.3171,
+      "step": 527
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4459749162197113,
+      "learning_rate": 5.2379095791187124e-05,
+      "loss": 0.3471,
+      "step": 528
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3901873230934143,
+      "learning_rate": 5.222571070368257e-05,
+      "loss": 0.2913,
+      "step": 529
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.38350939750671387,
+      "learning_rate": 5.2072304625964785e-05,
+      "loss": 0.2706,
+      "step": 530
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.39073866605758667,
+      "learning_rate": 5.191887900477443e-05,
+      "loss": 0.3448,
+      "step": 531
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.39826685190200806,
+      "learning_rate": 5.176543528703657e-05,
+      "loss": 0.3222,
+      "step": 532
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.37812623381614685,
+      "learning_rate": 5.161197491984684e-05,
+      "loss": 0.2869,
+      "step": 533
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4546664357185364,
+      "learning_rate": 5.1458499350458e-05,
+      "loss": 0.2279,
+      "step": 534
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4484722912311554,
+      "learning_rate": 5.130501002626609e-05,
+      "loss": 0.4678,
+      "step": 535
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.39242762327194214,
+      "learning_rate": 5.11515083947969e-05,
+      "loss": 0.2804,
+      "step": 536
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3988454341888428,
+      "learning_rate": 5.099799590369231e-05,
+      "loss": 0.3251,
+      "step": 537
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4300180673599243,
+      "learning_rate": 5.084447400069655e-05,
+      "loss": 0.3447,
+      "step": 538
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.38128119707107544,
+      "learning_rate": 5.069094413364272e-05,
+      "loss": 0.2844,
+      "step": 539
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4384288489818573,
+      "learning_rate": 5.053740775043892e-05,
+      "loss": 0.3447,
+      "step": 540
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4858492314815521,
+      "learning_rate": 5.038386629905475e-05,
+      "loss": 0.3538,
+      "step": 541
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.44837379455566406,
+      "learning_rate": 5.023032122750759e-05,
+      "loss": 0.3411,
+      "step": 542
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3672875165939331,
+      "learning_rate": 5.007677398384901e-05,
+      "loss": 0.2604,
+      "step": 543
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.40399667620658875,
+      "learning_rate": 4.9923226016151e-05,
+      "loss": 0.3242,
+      "step": 544
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4293772578239441,
+      "learning_rate": 4.976967877249242e-05,
+      "loss": 0.3445,
+      "step": 545
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.30639469623565674,
+      "learning_rate": 4.961613370094527e-05,
+      "loss": 0.242,
+      "step": 546
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4173681437969208,
+      "learning_rate": 4.9462592249561095e-05,
+      "loss": 0.3525,
+      "step": 547
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.39369767904281616,
+      "learning_rate": 4.9309055866357295e-05,
+      "loss": 0.2395,
+      "step": 548
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.41016408801078796,
+      "learning_rate": 4.915552599930345e-05,
+      "loss": 0.3568,
+      "step": 549
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.40518829226493835,
+      "learning_rate": 4.900200409630771e-05,
+      "loss": 0.2835,
+      "step": 550
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3586539328098297,
+      "learning_rate": 4.884849160520311e-05,
+      "loss": 0.238,
+      "step": 551
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.40043312311172485,
+      "learning_rate": 4.869498997373393e-05,
+      "loss": 0.2955,
+      "step": 552
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.45445960760116577,
+      "learning_rate": 4.8541500649542014e-05,
+      "loss": 0.3933,
+      "step": 553
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4135045111179352,
+      "learning_rate": 4.838802508015316e-05,
+      "loss": 0.2947,
+      "step": 554
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4568670690059662,
+      "learning_rate": 4.823456471296344e-05,
+      "loss": 0.3012,
+      "step": 555
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.34699201583862305,
+      "learning_rate": 4.8081120995225574e-05,
+      "loss": 0.2723,
+      "step": 556
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.44254976511001587,
+      "learning_rate": 4.792769537403523e-05,
+      "loss": 0.4082,
+      "step": 557
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4668785333633423,
+      "learning_rate": 4.777428929631743e-05,
+      "loss": 0.3876,
+      "step": 558
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4361911118030548,
+      "learning_rate": 4.762090420881289e-05,
+      "loss": 0.3692,
+      "step": 559
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5063223242759705,
+      "learning_rate": 4.746754155806437e-05,
+      "loss": 0.4413,
+      "step": 560
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3388671875,
+      "learning_rate": 4.731420279040303e-05,
+      "loss": 0.2972,
+      "step": 561
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.36246025562286377,
+      "learning_rate": 4.716088935193479e-05,
+      "loss": 0.1861,
+      "step": 562
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4286564588546753,
+      "learning_rate": 4.7007602688526695e-05,
+      "loss": 0.3404,
+      "step": 563
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3267059028148651,
+      "learning_rate": 4.6854344245793305e-05,
+      "loss": 0.2195,
+      "step": 564
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.34491968154907227,
+      "learning_rate": 4.6701115469083e-05,
+      "loss": 0.2614,
+      "step": 565
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.37428995966911316,
+      "learning_rate": 4.65479178034644e-05,
+      "loss": 0.1958,
+      "step": 566
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3770376741886139,
+      "learning_rate": 4.6394752693712734e-05,
+      "loss": 0.2851,
+      "step": 567
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4104520082473755,
+      "learning_rate": 4.624162158429619e-05,
+      "loss": 0.2633,
+      "step": 568
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4494938850402832,
+      "learning_rate": 4.608852591936231e-05,
+      "loss": 0.3675,
+      "step": 569
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4428771734237671,
+      "learning_rate": 4.593546714272438e-05,
+      "loss": 0.335,
+      "step": 570
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.45229870080947876,
+      "learning_rate": 4.578244669784777e-05,
+      "loss": 0.3534,
+      "step": 571
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.30882728099823,
+      "learning_rate": 4.562946602783636e-05,
+      "loss": 0.196,
+      "step": 572
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.47754567861557007,
+      "learning_rate": 4.547652657541897e-05,
+      "loss": 0.435,
+      "step": 573
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.42711323499679565,
+      "learning_rate": 4.532362978293564e-05,
+      "loss": 0.3714,
+      "step": 574
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.35923489928245544,
+      "learning_rate": 4.517077709232411e-05,
+      "loss": 0.241,
+      "step": 575
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.48945900797843933,
+      "learning_rate": 4.501796994510622e-05,
+      "loss": 0.3904,
+      "step": 576
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.27513808012008667,
+      "learning_rate": 4.4865209782374304e-05,
+      "loss": 0.1842,
+      "step": 577
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0968791246414185,
+      "learning_rate": 4.471249804477758e-05,
+      "loss": 0.2126,
+      "step": 578
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3747063875198364,
+      "learning_rate": 4.4559836172508565e-05,
+      "loss": 0.3206,
+      "step": 579
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8803998827934265,
+      "learning_rate": 4.4407225605289547e-05,
+      "loss": 0.231,
+      "step": 580
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.32417193055152893,
+      "learning_rate": 4.4254667782358924e-05,
+      "loss": 0.2558,
+      "step": 581
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4292859435081482,
+      "learning_rate": 4.410216414245771e-05,
+      "loss": 0.3587,
+      "step": 582
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4002678394317627,
+      "learning_rate": 4.394971612381591e-05,
+      "loss": 0.3028,
+      "step": 583
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.41407111287117004,
+      "learning_rate": 4.379732516413897e-05,
+      "loss": 0.2745,
+      "step": 584
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4084363281726837,
+      "learning_rate": 4.364499270059424e-05,
+      "loss": 0.3036,
+      "step": 585
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4028361141681671,
+      "learning_rate": 4.34927201697974e-05,
+      "loss": 0.3137,
+      "step": 586
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.39395204186439514,
+      "learning_rate": 4.334050900779893e-05,
+      "loss": 0.2781,
+      "step": 587
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4630826413631439,
+      "learning_rate": 4.3188360650070525e-05,
+      "loss": 0.3486,
+      "step": 588
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3602013885974884,
+      "learning_rate": 4.303627653149159e-05,
+      "loss": 0.2194,
+      "step": 589
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3775755763053894,
+      "learning_rate": 4.288425808633575e-05,
+      "loss": 0.2456,
+      "step": 590
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.403229683637619,
+      "learning_rate": 4.2732306748257224e-05,
+      "loss": 0.2866,
+      "step": 591
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3723888099193573,
+      "learning_rate": 4.258042395027738e-05,
+      "loss": 0.2374,
+      "step": 592
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.425342321395874,
+      "learning_rate": 4.2428611124771184e-05,
+      "loss": 0.2825,
+      "step": 593
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3840109407901764,
+      "learning_rate": 4.2276869703453724e-05,
+      "loss": 0.3005,
+      "step": 594
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4066047966480255,
+      "learning_rate": 4.21252011173667e-05,
+      "loss": 0.272,
+      "step": 595
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4041663706302643,
+      "learning_rate": 4.1973606796864884e-05,
+      "loss": 0.2784,
+      "step": 596
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.431432843208313,
+      "learning_rate": 4.182208817160269e-05,
+      "loss": 0.3663,
+      "step": 597
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.35299819707870483,
+      "learning_rate": 4.167064667052065e-05,
+      "loss": 0.2537,
+      "step": 598
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4183855354785919,
+      "learning_rate": 4.151928372183198e-05,
+      "loss": 0.306,
+      "step": 599
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.41894131898880005,
+      "learning_rate": 4.136800075300906e-05,
+      "loss": 0.3159,
+      "step": 600
+    },
+    {
+      "epoch": 0.57,
+      "eval_loss": 0.29318875074386597,
+      "eval_runtime": 354.6455,
+      "eval_samples_per_second": 0.962,
+      "eval_steps_per_second": 0.242,
+      "step": 600
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.47998136281967163,
+      "learning_rate": 4.1216799190770006e-05,
+      "loss": 0.4403,
+      "step": 601
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3485959470272064,
+      "learning_rate": 4.1065680461065194e-05,
+      "loss": 0.2448,
+      "step": 602
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3887343406677246,
+      "learning_rate": 4.0914645989063845e-05,
+      "loss": 0.2941,
+      "step": 603
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4195443093776703,
+      "learning_rate": 4.076369719914055e-05,
+      "loss": 0.2837,
+      "step": 604
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5062023401260376,
+      "learning_rate": 4.061283551486185e-05,
+      "loss": 0.4656,
+      "step": 605
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4645542502403259,
+      "learning_rate": 4.0462062358972796e-05,
+      "loss": 0.4063,
+      "step": 606
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4836640954017639,
+      "learning_rate": 4.03113791533836e-05,
+      "loss": 0.4252,
+      "step": 607
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.41753414273262024,
+      "learning_rate": 4.016078731915608e-05,
+      "loss": 0.2841,
+      "step": 608
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.46088331937789917,
+      "learning_rate": 4.001028827649046e-05,
+      "loss": 0.314,
+      "step": 609
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.38595324754714966,
+      "learning_rate": 3.9859883444711794e-05,
+      "loss": 0.2704,
+      "step": 610
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.35639718174934387,
+      "learning_rate": 3.970957424225666e-05,
+      "loss": 0.2531,
+      "step": 611
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3599381446838379,
+      "learning_rate": 3.955936208665979e-05,
+      "loss": 0.2416,
+      "step": 612
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.38243719935417175,
+      "learning_rate": 3.940924839454066e-05,
+      "loss": 0.2804,
+      "step": 613
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.562547504901886,
+      "learning_rate": 3.925923458159023e-05,
+      "loss": 0.3851,
+      "step": 614
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.37672048807144165,
+      "learning_rate": 3.9109322062557424e-05,
+      "loss": 0.2786,
+      "step": 615
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2807364761829376,
+      "learning_rate": 3.895951225123595e-05,
+      "loss": 0.1985,
+      "step": 616
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.38399168848991394,
+      "learning_rate": 3.880980656045087e-05,
+      "loss": 0.3069,
+      "step": 617
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3434379994869232,
+      "learning_rate": 3.866020640204533e-05,
+      "loss": 0.2676,
+      "step": 618
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3741794526576996,
+      "learning_rate": 3.851071318686721e-05,
+      "loss": 0.2641,
+      "step": 619
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3407473862171173,
+      "learning_rate": 3.8361328324755825e-05,
+      "loss": 0.2297,
+      "step": 620
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.42122921347618103,
+      "learning_rate": 3.821205322452863e-05,
+      "loss": 0.3497,
+      "step": 621
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4289397597312927,
+      "learning_rate": 3.8062889293967976e-05,
+      "loss": 0.4224,
+      "step": 622
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4152087867259979,
+      "learning_rate": 3.791383793980776e-05,
+      "loss": 0.3578,
+      "step": 623
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2982372045516968,
+      "learning_rate": 3.7764900567720196e-05,
+      "loss": 0.2088,
+      "step": 624
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4177871346473694,
+      "learning_rate": 3.761607858230257e-05,
+      "loss": 0.407,
+      "step": 625
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3586585521697998,
+      "learning_rate": 3.746737338706397e-05,
+      "loss": 0.2473,
+      "step": 626
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.41785377264022827,
+      "learning_rate": 3.7318786384412074e-05,
+      "loss": 0.3886,
+      "step": 627
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.38114410638809204,
+      "learning_rate": 3.71703189756399e-05,
+      "loss": 0.3263,
+      "step": 628
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3536858856678009,
+      "learning_rate": 3.70219725609126e-05,
+      "loss": 0.2757,
+      "step": 629
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.39638593792915344,
+      "learning_rate": 3.6873748539254246e-05,
+      "loss": 0.2713,
+      "step": 630
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.42872580885887146,
+      "learning_rate": 3.6725648308534696e-05,
+      "loss": 0.3934,
+      "step": 631
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.401853084564209,
+      "learning_rate": 3.65776732654563e-05,
+      "loss": 0.342,
+      "step": 632
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3954382836818695,
+      "learning_rate": 3.6429824805540815e-05,
+      "loss": 0.3667,
+      "step": 633
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3217025101184845,
+      "learning_rate": 3.628210432311621e-05,
+      "loss": 0.2171,
+      "step": 634
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.43747377395629883,
+      "learning_rate": 3.6134513211303556e-05,
+      "loss": 0.3761,
+      "step": 635
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.42601916193962097,
+      "learning_rate": 3.598705286200382e-05,
+      "loss": 0.303,
+      "step": 636
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.42282041907310486,
+      "learning_rate": 3.58397246658848e-05,
+      "loss": 0.3758,
+      "step": 637
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.45702841877937317,
+      "learning_rate": 3.5692530012367955e-05,
+      "loss": 0.3628,
+      "step": 638
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4157346487045288,
+      "learning_rate": 3.554547028961537e-05,
+      "loss": 0.3154,
+      "step": 639
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.35655519366264343,
+      "learning_rate": 3.5398546884516604e-05,
+      "loss": 0.227,
+      "step": 640
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4063996374607086,
+      "learning_rate": 3.5251761182675625e-05,
+      "loss": 0.3594,
+      "step": 641
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4284207224845886,
+      "learning_rate": 3.510511456839777e-05,
+      "loss": 0.3186,
+      "step": 642
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.44217658042907715,
+      "learning_rate": 3.495860842467664e-05,
+      "loss": 0.346,
+      "step": 643
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3692769408226013,
+      "learning_rate": 3.481224413318114e-05,
+      "loss": 0.3111,
+      "step": 644
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.43843981623649597,
+      "learning_rate": 3.4666023074242356e-05,
+      "loss": 0.3396,
+      "step": 645
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.46626362204551697,
+      "learning_rate": 3.4519946626840573e-05,
+      "loss": 0.3353,
+      "step": 646
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3999076783657074,
+      "learning_rate": 3.4374016168592296e-05,
+      "loss": 0.3004,
+      "step": 647
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4470156133174896,
+      "learning_rate": 3.422823307573722e-05,
+      "loss": 0.3498,
+      "step": 648
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.42150723934173584,
+      "learning_rate": 3.40825987231253e-05,
+      "loss": 0.3844,
+      "step": 649
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.4240174889564514,
+      "learning_rate": 3.393711448420372e-05,
+      "loss": 0.2996,
+      "step": 650
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.41879022121429443,
+      "learning_rate": 3.379178173100396e-05,
+      "loss": 0.2944,
+      "step": 651
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.33797869086265564,
+      "learning_rate": 3.364660183412892e-05,
+      "loss": 0.1914,
+      "step": 652
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.4611934721469879,
+      "learning_rate": 3.3501576162739904e-05,
+      "loss": 0.2862,
+      "step": 653
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3968198597431183,
+      "learning_rate": 3.335670608454376e-05,
+      "loss": 0.3456,
+      "step": 654
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.371667742729187,
+      "learning_rate": 3.321199296577998e-05,
+      "loss": 0.2035,
+      "step": 655
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.4197075664997101,
+      "learning_rate": 3.3067438171207766e-05,
+      "loss": 0.318,
+      "step": 656
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.43643325567245483,
+      "learning_rate": 3.292304306409325e-05,
+      "loss": 0.315,
+      "step": 657
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3940731883049011,
+      "learning_rate": 3.2778809006196564e-05,
+      "loss": 0.3001,
+      "step": 658
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3268234431743622,
+      "learning_rate": 3.263473735775899e-05,
+      "loss": 0.1886,
+      "step": 659
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3809857666492462,
+      "learning_rate": 3.249082947749019e-05,
+      "loss": 0.2497,
+      "step": 660
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4249383807182312,
+      "learning_rate": 3.234708672255538e-05,
+      "loss": 0.2939,
+      "step": 661
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.38137122988700867,
+      "learning_rate": 3.220351044856247e-05,
+      "loss": 0.2676,
+      "step": 662
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.39782699942588806,
+      "learning_rate": 3.206010200954935e-05,
+      "loss": 0.2731,
+      "step": 663
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.46480146050453186,
+      "learning_rate": 3.191686275797107e-05,
+      "loss": 0.3759,
+      "step": 664
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3921947479248047,
+      "learning_rate": 3.177379404468715e-05,
+      "loss": 0.2863,
+      "step": 665
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.31944262981414795,
+      "learning_rate": 3.163089721894876e-05,
+      "loss": 0.1938,
+      "step": 666
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.389593243598938,
+      "learning_rate": 3.148817362838606e-05,
+      "loss": 0.3199,
+      "step": 667
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4163493812084198,
+      "learning_rate": 3.134562461899545e-05,
+      "loss": 0.3146,
+      "step": 668
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3850823938846588,
+      "learning_rate": 3.120325153512687e-05,
+      "loss": 0.2587,
+      "step": 669
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3948986232280731,
+      "learning_rate": 3.10610557194712e-05,
+      "loss": 0.2779,
+      "step": 670
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.45257121324539185,
+      "learning_rate": 3.091903851304751e-05,
+      "loss": 0.4191,
+      "step": 671
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3454974293708801,
+      "learning_rate": 3.077720125519042e-05,
+      "loss": 0.2595,
+      "step": 672
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4489409923553467,
+      "learning_rate": 3.063554528353752e-05,
+      "loss": 0.3357,
+      "step": 673
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3785752058029175,
+      "learning_rate": 3.0494071934016737e-05,
+      "loss": 0.2534,
+      "step": 674
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4088801145553589,
+      "learning_rate": 3.03527825408337e-05,
+      "loss": 0.3411,
+      "step": 675
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3804587721824646,
+      "learning_rate": 3.0211678436459212e-05,
+      "loss": 0.2684,
+      "step": 676
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.40658900141716003,
+      "learning_rate": 3.0070760951616618e-05,
+      "loss": 0.3201,
+      "step": 677
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4057304561138153,
+      "learning_rate": 2.993003141526933e-05,
+      "loss": 0.3269,
+      "step": 678
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.33220186829566956,
+      "learning_rate": 2.978949115460824e-05,
+      "loss": 0.2162,
+      "step": 679
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4094476103782654,
+      "learning_rate": 2.9649141495039223e-05,
+      "loss": 0.2749,
+      "step": 680
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3836745023727417,
+      "learning_rate": 2.950898376017064e-05,
+      "loss": 0.281,
+      "step": 681
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3576979339122772,
+      "learning_rate": 2.9369019271800823e-05,
+      "loss": 0.2443,
+      "step": 682
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.42371636629104614,
+      "learning_rate": 2.9229249349905684e-05,
+      "loss": 0.3594,
+      "step": 683
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3623444736003876,
+      "learning_rate": 2.908967531262618e-05,
+      "loss": 0.2225,
+      "step": 684
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4112566113471985,
+      "learning_rate": 2.895029847625595e-05,
+      "loss": 0.2942,
+      "step": 685
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5042761564254761,
+      "learning_rate": 2.8811120155228844e-05,
+      "loss": 0.4271,
+      "step": 686
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.44093823432922363,
+      "learning_rate": 2.8672141662106578e-05,
+      "loss": 0.3363,
+      "step": 687
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4261564314365387,
+      "learning_rate": 2.853336430756631e-05,
+      "loss": 0.3465,
+      "step": 688
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4838970899581909,
+      "learning_rate": 2.8394789400388328e-05,
+      "loss": 0.3785,
+      "step": 689
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4244787395000458,
+      "learning_rate": 2.8256418247443662e-05,
+      "loss": 0.3105,
+      "step": 690
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.35669493675231934,
+      "learning_rate": 2.811825215368179e-05,
+      "loss": 0.2566,
+      "step": 691
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.44758477807044983,
+      "learning_rate": 2.798029242211828e-05,
+      "loss": 0.3657,
+      "step": 692
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4142071008682251,
+      "learning_rate": 2.7842540353822632e-05,
+      "loss": 0.2921,
+      "step": 693
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.37677180767059326,
+      "learning_rate": 2.770499724790584e-05,
+      "loss": 0.2649,
+      "step": 694
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4031088948249817,
+      "learning_rate": 2.7567664401508225e-05,
+      "loss": 0.3324,
+      "step": 695
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4529830813407898,
+      "learning_rate": 2.7430543109787222e-05,
+      "loss": 0.3907,
+      "step": 696
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.36361509561538696,
+      "learning_rate": 2.729363466590511e-05,
+      "loss": 0.226,
+      "step": 697
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3831719160079956,
+      "learning_rate": 2.7156940361016864e-05,
+      "loss": 0.2808,
+      "step": 698
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4037381708621979,
+      "learning_rate": 2.702046148425795e-05,
+      "loss": 0.275,
+      "step": 699
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4688885509967804,
+      "learning_rate": 2.688419932273219e-05,
+      "loss": 0.3733,
+      "step": 700
+    },
+    {
+      "epoch": 0.66,
+      "eval_loss": 0.28932783007621765,
+      "eval_runtime": 357.2617,
+      "eval_samples_per_second": 0.954,
+      "eval_steps_per_second": 0.241,
+      "step": 700
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.42587098479270935,
+      "learning_rate": 2.6748155161499567e-05,
+      "loss": 0.3448,
+      "step": 701
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.32856640219688416,
+      "learning_rate": 2.6612330283564223e-05,
+      "loss": 0.211,
+      "step": 702
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3772055506706238,
+      "learning_rate": 2.6476725969862227e-05,
+      "loss": 0.1888,
+      "step": 703
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.35770490765571594,
+      "learning_rate": 2.634134349924956e-05,
+      "loss": 0.2438,
+      "step": 704
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.4088924527168274,
+      "learning_rate": 2.6206184148490066e-05,
+      "loss": 0.3126,
+      "step": 705
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.4114530086517334,
+      "learning_rate": 2.6071249192243363e-05,
+      "loss": 0.316,
+      "step": 706
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.4412062168121338,
+      "learning_rate": 2.5936539903052892e-05,
+      "loss": 0.347,
+      "step": 707
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.4265690743923187,
+      "learning_rate": 2.580205755133384e-05,
+      "loss": 0.3956,
+      "step": 708
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.4463072717189789,
+      "learning_rate": 2.5667803405361213e-05,
+      "loss": 0.3905,
+      "step": 709
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.44927293062210083,
+      "learning_rate": 2.5533778731257824e-05,
+      "loss": 0.3912,
+      "step": 710
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3924589157104492,
+      "learning_rate": 2.5399984792982457e-05,
+      "loss": 0.2846,
+      "step": 711
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3993934690952301,
+      "learning_rate": 2.5266422852317795e-05,
+      "loss": 0.3045,
+      "step": 712
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4268989861011505,
+      "learning_rate": 2.513309416885865e-05,
+      "loss": 0.3611,
+      "step": 713
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.39551207423210144,
+      "learning_rate": 2.500000000000001e-05,
+      "loss": 0.2582,
+      "step": 714
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4030238091945648,
+      "learning_rate": 2.4867141600925214e-05,
+      "loss": 0.2922,
+      "step": 715
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4295490086078644,
+      "learning_rate": 2.4734520224594093e-05,
+      "loss": 0.3475,
+      "step": 716
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.42139291763305664,
+      "learning_rate": 2.4602137121731195e-05,
+      "loss": 0.319,
+      "step": 717
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3812490403652191,
+      "learning_rate": 2.44699935408139e-05,
+      "loss": 0.3031,
+      "step": 718
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.26581794023513794,
+      "learning_rate": 2.433809072806081e-05,
+      "loss": 0.1473,
+      "step": 719
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3923310935497284,
+      "learning_rate": 2.4206429927419794e-05,
+      "loss": 0.2733,
+      "step": 720
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4272439181804657,
+      "learning_rate": 2.40750123805564e-05,
+      "loss": 0.2695,
+      "step": 721
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4243634343147278,
+      "learning_rate": 2.3943839326842092e-05,
+      "loss": 0.311,
+      "step": 722
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3555092215538025,
+      "learning_rate": 2.3812912003342568e-05,
+      "loss": 0.2572,
+      "step": 723
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4561547040939331,
+      "learning_rate": 2.3682231644806108e-05,
+      "loss": 0.342,
+      "step": 724
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.39334869384765625,
+      "learning_rate": 2.3551799483651894e-05,
+      "loss": 0.2934,
+      "step": 725
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.49903395771980286,
+      "learning_rate": 2.342161674995843e-05,
+      "loss": 0.3789,
+      "step": 726
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.40439483523368835,
+      "learning_rate": 2.3291684671451903e-05,
+      "loss": 0.312,
+      "step": 727
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3952104151248932,
+      "learning_rate": 2.3162004473494657e-05,
+      "loss": 0.2506,
+      "step": 728
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3781839609146118,
+      "learning_rate": 2.3032577379073577e-05,
+      "loss": 0.309,
+      "step": 729
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3782980144023895,
+      "learning_rate": 2.2903404608788582e-05,
+      "loss": 0.2483,
+      "step": 730
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3966100215911865,
+      "learning_rate": 2.2774487380841115e-05,
+      "loss": 0.3199,
+      "step": 731
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4199528098106384,
+      "learning_rate": 2.2645826911022655e-05,
+      "loss": 0.3077,
+      "step": 732
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4090895652770996,
+      "learning_rate": 2.2517424412703252e-05,
+      "loss": 0.275,
+      "step": 733
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3550466001033783,
+      "learning_rate": 2.2389281096820075e-05,
+      "loss": 0.2582,
+      "step": 734
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4114989936351776,
+      "learning_rate": 2.2261398171865978e-05,
+      "loss": 0.3133,
+      "step": 735
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.39129117131233215,
+      "learning_rate": 2.2133776843878186e-05,
+      "loss": 0.2646,
+      "step": 736
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.41522538661956787,
+      "learning_rate": 2.2006418316426775e-05,
+      "loss": 0.2896,
+      "step": 737
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3981991112232208,
+      "learning_rate": 2.187932379060348e-05,
+      "loss": 0.3261,
+      "step": 738
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.37045934796333313,
+      "learning_rate": 2.1752494465010242e-05,
+      "loss": 0.2199,
+      "step": 739
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.41169455647468567,
+      "learning_rate": 2.1625931535747964e-05,
+      "loss": 0.3276,
+      "step": 740
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4321698844432831,
+      "learning_rate": 2.1499636196405228e-05,
+      "loss": 0.3621,
+      "step": 741
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.42883217334747314,
+      "learning_rate": 2.1373609638047032e-05,
+      "loss": 0.2832,
+      "step": 742
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.41223397850990295,
+      "learning_rate": 2.1247853049203543e-05,
+      "loss": 0.27,
+      "step": 743
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4243120849132538,
+      "learning_rate": 2.112236761585892e-05,
+      "loss": 0.3262,
+      "step": 744
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3997885286808014,
+      "learning_rate": 2.09971545214401e-05,
+      "loss": 0.3099,
+      "step": 745
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.415268212556839,
+      "learning_rate": 2.087221494680563e-05,
+      "loss": 0.3376,
+      "step": 746
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3074762523174286,
+      "learning_rate": 2.074755007023461e-05,
+      "loss": 0.2188,
+      "step": 747
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.30044907331466675,
+      "learning_rate": 2.0623161067415463e-05,
+      "loss": 0.1627,
+      "step": 748
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.42423152923583984,
+      "learning_rate": 2.049904911143492e-05,
+      "loss": 0.3046,
+      "step": 749
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.39862963557243347,
+      "learning_rate": 2.0375215372766944e-05,
+      "loss": 0.3095,
+      "step": 750
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.42895248532295227,
+      "learning_rate": 2.025166101926168e-05,
+      "loss": 0.4187,
+      "step": 751
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3250885605812073,
+      "learning_rate": 2.012838721613447e-05,
+      "loss": 0.2079,
+      "step": 752
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4108840227127075,
+      "learning_rate": 2.0005395125954812e-05,
+      "loss": 0.3812,
+      "step": 753
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.36078789830207825,
+      "learning_rate": 1.9882685908635462e-05,
+      "loss": 0.2741,
+      "step": 754
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.40727007389068604,
+      "learning_rate": 1.9760260721421426e-05,
+      "loss": 0.3124,
+      "step": 755
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.45711779594421387,
+      "learning_rate": 1.963812071887913e-05,
+      "loss": 0.3597,
+      "step": 756
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3516543507575989,
+      "learning_rate": 1.951626705288544e-05,
+      "loss": 0.3002,
+      "step": 757
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3649699091911316,
+      "learning_rate": 1.9394700872616855e-05,
+      "loss": 0.2371,
+      "step": 758
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.43116894364356995,
+      "learning_rate": 1.9273423324538658e-05,
+      "loss": 0.3675,
+      "step": 759
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.36592409014701843,
+      "learning_rate": 1.9152435552394104e-05,
+      "loss": 0.257,
+      "step": 760
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.33106040954589844,
+      "learning_rate": 1.903173869719362e-05,
+      "loss": 0.237,
+      "step": 761
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.34356361627578735,
+      "learning_rate": 1.891133389720407e-05,
+      "loss": 0.2748,
+      "step": 762
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.39263662695884705,
+      "learning_rate": 1.8791222287937982e-05,
+      "loss": 0.298,
+      "step": 763
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.36814969778060913,
+      "learning_rate": 1.8671405002142918e-05,
+      "loss": 0.2109,
+      "step": 764
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.39452245831489563,
+      "learning_rate": 1.855188316979068e-05,
+      "loss": 0.3128,
+      "step": 765
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.426152765750885,
+      "learning_rate": 1.8432657918066732e-05,
+      "loss": 0.3081,
+      "step": 766
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.4281598627567291,
+      "learning_rate": 1.831373037135955e-05,
+      "loss": 0.3807,
+      "step": 767
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.35662761330604553,
+      "learning_rate": 1.8195101651250017e-05,
+      "loss": 0.2135,
+      "step": 768
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.4004388153553009,
+      "learning_rate": 1.807677287650083e-05,
+      "loss": 0.2981,
+      "step": 769
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.39346322417259216,
+      "learning_rate": 1.7958745163045986e-05,
+      "loss": 0.2925,
+      "step": 770
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.4860340356826782,
+      "learning_rate": 1.7841019623980217e-05,
+      "loss": 0.3145,
+      "step": 771
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.34274688363075256,
+      "learning_rate": 1.77235973695485e-05,
+      "loss": 0.2131,
+      "step": 772
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.38533368706703186,
+      "learning_rate": 1.760647950713566e-05,
+      "loss": 0.3226,
+      "step": 773
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3744635581970215,
+      "learning_rate": 1.74896671412558e-05,
+      "loss": 0.2558,
+      "step": 774
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.44143036007881165,
+      "learning_rate": 1.737316137354197e-05,
+      "loss": 0.3752,
+      "step": 775
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3898691236972809,
+      "learning_rate": 1.725696330273575e-05,
+      "loss": 0.3164,
+      "step": 776
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.36931926012039185,
+      "learning_rate": 1.7141074024676912e-05,
+      "loss": 0.2835,
+      "step": 777
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.334431529045105,
+      "learning_rate": 1.7025494632293048e-05,
+      "loss": 0.1897,
+      "step": 778
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.4210609495639801,
+      "learning_rate": 1.6910226215589303e-05,
+      "loss": 0.3503,
+      "step": 779
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.42995116114616394,
+      "learning_rate": 1.6795269861638042e-05,
+      "loss": 0.3461,
+      "step": 780
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.37601590156555176,
+      "learning_rate": 1.6680626654568686e-05,
+      "loss": 0.2667,
+      "step": 781
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.46830493211746216,
+      "learning_rate": 1.656629767555739e-05,
+      "loss": 0.3874,
+      "step": 782
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.43335941433906555,
+      "learning_rate": 1.645228400281689e-05,
+      "loss": 0.2837,
+      "step": 783
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3984016180038452,
+      "learning_rate": 1.6338586711586356e-05,
+      "loss": 0.2518,
+      "step": 784
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.39415448904037476,
+      "learning_rate": 1.6225206874121218e-05,
+      "loss": 0.3141,
+      "step": 785
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.36676424741744995,
+      "learning_rate": 1.6112145559683055e-05,
+      "loss": 0.277,
+      "step": 786
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4425164759159088,
+      "learning_rate": 1.5999403834529548e-05,
+      "loss": 0.3931,
+      "step": 787
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3741757869720459,
+      "learning_rate": 1.5886982761904377e-05,
+      "loss": 0.2563,
+      "step": 788
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4397760331630707,
+      "learning_rate": 1.5774883402027208e-05,
+      "loss": 0.3206,
+      "step": 789
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4487156867980957,
+      "learning_rate": 1.5663106812083743e-05,
+      "loss": 0.3093,
+      "step": 790
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3566402494907379,
+      "learning_rate": 1.555165404621567e-05,
+      "loss": 0.2635,
+      "step": 791
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4316609799861908,
+      "learning_rate": 1.5440526155510763e-05,
+      "loss": 0.3665,
+      "step": 792
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.44090601801872253,
+      "learning_rate": 1.532972418799298e-05,
+      "loss": 0.3702,
+      "step": 793
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4205290973186493,
+      "learning_rate": 1.5219249188612556e-05,
+      "loss": 0.2842,
+      "step": 794
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4262665808200836,
+      "learning_rate": 1.5109102199236152e-05,
+      "loss": 0.3314,
+      "step": 795
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4841248095035553,
+      "learning_rate": 1.4999284258637053e-05,
+      "loss": 0.3435,
+      "step": 796
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.39925479888916016,
+      "learning_rate": 1.488979640248534e-05,
+      "loss": 0.2898,
+      "step": 797
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3584456443786621,
+      "learning_rate": 1.4780639663338125e-05,
+      "loss": 0.2573,
+      "step": 798
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3883351981639862,
+      "learning_rate": 1.4671815070629868e-05,
+      "loss": 0.2335,
+      "step": 799
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3370121121406555,
+      "learning_rate": 1.4563323650662586e-05,
+      "loss": 0.1963,
+      "step": 800
+    },
+    {
+      "epoch": 0.76,
+      "eval_loss": 0.2855388820171356,
+      "eval_runtime": 363.3027,
+      "eval_samples_per_second": 0.939,
+      "eval_steps_per_second": 0.237,
+      "step": 800
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3510516285896301,
+      "learning_rate": 1.4455166426596223e-05,
+      "loss": 0.2067,
+      "step": 801
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.4687687158584595,
+      "learning_rate": 1.434734441843899e-05,
+      "loss": 0.4754,
+      "step": 802
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.37843263149261475,
+      "learning_rate": 1.4239858643037751e-05,
+      "loss": 0.2939,
+      "step": 803
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.36956560611724854,
+      "learning_rate": 1.4132710114068426e-05,
+      "loss": 0.2678,
+      "step": 804
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3482286036014557,
+      "learning_rate": 1.4025899842026442e-05,
+      "loss": 0.1835,
+      "step": 805
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3551994860172272,
+      "learning_rate": 1.3919428834217163e-05,
+      "loss": 0.2146,
+      "step": 806
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3674173057079315,
+      "learning_rate": 1.3813298094746491e-05,
+      "loss": 0.2379,
+      "step": 807
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3596320152282715,
+      "learning_rate": 1.3707508624511262e-05,
+      "loss": 0.2063,
+      "step": 808
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.4381248950958252,
+      "learning_rate": 1.36020614211899e-05,
+      "loss": 0.3591,
+      "step": 809
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.37001508474349976,
+      "learning_rate": 1.349695747923298e-05,
+      "loss": 0.2559,
+      "step": 810
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.39004752039909363,
+      "learning_rate": 1.339219778985385e-05,
+      "loss": 0.282,
+      "step": 811
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.4031607210636139,
+      "learning_rate": 1.3287783341019278e-05,
+      "loss": 0.2832,
+      "step": 812
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.41924864053726196,
+      "learning_rate": 1.3183715117440142e-05,
+      "loss": 0.3279,
+      "step": 813
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.44648006558418274,
+      "learning_rate": 1.3079994100562159e-05,
+      "loss": 0.3121,
+      "step": 814
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.40512123703956604,
+      "learning_rate": 1.2976621268556571e-05,
+      "loss": 0.3256,
+      "step": 815
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.38028234243392944,
+      "learning_rate": 1.2873597596311027e-05,
+      "loss": 0.2781,
+      "step": 816
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.44473540782928467,
+      "learning_rate": 1.2770924055420258e-05,
+      "loss": 0.3052,
+      "step": 817
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.42124703526496887,
+      "learning_rate": 1.2668601614177017e-05,
+      "loss": 0.3425,
+      "step": 818
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3869301676750183,
+      "learning_rate": 1.2566631237562893e-05,
+      "loss": 0.251,
+      "step": 819
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.37112957239151,
+      "learning_rate": 1.246501388723923e-05,
+      "loss": 0.2571,
+      "step": 820
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3932996988296509,
+      "learning_rate": 1.2363750521538064e-05,
+      "loss": 0.265,
+      "step": 821
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.4063304662704468,
+      "learning_rate": 1.2262842095453065e-05,
+      "loss": 0.3754,
+      "step": 822
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.41032230854034424,
+      "learning_rate": 1.2162289560630524e-05,
+      "loss": 0.2949,
+      "step": 823
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.4473836421966553,
+      "learning_rate": 1.2062093865360458e-05,
+      "loss": 0.3533,
+      "step": 824
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3655039072036743,
+      "learning_rate": 1.1962255954567537e-05,
+      "loss": 0.2418,
+      "step": 825
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3819264769554138,
+      "learning_rate": 1.1862776769802276e-05,
+      "loss": 0.2777,
+      "step": 826
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.41048842668533325,
+      "learning_rate": 1.1763657249232107e-05,
+      "loss": 0.3412,
+      "step": 827
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3315446078777313,
+      "learning_rate": 1.1664898327632551e-05,
+      "loss": 0.2357,
+      "step": 828
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.41815823316574097,
+      "learning_rate": 1.1566500936378388e-05,
+      "loss": 0.3432,
+      "step": 829
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5035361647605896,
+      "learning_rate": 1.146846600343488e-05,
+      "loss": 0.3579,
+      "step": 830
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.44803521037101746,
+      "learning_rate": 1.1370794453349037e-05,
+      "loss": 0.4041,
+      "step": 831
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5397711396217346,
+      "learning_rate": 1.1273487207240847e-05,
+      "loss": 0.2778,
+      "step": 832
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6141150593757629,
+      "learning_rate": 1.1176545182794674e-05,
+      "loss": 0.1827,
+      "step": 833
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3853137493133545,
+      "learning_rate": 1.1079969294250514e-05,
+      "loss": 0.2774,
+      "step": 834
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.40370574593544006,
+      "learning_rate": 1.0983760452395413e-05,
+      "loss": 0.3269,
+      "step": 835
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.4078752100467682,
+      "learning_rate": 1.0887919564554894e-05,
+      "loss": 0.3288,
+      "step": 836
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.44073888659477234,
+      "learning_rate": 1.079244753458437e-05,
+      "loss": 0.3352,
+      "step": 837
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.39730504155158997,
+      "learning_rate": 1.0697345262860636e-05,
+      "loss": 0.3329,
+      "step": 838
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.39584875106811523,
+      "learning_rate": 1.0602613646273374e-05,
+      "loss": 0.357,
+      "step": 839
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.48111480474472046,
+      "learning_rate": 1.0508253578216693e-05,
+      "loss": 0.3773,
+      "step": 840
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4419292211532593,
+      "learning_rate": 1.0414265948580692e-05,
+      "loss": 0.3294,
+      "step": 841
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4092678725719452,
+      "learning_rate": 1.032065164374313e-05,
+      "loss": 0.2508,
+      "step": 842
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.33298301696777344,
+      "learning_rate": 1.0227411546560962e-05,
+      "loss": 0.2546,
+      "step": 843
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3825547993183136,
+      "learning_rate": 1.01345465363621e-05,
+      "loss": 0.2984,
+      "step": 844
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.421617716550827,
+      "learning_rate": 1.0042057488937067e-05,
+      "loss": 0.2979,
+      "step": 845
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.41726574301719666,
+      "learning_rate": 9.949945276530781e-06,
+      "loss": 0.3864,
+      "step": 846
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.32341477274894714,
+      "learning_rate": 9.858210767834292e-06,
+      "loss": 0.2488,
+      "step": 847
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4241317808628082,
+      "learning_rate": 9.766854827976617e-06,
+      "loss": 0.2886,
+      "step": 848
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3773423135280609,
+      "learning_rate": 9.675878318516546e-06,
+      "loss": 0.2181,
+      "step": 849
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5464656949043274,
+      "learning_rate": 9.58528209743459e-06,
+      "loss": 0.4018,
+      "step": 850
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3950820565223694,
+      "learning_rate": 9.495067019124792e-06,
+      "loss": 0.3146,
+      "step": 851
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.414507657289505,
+      "learning_rate": 9.405233934386726e-06,
+      "loss": 0.2877,
+      "step": 852
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3792780935764313,
+      "learning_rate": 9.315783690417479e-06,
+      "loss": 0.2463,
+      "step": 853
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.39386847615242004,
+      "learning_rate": 9.226717130803636e-06,
+      "loss": 0.2641,
+      "step": 854
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3786526322364807,
+      "learning_rate": 9.138035095513336e-06,
+      "loss": 0.3282,
+      "step": 855
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.31766387820243835,
+      "learning_rate": 9.049738420888349e-06,
+      "loss": 0.2519,
+      "step": 856
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.4008125960826874,
+      "learning_rate": 8.961827939636196e-06,
+      "loss": 0.2748,
+      "step": 857
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3055739402770996,
+      "learning_rate": 8.87430448082227e-06,
+      "loss": 0.1406,
+      "step": 858
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3408359885215759,
+      "learning_rate": 8.787168869862066e-06,
+      "loss": 0.228,
+      "step": 859
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.408791184425354,
+      "learning_rate": 8.700421928513352e-06,
+      "loss": 0.3971,
+      "step": 860
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.4055376350879669,
+      "learning_rate": 8.614064474868421e-06,
+      "loss": 0.2782,
+      "step": 861
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.40838509798049927,
+      "learning_rate": 8.528097323346407e-06,
+      "loss": 0.3225,
+      "step": 862
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.31662386655807495,
+      "learning_rate": 8.442521284685573e-06,
+      "loss": 0.1899,
+      "step": 863
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.35482585430145264,
+      "learning_rate": 8.357337165935675e-06,
+      "loss": 0.2719,
+      "step": 864
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.446452260017395,
+      "learning_rate": 8.27254577045039e-06,
+      "loss": 0.3415,
+      "step": 865
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3713141679763794,
+      "learning_rate": 8.188147897879667e-06,
+      "loss": 0.2239,
+      "step": 866
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.36580514907836914,
+      "learning_rate": 8.104144344162229e-06,
+      "loss": 0.2501,
+      "step": 867
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.34969714283943176,
+      "learning_rate": 8.02053590151805e-06,
+      "loss": 0.2573,
+      "step": 868
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.40465623140335083,
+      "learning_rate": 7.937323358440935e-06,
+      "loss": 0.2663,
+      "step": 869
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.31615254282951355,
+      "learning_rate": 7.854507499691005e-06,
+      "loss": 0.1639,
+      "step": 870
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3906609117984772,
+      "learning_rate": 7.772089106287344e-06,
+      "loss": 0.3168,
+      "step": 871
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.4438879191875458,
+      "learning_rate": 7.690068955500624e-06,
+      "loss": 0.3684,
+      "step": 872
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.40330731868743896,
+      "learning_rate": 7.608447820845771e-06,
+      "loss": 0.2442,
+      "step": 873
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.37720128893852234,
+      "learning_rate": 7.527226472074678e-06,
+      "loss": 0.2629,
+      "step": 874
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.4088658392429352,
+      "learning_rate": 7.446405675168938e-06,
+      "loss": 0.3297,
+      "step": 875
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3644494414329529,
+      "learning_rate": 7.365986192332625e-06,
+      "loss": 0.191,
+      "step": 876
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.37942373752593994,
+      "learning_rate": 7.285968781985092e-06,
+      "loss": 0.2569,
+      "step": 877
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.33419913053512573,
+      "learning_rate": 7.206354198753862e-06,
+      "loss": 0.1993,
+      "step": 878
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.4307936429977417,
+      "learning_rate": 7.127143193467445e-06,
+      "loss": 0.3552,
+      "step": 879
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.47775211930274963,
+      "learning_rate": 7.048336513148307e-06,
+      "loss": 0.3747,
+      "step": 880
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5507100820541382,
+      "learning_rate": 6.969934901005809e-06,
+      "loss": 0.4254,
+      "step": 881
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4048221707344055,
+      "learning_rate": 6.8919390964291895e-06,
+      "loss": 0.3327,
+      "step": 882
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.38773393630981445,
+      "learning_rate": 6.814349834980621e-06,
+      "loss": 0.2921,
+      "step": 883
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4329793453216553,
+      "learning_rate": 6.7371678483882264e-06,
+      "loss": 0.4112,
+      "step": 884
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.36606279015541077,
+      "learning_rate": 6.660393864539222e-06,
+      "loss": 0.2651,
+      "step": 885
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3340979516506195,
+      "learning_rate": 6.58402860747302e-06,
+      "loss": 0.2025,
+      "step": 886
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3961739242076874,
+      "learning_rate": 6.508072797374454e-06,
+      "loss": 0.3226,
+      "step": 887
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4448009729385376,
+      "learning_rate": 6.4325271505669025e-06,
+      "loss": 0.3162,
+      "step": 888
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.41511788964271545,
+      "learning_rate": 6.3573923795056e-06,
+      "loss": 0.3132,
+      "step": 889
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4135601222515106,
+      "learning_rate": 6.282669192770896e-06,
+      "loss": 0.3333,
+      "step": 890
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3869798481464386,
+      "learning_rate": 6.208358295061572e-06,
+      "loss": 0.2935,
+      "step": 891
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.36579182744026184,
+      "learning_rate": 6.134460387188206e-06,
+      "loss": 0.2528,
+      "step": 892
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3800322115421295,
+      "learning_rate": 6.060976166066546e-06,
+      "loss": 0.2636,
+      "step": 893
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.45664429664611816,
+      "learning_rate": 5.9879063247109405e-06,
+      "loss": 0.3282,
+      "step": 894
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.4228927195072174,
+      "learning_rate": 5.915251552227829e-06,
+      "loss": 0.2821,
+      "step": 895
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.4840869903564453,
+      "learning_rate": 5.8430125338092115e-06,
+      "loss": 0.3798,
+      "step": 896
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.36882513761520386,
+      "learning_rate": 5.771189950726191e-06,
+      "loss": 0.2251,
+      "step": 897
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.4531310200691223,
+      "learning_rate": 5.699784480322568e-06,
+      "loss": 0.3698,
+      "step": 898
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.40636736154556274,
+      "learning_rate": 5.628796796008434e-06,
+      "loss": 0.3134,
+      "step": 899
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.38329729437828064,
+      "learning_rate": 5.558227567253832e-06,
+      "loss": 0.2838,
+      "step": 900
+    },
+    {
+      "epoch": 0.85,
+      "eval_loss": 0.28128230571746826,
+      "eval_runtime": 355.356,
+      "eval_samples_per_second": 0.96,
+      "eval_steps_per_second": 0.242,
+      "step": 900
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.4551596939563751,
+      "learning_rate": 5.488077459582425e-06,
+      "loss": 0.3341,
+      "step": 901
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3780941069126129,
+      "learning_rate": 5.418347134565249e-06,
+      "loss": 0.2976,
+      "step": 902
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.36819714307785034,
+      "learning_rate": 5.349037249814442e-06,
+      "loss": 0.2837,
+      "step": 903
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.4440470337867737,
+      "learning_rate": 5.28014845897708e-06,
+      "loss": 0.3038,
+      "step": 904
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.44559532403945923,
+      "learning_rate": 5.211681411728969e-06,
+      "loss": 0.3274,
+      "step": 905
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.38445454835891724,
+      "learning_rate": 5.143636753768549e-06,
+      "loss": 0.2386,
+      "step": 906
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.37368711829185486,
+      "learning_rate": 5.076015126810784e-06,
+      "loss": 0.2969,
+      "step": 907
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.40762612223625183,
+      "learning_rate": 5.008817168581137e-06,
+      "loss": 0.2762,
+      "step": 908
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.43456774950027466,
+      "learning_rate": 4.94204351280953e-06,
+      "loss": 0.3225,
+      "step": 909
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.4321666955947876,
+      "learning_rate": 4.8756947892243725e-06,
+      "loss": 0.3183,
+      "step": 910
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.397964745759964,
+      "learning_rate": 4.809771623546627e-06,
+      "loss": 0.2195,
+      "step": 911
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.33979281783103943,
+      "learning_rate": 4.744274637483936e-06,
+      "loss": 0.2079,
+      "step": 912
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.34562093019485474,
+      "learning_rate": 4.6792044487247e-06,
+      "loss": 0.2837,
+      "step": 913
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.39363908767700195,
+      "learning_rate": 4.614561670932288e-06,
+      "loss": 0.2746,
+      "step": 914
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.38768595457077026,
+      "learning_rate": 4.550346913739256e-06,
+      "loss": 0.2339,
+      "step": 915
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.38382333517074585,
+      "learning_rate": 4.486560782741578e-06,
+      "loss": 0.2222,
+      "step": 916
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.33442220091819763,
+      "learning_rate": 4.423203879492943e-06,
+      "loss": 0.2147,
+      "step": 917
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4514220356941223,
+      "learning_rate": 4.360276801499086e-06,
+      "loss": 0.3516,
+      "step": 918
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4304735064506531,
+      "learning_rate": 4.29778014221216e-06,
+      "loss": 0.2723,
+      "step": 919
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4404843747615814,
+      "learning_rate": 4.2357144910251e-06,
+      "loss": 0.3078,
+      "step": 920
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4656025767326355,
+      "learning_rate": 4.174080433266136e-06,
+      "loss": 0.3373,
+      "step": 921
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.44521740078926086,
+      "learning_rate": 4.112878550193194e-06,
+      "loss": 0.3176,
+      "step": 922
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.34908926486968994,
+      "learning_rate": 4.05210941898847e-06,
+      "loss": 0.2233,
+      "step": 923
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.4160226881504059,
+      "learning_rate": 3.991773612752952e-06,
+      "loss": 0.2466,
+      "step": 924
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.4414418935775757,
+      "learning_rate": 3.9318717005010495e-06,
+      "loss": 0.319,
+      "step": 925
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.39007139205932617,
+      "learning_rate": 3.872404247155193e-06,
+      "loss": 0.3179,
+      "step": 926
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.42397183179855347,
+      "learning_rate": 3.8133718135405285e-06,
+      "loss": 0.3541,
+      "step": 927
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.4163234233856201,
+      "learning_rate": 3.754774956379614e-06,
+      "loss": 0.2882,
+      "step": 928
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.4577482044696808,
+      "learning_rate": 3.696614228287187e-06,
+      "loss": 0.3111,
+      "step": 929
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.44969362020492554,
+      "learning_rate": 3.6388901777649477e-06,
+      "loss": 0.3241,
+      "step": 930
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.44905760884284973,
+      "learning_rate": 3.581603349196372e-06,
+      "loss": 0.2911,
+      "step": 931
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.47566720843315125,
+      "learning_rate": 3.5247542828415747e-06,
+      "loss": 0.3969,
+      "step": 932
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.389046847820282,
+      "learning_rate": 3.468343514832251e-06,
+      "loss": 0.2412,
+      "step": 933
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3643287718296051,
+      "learning_rate": 3.4123715771665786e-06,
+      "loss": 0.2113,
+      "step": 934
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.40488845109939575,
+      "learning_rate": 3.356838997704226e-06,
+      "loss": 0.328,
+      "step": 935
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3876403570175171,
+      "learning_rate": 3.3017463001613626e-06,
+      "loss": 0.2549,
+      "step": 936
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.33114027976989746,
+      "learning_rate": 3.2470940041057106e-06,
+      "loss": 0.2106,
+      "step": 937
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3751782178878784,
+      "learning_rate": 3.1928826249516987e-06,
+      "loss": 0.2123,
+      "step": 938
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.4227389097213745,
+      "learning_rate": 3.139112673955513e-06,
+      "loss": 0.2857,
+      "step": 939
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.39030423760414124,
+      "learning_rate": 3.08578465821035e-06,
+      "loss": 0.3081,
+      "step": 940
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.37987640500068665,
+      "learning_rate": 3.0328990806415934e-06,
+      "loss": 0.2882,
+      "step": 941
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.4686523377895355,
+      "learning_rate": 2.9804564400020994e-06,
+      "loss": 0.3429,
+      "step": 942
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.415265828371048,
+      "learning_rate": 2.9284572308674596e-06,
+      "loss": 0.3369,
+      "step": 943
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.4172137677669525,
+      "learning_rate": 2.8769019436313715e-06,
+      "loss": 0.3164,
+      "step": 944
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3697200417518616,
+      "learning_rate": 2.825791064500993e-06,
+      "loss": 0.2794,
+      "step": 945
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4631030261516571,
+      "learning_rate": 2.7751250754923573e-06,
+      "loss": 0.305,
+      "step": 946
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3833356201648712,
+      "learning_rate": 2.7249044544258363e-06,
+      "loss": 0.3149,
+      "step": 947
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.45039981603622437,
+      "learning_rate": 2.675129674921639e-06,
+      "loss": 0.4109,
+      "step": 948
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5102495551109314,
+      "learning_rate": 2.625801206395312e-06,
+      "loss": 0.394,
+      "step": 949
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.39149990677833557,
+      "learning_rate": 2.576919514053355e-06,
+      "loss": 0.2679,
+      "step": 950
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.47618532180786133,
+      "learning_rate": 2.528485058888813e-06,
+      "loss": 0.3775,
+      "step": 951
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4113098382949829,
+      "learning_rate": 2.48049829767692e-06,
+      "loss": 0.3306,
+      "step": 952
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5059715509414673,
+      "learning_rate": 2.4329596829708144e-06,
+      "loss": 0.4063,
+      "step": 953
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.37893015146255493,
+      "learning_rate": 2.385869663097251e-06,
+      "loss": 0.2864,
+      "step": 954
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.45276299118995667,
+      "learning_rate": 2.3392286821523724e-06,
+      "loss": 0.3486,
+      "step": 955
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.36547499895095825,
+      "learning_rate": 2.2930371799975594e-06,
+      "loss": 0.2141,
+      "step": 956
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3953619599342346,
+      "learning_rate": 2.2472955922552163e-06,
+      "loss": 0.2577,
+      "step": 957
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.42239704728126526,
+      "learning_rate": 2.202004350304715e-06,
+      "loss": 0.3486,
+      "step": 958
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.4234963655471802,
+      "learning_rate": 2.157163881278312e-06,
+      "loss": 0.3199,
+      "step": 959
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.4011766016483307,
+      "learning_rate": 2.112774608057111e-06,
+      "loss": 0.293,
+      "step": 960
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.42703816294670105,
+      "learning_rate": 2.068836949267089e-06,
+      "loss": 0.3398,
+      "step": 961
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.44674062728881836,
+      "learning_rate": 2.0253513192751373e-06,
+      "loss": 0.2782,
+      "step": 962
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3943072557449341,
+      "learning_rate": 1.982318128185151e-06,
+      "loss": 0.2976,
+      "step": 963
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.43310999870300293,
+      "learning_rate": 1.9397377818341944e-06,
+      "loss": 0.3192,
+      "step": 964
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.4366069734096527,
+      "learning_rate": 1.8976106817886196e-06,
+      "loss": 0.3672,
+      "step": 965
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4421326816082001,
+      "learning_rate": 1.855937225340315e-06,
+      "loss": 0.2958,
+      "step": 966
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3994470536708832,
+      "learning_rate": 1.8147178055029579e-06,
+      "loss": 0.2781,
+      "step": 967
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3471664786338806,
+      "learning_rate": 1.7739528110083004e-06,
+      "loss": 0.2547,
+      "step": 968
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4269879162311554,
+      "learning_rate": 1.7336426263024897e-06,
+      "loss": 0.3138,
+      "step": 969
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.32360756397247314,
+      "learning_rate": 1.6937876315424705e-06,
+      "loss": 0.2062,
+      "step": 970
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.371646523475647,
+      "learning_rate": 1.6543882025923886e-06,
+      "loss": 0.2664,
+      "step": 971
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.44732993841171265,
+      "learning_rate": 1.6154447110200255e-06,
+      "loss": 0.3455,
+      "step": 972
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3923274278640747,
+      "learning_rate": 1.5769575240933422e-06,
+      "loss": 0.2261,
+      "step": 973
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5498822927474976,
+      "learning_rate": 1.5389270047769578e-06,
+      "loss": 0.4919,
+      "step": 974
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.39652010798454285,
+      "learning_rate": 1.5013535117287647e-06,
+      "loss": 0.2988,
+      "step": 975
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.36487555503845215,
+      "learning_rate": 1.4642373992965364e-06,
+      "loss": 0.2046,
+      "step": 976
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.37091100215911865,
+      "learning_rate": 1.4275790175145908e-06,
+      "loss": 0.2323,
+      "step": 977
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3991571068763733,
+      "learning_rate": 1.3913787121004719e-06,
+      "loss": 0.259,
+      "step": 978
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.4371541142463684,
+      "learning_rate": 1.3556368244517114e-06,
+      "loss": 0.3075,
+      "step": 979
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.43370938301086426,
+      "learning_rate": 1.3203536916425841e-06,
+      "loss": 0.2852,
+      "step": 980
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3993806838989258,
+      "learning_rate": 1.2855296464209688e-06,
+      "loss": 0.305,
+      "step": 981
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.45992836356163025,
+      "learning_rate": 1.2511650172051637e-06,
+      "loss": 0.4196,
+      "step": 982
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.4112424850463867,
+      "learning_rate": 1.217260128080816e-06,
+      "loss": 0.2917,
+      "step": 983
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3693508803844452,
+      "learning_rate": 1.1838152987978578e-06,
+      "loss": 0.2799,
+      "step": 984
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.43178197741508484,
+      "learning_rate": 1.1508308447674975e-06,
+      "loss": 0.3036,
+      "step": 985
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.39910873770713806,
+      "learning_rate": 1.1183070770592441e-06,
+      "loss": 0.2914,
+      "step": 986
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.37289297580718994,
+      "learning_rate": 1.0862443023979651e-06,
+      "loss": 0.2589,
+      "step": 987
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.33426058292388916,
+      "learning_rate": 1.0546428231609895e-06,
+      "loss": 0.2327,
+      "step": 988
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.507800281047821,
+      "learning_rate": 1.0235029373752758e-06,
+      "loss": 0.4006,
+      "step": 989
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.4082355499267578,
+      "learning_rate": 9.928249387145982e-07,
+      "loss": 0.2712,
+      "step": 990
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3715406358242035,
+      "learning_rate": 9.6260911649676e-07,
+      "loss": 0.2136,
+      "step": 991
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3425183594226837,
+      "learning_rate": 9.32855755680867e-07,
+      "loss": 0.2196,
+      "step": 992
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.39396902918815613,
+      "learning_rate": 9.035651368646648e-07,
+      "loss": 0.309,
+      "step": 993
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.48934268951416016,
+      "learning_rate": 8.747375362818666e-07,
+      "loss": 0.4108,
+      "step": 994
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.473580539226532,
+      "learning_rate": 8.463732257995571e-07,
+      "loss": 0.3291,
+      "step": 995
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3154013752937317,
+      "learning_rate": 8.184724729156379e-07,
+      "loss": 0.2089,
+      "step": 996
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.43805092573165894,
+      "learning_rate": 7.910355407562742e-07,
+      "loss": 0.2972,
+      "step": 997
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.409624308347702,
+      "learning_rate": 7.640626880734581e-07,
+      "loss": 0.225,
+      "step": 998
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3110271096229553,
+      "learning_rate": 7.375541692425326e-07,
+      "loss": 0.2131,
+      "step": 999
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.42584216594696045,
+      "learning_rate": 7.115102342598101e-07,
+      "loss": 0.3358,
+      "step": 1000
+    },
+    {
+      "epoch": 0.95,
+      "eval_loss": 0.2798261344432831,
+      "eval_runtime": 356.6759,
+      "eval_samples_per_second": 0.956,
+      "eval_steps_per_second": 0.241,
+      "step": 1000
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.41088831424713135,
+      "learning_rate": 6.859311287402081e-07,
+      "loss": 0.2489,
+      "step": 1001
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.32438158988952637,
+      "learning_rate": 6.608170939149283e-07,
+      "loss": 0.1902,
+      "step": 1002
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.39774736762046814,
+      "learning_rate": 6.361683666291973e-07,
+      "loss": 0.3225,
+      "step": 1003
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.47220176458358765,
+      "learning_rate": 6.119851793400189e-07,
+      "loss": 0.3106,
+      "step": 1004
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.4132622182369232,
+      "learning_rate": 5.882677601139919e-07,
+      "loss": 0.27,
+      "step": 1005
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.42977580428123474,
+      "learning_rate": 5.650163326251345e-07,
+      "loss": 0.2634,
+      "step": 1006
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.4762022793292999,
+      "learning_rate": 5.422311161528193e-07,
+      "loss": 0.3769,
+      "step": 1007
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.41503340005874634,
+      "learning_rate": 5.199123255796634e-07,
+      "loss": 0.2625,
+      "step": 1008
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.31677284836769104,
+      "learning_rate": 4.980601713895305e-07,
+      "loss": 0.1755,
+      "step": 1009
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.410304456949234,
+      "learning_rate": 4.7667485966552685e-07,
+      "loss": 0.2458,
+      "step": 1010
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3825591802597046,
+      "learning_rate": 4.55756592088058e-07,
+      "loss": 0.3268,
+      "step": 1011
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.41916465759277344,
+      "learning_rate": 4.353055659329419e-07,
+      "loss": 0.3815,
+      "step": 1012
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4557468295097351,
+      "learning_rate": 4.153219740695435e-07,
+      "loss": 0.3853,
+      "step": 1013
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.48118120431900024,
+      "learning_rate": 3.958060049589485e-07,
+      "loss": 0.3621,
+      "step": 1014
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3555266261100769,
+      "learning_rate": 3.767578426521923e-07,
+      "loss": 0.2468,
+      "step": 1015
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.35945066809654236,
+      "learning_rate": 3.5817766678850615e-07,
+      "loss": 0.2554,
+      "step": 1016
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3540194034576416,
+      "learning_rate": 3.4006565259366295e-07,
+      "loss": 0.2505,
+      "step": 1017
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.38890886306762695,
+      "learning_rate": 3.224219708782894e-07,
+      "loss": 0.2866,
+      "step": 1018
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.48827308416366577,
+      "learning_rate": 3.052467880362675e-07,
+      "loss": 0.3987,
+      "step": 1019
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.38838255405426025,
+      "learning_rate": 2.88540266043158e-07,
+      "loss": 0.3078,
+      "step": 1020
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4259321093559265,
+      "learning_rate": 2.723025624546849e-07,
+      "loss": 0.3319,
+      "step": 1021
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.380047082901001,
+      "learning_rate": 2.5653383040524227e-07,
+      "loss": 0.3087,
+      "step": 1022
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.41765642166137695,
+      "learning_rate": 2.412342186064565e-07,
+      "loss": 0.2376,
+      "step": 1023
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.40853816270828247,
+      "learning_rate": 2.2640387134577058e-07,
+      "loss": 0.3648,
+      "step": 1024
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3422960340976715,
+      "learning_rate": 2.1204292848509555e-07,
+      "loss": 0.2396,
+      "step": 1025
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4270833134651184,
+      "learning_rate": 1.98151525459489e-07,
+      "loss": 0.2757,
+      "step": 1026
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.36285123229026794,
+      "learning_rate": 1.8472979327587292e-07,
+      "loss": 0.2177,
+      "step": 1027
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.45351192355155945,
+      "learning_rate": 1.717778585118013e-07,
+      "loss": 0.3492,
+      "step": 1028
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.3931853473186493,
+      "learning_rate": 1.5929584331427216e-07,
+      "loss": 0.3072,
+      "step": 1029
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5220822095870972,
+      "learning_rate": 1.4728386539856754e-07,
+      "loss": 0.3564,
+      "step": 1030
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.4573669731616974,
+      "learning_rate": 1.3574203804713747e-07,
+      "loss": 0.4152,
+      "step": 1031
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.3876321613788605,
+      "learning_rate": 1.246704701085566e-07,
+      "loss": 0.2819,
+      "step": 1032
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.3961232900619507,
+      "learning_rate": 1.1406926599646373e-07,
+      "loss": 0.2839,
+      "step": 1033
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.3738446533679962,
+      "learning_rate": 1.0393852568860719e-07,
+      "loss": 0.238,
+      "step": 1034
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2902642488479614,
+      "learning_rate": 9.427834472588992e-08,
+      "loss": 0.15,
+      "step": 1035
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.42457300424575806,
+      "learning_rate": 8.508881421145365e-08,
+      "loss": 0.2725,
+      "step": 1036
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.37138867378234863,
+      "learning_rate": 7.637002080985168e-08,
+      "loss": 0.2357,
+      "step": 1037
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.41626298427581787,
+      "learning_rate": 6.812204674619959e-08,
+      "loss": 0.3258,
+      "step": 1038
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.3772933781147003,
+      "learning_rate": 6.034496980542037e-08,
+      "loss": 0.2826,
+      "step": 1039
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3898661434650421,
+      "learning_rate": 5.303886333151153e-08,
+      "loss": 0.2752,
+      "step": 1040
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.38212069869041443,
+      "learning_rate": 4.620379622682358e-08,
+      "loss": 0.2745,
+      "step": 1041
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.40106073021888733,
+      "learning_rate": 3.9839832951465986e-08,
+      "loss": 0.2529,
+      "step": 1042
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3766931891441345,
+      "learning_rate": 3.394703352263551e-08,
+      "loss": 0.2692,
+      "step": 1043
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.4212016761302948,
+      "learning_rate": 2.8525453514099965e-08,
+      "loss": 0.3412,
+      "step": 1044
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.35428979992866516,
+      "learning_rate": 2.3575144055643094e-08,
+      "loss": 0.1975,
+      "step": 1045
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3906976580619812,
+      "learning_rate": 1.9096151832609375e-08,
+      "loss": 0.2476,
+      "step": 1046
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3801487684249878,
+      "learning_rate": 1.5088519085437734e-08,
+      "loss": 0.2996,
+      "step": 1047
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3645074963569641,
+      "learning_rate": 1.1552283609272962e-08,
+      "loss": 0.2287,
+      "step": 1048
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.38731151819229126,
+      "learning_rate": 8.487478753615997e-09,
+      "loss": 0.3338,
+      "step": 1049
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.4575386941432953,
+      "learning_rate": 5.894133422001957e-09,
+      "loss": 0.3718,
+      "step": 1050
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.33703967928886414,
+      "learning_rate": 3.772272071722593e-09,
+      "loss": 0.2247,
+      "step": 1051
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.3912496864795685,
+      "learning_rate": 2.1219147136264382e-09,
+      "loss": 0.2658,
+      "step": 1052
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.4263685643672943,
+      "learning_rate": 9.43076911874563e-10,
+      "loss": 0.3598,
+      "step": 1053
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.4310249090194702,
+      "learning_rate": 2.3576978384065584e-10,
+      "loss": 0.3208,
+      "step": 1054
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.3546713888645172,
+      "learning_rate": 0.0,
+      "loss": 0.2585,
+      "step": 1055
+    },
+    {
+      "epoch": 1.0,
+      "step": 1055,
+      "total_flos": 1.1584683941193318e+17,
+      "train_loss": 0.31262567146412007,
+      "train_runtime": 54950.9684,
+      "train_samples_per_second": 0.615,
+      "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
+}