diff --git a/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa_shuffled-streaming_baseline_pretrained_lora_agg_fc4_s2048/README.md b/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa_shuffled-streaming_baseline_pretrained_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_pretrained_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_pretrained_lora_agg_fc4_s2048/adapter_config.json b/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa_shuffled-streaming_baseline_pretrained_lora_agg_fc4_s2048/adapter_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..d21f0b363f7ec05afcd3d122c845f6472fd089c1
--- /dev/null
+++ b/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa_shuffled-streaming_baseline_pretrained_lora_agg_fc4_s2048/adapter_config.json
@@ -0,0 +1,26 @@
+{
+  "auto_mapping": null,
+  "base_model_name_or_path": "/data/wiedmann/hub/models--lmms-lab--llava-onevision-qwen2-7b-ov/snapshots/0b07bf7565e244cf4f39982249eafe8cd799d6dd",
+  "bias": "none",
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "lora_alpha": 256,
+  "lora_dropout": 0.05,
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "r": 128,
+  "revision": null,
+  "target_modules": [
+    "o_proj",
+    "up_proj",
+    "gate_proj",
+    "k_proj",
+    "down_proj",
+    "q_proj",
+    "v_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_pretrained_lora_agg_fc4_s2048/adapter_model.bin b/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa_shuffled-streaming_baseline_pretrained_lora_agg_fc4_s2048/adapter_model.bin
new file mode 100644
index 0000000000000000000000000000000000000000..8aeff80d9a783fdabf5c97e332f75c08b906d110
--- /dev/null
+++ b/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa_shuffled-streaming_baseline_pretrained_lora_agg_fc4_s2048/adapter_model.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5db022f2cbbf05596d57b79bd4023419084337a898fa97c1295b4a7e3e7e1379
+size 692127130
diff --git a/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa_shuffled-streaming_baseline_pretrained_lora_agg_fc4_s2048/config.json b/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa_shuffled-streaming_baseline_pretrained_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_pretrained_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_pretrained_lora_agg_fc4_s2048/generation_config.json b/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa_shuffled-streaming_baseline_pretrained_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_pretrained_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_pretrained_lora_agg_fc4_s2048/non_lora_trainables.bin b/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa_shuffled-streaming_baseline_pretrained_lora_agg_fc4_s2048/non_lora_trainables.bin
new file mode 100644
index 0000000000000000000000000000000000000000..2a328a72851469dfe91148a2cc8c69e35474f007
--- /dev/null
+++ b/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa_shuffled-streaming_baseline_pretrained_lora_agg_fc4_s2048/non_lora_trainables.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ffcc646274e0f2464e989ab1c8e0bf112c314d9dbf7b12995122f15c0ae0a5ab
+size 211441788
diff --git a/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa_shuffled-streaming_baseline_pretrained_lora_agg_fc4_s2048/trainer_state.json b/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa_shuffled-streaming_baseline_pretrained_lora_agg_fc4_s2048/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..357ad7886c04baa4ef49a015c177116069c4c76a
--- /dev/null
+++ b/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa_shuffled-streaming_baseline_pretrained_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": 8.292136192321777,
+      "learning_rate": 3.125e-06,
+      "loss": 1.1157,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.768860816955566,
+      "learning_rate": 6.25e-06,
+      "loss": 1.0485,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.35648250579834,
+      "learning_rate": 9.375000000000001e-06,
+      "loss": 1.151,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.021062850952148,
+      "learning_rate": 1.25e-05,
+      "loss": 1.1086,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.975555419921875,
+      "learning_rate": 1.5625e-05,
+      "loss": 0.8496,
+      "step": 5
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.289080858230591,
+      "learning_rate": 1.8750000000000002e-05,
+      "loss": 0.7093,
+      "step": 6
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0383782386779785,
+      "learning_rate": 2.1875e-05,
+      "loss": 0.5854,
+      "step": 7
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6611881256103516,
+      "learning_rate": 2.5e-05,
+      "loss": 0.4765,
+      "step": 8
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.040956974029541,
+      "learning_rate": 2.8125000000000003e-05,
+      "loss": 0.4001,
+      "step": 9
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.1689270734786987,
+      "learning_rate": 3.125e-05,
+      "loss": 0.3752,
+      "step": 10
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8431040048599243,
+      "learning_rate": 3.4375e-05,
+      "loss": 0.3271,
+      "step": 11
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7128012776374817,
+      "learning_rate": 3.7500000000000003e-05,
+      "loss": 0.2895,
+      "step": 12
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7321616411209106,
+      "learning_rate": 4.0625000000000005e-05,
+      "loss": 0.3612,
+      "step": 13
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7762096524238586,
+      "learning_rate": 4.375e-05,
+      "loss": 0.3811,
+      "step": 14
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8134955167770386,
+      "learning_rate": 4.6875e-05,
+      "loss": 0.5429,
+      "step": 15
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7506279349327087,
+      "learning_rate": 5e-05,
+      "loss": 0.398,
+      "step": 16
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6209344267845154,
+      "learning_rate": 5.3125000000000004e-05,
+      "loss": 0.4474,
+      "step": 17
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6468816995620728,
+      "learning_rate": 5.6250000000000005e-05,
+      "loss": 0.2964,
+      "step": 18
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5138421654701233,
+      "learning_rate": 5.9375e-05,
+      "loss": 0.3583,
+      "step": 19
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.48039573431015015,
+      "learning_rate": 6.25e-05,
+      "loss": 0.2551,
+      "step": 20
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5722964406013489,
+      "learning_rate": 6.562500000000001e-05,
+      "loss": 0.3795,
+      "step": 21
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.4415222704410553,
+      "learning_rate": 6.875e-05,
+      "loss": 0.1695,
+      "step": 22
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.49626463651657104,
+      "learning_rate": 7.1875e-05,
+      "loss": 0.2995,
+      "step": 23
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7289572954177856,
+      "learning_rate": 7.500000000000001e-05,
+      "loss": 0.4783,
+      "step": 24
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.48712146282196045,
+      "learning_rate": 7.8125e-05,
+      "loss": 0.2244,
+      "step": 25
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5903283357620239,
+      "learning_rate": 8.125000000000001e-05,
+      "loss": 0.3785,
+      "step": 26
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6311174631118774,
+      "learning_rate": 8.4375e-05,
+      "loss": 0.3707,
+      "step": 27
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5376498103141785,
+      "learning_rate": 8.75e-05,
+      "loss": 0.3316,
+      "step": 28
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5353296995162964,
+      "learning_rate": 9.062500000000001e-05,
+      "loss": 0.3356,
+      "step": 29
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5755311846733093,
+      "learning_rate": 9.375e-05,
+      "loss": 0.3016,
+      "step": 30
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5438194274902344,
+      "learning_rate": 9.687500000000001e-05,
+      "loss": 0.3049,
+      "step": 31
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.717596173286438,
+      "learning_rate": 0.0001,
+      "loss": 0.3356,
+      "step": 32
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6170371770858765,
+      "learning_rate": 9.999976423021617e-05,
+      "loss": 0.3748,
+      "step": 33
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5481171011924744,
+      "learning_rate": 9.999905692308813e-05,
+      "loss": 0.3752,
+      "step": 34
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6693851947784424,
+      "learning_rate": 9.999787808528638e-05,
+      "loss": 0.4887,
+      "step": 35
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4554136395454407,
+      "learning_rate": 9.999622772792828e-05,
+      "loss": 0.29,
+      "step": 36
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.48423120379447937,
+      "learning_rate": 9.9994105866578e-05,
+      "loss": 0.3779,
+      "step": 37
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5335750579833984,
+      "learning_rate": 9.999151252124639e-05,
+      "loss": 0.3618,
+      "step": 38
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.43445178866386414,
+      "learning_rate": 9.998844771639073e-05,
+      "loss": 0.2841,
+      "step": 39
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3957979381084442,
+      "learning_rate": 9.998491148091457e-05,
+      "loss": 0.3117,
+      "step": 40
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5982969999313354,
+      "learning_rate": 9.99809038481674e-05,
+      "loss": 0.4339,
+      "step": 41
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.49998074769973755,
+      "learning_rate": 9.997642485594436e-05,
+      "loss": 0.3395,
+      "step": 42
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4327642321586609,
+      "learning_rate": 9.99714745464859e-05,
+      "loss": 0.3582,
+      "step": 43
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4588583707809448,
+      "learning_rate": 9.996605296647737e-05,
+      "loss": 0.293,
+      "step": 44
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5169981122016907,
+      "learning_rate": 9.996016016704854e-05,
+      "loss": 0.4981,
+      "step": 45
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.40961846709251404,
+      "learning_rate": 9.995379620377318e-05,
+      "loss": 0.2724,
+      "step": 46
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3833279013633728,
+      "learning_rate": 9.99469611366685e-05,
+      "loss": 0.2545,
+      "step": 47
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4494597911834717,
+      "learning_rate": 9.993965503019457e-05,
+      "loss": 0.4003,
+      "step": 48
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4212416410446167,
+      "learning_rate": 9.99318779532538e-05,
+      "loss": 0.3389,
+      "step": 49
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.47915512323379517,
+      "learning_rate": 9.992362997919016e-05,
+      "loss": 0.3264,
+      "step": 50
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4876742660999298,
+      "learning_rate": 9.991491118578856e-05,
+      "loss": 0.3748,
+      "step": 51
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4461846351623535,
+      "learning_rate": 9.990572165527413e-05,
+      "loss": 0.3409,
+      "step": 52
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5086767673492432,
+      "learning_rate": 9.98960614743114e-05,
+      "loss": 0.4353,
+      "step": 53
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5064837336540222,
+      "learning_rate": 9.988593073400354e-05,
+      "loss": 0.4387,
+      "step": 54
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.45686185359954834,
+      "learning_rate": 9.987532952989145e-05,
+      "loss": 0.3752,
+      "step": 55
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4261355996131897,
+      "learning_rate": 9.986425796195287e-05,
+      "loss": 0.2756,
+      "step": 56
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4915623664855957,
+      "learning_rate": 9.985271613460144e-05,
+      "loss": 0.4274,
+      "step": 57
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.396177738904953,
+      "learning_rate": 9.984070415668574e-05,
+      "loss": 0.2431,
+      "step": 58
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.48454904556274414,
+      "learning_rate": 9.982822214148821e-05,
+      "loss": 0.4181,
+      "step": 59
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4290277361869812,
+      "learning_rate": 9.981527020672413e-05,
+      "loss": 0.3082,
+      "step": 60
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4694831371307373,
+      "learning_rate": 9.980184847454052e-05,
+      "loss": 0.3985,
+      "step": 61
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4528447091579437,
+      "learning_rate": 9.978795707151492e-05,
+      "loss": 0.3157,
+      "step": 62
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4043198227882385,
+      "learning_rate": 9.977359612865423e-05,
+      "loss": 0.3046,
+      "step": 63
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.44700002670288086,
+      "learning_rate": 9.975876578139355e-05,
+      "loss": 0.3439,
+      "step": 64
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.475717693567276,
+      "learning_rate": 9.974346616959476e-05,
+      "loss": 0.4482,
+      "step": 65
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.398124635219574,
+      "learning_rate": 9.972769743754531e-05,
+      "loss": 0.2932,
+      "step": 66
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.36214354634284973,
+      "learning_rate": 9.971145973395684e-05,
+      "loss": 0.2426,
+      "step": 67
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3866325914859772,
+      "learning_rate": 9.969475321196373e-05,
+      "loss": 0.2148,
+      "step": 68
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4149729013442993,
+      "learning_rate": 9.96775780291217e-05,
+      "loss": 0.2969,
+      "step": 69
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4220738410949707,
+      "learning_rate": 9.965993434740633e-05,
+      "loss": 0.2893,
+      "step": 70
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5020914673805237,
+      "learning_rate": 9.96418223332115e-05,
+      "loss": 0.348,
+      "step": 71
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.40886518359184265,
+      "learning_rate": 9.962324215734782e-05,
+      "loss": 0.2084,
+      "step": 72
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5200652480125427,
+      "learning_rate": 9.960419399504106e-05,
+      "loss": 0.3925,
+      "step": 73
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4241352379322052,
+      "learning_rate": 9.958467802593046e-05,
+      "loss": 0.2713,
+      "step": 74
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4536287486553192,
+      "learning_rate": 9.956469443406706e-05,
+      "loss": 0.3124,
+      "step": 75
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3906016945838928,
+      "learning_rate": 9.954424340791196e-05,
+      "loss": 0.2938,
+      "step": 76
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5077170133590698,
+      "learning_rate": 9.952332514033447e-05,
+      "loss": 0.3497,
+      "step": 77
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4249085783958435,
+      "learning_rate": 9.950193982861047e-05,
+      "loss": 0.39,
+      "step": 78
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4634329676628113,
+      "learning_rate": 9.948008767442034e-05,
+      "loss": 0.3371,
+      "step": 79
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.396329402923584,
+      "learning_rate": 9.94577688838472e-05,
+      "loss": 0.2879,
+      "step": 80
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.42030784487724304,
+      "learning_rate": 9.943498366737487e-05,
+      "loss": 0.357,
+      "step": 81
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.432246595621109,
+      "learning_rate": 9.941173223988602e-05,
+      "loss": 0.3056,
+      "step": 82
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.48889413475990295,
+      "learning_rate": 9.938801482065998e-05,
+      "loss": 0.3409,
+      "step": 83
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4619070887565613,
+      "learning_rate": 9.93638316333708e-05,
+      "loss": 0.3874,
+      "step": 84
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4401371479034424,
+      "learning_rate": 9.933918290608509e-05,
+      "loss": 0.3222,
+      "step": 85
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.44268497824668884,
+      "learning_rate": 9.93140688712598e-05,
+      "loss": 0.3479,
+      "step": 86
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.48481079936027527,
+      "learning_rate": 9.928848976574019e-05,
+      "loss": 0.4132,
+      "step": 87
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.41030359268188477,
+      "learning_rate": 9.926244583075748e-05,
+      "loss": 0.328,
+      "step": 88
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4612218141555786,
+      "learning_rate": 9.923593731192655e-05,
+      "loss": 0.3758,
+      "step": 89
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3865838646888733,
+      "learning_rate": 9.920896445924373e-05,
+      "loss": 0.2874,
+      "step": 90
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.37367960810661316,
+      "learning_rate": 9.918152752708438e-05,
+      "loss": 0.1976,
+      "step": 91
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4356864094734192,
+      "learning_rate": 9.915362677420044e-05,
+      "loss": 0.3609,
+      "step": 92
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4401768445968628,
+      "learning_rate": 9.912526246371814e-05,
+      "loss": 0.4305,
+      "step": 93
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4593706429004669,
+      "learning_rate": 9.909643486313533e-05,
+      "loss": 0.3576,
+      "step": 94
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.46525055170059204,
+      "learning_rate": 9.906714424431913e-05,
+      "loss": 0.3811,
+      "step": 95
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4581677317619324,
+      "learning_rate": 9.903739088350325e-05,
+      "loss": 0.2699,
+      "step": 96
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.46122807264328003,
+      "learning_rate": 9.900717506128541e-05,
+      "loss": 0.3015,
+      "step": 97
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4126625061035156,
+      "learning_rate": 9.897649706262473e-05,
+      "loss": 0.3057,
+      "step": 98
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5149990320205688,
+      "learning_rate": 9.894535717683902e-05,
+      "loss": 0.3757,
+      "step": 99
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.43214476108551025,
+      "learning_rate": 9.891375569760204e-05,
+      "loss": 0.3422,
+      "step": 100
+    },
+    {
+      "epoch": 0.09,
+      "eval_loss": 0.313110888004303,
+      "eval_runtime": 369.003,
+      "eval_samples_per_second": 0.924,
+      "eval_steps_per_second": 0.233,
+      "step": 100
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.46055516600608826,
+      "learning_rate": 9.888169292294076e-05,
+      "loss": 0.2952,
+      "step": 101
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.45672428607940674,
+      "learning_rate": 9.884916915523251e-05,
+      "loss": 0.2906,
+      "step": 102
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5150719881057739,
+      "learning_rate": 9.881618470120215e-05,
+      "loss": 0.3568,
+      "step": 103
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.48563671112060547,
+      "learning_rate": 9.87827398719192e-05,
+      "loss": 0.384,
+      "step": 104
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.45496055483818054,
+      "learning_rate": 9.874883498279484e-05,
+      "loss": 0.2662,
+      "step": 105
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4402937889099121,
+      "learning_rate": 9.871447035357903e-05,
+      "loss": 0.3177,
+      "step": 106
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.41507667303085327,
+      "learning_rate": 9.867964630835743e-05,
+      "loss": 0.2758,
+      "step": 107
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3731890022754669,
+      "learning_rate": 9.864436317554831e-05,
+      "loss": 0.276,
+      "step": 108
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4177415072917938,
+      "learning_rate": 9.860862128789953e-05,
+      "loss": 0.2834,
+      "step": 109
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4130752384662628,
+      "learning_rate": 9.857242098248542e-05,
+      "loss": 0.2267,
+      "step": 110
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.43862080574035645,
+      "learning_rate": 9.853576260070347e-05,
+      "loss": 0.3281,
+      "step": 111
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.39986515045166016,
+      "learning_rate": 9.849864648827125e-05,
+      "loss": 0.2788,
+      "step": 112
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.47540712356567383,
+      "learning_rate": 9.846107299522304e-05,
+      "loss": 0.3813,
+      "step": 113
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5130903124809265,
+      "learning_rate": 9.842304247590667e-05,
+      "loss": 0.4247,
+      "step": 114
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4151199460029602,
+      "learning_rate": 9.838455528897998e-05,
+      "loss": 0.2699,
+      "step": 115
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.424902081489563,
+      "learning_rate": 9.834561179740762e-05,
+      "loss": 0.29,
+      "step": 116
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.45746472477912903,
+      "learning_rate": 9.830621236845754e-05,
+      "loss": 0.3062,
+      "step": 117
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4259563088417053,
+      "learning_rate": 9.826635737369751e-05,
+      "loss": 0.3162,
+      "step": 118
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4131220877170563,
+      "learning_rate": 9.82260471889917e-05,
+      "loss": 0.2649,
+      "step": 119
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.41821715235710144,
+      "learning_rate": 9.818528219449705e-05,
+      "loss": 0.2605,
+      "step": 120
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4035808742046356,
+      "learning_rate": 9.814406277465968e-05,
+      "loss": 0.2577,
+      "step": 121
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4632991552352905,
+      "learning_rate": 9.810238931821138e-05,
+      "loss": 0.3359,
+      "step": 122
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.48431235551834106,
+      "learning_rate": 9.806026221816581e-05,
+      "loss": 0.4035,
+      "step": 123
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.42210009694099426,
+      "learning_rate": 9.801768187181486e-05,
+      "loss": 0.3626,
+      "step": 124
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.39771971106529236,
+      "learning_rate": 9.797464868072488e-05,
+      "loss": 0.2627,
+      "step": 125
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.45902183651924133,
+      "learning_rate": 9.793116305073292e-05,
+      "loss": 0.3434,
+      "step": 126
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4931783974170685,
+      "learning_rate": 9.788722539194291e-05,
+      "loss": 0.388,
+      "step": 127
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4334816634654999,
+      "learning_rate": 9.784283611872169e-05,
+      "loss": 0.2955,
+      "step": 128
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.42453140020370483,
+      "learning_rate": 9.77979956496953e-05,
+      "loss": 0.3484,
+      "step": 129
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.48125508427619934,
+      "learning_rate": 9.77527044077448e-05,
+      "loss": 0.4545,
+      "step": 130
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.47027331590652466,
+      "learning_rate": 9.770696282000244e-05,
+      "loss": 0.3578,
+      "step": 131
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4800739884376526,
+      "learning_rate": 9.766077131784763e-05,
+      "loss": 0.4042,
+      "step": 132
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.405161052942276,
+      "learning_rate": 9.761413033690276e-05,
+      "loss": 0.3029,
+      "step": 133
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4392142593860626,
+      "learning_rate": 9.756704031702918e-05,
+      "loss": 0.3759,
+      "step": 134
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3999072015285492,
+      "learning_rate": 9.751950170232309e-05,
+      "loss": 0.2848,
+      "step": 135
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4366023540496826,
+      "learning_rate": 9.74715149411112e-05,
+      "loss": 0.3292,
+      "step": 136
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.46707382798194885,
+      "learning_rate": 9.742308048594665e-05,
+      "loss": 0.3802,
+      "step": 137
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.2948401868343353,
+      "learning_rate": 9.73741987936047e-05,
+      "loss": 0.1729,
+      "step": 138
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3631648123264313,
+      "learning_rate": 9.732487032507837e-05,
+      "loss": 0.2331,
+      "step": 139
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.49262353777885437,
+      "learning_rate": 9.727509554557417e-05,
+      "loss": 0.3983,
+      "step": 140
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4192075729370117,
+      "learning_rate": 9.722487492450764e-05,
+      "loss": 0.2812,
+      "step": 141
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4935896396636963,
+      "learning_rate": 9.717420893549902e-05,
+      "loss": 0.3661,
+      "step": 142
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.47983360290527344,
+      "learning_rate": 9.712309805636863e-05,
+      "loss": 0.3091,
+      "step": 143
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5539906024932861,
+      "learning_rate": 9.707154276913256e-05,
+      "loss": 0.3764,
+      "step": 144
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.463460236787796,
+      "learning_rate": 9.701954355999792e-05,
+      "loss": 0.3666,
+      "step": 145
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4963212311267853,
+      "learning_rate": 9.696710091935841e-05,
+      "loss": 0.3369,
+      "step": 146
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.39869871735572815,
+      "learning_rate": 9.691421534178966e-05,
+      "loss": 0.2827,
+      "step": 147
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4113066792488098,
+      "learning_rate": 9.686088732604449e-05,
+      "loss": 0.305,
+      "step": 148
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.34672507643699646,
+      "learning_rate": 9.68071173750483e-05,
+      "loss": 0.2199,
+      "step": 149
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4519072473049164,
+      "learning_rate": 9.675290599589429e-05,
+      "loss": 0.3698,
+      "step": 150
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.45395800471305847,
+      "learning_rate": 9.669825369983865e-05,
+      "loss": 0.2844,
+      "step": 151
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4159053862094879,
+      "learning_rate": 9.664316100229578e-05,
+      "loss": 0.3359,
+      "step": 152
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4960489869117737,
+      "learning_rate": 9.658762842283343e-05,
+      "loss": 0.3636,
+      "step": 153
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.35196053981781006,
+      "learning_rate": 9.653165648516776e-05,
+      "loss": 0.2557,
+      "step": 154
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5273671746253967,
+      "learning_rate": 9.647524571715843e-05,
+      "loss": 0.3716,
+      "step": 155
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.38996317982673645,
+      "learning_rate": 9.641839665080363e-05,
+      "loss": 0.301,
+      "step": 156
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.41907793283462524,
+      "learning_rate": 9.636110982223505e-05,
+      "loss": 0.2864,
+      "step": 157
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.41964811086654663,
+      "learning_rate": 9.630338577171282e-05,
+      "loss": 0.3188,
+      "step": 158
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3815939426422119,
+      "learning_rate": 9.624522504362039e-05,
+      "loss": 0.2946,
+      "step": 159
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4507930874824524,
+      "learning_rate": 9.618662818645948e-05,
+      "loss": 0.334,
+      "step": 160
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.33728739619255066,
+      "learning_rate": 9.612759575284483e-05,
+      "loss": 0.2144,
+      "step": 161
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4980108439922333,
+      "learning_rate": 9.606812829949896e-05,
+      "loss": 0.31,
+      "step": 162
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3999607563018799,
+      "learning_rate": 9.600822638724705e-05,
+      "loss": 0.2857,
+      "step": 163
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3947831094264984,
+      "learning_rate": 9.594789058101153e-05,
+      "loss": 0.2611,
+      "step": 164
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.39525744318962097,
+      "learning_rate": 9.588712144980681e-05,
+      "loss": 0.3172,
+      "step": 165
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4430619776248932,
+      "learning_rate": 9.582591956673387e-05,
+      "loss": 0.3158,
+      "step": 166
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.43560928106307983,
+      "learning_rate": 9.576428550897489e-05,
+      "loss": 0.2594,
+      "step": 167
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.45979127287864685,
+      "learning_rate": 9.570221985778785e-05,
+      "loss": 0.3452,
+      "step": 168
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.429950088262558,
+      "learning_rate": 9.563972319850092e-05,
+      "loss": 0.3118,
+      "step": 169
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.2910686433315277,
+      "learning_rate": 9.557679612050708e-05,
+      "loss": 0.2042,
+      "step": 170
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4286372661590576,
+      "learning_rate": 9.551343921725843e-05,
+      "loss": 0.3646,
+      "step": 171
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.524486780166626,
+      "learning_rate": 9.544965308626074e-05,
+      "loss": 0.4599,
+      "step": 172
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.35753336548805237,
+      "learning_rate": 9.538543832906773e-05,
+      "loss": 0.2574,
+      "step": 173
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.42938441038131714,
+      "learning_rate": 9.53207955512753e-05,
+      "loss": 0.3103,
+      "step": 174
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.45574262738227844,
+      "learning_rate": 9.525572536251607e-05,
+      "loss": 0.4114,
+      "step": 175
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.44598862528800964,
+      "learning_rate": 9.519022837645337e-05,
+      "loss": 0.3349,
+      "step": 176
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.42717549204826355,
+      "learning_rate": 9.512430521077565e-05,
+      "loss": 0.2905,
+      "step": 177
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.42020952701568604,
+      "learning_rate": 9.505795648719048e-05,
+      "loss": 0.3376,
+      "step": 178
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.473145455121994,
+      "learning_rate": 9.499118283141887e-05,
+      "loss": 0.3698,
+      "step": 179
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3857220709323883,
+      "learning_rate": 9.492398487318921e-05,
+      "loss": 0.3186,
+      "step": 180
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3907882273197174,
+      "learning_rate": 9.485636324623146e-05,
+      "loss": 0.2562,
+      "step": 181
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.41519075632095337,
+      "learning_rate": 9.478831858827104e-05,
+      "loss": 0.3563,
+      "step": 182
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.41073116660118103,
+      "learning_rate": 9.471985154102292e-05,
+      "loss": 0.3255,
+      "step": 183
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7622061371803284,
+      "learning_rate": 9.465096275018556e-05,
+      "loss": 0.422,
+      "step": 184
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.6388115882873535,
+      "learning_rate": 9.458165286543476e-05,
+      "loss": 0.3614,
+      "step": 185
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.385911226272583,
+      "learning_rate": 9.451192254041758e-05,
+      "loss": 0.3682,
+      "step": 186
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5957494378089905,
+      "learning_rate": 9.444177243274618e-05,
+      "loss": 0.4565,
+      "step": 187
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.45678314566612244,
+      "learning_rate": 9.437120320399158e-05,
+      "loss": 0.3469,
+      "step": 188
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4074920117855072,
+      "learning_rate": 9.430021551967744e-05,
+      "loss": 0.2817,
+      "step": 189
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.38819795846939087,
+      "learning_rate": 9.422881004927381e-05,
+      "loss": 0.2778,
+      "step": 190
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.41338834166526794,
+      "learning_rate": 9.415698746619079e-05,
+      "loss": 0.3307,
+      "step": 191
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4173876941204071,
+      "learning_rate": 9.408474844777217e-05,
+      "loss": 0.2614,
+      "step": 192
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.48882555961608887,
+      "learning_rate": 9.401209367528908e-05,
+      "loss": 0.4768,
+      "step": 193
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4182356894016266,
+      "learning_rate": 9.393902383393347e-05,
+      "loss": 0.3074,
+      "step": 194
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5591652393341064,
+      "learning_rate": 9.386553961281179e-05,
+      "loss": 0.4226,
+      "step": 195
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.44878554344177246,
+      "learning_rate": 9.379164170493844e-05,
+      "loss": 0.3831,
+      "step": 196
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.43590179085731506,
+      "learning_rate": 9.371733080722911e-05,
+      "loss": 0.3421,
+      "step": 197
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3382357358932495,
+      "learning_rate": 9.36426076204944e-05,
+      "loss": 0.2252,
+      "step": 198
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.408507764339447,
+      "learning_rate": 9.35674728494331e-05,
+      "loss": 0.3549,
+      "step": 199
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.38054025173187256,
+      "learning_rate": 9.349192720262555e-05,
+      "loss": 0.2996,
+      "step": 200
+    },
+    {
+      "epoch": 0.19,
+      "eval_loss": 0.30657097697257996,
+      "eval_runtime": 369.0611,
+      "eval_samples_per_second": 0.924,
+      "eval_steps_per_second": 0.233,
+      "step": 200
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3924165964126587,
+      "learning_rate": 9.341597139252698e-05,
+      "loss": 0.3209,
+      "step": 201
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4241688549518585,
+      "learning_rate": 9.333960613546079e-05,
+      "loss": 0.3052,
+      "step": 202
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3696678578853607,
+      "learning_rate": 9.326283215161178e-05,
+      "loss": 0.2661,
+      "step": 203
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3506186604499817,
+      "learning_rate": 9.31856501650194e-05,
+      "loss": 0.1653,
+      "step": 204
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4023956060409546,
+      "learning_rate": 9.310806090357082e-05,
+      "loss": 0.3561,
+      "step": 205
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3326328694820404,
+      "learning_rate": 9.303006509899418e-05,
+      "loss": 0.1831,
+      "step": 206
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.49635419249534607,
+      "learning_rate": 9.29516634868517e-05,
+      "loss": 0.3466,
+      "step": 207
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5020710229873657,
+      "learning_rate": 9.287285680653254e-05,
+      "loss": 0.3738,
+      "step": 208
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.41380178928375244,
+      "learning_rate": 9.279364580124614e-05,
+      "loss": 0.3103,
+      "step": 209
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3961673378944397,
+      "learning_rate": 9.27140312180149e-05,
+      "loss": 0.2307,
+      "step": 210
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5154891610145569,
+      "learning_rate": 9.263401380766739e-05,
+      "loss": 0.4062,
+      "step": 211
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4557787775993347,
+      "learning_rate": 9.255359432483105e-05,
+      "loss": 0.3533,
+      "step": 212
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3933027386665344,
+      "learning_rate": 9.247277352792533e-05,
+      "loss": 0.2485,
+      "step": 213
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.48823875188827515,
+      "learning_rate": 9.239155217915422e-05,
+      "loss": 0.3584,
+      "step": 214
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.37317004799842834,
+      "learning_rate": 9.230993104449939e-05,
+      "loss": 0.2333,
+      "step": 215
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.40248480439186096,
+      "learning_rate": 9.222791089371265e-05,
+      "loss": 0.3014,
+      "step": 216
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.38044705986976624,
+      "learning_rate": 9.214549250030899e-05,
+      "loss": 0.2735,
+      "step": 217
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4306708574295044,
+      "learning_rate": 9.206267664155907e-05,
+      "loss": 0.3911,
+      "step": 218
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.348537802696228,
+      "learning_rate": 9.197946409848194e-05,
+      "loss": 0.2326,
+      "step": 219
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3672493100166321,
+      "learning_rate": 9.189585565583778e-05,
+      "loss": 0.2972,
+      "step": 220
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.39911115169525146,
+      "learning_rate": 9.181185210212034e-05,
+      "loss": 0.3269,
+      "step": 221
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.37353402376174927,
+      "learning_rate": 9.172745422954961e-05,
+      "loss": 0.2626,
+      "step": 222
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3879914879798889,
+      "learning_rate": 9.164266283406433e-05,
+      "loss": 0.3379,
+      "step": 223
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.39337465167045593,
+      "learning_rate": 9.155747871531444e-05,
+      "loss": 0.323,
+      "step": 224
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.44832608103752136,
+      "learning_rate": 9.14719026766536e-05,
+      "loss": 0.3719,
+      "step": 225
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4089319109916687,
+      "learning_rate": 9.13859355251316e-05,
+      "loss": 0.2945,
+      "step": 226
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3743653893470764,
+      "learning_rate": 9.129957807148666e-05,
+      "loss": 0.2273,
+      "step": 227
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4819430112838745,
+      "learning_rate": 9.121283113013795e-05,
+      "loss": 0.4118,
+      "step": 228
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.374355286359787,
+      "learning_rate": 9.112569551917774e-05,
+      "loss": 0.2561,
+      "step": 229
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4936485290527344,
+      "learning_rate": 9.103817206036382e-05,
+      "loss": 0.3828,
+      "step": 230
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.46877849102020264,
+      "learning_rate": 9.095026157911166e-05,
+      "loss": 0.3409,
+      "step": 231
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.43224769830703735,
+      "learning_rate": 9.086196490448668e-05,
+      "loss": 0.2612,
+      "step": 232
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.45080262422561646,
+      "learning_rate": 9.077328286919638e-05,
+      "loss": 0.3607,
+      "step": 233
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4469703137874603,
+      "learning_rate": 9.068421630958254e-05,
+      "loss": 0.378,
+      "step": 234
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.349186509847641,
+      "learning_rate": 9.059476606561328e-05,
+      "loss": 0.2255,
+      "step": 235
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4107833802700043,
+      "learning_rate": 9.050493298087523e-05,
+      "loss": 0.3467,
+      "step": 236
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.39175042510032654,
+      "learning_rate": 9.041471790256542e-05,
+      "loss": 0.3095,
+      "step": 237
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.38782957196235657,
+      "learning_rate": 9.032412168148345e-05,
+      "loss": 0.3167,
+      "step": 238
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4580951929092407,
+      "learning_rate": 9.02331451720234e-05,
+      "loss": 0.3431,
+      "step": 239
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.35884562134742737,
+      "learning_rate": 9.014178923216572e-05,
+      "loss": 0.2397,
+      "step": 240
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.38495123386383057,
+      "learning_rate": 9.005005472346924e-05,
+      "loss": 0.2561,
+      "step": 241
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.44250035285949707,
+      "learning_rate": 8.995794251106295e-05,
+      "loss": 0.3898,
+      "step": 242
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.40027764439582825,
+      "learning_rate": 8.986545346363792e-05,
+      "loss": 0.3133,
+      "step": 243
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4065520167350769,
+      "learning_rate": 8.977258845343905e-05,
+      "loss": 0.3379,
+      "step": 244
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.37577980756759644,
+      "learning_rate": 8.967934835625689e-05,
+      "loss": 0.2923,
+      "step": 245
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4770040214061737,
+      "learning_rate": 8.958573405141932e-05,
+      "loss": 0.3388,
+      "step": 246
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4799995720386505,
+      "learning_rate": 8.949174642178332e-05,
+      "loss": 0.3908,
+      "step": 247
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4516853094100952,
+      "learning_rate": 8.939738635372664e-05,
+      "loss": 0.3261,
+      "step": 248
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.46554887294769287,
+      "learning_rate": 8.930265473713938e-05,
+      "loss": 0.37,
+      "step": 249
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.44500887393951416,
+      "learning_rate": 8.920755246541563e-05,
+      "loss": 0.3501,
+      "step": 250
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4066321849822998,
+      "learning_rate": 8.911208043544513e-05,
+      "loss": 0.3176,
+      "step": 251
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3847165107727051,
+      "learning_rate": 8.90162395476046e-05,
+      "loss": 0.2943,
+      "step": 252
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5154759287834167,
+      "learning_rate": 8.89200307057495e-05,
+      "loss": 0.4593,
+      "step": 253
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4412396550178528,
+      "learning_rate": 8.882345481720533e-05,
+      "loss": 0.3951,
+      "step": 254
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3995065689086914,
+      "learning_rate": 8.872651279275916e-05,
+      "loss": 0.3207,
+      "step": 255
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4856877028942108,
+      "learning_rate": 8.862920554665098e-05,
+      "loss": 0.3392,
+      "step": 256
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.397654265165329,
+      "learning_rate": 8.853153399656513e-05,
+      "loss": 0.3074,
+      "step": 257
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.45534300804138184,
+      "learning_rate": 8.843349906362163e-05,
+      "loss": 0.3804,
+      "step": 258
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4625827372074127,
+      "learning_rate": 8.833510167236747e-05,
+      "loss": 0.4009,
+      "step": 259
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.39180710911750793,
+      "learning_rate": 8.823634275076791e-05,
+      "loss": 0.3281,
+      "step": 260
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4082053601741791,
+      "learning_rate": 8.813722323019773e-05,
+      "loss": 0.3767,
+      "step": 261
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.35721302032470703,
+      "learning_rate": 8.803774404543247e-05,
+      "loss": 0.2541,
+      "step": 262
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3870796859264374,
+      "learning_rate": 8.793790613463955e-05,
+      "loss": 0.3266,
+      "step": 263
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.31306833028793335,
+      "learning_rate": 8.783771043936948e-05,
+      "loss": 0.2212,
+      "step": 264
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3992561995983124,
+      "learning_rate": 8.773715790454695e-05,
+      "loss": 0.2923,
+      "step": 265
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3643523156642914,
+      "learning_rate": 8.763624947846195e-05,
+      "loss": 0.2232,
+      "step": 266
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4733940362930298,
+      "learning_rate": 8.753498611276079e-05,
+      "loss": 0.3575,
+      "step": 267
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4740948975086212,
+      "learning_rate": 8.743336876243712e-05,
+      "loss": 0.4098,
+      "step": 268
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5345129370689392,
+      "learning_rate": 8.7331398385823e-05,
+      "loss": 0.4515,
+      "step": 269
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.48729366064071655,
+      "learning_rate": 8.722907594457974e-05,
+      "loss": 0.3894,
+      "step": 270
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4488878846168518,
+      "learning_rate": 8.712640240368899e-05,
+      "loss": 0.3907,
+      "step": 271
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4957049489021301,
+      "learning_rate": 8.702337873144343e-05,
+      "loss": 0.3139,
+      "step": 272
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.39673352241516113,
+      "learning_rate": 8.692000589943785e-05,
+      "loss": 0.2665,
+      "step": 273
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4806172549724579,
+      "learning_rate": 8.681628488255986e-05,
+      "loss": 0.3941,
+      "step": 274
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.382018506526947,
+      "learning_rate": 8.671221665898073e-05,
+      "loss": 0.2774,
+      "step": 275
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.42508023977279663,
+      "learning_rate": 8.660780221014616e-05,
+      "loss": 0.3222,
+      "step": 276
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.42183950543403625,
+      "learning_rate": 8.650304252076703e-05,
+      "loss": 0.304,
+      "step": 277
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3877944350242615,
+      "learning_rate": 8.639793857881011e-05,
+      "loss": 0.265,
+      "step": 278
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.535112202167511,
+      "learning_rate": 8.629249137548874e-05,
+      "loss": 0.3867,
+      "step": 279
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.40608519315719604,
+      "learning_rate": 8.618670190525352e-05,
+      "loss": 0.2603,
+      "step": 280
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4165409207344055,
+      "learning_rate": 8.608057116578283e-05,
+      "loss": 0.2896,
+      "step": 281
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4698445796966553,
+      "learning_rate": 8.597410015797358e-05,
+      "loss": 0.4019,
+      "step": 282
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.38493677973747253,
+      "learning_rate": 8.586728988593158e-05,
+      "loss": 0.2312,
+      "step": 283
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4839042127132416,
+      "learning_rate": 8.576014135696226e-05,
+      "loss": 0.3249,
+      "step": 284
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4200250804424286,
+      "learning_rate": 8.565265558156101e-05,
+      "loss": 0.2803,
+      "step": 285
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5181607604026794,
+      "learning_rate": 8.554483357340379e-05,
+      "loss": 0.3557,
+      "step": 286
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3874339163303375,
+      "learning_rate": 8.543667634933742e-05,
+      "loss": 0.2463,
+      "step": 287
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.40273764729499817,
+      "learning_rate": 8.532818492937014e-05,
+      "loss": 0.2951,
+      "step": 288
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.42094656825065613,
+      "learning_rate": 8.521936033666187e-05,
+      "loss": 0.3021,
+      "step": 289
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.41578373312950134,
+      "learning_rate": 8.511020359751466e-05,
+      "loss": 0.3096,
+      "step": 290
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3404378294944763,
+      "learning_rate": 8.500071574136295e-05,
+      "loss": 0.2092,
+      "step": 291
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4218709468841553,
+      "learning_rate": 8.489089780076386e-05,
+      "loss": 0.2908,
+      "step": 292
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.44549334049224854,
+      "learning_rate": 8.478075081138745e-05,
+      "loss": 0.298,
+      "step": 293
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5066516399383545,
+      "learning_rate": 8.467027581200702e-05,
+      "loss": 0.406,
+      "step": 294
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.475799024105072,
+      "learning_rate": 8.455947384448924e-05,
+      "loss": 0.3564,
+      "step": 295
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3603907823562622,
+      "learning_rate": 8.444834595378434e-05,
+      "loss": 0.2524,
+      "step": 296
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.41698744893074036,
+      "learning_rate": 8.433689318791626e-05,
+      "loss": 0.3398,
+      "step": 297
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.40112778544425964,
+      "learning_rate": 8.422511659797279e-05,
+      "loss": 0.2993,
+      "step": 298
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4104403853416443,
+      "learning_rate": 8.411301723809563e-05,
+      "loss": 0.2949,
+      "step": 299
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4038005471229553,
+      "learning_rate": 8.400059616547047e-05,
+      "loss": 0.3284,
+      "step": 300
+    },
+    {
+      "epoch": 0.28,
+      "eval_loss": 0.3049333393573761,
+      "eval_runtime": 365.0811,
+      "eval_samples_per_second": 0.934,
+      "eval_steps_per_second": 0.236,
+      "step": 300
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4575110375881195,
+      "learning_rate": 8.388785444031695e-05,
+      "loss": 0.3907,
+      "step": 301
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3961484730243683,
+      "learning_rate": 8.377479312587879e-05,
+      "loss": 0.2742,
+      "step": 302
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.41058430075645447,
+      "learning_rate": 8.366141328841366e-05,
+      "loss": 0.3352,
+      "step": 303
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.35100504755973816,
+      "learning_rate": 8.354771599718312e-05,
+      "loss": 0.2683,
+      "step": 304
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4269458055496216,
+      "learning_rate": 8.343370232444261e-05,
+      "loss": 0.3509,
+      "step": 305
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.37469661235809326,
+      "learning_rate": 8.331937334543132e-05,
+      "loss": 0.2968,
+      "step": 306
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4773250222206116,
+      "learning_rate": 8.320473013836196e-05,
+      "loss": 0.4337,
+      "step": 307
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.44496840238571167,
+      "learning_rate": 8.308977378441072e-05,
+      "loss": 0.4072,
+      "step": 308
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3937629461288452,
+      "learning_rate": 8.297450536770697e-05,
+      "loss": 0.3005,
+      "step": 309
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3999443054199219,
+      "learning_rate": 8.28589259753231e-05,
+      "loss": 0.2939,
+      "step": 310
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.40826454758644104,
+      "learning_rate": 8.274303669726426e-05,
+      "loss": 0.335,
+      "step": 311
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3842678964138031,
+      "learning_rate": 8.262683862645804e-05,
+      "loss": 0.3164,
+      "step": 312
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.445785790681839,
+      "learning_rate": 8.251033285874421e-05,
+      "loss": 0.3551,
+      "step": 313
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.42152997851371765,
+      "learning_rate": 8.239352049286435e-05,
+      "loss": 0.3217,
+      "step": 314
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.40929335355758667,
+      "learning_rate": 8.22764026304515e-05,
+      "loss": 0.3344,
+      "step": 315
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.43447253108024597,
+      "learning_rate": 8.21589803760198e-05,
+      "loss": 0.3519,
+      "step": 316
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3827757239341736,
+      "learning_rate": 8.204125483695403e-05,
+      "loss": 0.3186,
+      "step": 317
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.39438971877098083,
+      "learning_rate": 8.192322712349917e-05,
+      "loss": 0.2675,
+      "step": 318
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4137861728668213,
+      "learning_rate": 8.180489834875e-05,
+      "loss": 0.3188,
+      "step": 319
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4080106317996979,
+      "learning_rate": 8.168626962864045e-05,
+      "loss": 0.2692,
+      "step": 320
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4647471606731415,
+      "learning_rate": 8.156734208193327e-05,
+      "loss": 0.3092,
+      "step": 321
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.40783581137657166,
+      "learning_rate": 8.144811683020932e-05,
+      "loss": 0.2911,
+      "step": 322
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.43954282999038696,
+      "learning_rate": 8.132859499785707e-05,
+      "loss": 0.3518,
+      "step": 323
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.40917956829071045,
+      "learning_rate": 8.120877771206201e-05,
+      "loss": 0.3135,
+      "step": 324
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.49050939083099365,
+      "learning_rate": 8.108866610279593e-05,
+      "loss": 0.3972,
+      "step": 325
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.44175946712493896,
+      "learning_rate": 8.096826130280639e-05,
+      "loss": 0.3106,
+      "step": 326
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4395340085029602,
+      "learning_rate": 8.08475644476059e-05,
+      "loss": 0.3583,
+      "step": 327
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.41783055663108826,
+      "learning_rate": 8.072657667546136e-05,
+      "loss": 0.3697,
+      "step": 328
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4115869402885437,
+      "learning_rate": 8.060529912738315e-05,
+      "loss": 0.3664,
+      "step": 329
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4332055151462555,
+      "learning_rate": 8.048373294711455e-05,
+      "loss": 0.3316,
+      "step": 330
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4499228000640869,
+      "learning_rate": 8.036187928112087e-05,
+      "loss": 0.3254,
+      "step": 331
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.47057217359542847,
+      "learning_rate": 8.023973927857857e-05,
+      "loss": 0.3962,
+      "step": 332
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5100306272506714,
+      "learning_rate": 8.011731409136454e-05,
+      "loss": 0.3936,
+      "step": 333
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.43306952714920044,
+      "learning_rate": 7.99946048740452e-05,
+      "loss": 0.2492,
+      "step": 334
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4529989957809448,
+      "learning_rate": 7.987161278386554e-05,
+      "loss": 0.3656,
+      "step": 335
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.40496325492858887,
+      "learning_rate": 7.974833898073832e-05,
+      "loss": 0.3076,
+      "step": 336
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.45533546805381775,
+      "learning_rate": 7.962478462723307e-05,
+      "loss": 0.3702,
+      "step": 337
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.43411317467689514,
+      "learning_rate": 7.950095088856508e-05,
+      "loss": 0.3568,
+      "step": 338
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.35111913084983826,
+      "learning_rate": 7.937683893258454e-05,
+      "loss": 0.2242,
+      "step": 339
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.47751128673553467,
+      "learning_rate": 7.925244992976538e-05,
+      "loss": 0.4089,
+      "step": 340
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3811795115470886,
+      "learning_rate": 7.912778505319436e-05,
+      "loss": 0.3271,
+      "step": 341
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.33011725544929504,
+      "learning_rate": 7.900284547855991e-05,
+      "loss": 0.197,
+      "step": 342
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.501883327960968,
+      "learning_rate": 7.88776323841411e-05,
+      "loss": 0.3669,
+      "step": 343
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1279911994934082,
+      "learning_rate": 7.875214695079647e-05,
+      "loss": 0.3927,
+      "step": 344
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.33105990290641785,
+      "learning_rate": 7.862639036195298e-05,
+      "loss": 0.2605,
+      "step": 345
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.49403706192970276,
+      "learning_rate": 7.850036380359478e-05,
+      "loss": 0.347,
+      "step": 346
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.42025765776634216,
+      "learning_rate": 7.837406846425204e-05,
+      "loss": 0.3771,
+      "step": 347
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.44321537017822266,
+      "learning_rate": 7.824750553498976e-05,
+      "loss": 0.3933,
+      "step": 348
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4814360737800598,
+      "learning_rate": 7.812067620939652e-05,
+      "loss": 0.3569,
+      "step": 349
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.38922053575515747,
+      "learning_rate": 7.799358168357323e-05,
+      "loss": 0.2926,
+      "step": 350
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5561249852180481,
+      "learning_rate": 7.786622315612183e-05,
+      "loss": 0.3968,
+      "step": 351
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.37317320704460144,
+      "learning_rate": 7.773860182813403e-05,
+      "loss": 0.27,
+      "step": 352
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4488440155982971,
+      "learning_rate": 7.761071890317994e-05,
+      "loss": 0.3418,
+      "step": 353
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3837953209877014,
+      "learning_rate": 7.748257558729677e-05,
+      "loss": 0.3164,
+      "step": 354
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4202031195163727,
+      "learning_rate": 7.735417308897736e-05,
+      "loss": 0.3659,
+      "step": 355
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.37304744124412537,
+      "learning_rate": 7.72255126191589e-05,
+      "loss": 0.3049,
+      "step": 356
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.48277875781059265,
+      "learning_rate": 7.709659539121144e-05,
+      "loss": 0.3293,
+      "step": 357
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3821432292461395,
+      "learning_rate": 7.696742262092644e-05,
+      "loss": 0.3001,
+      "step": 358
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3483389914035797,
+      "learning_rate": 7.683799552650534e-05,
+      "loss": 0.2354,
+      "step": 359
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.33866173028945923,
+      "learning_rate": 7.67083153285481e-05,
+      "loss": 0.2413,
+      "step": 360
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.37959256768226624,
+      "learning_rate": 7.657838325004159e-05,
+      "loss": 0.2585,
+      "step": 361
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4012792706489563,
+      "learning_rate": 7.644820051634812e-05,
+      "loss": 0.3268,
+      "step": 362
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.41745442152023315,
+      "learning_rate": 7.63177683551939e-05,
+      "loss": 0.2849,
+      "step": 363
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.41246259212493896,
+      "learning_rate": 7.618708799665744e-05,
+      "loss": 0.2823,
+      "step": 364
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4160042107105255,
+      "learning_rate": 7.605616067315793e-05,
+      "loss": 0.3565,
+      "step": 365
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.376625657081604,
+      "learning_rate": 7.592498761944362e-05,
+      "loss": 0.2222,
+      "step": 366
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3057038486003876,
+      "learning_rate": 7.579357007258022e-05,
+      "loss": 0.1671,
+      "step": 367
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.401427686214447,
+      "learning_rate": 7.56619092719392e-05,
+      "loss": 0.266,
+      "step": 368
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3868807256221771,
+      "learning_rate": 7.553000645918611e-05,
+      "loss": 0.1989,
+      "step": 369
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3515366315841675,
+      "learning_rate": 7.539786287826885e-05,
+      "loss": 0.2194,
+      "step": 370
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3604734539985657,
+      "learning_rate": 7.526547977540592e-05,
+      "loss": 0.219,
+      "step": 371
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.40912362933158875,
+      "learning_rate": 7.51328583990748e-05,
+      "loss": 0.2583,
+      "step": 372
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4323633313179016,
+      "learning_rate": 7.500000000000001e-05,
+      "loss": 0.2902,
+      "step": 373
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.39483320713043213,
+      "learning_rate": 7.486690583114136e-05,
+      "loss": 0.2161,
+      "step": 374
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5061144828796387,
+      "learning_rate": 7.473357714768222e-05,
+      "loss": 0.3619,
+      "step": 375
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5349733233451843,
+      "learning_rate": 7.460001520701755e-05,
+      "loss": 0.4505,
+      "step": 376
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4656641185283661,
+      "learning_rate": 7.446622126874218e-05,
+      "loss": 0.3452,
+      "step": 377
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4953327178955078,
+      "learning_rate": 7.433219659463881e-05,
+      "loss": 0.3748,
+      "step": 378
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.44388875365257263,
+      "learning_rate": 7.41979424486662e-05,
+      "loss": 0.3572,
+      "step": 379
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.43492433428764343,
+      "learning_rate": 7.406346009694713e-05,
+      "loss": 0.278,
+      "step": 380
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3651255667209625,
+      "learning_rate": 7.392875080775665e-05,
+      "loss": 0.2778,
+      "step": 381
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.44651955366134644,
+      "learning_rate": 7.379381585150996e-05,
+      "loss": 0.3139,
+      "step": 382
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3903564214706421,
+      "learning_rate": 7.365865650075046e-05,
+      "loss": 0.2858,
+      "step": 383
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3159720003604889,
+      "learning_rate": 7.35232740301378e-05,
+      "loss": 0.2257,
+      "step": 384
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3815470337867737,
+      "learning_rate": 7.338766971643579e-05,
+      "loss": 0.3143,
+      "step": 385
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.42418259382247925,
+      "learning_rate": 7.325184483850042e-05,
+      "loss": 0.3135,
+      "step": 386
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.35933899879455566,
+      "learning_rate": 7.311580067726783e-05,
+      "loss": 0.2251,
+      "step": 387
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3698961138725281,
+      "learning_rate": 7.297953851574206e-05,
+      "loss": 0.2823,
+      "step": 388
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.48266690969467163,
+      "learning_rate": 7.284305963898314e-05,
+      "loss": 0.4571,
+      "step": 389
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.43520501255989075,
+      "learning_rate": 7.27063653340949e-05,
+      "loss": 0.3461,
+      "step": 390
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4097878932952881,
+      "learning_rate": 7.25694568902128e-05,
+      "loss": 0.3374,
+      "step": 391
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.27015575766563416,
+      "learning_rate": 7.243233559849179e-05,
+      "loss": 0.2034,
+      "step": 392
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.33289438486099243,
+      "learning_rate": 7.229500275209418e-05,
+      "loss": 0.2546,
+      "step": 393
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.38612377643585205,
+      "learning_rate": 7.215745964617737e-05,
+      "loss": 0.282,
+      "step": 394
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3781869113445282,
+      "learning_rate": 7.201970757788172e-05,
+      "loss": 0.3376,
+      "step": 395
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3694140315055847,
+      "learning_rate": 7.188174784631824e-05,
+      "loss": 0.232,
+      "step": 396
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3989308178424835,
+      "learning_rate": 7.174358175255636e-05,
+      "loss": 0.2974,
+      "step": 397
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.398527592420578,
+      "learning_rate": 7.160521059961169e-05,
+      "loss": 0.2793,
+      "step": 398
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3028777539730072,
+      "learning_rate": 7.146663569243371e-05,
+      "loss": 0.2258,
+      "step": 399
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4481874108314514,
+      "learning_rate": 7.132785833789344e-05,
+      "loss": 0.3966,
+      "step": 400
+    },
+    {
+      "epoch": 0.38,
+      "eval_loss": 0.30074459314346313,
+      "eval_runtime": 361.5641,
+      "eval_samples_per_second": 0.943,
+      "eval_steps_per_second": 0.238,
+      "step": 400
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.36450183391571045,
+      "learning_rate": 7.118887984477116e-05,
+      "loss": 0.2849,
+      "step": 401
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.35163772106170654,
+      "learning_rate": 7.104970152374405e-05,
+      "loss": 0.3068,
+      "step": 402
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.37590691447257996,
+      "learning_rate": 7.091032468737381e-05,
+      "loss": 0.2877,
+      "step": 403
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.40487679839134216,
+      "learning_rate": 7.077075065009433e-05,
+      "loss": 0.3014,
+      "step": 404
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3623451888561249,
+      "learning_rate": 7.063098072819919e-05,
+      "loss": 0.3063,
+      "step": 405
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.33322402834892273,
+      "learning_rate": 7.049101623982937e-05,
+      "loss": 0.2373,
+      "step": 406
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.39618080854415894,
+      "learning_rate": 7.03508585049608e-05,
+      "loss": 0.2652,
+      "step": 407
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4246986210346222,
+      "learning_rate": 7.021050884539177e-05,
+      "loss": 0.3303,
+      "step": 408
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.38274893164634705,
+      "learning_rate": 7.006996858473069e-05,
+      "loss": 0.2779,
+      "step": 409
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.36303016543388367,
+      "learning_rate": 6.99292390483834e-05,
+      "loss": 0.2352,
+      "step": 410
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.33929991722106934,
+      "learning_rate": 6.97883215635408e-05,
+      "loss": 0.2659,
+      "step": 411
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4917987585067749,
+      "learning_rate": 6.964721745916631e-05,
+      "loss": 0.3392,
+      "step": 412
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3732040822505951,
+      "learning_rate": 6.950592806598327e-05,
+      "loss": 0.2778,
+      "step": 413
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3712957799434662,
+      "learning_rate": 6.936445471646249e-05,
+      "loss": 0.2707,
+      "step": 414
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.38530233502388,
+      "learning_rate": 6.922279874480959e-05,
+      "loss": 0.3404,
+      "step": 415
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.39326515793800354,
+      "learning_rate": 6.908096148695251e-05,
+      "loss": 0.3183,
+      "step": 416
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3737872242927551,
+      "learning_rate": 6.89389442805288e-05,
+      "loss": 0.2774,
+      "step": 417
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4002056419849396,
+      "learning_rate": 6.879674846487314e-05,
+      "loss": 0.3047,
+      "step": 418
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.42205730080604553,
+      "learning_rate": 6.865437538100457e-05,
+      "loss": 0.3397,
+      "step": 419
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3429977595806122,
+      "learning_rate": 6.851182637161396e-05,
+      "loss": 0.2317,
+      "step": 420
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3894031047821045,
+      "learning_rate": 6.836910278105124e-05,
+      "loss": 0.3,
+      "step": 421
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.43323320150375366,
+      "learning_rate": 6.822620595531286e-05,
+      "loss": 0.3458,
+      "step": 422
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.39508873224258423,
+      "learning_rate": 6.808313724202894e-05,
+      "loss": 0.2871,
+      "step": 423
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3460524082183838,
+      "learning_rate": 6.793989799045066e-05,
+      "loss": 0.2301,
+      "step": 424
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4365580677986145,
+      "learning_rate": 6.779648955143754e-05,
+      "loss": 0.3176,
+      "step": 425
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4064093828201294,
+      "learning_rate": 6.765291327744463e-05,
+      "loss": 0.3319,
+      "step": 426
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.36963775753974915,
+      "learning_rate": 6.750917052250982e-05,
+      "loss": 0.2946,
+      "step": 427
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.39917299151420593,
+      "learning_rate": 6.736526264224101e-05,
+      "loss": 0.3094,
+      "step": 428
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.40390291810035706,
+      "learning_rate": 6.722119099380345e-05,
+      "loss": 0.3376,
+      "step": 429
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3373975157737732,
+      "learning_rate": 6.707695693590676e-05,
+      "loss": 0.2105,
+      "step": 430
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3403448760509491,
+      "learning_rate": 6.693256182879225e-05,
+      "loss": 0.2813,
+      "step": 431
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4192211329936981,
+      "learning_rate": 6.678800703422003e-05,
+      "loss": 0.3648,
+      "step": 432
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4666745364665985,
+      "learning_rate": 6.664329391545626e-05,
+      "loss": 0.3932,
+      "step": 433
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4507717788219452,
+      "learning_rate": 6.64984238372601e-05,
+      "loss": 0.432,
+      "step": 434
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3652133047580719,
+      "learning_rate": 6.635339816587109e-05,
+      "loss": 0.2959,
+      "step": 435
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4909898042678833,
+      "learning_rate": 6.620821826899606e-05,
+      "loss": 0.3865,
+      "step": 436
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.42049509286880493,
+      "learning_rate": 6.606288551579629e-05,
+      "loss": 0.343,
+      "step": 437
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.46595799922943115,
+      "learning_rate": 6.59174012768747e-05,
+      "loss": 0.3877,
+      "step": 438
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3851853609085083,
+      "learning_rate": 6.577176692426279e-05,
+      "loss": 0.3419,
+      "step": 439
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.40060752630233765,
+      "learning_rate": 6.562598383140772e-05,
+      "loss": 0.3222,
+      "step": 440
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3795297145843506,
+      "learning_rate": 6.548005337315944e-05,
+      "loss": 0.2921,
+      "step": 441
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3178587853908539,
+      "learning_rate": 6.533397692575766e-05,
+      "loss": 0.1797,
+      "step": 442
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5004703402519226,
+      "learning_rate": 6.518775586681887e-05,
+      "loss": 0.434,
+      "step": 443
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4011465013027191,
+      "learning_rate": 6.504139157532337e-05,
+      "loss": 0.2831,
+      "step": 444
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3425051271915436,
+      "learning_rate": 6.489488543160224e-05,
+      "loss": 0.2245,
+      "step": 445
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.36990055441856384,
+      "learning_rate": 6.47482388173244e-05,
+      "loss": 0.2414,
+      "step": 446
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4179847836494446,
+      "learning_rate": 6.460145311548341e-05,
+      "loss": 0.3736,
+      "step": 447
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.45481300354003906,
+      "learning_rate": 6.445452971038464e-05,
+      "loss": 0.3256,
+      "step": 448
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4463241398334503,
+      "learning_rate": 6.430746998763204e-05,
+      "loss": 0.3756,
+      "step": 449
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.41653648018836975,
+      "learning_rate": 6.416027533411519e-05,
+      "loss": 0.3801,
+      "step": 450
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6695518493652344,
+      "learning_rate": 6.401294713799618e-05,
+      "loss": 0.4388,
+      "step": 451
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3954789936542511,
+      "learning_rate": 6.386548678869644e-05,
+      "loss": 0.3045,
+      "step": 452
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.336781769990921,
+      "learning_rate": 6.37178956768838e-05,
+      "loss": 0.2719,
+      "step": 453
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.39205390214920044,
+      "learning_rate": 6.35701751944592e-05,
+      "loss": 0.3162,
+      "step": 454
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.39949509501457214,
+      "learning_rate": 6.342232673454371e-05,
+      "loss": 0.341,
+      "step": 455
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.31091591715812683,
+      "learning_rate": 6.32743516914653e-05,
+      "loss": 0.2077,
+      "step": 456
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.43705612421035767,
+      "learning_rate": 6.312625146074575e-05,
+      "loss": 0.3551,
+      "step": 457
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3647152781486511,
+      "learning_rate": 6.297802743908741e-05,
+      "loss": 0.2654,
+      "step": 458
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.37313953042030334,
+      "learning_rate": 6.282968102436009e-05,
+      "loss": 0.3007,
+      "step": 459
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3366411328315735,
+      "learning_rate": 6.268121361558792e-05,
+      "loss": 0.2329,
+      "step": 460
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3827972114086151,
+      "learning_rate": 6.253262661293604e-05,
+      "loss": 0.3469,
+      "step": 461
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.37852704524993896,
+      "learning_rate": 6.238392141769743e-05,
+      "loss": 0.2478,
+      "step": 462
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4133304953575134,
+      "learning_rate": 6.22350994322798e-05,
+      "loss": 0.3937,
+      "step": 463
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.40183690190315247,
+      "learning_rate": 6.208616206019224e-05,
+      "loss": 0.266,
+      "step": 464
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3527545928955078,
+      "learning_rate": 6.193711070603202e-05,
+      "loss": 0.2564,
+      "step": 465
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3591715693473816,
+      "learning_rate": 6.178794677547137e-05,
+      "loss": 0.2786,
+      "step": 466
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.41740310192108154,
+      "learning_rate": 6.163867167524419e-05,
+      "loss": 0.3264,
+      "step": 467
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.43474718928337097,
+      "learning_rate": 6.14892868131328e-05,
+      "loss": 0.2989,
+      "step": 468
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.35294049978256226,
+      "learning_rate": 6.133979359795468e-05,
+      "loss": 0.2708,
+      "step": 469
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3931112587451935,
+      "learning_rate": 6.119019343954914e-05,
+      "loss": 0.2916,
+      "step": 470
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.37170812487602234,
+      "learning_rate": 6.104048774876406e-05,
+      "loss": 0.308,
+      "step": 471
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.40284213423728943,
+      "learning_rate": 6.0890677937442574e-05,
+      "loss": 0.3623,
+      "step": 472
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5154820084571838,
+      "learning_rate": 6.074076541840977e-05,
+      "loss": 0.4817,
+      "step": 473
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.36041492223739624,
+      "learning_rate": 6.059075160545933e-05,
+      "loss": 0.236,
+      "step": 474
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3620935380458832,
+      "learning_rate": 6.0440637913340224e-05,
+      "loss": 0.237,
+      "step": 475
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3835925757884979,
+      "learning_rate": 6.029042575774334e-05,
+      "loss": 0.3022,
+      "step": 476
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4346061646938324,
+      "learning_rate": 6.0140116555288204e-05,
+      "loss": 0.3342,
+      "step": 477
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.35895106196403503,
+      "learning_rate": 5.998971172350953e-05,
+      "loss": 0.3169,
+      "step": 478
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.33846020698547363,
+      "learning_rate": 5.9839212680843925e-05,
+      "loss": 0.261,
+      "step": 479
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.44269314408302307,
+      "learning_rate": 5.968862084661643e-05,
+      "loss": 0.3722,
+      "step": 480
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3683073818683624,
+      "learning_rate": 5.953793764102722e-05,
+      "loss": 0.3025,
+      "step": 481
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.374986857175827,
+      "learning_rate": 5.938716448513818e-05,
+      "loss": 0.283,
+      "step": 482
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5157366991043091,
+      "learning_rate": 5.923630280085948e-05,
+      "loss": 0.4559,
+      "step": 483
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4158179759979248,
+      "learning_rate": 5.908535401093618e-05,
+      "loss": 0.3686,
+      "step": 484
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3556183874607086,
+      "learning_rate": 5.8934319538934824e-05,
+      "loss": 0.2921,
+      "step": 485
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.35228267312049866,
+      "learning_rate": 5.878320080923001e-05,
+      "loss": 0.2898,
+      "step": 486
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.45697152614593506,
+      "learning_rate": 5.863199924699095e-05,
+      "loss": 0.2984,
+      "step": 487
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4791191518306732,
+      "learning_rate": 5.848071627816803e-05,
+      "loss": 0.4247,
+      "step": 488
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.42639708518981934,
+      "learning_rate": 5.8329353329479366e-05,
+      "loss": 0.3479,
+      "step": 489
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.33568671345710754,
+      "learning_rate": 5.8177911828397336e-05,
+      "loss": 0.2478,
+      "step": 490
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4149231016635895,
+      "learning_rate": 5.802639320313514e-05,
+      "loss": 0.3739,
+      "step": 491
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.41260474920272827,
+      "learning_rate": 5.787479888263333e-05,
+      "loss": 0.3268,
+      "step": 492
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4186526834964752,
+      "learning_rate": 5.77231302965463e-05,
+      "loss": 0.2971,
+      "step": 493
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4042125940322876,
+      "learning_rate": 5.757138887522884e-05,
+      "loss": 0.2901,
+      "step": 494
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.348179429769516,
+      "learning_rate": 5.7419576049722646e-05,
+      "loss": 0.2455,
+      "step": 495
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3648551106452942,
+      "learning_rate": 5.7267693251742794e-05,
+      "loss": 0.2873,
+      "step": 496
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4363463521003723,
+      "learning_rate": 5.7115741913664264e-05,
+      "loss": 0.2984,
+      "step": 497
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4140514135360718,
+      "learning_rate": 5.696372346850841e-05,
+      "loss": 0.3863,
+      "step": 498
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.405406653881073,
+      "learning_rate": 5.68116393499295e-05,
+      "loss": 0.3266,
+      "step": 499
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.35093408823013306,
+      "learning_rate": 5.6659490992201094e-05,
+      "loss": 0.2607,
+      "step": 500
+    },
+    {
+      "epoch": 0.47,
+      "eval_loss": 0.2947947680950165,
+      "eval_runtime": 365.5804,
+      "eval_samples_per_second": 0.933,
+      "eval_steps_per_second": 0.235,
+      "step": 500
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4803601801395416,
+      "learning_rate": 5.650727983020262e-05,
+      "loss": 0.4519,
+      "step": 501
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3370169401168823,
+      "learning_rate": 5.635500729940577e-05,
+      "loss": 0.2145,
+      "step": 502
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.43801531195640564,
+      "learning_rate": 5.620267483586105e-05,
+      "loss": 0.3777,
+      "step": 503
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.42908695340156555,
+      "learning_rate": 5.6050283876184116e-05,
+      "loss": 0.3662,
+      "step": 504
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.47098079323768616,
+      "learning_rate": 5.5897835857542317e-05,
+      "loss": 0.3639,
+      "step": 505
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.41761061549186707,
+      "learning_rate": 5.574533221764109e-05,
+      "loss": 0.2686,
+      "step": 506
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3722454011440277,
+      "learning_rate": 5.5592774394710465e-05,
+      "loss": 0.2838,
+      "step": 507
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.42140689492225647,
+      "learning_rate": 5.5440163827491454e-05,
+      "loss": 0.2906,
+      "step": 508
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3904332220554352,
+      "learning_rate": 5.5287501955222444e-05,
+      "loss": 0.2729,
+      "step": 509
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3544531464576721,
+      "learning_rate": 5.513479021762572e-05,
+      "loss": 0.2917,
+      "step": 510
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3679385781288147,
+      "learning_rate": 5.4982030054893785e-05,
+      "loss": 0.23,
+      "step": 511
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.39986100792884827,
+      "learning_rate": 5.482922290767589e-05,
+      "loss": 0.2592,
+      "step": 512
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.37825778126716614,
+      "learning_rate": 5.467637021706438e-05,
+      "loss": 0.2888,
+      "step": 513
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5194998383522034,
+      "learning_rate": 5.452347342458104e-05,
+      "loss": 0.3877,
+      "step": 514
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.39368703961372375,
+      "learning_rate": 5.437053397216364e-05,
+      "loss": 0.3068,
+      "step": 515
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4012013375759125,
+      "learning_rate": 5.4217553302152237e-05,
+      "loss": 0.3196,
+      "step": 516
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4755167067050934,
+      "learning_rate": 5.406453285727564e-05,
+      "loss": 0.3527,
+      "step": 517
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4077131152153015,
+      "learning_rate": 5.39114740806377e-05,
+      "loss": 0.279,
+      "step": 518
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.34338611364364624,
+      "learning_rate": 5.3758378415703825e-05,
+      "loss": 0.2479,
+      "step": 519
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4707435667514801,
+      "learning_rate": 5.360524730628728e-05,
+      "loss": 0.3983,
+      "step": 520
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.39420026540756226,
+      "learning_rate": 5.345208219653561e-05,
+      "loss": 0.2878,
+      "step": 521
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4009648561477661,
+      "learning_rate": 5.3298884530917014e-05,
+      "loss": 0.3348,
+      "step": 522
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.36652106046676636,
+      "learning_rate": 5.3145655754206714e-05,
+      "loss": 0.2595,
+      "step": 523
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3267623484134674,
+      "learning_rate": 5.2992397311473316e-05,
+      "loss": 0.2109,
+      "step": 524
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4243021607398987,
+      "learning_rate": 5.2839110648065214e-05,
+      "loss": 0.3234,
+      "step": 525
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4481864869594574,
+      "learning_rate": 5.2685797209596976e-05,
+      "loss": 0.3215,
+      "step": 526
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4373661279678345,
+      "learning_rate": 5.2532458441935636e-05,
+      "loss": 0.3156,
+      "step": 527
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.45050036907196045,
+      "learning_rate": 5.2379095791187124e-05,
+      "loss": 0.3466,
+      "step": 528
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.38725122809410095,
+      "learning_rate": 5.222571070368257e-05,
+      "loss": 0.2875,
+      "step": 529
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.39766013622283936,
+      "learning_rate": 5.2072304625964785e-05,
+      "loss": 0.2688,
+      "step": 530
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.39241930842399597,
+      "learning_rate": 5.191887900477443e-05,
+      "loss": 0.3398,
+      "step": 531
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.39578163623809814,
+      "learning_rate": 5.176543528703657e-05,
+      "loss": 0.3156,
+      "step": 532
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.37313225865364075,
+      "learning_rate": 5.161197491984684e-05,
+      "loss": 0.2858,
+      "step": 533
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.40683722496032715,
+      "learning_rate": 5.1458499350458e-05,
+      "loss": 0.2229,
+      "step": 534
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4621437191963196,
+      "learning_rate": 5.130501002626609e-05,
+      "loss": 0.4654,
+      "step": 535
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4117042124271393,
+      "learning_rate": 5.11515083947969e-05,
+      "loss": 0.2869,
+      "step": 536
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4133632183074951,
+      "learning_rate": 5.099799590369231e-05,
+      "loss": 0.3233,
+      "step": 537
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4366859197616577,
+      "learning_rate": 5.084447400069655e-05,
+      "loss": 0.3406,
+      "step": 538
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3743889629840851,
+      "learning_rate": 5.069094413364272e-05,
+      "loss": 0.2795,
+      "step": 539
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4406284987926483,
+      "learning_rate": 5.053740775043892e-05,
+      "loss": 0.3398,
+      "step": 540
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4889068305492401,
+      "learning_rate": 5.038386629905475e-05,
+      "loss": 0.3474,
+      "step": 541
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.45164594054222107,
+      "learning_rate": 5.023032122750759e-05,
+      "loss": 0.3372,
+      "step": 542
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3668400049209595,
+      "learning_rate": 5.007677398384901e-05,
+      "loss": 0.2586,
+      "step": 543
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.41004592180252075,
+      "learning_rate": 4.9923226016151e-05,
+      "loss": 0.3249,
+      "step": 544
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.40927720069885254,
+      "learning_rate": 4.976967877249242e-05,
+      "loss": 0.3432,
+      "step": 545
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.30458950996398926,
+      "learning_rate": 4.961613370094527e-05,
+      "loss": 0.2404,
+      "step": 546
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4174305200576782,
+      "learning_rate": 4.9462592249561095e-05,
+      "loss": 0.3525,
+      "step": 547
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4010785222053528,
+      "learning_rate": 4.9309055866357295e-05,
+      "loss": 0.2424,
+      "step": 548
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4169844388961792,
+      "learning_rate": 4.915552599930345e-05,
+      "loss": 0.3459,
+      "step": 549
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.39555662870407104,
+      "learning_rate": 4.900200409630771e-05,
+      "loss": 0.2864,
+      "step": 550
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.36934277415275574,
+      "learning_rate": 4.884849160520311e-05,
+      "loss": 0.236,
+      "step": 551
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.40973079204559326,
+      "learning_rate": 4.869498997373393e-05,
+      "loss": 0.2936,
+      "step": 552
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4465014636516571,
+      "learning_rate": 4.8541500649542014e-05,
+      "loss": 0.3892,
+      "step": 553
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4180162847042084,
+      "learning_rate": 4.838802508015316e-05,
+      "loss": 0.2944,
+      "step": 554
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4353102743625641,
+      "learning_rate": 4.823456471296344e-05,
+      "loss": 0.2907,
+      "step": 555
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3468726575374603,
+      "learning_rate": 4.8081120995225574e-05,
+      "loss": 0.2736,
+      "step": 556
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4479537308216095,
+      "learning_rate": 4.792769537403523e-05,
+      "loss": 0.3997,
+      "step": 557
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4738159775733948,
+      "learning_rate": 4.777428929631743e-05,
+      "loss": 0.3861,
+      "step": 558
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.43831104040145874,
+      "learning_rate": 4.762090420881289e-05,
+      "loss": 0.3637,
+      "step": 559
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4980882406234741,
+      "learning_rate": 4.746754155806437e-05,
+      "loss": 0.4386,
+      "step": 560
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.34668153524398804,
+      "learning_rate": 4.731420279040303e-05,
+      "loss": 0.2964,
+      "step": 561
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.35641565918922424,
+      "learning_rate": 4.716088935193479e-05,
+      "loss": 0.1803,
+      "step": 562
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4649956524372101,
+      "learning_rate": 4.7007602688526695e-05,
+      "loss": 0.3418,
+      "step": 563
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.32428959012031555,
+      "learning_rate": 4.6854344245793305e-05,
+      "loss": 0.2125,
+      "step": 564
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.37127062678337097,
+      "learning_rate": 4.6701115469083e-05,
+      "loss": 0.261,
+      "step": 565
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4087611734867096,
+      "learning_rate": 4.65479178034644e-05,
+      "loss": 0.1994,
+      "step": 566
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.48373618721961975,
+      "learning_rate": 4.6394752693712734e-05,
+      "loss": 0.2926,
+      "step": 567
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.42548930644989014,
+      "learning_rate": 4.624162158429619e-05,
+      "loss": 0.2685,
+      "step": 568
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4652377963066101,
+      "learning_rate": 4.608852591936231e-05,
+      "loss": 0.368,
+      "step": 569
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.42774638533592224,
+      "learning_rate": 4.593546714272438e-05,
+      "loss": 0.3386,
+      "step": 570
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.46106675267219543,
+      "learning_rate": 4.578244669784777e-05,
+      "loss": 0.3552,
+      "step": 571
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3079265058040619,
+      "learning_rate": 4.562946602783636e-05,
+      "loss": 0.1998,
+      "step": 572
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.486830472946167,
+      "learning_rate": 4.547652657541897e-05,
+      "loss": 0.441,
+      "step": 573
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.41522058844566345,
+      "learning_rate": 4.532362978293564e-05,
+      "loss": 0.372,
+      "step": 574
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3492695987224579,
+      "learning_rate": 4.517077709232411e-05,
+      "loss": 0.2353,
+      "step": 575
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.47843942046165466,
+      "learning_rate": 4.501796994510622e-05,
+      "loss": 0.3818,
+      "step": 576
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2723293900489807,
+      "learning_rate": 4.4865209782374304e-05,
+      "loss": 0.1899,
+      "step": 577
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3466640114784241,
+      "learning_rate": 4.471249804477758e-05,
+      "loss": 0.2106,
+      "step": 578
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.37762540578842163,
+      "learning_rate": 4.4559836172508565e-05,
+      "loss": 0.3165,
+      "step": 579
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3390047550201416,
+      "learning_rate": 4.4407225605289547e-05,
+      "loss": 0.2269,
+      "step": 580
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3165014684200287,
+      "learning_rate": 4.4254667782358924e-05,
+      "loss": 0.2524,
+      "step": 581
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4255695939064026,
+      "learning_rate": 4.410216414245771e-05,
+      "loss": 0.3611,
+      "step": 582
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3833935558795929,
+      "learning_rate": 4.394971612381591e-05,
+      "loss": 0.2933,
+      "step": 583
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.40104764699935913,
+      "learning_rate": 4.379732516413897e-05,
+      "loss": 0.2757,
+      "step": 584
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4127373695373535,
+      "learning_rate": 4.364499270059424e-05,
+      "loss": 0.3023,
+      "step": 585
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3981027603149414,
+      "learning_rate": 4.34927201697974e-05,
+      "loss": 0.314,
+      "step": 586
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4027072489261627,
+      "learning_rate": 4.334050900779893e-05,
+      "loss": 0.2792,
+      "step": 587
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4657103717327118,
+      "learning_rate": 4.3188360650070525e-05,
+      "loss": 0.3439,
+      "step": 588
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3677713871002197,
+      "learning_rate": 4.303627653149159e-05,
+      "loss": 0.2206,
+      "step": 589
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.37127944827079773,
+      "learning_rate": 4.288425808633575e-05,
+      "loss": 0.2378,
+      "step": 590
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.39544370770454407,
+      "learning_rate": 4.2732306748257224e-05,
+      "loss": 0.2813,
+      "step": 591
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3750370740890503,
+      "learning_rate": 4.258042395027738e-05,
+      "loss": 0.2325,
+      "step": 592
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4383431077003479,
+      "learning_rate": 4.2428611124771184e-05,
+      "loss": 0.2814,
+      "step": 593
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3971858024597168,
+      "learning_rate": 4.2276869703453724e-05,
+      "loss": 0.2972,
+      "step": 594
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4210115671157837,
+      "learning_rate": 4.21252011173667e-05,
+      "loss": 0.269,
+      "step": 595
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.41311362385749817,
+      "learning_rate": 4.1973606796864884e-05,
+      "loss": 0.2741,
+      "step": 596
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4613640606403351,
+      "learning_rate": 4.182208817160269e-05,
+      "loss": 0.3679,
+      "step": 597
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.36621612310409546,
+      "learning_rate": 4.167064667052065e-05,
+      "loss": 0.2462,
+      "step": 598
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.41440847516059875,
+      "learning_rate": 4.151928372183198e-05,
+      "loss": 0.3078,
+      "step": 599
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.42148759961128235,
+      "learning_rate": 4.136800075300906e-05,
+      "loss": 0.31,
+      "step": 600
+    },
+    {
+      "epoch": 0.57,
+      "eval_loss": 0.28922584652900696,
+      "eval_runtime": 368.1185,
+      "eval_samples_per_second": 0.926,
+      "eval_steps_per_second": 0.234,
+      "step": 600
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4705705940723419,
+      "learning_rate": 4.1216799190770006e-05,
+      "loss": 0.4373,
+      "step": 601
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3295850455760956,
+      "learning_rate": 4.1065680461065194e-05,
+      "loss": 0.2353,
+      "step": 602
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.382693886756897,
+      "learning_rate": 4.0914645989063845e-05,
+      "loss": 0.2847,
+      "step": 603
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4192850887775421,
+      "learning_rate": 4.076369719914055e-05,
+      "loss": 0.2841,
+      "step": 604
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5032066106796265,
+      "learning_rate": 4.061283551486185e-05,
+      "loss": 0.46,
+      "step": 605
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.45867612957954407,
+      "learning_rate": 4.0462062358972796e-05,
+      "loss": 0.3999,
+      "step": 606
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4792550802230835,
+      "learning_rate": 4.03113791533836e-05,
+      "loss": 0.4241,
+      "step": 607
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.416970819234848,
+      "learning_rate": 4.016078731915608e-05,
+      "loss": 0.2838,
+      "step": 608
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.43371081352233887,
+      "learning_rate": 4.001028827649046e-05,
+      "loss": 0.3108,
+      "step": 609
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3693605959415436,
+      "learning_rate": 3.9859883444711794e-05,
+      "loss": 0.2655,
+      "step": 610
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3386535942554474,
+      "learning_rate": 3.970957424225666e-05,
+      "loss": 0.2461,
+      "step": 611
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.350262314081192,
+      "learning_rate": 3.955936208665979e-05,
+      "loss": 0.2369,
+      "step": 612
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3861701488494873,
+      "learning_rate": 3.940924839454066e-05,
+      "loss": 0.2805,
+      "step": 613
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4268942177295685,
+      "learning_rate": 3.925923458159023e-05,
+      "loss": 0.3769,
+      "step": 614
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3818383514881134,
+      "learning_rate": 3.9109322062557424e-05,
+      "loss": 0.2825,
+      "step": 615
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2846945822238922,
+      "learning_rate": 3.895951225123595e-05,
+      "loss": 0.1928,
+      "step": 616
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.38463255763053894,
+      "learning_rate": 3.880980656045087e-05,
+      "loss": 0.3015,
+      "step": 617
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3431190252304077,
+      "learning_rate": 3.866020640204533e-05,
+      "loss": 0.2648,
+      "step": 618
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3795267641544342,
+      "learning_rate": 3.851071318686721e-05,
+      "loss": 0.2535,
+      "step": 619
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.35856059193611145,
+      "learning_rate": 3.8361328324755825e-05,
+      "loss": 0.236,
+      "step": 620
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.42728835344314575,
+      "learning_rate": 3.821205322452863e-05,
+      "loss": 0.3406,
+      "step": 621
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4889734983444214,
+      "learning_rate": 3.8062889293967976e-05,
+      "loss": 0.4123,
+      "step": 622
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.401030957698822,
+      "learning_rate": 3.791383793980776e-05,
+      "loss": 0.3424,
+      "step": 623
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3227827548980713,
+      "learning_rate": 3.7764900567720196e-05,
+      "loss": 0.2056,
+      "step": 624
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.41959959268569946,
+      "learning_rate": 3.761607858230257e-05,
+      "loss": 0.3943,
+      "step": 625
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3688479959964752,
+      "learning_rate": 3.746737338706397e-05,
+      "loss": 0.2502,
+      "step": 626
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.40960216522216797,
+      "learning_rate": 3.7318786384412074e-05,
+      "loss": 0.3792,
+      "step": 627
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4024848937988281,
+      "learning_rate": 3.71703189756399e-05,
+      "loss": 0.3064,
+      "step": 628
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.37706688046455383,
+      "learning_rate": 3.70219725609126e-05,
+      "loss": 0.2769,
+      "step": 629
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.39217904210090637,
+      "learning_rate": 3.6873748539254246e-05,
+      "loss": 0.267,
+      "step": 630
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.43298712372779846,
+      "learning_rate": 3.6725648308534696e-05,
+      "loss": 0.3874,
+      "step": 631
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3942379653453827,
+      "learning_rate": 3.65776732654563e-05,
+      "loss": 0.3389,
+      "step": 632
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.41072559356689453,
+      "learning_rate": 3.6429824805540815e-05,
+      "loss": 0.3579,
+      "step": 633
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.32904714345932007,
+      "learning_rate": 3.628210432311621e-05,
+      "loss": 0.2145,
+      "step": 634
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.49629005789756775,
+      "learning_rate": 3.6134513211303556e-05,
+      "loss": 0.3758,
+      "step": 635
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4583534002304077,
+      "learning_rate": 3.598705286200382e-05,
+      "loss": 0.2954,
+      "step": 636
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4292958080768585,
+      "learning_rate": 3.58397246658848e-05,
+      "loss": 0.3673,
+      "step": 637
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.46979236602783203,
+      "learning_rate": 3.5692530012367955e-05,
+      "loss": 0.3616,
+      "step": 638
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4094065725803375,
+      "learning_rate": 3.554547028961537e-05,
+      "loss": 0.3011,
+      "step": 639
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3580895960330963,
+      "learning_rate": 3.5398546884516604e-05,
+      "loss": 0.2205,
+      "step": 640
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4058517813682556,
+      "learning_rate": 3.5251761182675625e-05,
+      "loss": 0.3528,
+      "step": 641
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4256916344165802,
+      "learning_rate": 3.510511456839777e-05,
+      "loss": 0.3135,
+      "step": 642
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4722270965576172,
+      "learning_rate": 3.495860842467664e-05,
+      "loss": 0.349,
+      "step": 643
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.36562517285346985,
+      "learning_rate": 3.481224413318114e-05,
+      "loss": 0.3113,
+      "step": 644
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4289080798625946,
+      "learning_rate": 3.4666023074242356e-05,
+      "loss": 0.3301,
+      "step": 645
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.43501877784729004,
+      "learning_rate": 3.4519946626840573e-05,
+      "loss": 0.3298,
+      "step": 646
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.40116801857948303,
+      "learning_rate": 3.4374016168592296e-05,
+      "loss": 0.2993,
+      "step": 647
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4554198086261749,
+      "learning_rate": 3.422823307573722e-05,
+      "loss": 0.3479,
+      "step": 648
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.45405974984169006,
+      "learning_rate": 3.40825987231253e-05,
+      "loss": 0.3867,
+      "step": 649
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.4486112892627716,
+      "learning_rate": 3.393711448420372e-05,
+      "loss": 0.3022,
+      "step": 650
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.41444268822669983,
+      "learning_rate": 3.379178173100396e-05,
+      "loss": 0.2889,
+      "step": 651
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3450157344341278,
+      "learning_rate": 3.364660183412892e-05,
+      "loss": 0.1888,
+      "step": 652
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.4667302966117859,
+      "learning_rate": 3.3501576162739904e-05,
+      "loss": 0.2879,
+      "step": 653
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.4100315570831299,
+      "learning_rate": 3.335670608454376e-05,
+      "loss": 0.3418,
+      "step": 654
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3660954535007477,
+      "learning_rate": 3.321199296577998e-05,
+      "loss": 0.2039,
+      "step": 655
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.4195208251476288,
+      "learning_rate": 3.3067438171207766e-05,
+      "loss": 0.313,
+      "step": 656
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.4228833019733429,
+      "learning_rate": 3.292304306409325e-05,
+      "loss": 0.3043,
+      "step": 657
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.39977604150772095,
+      "learning_rate": 3.2778809006196564e-05,
+      "loss": 0.2965,
+      "step": 658
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3272872567176819,
+      "learning_rate": 3.263473735775899e-05,
+      "loss": 0.1833,
+      "step": 659
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.37454575300216675,
+      "learning_rate": 3.249082947749019e-05,
+      "loss": 0.2397,
+      "step": 660
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.41179579496383667,
+      "learning_rate": 3.234708672255538e-05,
+      "loss": 0.2888,
+      "step": 661
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3637540936470032,
+      "learning_rate": 3.220351044856247e-05,
+      "loss": 0.2515,
+      "step": 662
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3879622220993042,
+      "learning_rate": 3.206010200954935e-05,
+      "loss": 0.2669,
+      "step": 663
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4636165201663971,
+      "learning_rate": 3.191686275797107e-05,
+      "loss": 0.3719,
+      "step": 664
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4034343957901001,
+      "learning_rate": 3.177379404468715e-05,
+      "loss": 0.285,
+      "step": 665
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3273599445819855,
+      "learning_rate": 3.163089721894876e-05,
+      "loss": 0.187,
+      "step": 666
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.41577887535095215,
+      "learning_rate": 3.148817362838606e-05,
+      "loss": 0.3153,
+      "step": 667
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4373845160007477,
+      "learning_rate": 3.134562461899545e-05,
+      "loss": 0.316,
+      "step": 668
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3871656060218811,
+      "learning_rate": 3.120325153512687e-05,
+      "loss": 0.2523,
+      "step": 669
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.40141761302948,
+      "learning_rate": 3.10610557194712e-05,
+      "loss": 0.2814,
+      "step": 670
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.45727792382240295,
+      "learning_rate": 3.091903851304751e-05,
+      "loss": 0.4158,
+      "step": 671
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3446201980113983,
+      "learning_rate": 3.077720125519042e-05,
+      "loss": 0.2537,
+      "step": 672
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.45891639590263367,
+      "learning_rate": 3.063554528353752e-05,
+      "loss": 0.3315,
+      "step": 673
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3911459445953369,
+      "learning_rate": 3.0494071934016737e-05,
+      "loss": 0.2481,
+      "step": 674
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4104984998703003,
+      "learning_rate": 3.03527825408337e-05,
+      "loss": 0.3361,
+      "step": 675
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4029580056667328,
+      "learning_rate": 3.0211678436459212e-05,
+      "loss": 0.2691,
+      "step": 676
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.41102248430252075,
+      "learning_rate": 3.0070760951616618e-05,
+      "loss": 0.3104,
+      "step": 677
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.41624414920806885,
+      "learning_rate": 2.993003141526933e-05,
+      "loss": 0.325,
+      "step": 678
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.34995174407958984,
+      "learning_rate": 2.978949115460824e-05,
+      "loss": 0.2104,
+      "step": 679
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4261603355407715,
+      "learning_rate": 2.9649141495039223e-05,
+      "loss": 0.2705,
+      "step": 680
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.40094584226608276,
+      "learning_rate": 2.950898376017064e-05,
+      "loss": 0.2832,
+      "step": 681
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.35921230912208557,
+      "learning_rate": 2.9369019271800823e-05,
+      "loss": 0.2287,
+      "step": 682
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4330695569515228,
+      "learning_rate": 2.9229249349905684e-05,
+      "loss": 0.3502,
+      "step": 683
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3757020831108093,
+      "learning_rate": 2.908967531262618e-05,
+      "loss": 0.2114,
+      "step": 684
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4223138689994812,
+      "learning_rate": 2.895029847625595e-05,
+      "loss": 0.2992,
+      "step": 685
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5096969604492188,
+      "learning_rate": 2.8811120155228844e-05,
+      "loss": 0.4279,
+      "step": 686
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4671712815761566,
+      "learning_rate": 2.8672141662106578e-05,
+      "loss": 0.3355,
+      "step": 687
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4327547550201416,
+      "learning_rate": 2.853336430756631e-05,
+      "loss": 0.3485,
+      "step": 688
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5177743434906006,
+      "learning_rate": 2.8394789400388328e-05,
+      "loss": 0.3823,
+      "step": 689
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.47393983602523804,
+      "learning_rate": 2.8256418247443662e-05,
+      "loss": 0.3008,
+      "step": 690
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3744626045227051,
+      "learning_rate": 2.811825215368179e-05,
+      "loss": 0.2546,
+      "step": 691
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.45847421884536743,
+      "learning_rate": 2.798029242211828e-05,
+      "loss": 0.3683,
+      "step": 692
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4064277708530426,
+      "learning_rate": 2.7842540353822632e-05,
+      "loss": 0.289,
+      "step": 693
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.38756507635116577,
+      "learning_rate": 2.770499724790584e-05,
+      "loss": 0.271,
+      "step": 694
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4184701144695282,
+      "learning_rate": 2.7567664401508225e-05,
+      "loss": 0.3339,
+      "step": 695
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.45738500356674194,
+      "learning_rate": 2.7430543109787222e-05,
+      "loss": 0.3848,
+      "step": 696
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3646864593029022,
+      "learning_rate": 2.729363466590511e-05,
+      "loss": 0.2209,
+      "step": 697
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3928185999393463,
+      "learning_rate": 2.7156940361016864e-05,
+      "loss": 0.2751,
+      "step": 698
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3983188569545746,
+      "learning_rate": 2.702046148425795e-05,
+      "loss": 0.2703,
+      "step": 699
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.45070815086364746,
+      "learning_rate": 2.688419932273219e-05,
+      "loss": 0.3671,
+      "step": 700
+    },
+    {
+      "epoch": 0.66,
+      "eval_loss": 0.28604161739349365,
+      "eval_runtime": 363.145,
+      "eval_samples_per_second": 0.939,
+      "eval_steps_per_second": 0.237,
+      "step": 700
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4320833384990692,
+      "learning_rate": 2.6748155161499567e-05,
+      "loss": 0.3442,
+      "step": 701
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3390715718269348,
+      "learning_rate": 2.6612330283564223e-05,
+      "loss": 0.2091,
+      "step": 702
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3587324023246765,
+      "learning_rate": 2.6476725969862227e-05,
+      "loss": 0.1811,
+      "step": 703
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3428841531276703,
+      "learning_rate": 2.634134349924956e-05,
+      "loss": 0.2343,
+      "step": 704
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.39645302295684814,
+      "learning_rate": 2.6206184148490066e-05,
+      "loss": 0.3027,
+      "step": 705
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.4043754041194916,
+      "learning_rate": 2.6071249192243363e-05,
+      "loss": 0.3092,
+      "step": 706
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.4528702199459076,
+      "learning_rate": 2.5936539903052892e-05,
+      "loss": 0.3432,
+      "step": 707
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.4520406126976013,
+      "learning_rate": 2.580205755133384e-05,
+      "loss": 0.3998,
+      "step": 708
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.44080615043640137,
+      "learning_rate": 2.5667803405361213e-05,
+      "loss": 0.3888,
+      "step": 709
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.46568986773490906,
+      "learning_rate": 2.5533778731257824e-05,
+      "loss": 0.3841,
+      "step": 710
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3910614252090454,
+      "learning_rate": 2.5399984792982457e-05,
+      "loss": 0.2722,
+      "step": 711
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.41148898005485535,
+      "learning_rate": 2.5266422852317795e-05,
+      "loss": 0.2991,
+      "step": 712
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.44081127643585205,
+      "learning_rate": 2.513309416885865e-05,
+      "loss": 0.3588,
+      "step": 713
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4219505786895752,
+      "learning_rate": 2.500000000000001e-05,
+      "loss": 0.2559,
+      "step": 714
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4109588861465454,
+      "learning_rate": 2.4867141600925214e-05,
+      "loss": 0.2894,
+      "step": 715
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.41026967763900757,
+      "learning_rate": 2.4734520224594093e-05,
+      "loss": 0.3407,
+      "step": 716
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4314819872379303,
+      "learning_rate": 2.4602137121731195e-05,
+      "loss": 0.318,
+      "step": 717
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4028915762901306,
+      "learning_rate": 2.44699935408139e-05,
+      "loss": 0.3007,
+      "step": 718
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.27764901518821716,
+      "learning_rate": 2.433809072806081e-05,
+      "loss": 0.142,
+      "step": 719
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.39568236470222473,
+      "learning_rate": 2.4206429927419794e-05,
+      "loss": 0.2643,
+      "step": 720
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.44984322786331177,
+      "learning_rate": 2.40750123805564e-05,
+      "loss": 0.2635,
+      "step": 721
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4358389377593994,
+      "learning_rate": 2.3943839326842092e-05,
+      "loss": 0.307,
+      "step": 722
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.378313809633255,
+      "learning_rate": 2.3812912003342568e-05,
+      "loss": 0.2527,
+      "step": 723
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.457180380821228,
+      "learning_rate": 2.3682231644806108e-05,
+      "loss": 0.3352,
+      "step": 724
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3841799795627594,
+      "learning_rate": 2.3551799483651894e-05,
+      "loss": 0.2847,
+      "step": 725
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5085780620574951,
+      "learning_rate": 2.342161674995843e-05,
+      "loss": 0.3706,
+      "step": 726
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4154749810695648,
+      "learning_rate": 2.3291684671451903e-05,
+      "loss": 0.3118,
+      "step": 727
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3889504075050354,
+      "learning_rate": 2.3162004473494657e-05,
+      "loss": 0.2459,
+      "step": 728
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.38008660078048706,
+      "learning_rate": 2.3032577379073577e-05,
+      "loss": 0.2991,
+      "step": 729
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.41919273138046265,
+      "learning_rate": 2.2903404608788582e-05,
+      "loss": 0.2411,
+      "step": 730
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4136778712272644,
+      "learning_rate": 2.2774487380841115e-05,
+      "loss": 0.3154,
+      "step": 731
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4289793372154236,
+      "learning_rate": 2.2645826911022655e-05,
+      "loss": 0.307,
+      "step": 732
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4092271029949188,
+      "learning_rate": 2.2517424412703252e-05,
+      "loss": 0.274,
+      "step": 733
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3704524338245392,
+      "learning_rate": 2.2389281096820075e-05,
+      "loss": 0.2527,
+      "step": 734
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4273086190223694,
+      "learning_rate": 2.2261398171865978e-05,
+      "loss": 0.3083,
+      "step": 735
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.408509761095047,
+      "learning_rate": 2.2133776843878186e-05,
+      "loss": 0.2649,
+      "step": 736
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.42201802134513855,
+      "learning_rate": 2.2006418316426775e-05,
+      "loss": 0.2878,
+      "step": 737
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4003610610961914,
+      "learning_rate": 2.187932379060348e-05,
+      "loss": 0.316,
+      "step": 738
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4023764431476593,
+      "learning_rate": 2.1752494465010242e-05,
+      "loss": 0.2217,
+      "step": 739
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4181244373321533,
+      "learning_rate": 2.1625931535747964e-05,
+      "loss": 0.3216,
+      "step": 740
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4374115765094757,
+      "learning_rate": 2.1499636196405228e-05,
+      "loss": 0.3556,
+      "step": 741
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.42732319235801697,
+      "learning_rate": 2.1373609638047032e-05,
+      "loss": 0.2725,
+      "step": 742
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4231344163417816,
+      "learning_rate": 2.1247853049203543e-05,
+      "loss": 0.2665,
+      "step": 743
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.43570348620414734,
+      "learning_rate": 2.112236761585892e-05,
+      "loss": 0.3193,
+      "step": 744
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.40919575095176697,
+      "learning_rate": 2.09971545214401e-05,
+      "loss": 0.2973,
+      "step": 745
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4275833070278168,
+      "learning_rate": 2.087221494680563e-05,
+      "loss": 0.3297,
+      "step": 746
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3157639503479004,
+      "learning_rate": 2.074755007023461e-05,
+      "loss": 0.2182,
+      "step": 747
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3256014883518219,
+      "learning_rate": 2.0623161067415463e-05,
+      "loss": 0.1635,
+      "step": 748
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.43565601110458374,
+      "learning_rate": 2.049904911143492e-05,
+      "loss": 0.3097,
+      "step": 749
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.41193947196006775,
+      "learning_rate": 2.0375215372766944e-05,
+      "loss": 0.306,
+      "step": 750
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.43373343348503113,
+      "learning_rate": 2.025166101926168e-05,
+      "loss": 0.4039,
+      "step": 751
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3443840444087982,
+      "learning_rate": 2.012838721613447e-05,
+      "loss": 0.2038,
+      "step": 752
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.43898192048072815,
+      "learning_rate": 2.0005395125954812e-05,
+      "loss": 0.3767,
+      "step": 753
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.38464224338531494,
+      "learning_rate": 1.9882685908635462e-05,
+      "loss": 0.2612,
+      "step": 754
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.40831586718559265,
+      "learning_rate": 1.9760260721421426e-05,
+      "loss": 0.3061,
+      "step": 755
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.480909138917923,
+      "learning_rate": 1.963812071887913e-05,
+      "loss": 0.3539,
+      "step": 756
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.366283118724823,
+      "learning_rate": 1.951626705288544e-05,
+      "loss": 0.2991,
+      "step": 757
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.42517396807670593,
+      "learning_rate": 1.9394700872616855e-05,
+      "loss": 0.2383,
+      "step": 758
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.44554004073143005,
+      "learning_rate": 1.9273423324538658e-05,
+      "loss": 0.3578,
+      "step": 759
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3832852244377136,
+      "learning_rate": 1.9152435552394104e-05,
+      "loss": 0.2546,
+      "step": 760
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.33742815256118774,
+      "learning_rate": 1.903173869719362e-05,
+      "loss": 0.2239,
+      "step": 761
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.36180728673934937,
+      "learning_rate": 1.891133389720407e-05,
+      "loss": 0.2625,
+      "step": 762
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.4174533188343048,
+      "learning_rate": 1.8791222287937982e-05,
+      "loss": 0.2926,
+      "step": 763
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3777328133583069,
+      "learning_rate": 1.8671405002142918e-05,
+      "loss": 0.2122,
+      "step": 764
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.39400067925453186,
+      "learning_rate": 1.855188316979068e-05,
+      "loss": 0.2878,
+      "step": 765
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.46404820680618286,
+      "learning_rate": 1.8432657918066732e-05,
+      "loss": 0.3093,
+      "step": 766
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.4356730580329895,
+      "learning_rate": 1.831373037135955e-05,
+      "loss": 0.3695,
+      "step": 767
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.35706648230552673,
+      "learning_rate": 1.8195101651250017e-05,
+      "loss": 0.2093,
+      "step": 768
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7616677284240723,
+      "learning_rate": 1.807677287650083e-05,
+      "loss": 0.285,
+      "step": 769
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3962506055831909,
+      "learning_rate": 1.7958745163045986e-05,
+      "loss": 0.2831,
+      "step": 770
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.4961659610271454,
+      "learning_rate": 1.7841019623980217e-05,
+      "loss": 0.2969,
+      "step": 771
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.36217957735061646,
+      "learning_rate": 1.77235973695485e-05,
+      "loss": 0.2028,
+      "step": 772
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.43254271149635315,
+      "learning_rate": 1.760647950713566e-05,
+      "loss": 0.3217,
+      "step": 773
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3909856379032135,
+      "learning_rate": 1.74896671412558e-05,
+      "loss": 0.2526,
+      "step": 774
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.46320241689682007,
+      "learning_rate": 1.737316137354197e-05,
+      "loss": 0.3786,
+      "step": 775
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.42587563395500183,
+      "learning_rate": 1.725696330273575e-05,
+      "loss": 0.3198,
+      "step": 776
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3970955014228821,
+      "learning_rate": 1.7141074024676912e-05,
+      "loss": 0.2809,
+      "step": 777
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3826169967651367,
+      "learning_rate": 1.7025494632293048e-05,
+      "loss": 0.1914,
+      "step": 778
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5121017098426819,
+      "learning_rate": 1.6910226215589303e-05,
+      "loss": 0.3507,
+      "step": 779
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.4497103989124298,
+      "learning_rate": 1.6795269861638042e-05,
+      "loss": 0.3375,
+      "step": 780
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.39944735169410706,
+      "learning_rate": 1.6680626654568686e-05,
+      "loss": 0.261,
+      "step": 781
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5037726759910583,
+      "learning_rate": 1.656629767555739e-05,
+      "loss": 0.3781,
+      "step": 782
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.42347124218940735,
+      "learning_rate": 1.645228400281689e-05,
+      "loss": 0.279,
+      "step": 783
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.4010707437992096,
+      "learning_rate": 1.6338586711586356e-05,
+      "loss": 0.2316,
+      "step": 784
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.381178617477417,
+      "learning_rate": 1.6225206874121218e-05,
+      "loss": 0.3069,
+      "step": 785
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.38435810804367065,
+      "learning_rate": 1.6112145559683055e-05,
+      "loss": 0.2748,
+      "step": 786
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.46212464570999146,
+      "learning_rate": 1.5999403834529548e-05,
+      "loss": 0.3888,
+      "step": 787
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3631595969200134,
+      "learning_rate": 1.5886982761904377e-05,
+      "loss": 0.2455,
+      "step": 788
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4433256685733795,
+      "learning_rate": 1.5774883402027208e-05,
+      "loss": 0.3044,
+      "step": 789
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.48591962456703186,
+      "learning_rate": 1.5663106812083743e-05,
+      "loss": 0.3133,
+      "step": 790
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.36325785517692566,
+      "learning_rate": 1.555165404621567e-05,
+      "loss": 0.2524,
+      "step": 791
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.441135436296463,
+      "learning_rate": 1.5440526155510763e-05,
+      "loss": 0.3638,
+      "step": 792
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4525882601737976,
+      "learning_rate": 1.532972418799298e-05,
+      "loss": 0.3699,
+      "step": 793
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4437645971775055,
+      "learning_rate": 1.5219249188612556e-05,
+      "loss": 0.2911,
+      "step": 794
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.44194477796554565,
+      "learning_rate": 1.5109102199236152e-05,
+      "loss": 0.3282,
+      "step": 795
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5121310353279114,
+      "learning_rate": 1.4999284258637053e-05,
+      "loss": 0.3422,
+      "step": 796
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.4352394640445709,
+      "learning_rate": 1.488979640248534e-05,
+      "loss": 0.2937,
+      "step": 797
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.34901687502861023,
+      "learning_rate": 1.4780639663338125e-05,
+      "loss": 0.2421,
+      "step": 798
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.4127504229545593,
+      "learning_rate": 1.4671815070629868e-05,
+      "loss": 0.2178,
+      "step": 799
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.36559101939201355,
+      "learning_rate": 1.4563323650662586e-05,
+      "loss": 0.1931,
+      "step": 800
+    },
+    {
+      "epoch": 0.76,
+      "eval_loss": 0.2789733111858368,
+      "eval_runtime": 363.429,
+      "eval_samples_per_second": 0.938,
+      "eval_steps_per_second": 0.237,
+      "step": 800
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3729483187198639,
+      "learning_rate": 1.4455166426596223e-05,
+      "loss": 0.2045,
+      "step": 801
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5119748115539551,
+      "learning_rate": 1.434734441843899e-05,
+      "loss": 0.4742,
+      "step": 802
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.4015452563762665,
+      "learning_rate": 1.4239858643037751e-05,
+      "loss": 0.2901,
+      "step": 803
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3744587004184723,
+      "learning_rate": 1.4132710114068426e-05,
+      "loss": 0.2608,
+      "step": 804
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3418838679790497,
+      "learning_rate": 1.4025899842026442e-05,
+      "loss": 0.172,
+      "step": 805
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.43584105372428894,
+      "learning_rate": 1.3919428834217163e-05,
+      "loss": 0.2127,
+      "step": 806
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3847789466381073,
+      "learning_rate": 1.3813298094746491e-05,
+      "loss": 0.2366,
+      "step": 807
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3763447105884552,
+      "learning_rate": 1.3707508624511262e-05,
+      "loss": 0.2055,
+      "step": 808
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.46674445271492004,
+      "learning_rate": 1.36020614211899e-05,
+      "loss": 0.3577,
+      "step": 809
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.37254050374031067,
+      "learning_rate": 1.349695747923298e-05,
+      "loss": 0.2526,
+      "step": 810
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3784024119377136,
+      "learning_rate": 1.339219778985385e-05,
+      "loss": 0.274,
+      "step": 811
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.41034582257270813,
+      "learning_rate": 1.3287783341019278e-05,
+      "loss": 0.2769,
+      "step": 812
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.426227867603302,
+      "learning_rate": 1.3183715117440142e-05,
+      "loss": 0.3297,
+      "step": 813
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.4498612880706787,
+      "learning_rate": 1.3079994100562159e-05,
+      "loss": 0.3108,
+      "step": 814
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.4151266813278198,
+      "learning_rate": 1.2976621268556571e-05,
+      "loss": 0.3209,
+      "step": 815
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.38638436794281006,
+      "learning_rate": 1.2873597596311027e-05,
+      "loss": 0.2732,
+      "step": 816
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.4632682502269745,
+      "learning_rate": 1.2770924055420258e-05,
+      "loss": 0.3059,
+      "step": 817
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.41768112778663635,
+      "learning_rate": 1.2668601614177017e-05,
+      "loss": 0.338,
+      "step": 818
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.39693892002105713,
+      "learning_rate": 1.2566631237562893e-05,
+      "loss": 0.2435,
+      "step": 819
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.39790940284729004,
+      "learning_rate": 1.246501388723923e-05,
+      "loss": 0.2508,
+      "step": 820
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3855772614479065,
+      "learning_rate": 1.2363750521538064e-05,
+      "loss": 0.2554,
+      "step": 821
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.43781426548957825,
+      "learning_rate": 1.2262842095453065e-05,
+      "loss": 0.3742,
+      "step": 822
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.4185785949230194,
+      "learning_rate": 1.2162289560630524e-05,
+      "loss": 0.2896,
+      "step": 823
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.4542255401611328,
+      "learning_rate": 1.2062093865360458e-05,
+      "loss": 0.3461,
+      "step": 824
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.38618943095207214,
+      "learning_rate": 1.1962255954567537e-05,
+      "loss": 0.2373,
+      "step": 825
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.38416847586631775,
+      "learning_rate": 1.1862776769802276e-05,
+      "loss": 0.269,
+      "step": 826
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.4133259952068329,
+      "learning_rate": 1.1763657249232107e-05,
+      "loss": 0.3293,
+      "step": 827
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3361450135707855,
+      "learning_rate": 1.1664898327632551e-05,
+      "loss": 0.234,
+      "step": 828
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.44121041893959045,
+      "learning_rate": 1.1566500936378388e-05,
+      "loss": 0.326,
+      "step": 829
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5260511040687561,
+      "learning_rate": 1.146846600343488e-05,
+      "loss": 0.3423,
+      "step": 830
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.4634191691875458,
+      "learning_rate": 1.1370794453349037e-05,
+      "loss": 0.3999,
+      "step": 831
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5456900000572205,
+      "learning_rate": 1.1273487207240847e-05,
+      "loss": 0.2847,
+      "step": 832
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3128466010093689,
+      "learning_rate": 1.1176545182794674e-05,
+      "loss": 0.1757,
+      "step": 833
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3990103006362915,
+      "learning_rate": 1.1079969294250514e-05,
+      "loss": 0.2695,
+      "step": 834
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.41605833172798157,
+      "learning_rate": 1.0983760452395413e-05,
+      "loss": 0.3151,
+      "step": 835
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5102483034133911,
+      "learning_rate": 1.0887919564554894e-05,
+      "loss": 0.3223,
+      "step": 836
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.4513149559497833,
+      "learning_rate": 1.079244753458437e-05,
+      "loss": 0.3267,
+      "step": 837
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.4077318012714386,
+      "learning_rate": 1.0697345262860636e-05,
+      "loss": 0.3257,
+      "step": 838
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.444173663854599,
+      "learning_rate": 1.0602613646273374e-05,
+      "loss": 0.3632,
+      "step": 839
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.49763721227645874,
+      "learning_rate": 1.0508253578216693e-05,
+      "loss": 0.3773,
+      "step": 840
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4388090968132019,
+      "learning_rate": 1.0414265948580692e-05,
+      "loss": 0.3191,
+      "step": 841
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.42799705266952515,
+      "learning_rate": 1.032065164374313e-05,
+      "loss": 0.2509,
+      "step": 842
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3538854718208313,
+      "learning_rate": 1.0227411546560962e-05,
+      "loss": 0.2537,
+      "step": 843
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4066655933856964,
+      "learning_rate": 1.01345465363621e-05,
+      "loss": 0.2966,
+      "step": 844
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4188317060470581,
+      "learning_rate": 1.0042057488937067e-05,
+      "loss": 0.2803,
+      "step": 845
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.44724395871162415,
+      "learning_rate": 9.949945276530781e-06,
+      "loss": 0.3817,
+      "step": 846
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3425678014755249,
+      "learning_rate": 9.858210767834292e-06,
+      "loss": 0.2459,
+      "step": 847
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.44858115911483765,
+      "learning_rate": 9.766854827976617e-06,
+      "loss": 0.2931,
+      "step": 848
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.41624724864959717,
+      "learning_rate": 9.675878318516546e-06,
+      "loss": 0.2119,
+      "step": 849
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5497798323631287,
+      "learning_rate": 9.58528209743459e-06,
+      "loss": 0.3921,
+      "step": 850
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.42344599962234497,
+      "learning_rate": 9.495067019124792e-06,
+      "loss": 0.3108,
+      "step": 851
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.442204087972641,
+      "learning_rate": 9.405233934386726e-06,
+      "loss": 0.2778,
+      "step": 852
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.38365742564201355,
+      "learning_rate": 9.315783690417479e-06,
+      "loss": 0.2398,
+      "step": 853
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.4032128155231476,
+      "learning_rate": 9.226717130803636e-06,
+      "loss": 0.2589,
+      "step": 854
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.41195717453956604,
+      "learning_rate": 9.138035095513336e-06,
+      "loss": 0.3228,
+      "step": 855
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3214769959449768,
+      "learning_rate": 9.049738420888349e-06,
+      "loss": 0.2425,
+      "step": 856
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.417533278465271,
+      "learning_rate": 8.961827939636196e-06,
+      "loss": 0.2644,
+      "step": 857
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3346613347530365,
+      "learning_rate": 8.87430448082227e-06,
+      "loss": 0.1383,
+      "step": 858
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3608648180961609,
+      "learning_rate": 8.787168869862066e-06,
+      "loss": 0.2268,
+      "step": 859
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.4118165373802185,
+      "learning_rate": 8.700421928513352e-06,
+      "loss": 0.3842,
+      "step": 860
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.4299393594264984,
+      "learning_rate": 8.614064474868421e-06,
+      "loss": 0.2688,
+      "step": 861
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.47080379724502563,
+      "learning_rate": 8.528097323346407e-06,
+      "loss": 0.3258,
+      "step": 862
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.33820775151252747,
+      "learning_rate": 8.442521284685573e-06,
+      "loss": 0.1947,
+      "step": 863
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.39227184653282166,
+      "learning_rate": 8.357337165935675e-06,
+      "loss": 0.2687,
+      "step": 864
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.44301146268844604,
+      "learning_rate": 8.27254577045039e-06,
+      "loss": 0.3271,
+      "step": 865
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.38329029083251953,
+      "learning_rate": 8.188147897879667e-06,
+      "loss": 0.2209,
+      "step": 866
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3819541931152344,
+      "learning_rate": 8.104144344162229e-06,
+      "loss": 0.2434,
+      "step": 867
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3489638566970825,
+      "learning_rate": 8.02053590151805e-06,
+      "loss": 0.2541,
+      "step": 868
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.44321131706237793,
+      "learning_rate": 7.937323358440935e-06,
+      "loss": 0.2676,
+      "step": 869
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3169786036014557,
+      "learning_rate": 7.854507499691005e-06,
+      "loss": 0.1555,
+      "step": 870
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.4151139259338379,
+      "learning_rate": 7.772089106287344e-06,
+      "loss": 0.3088,
+      "step": 871
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.4396266043186188,
+      "learning_rate": 7.690068955500624e-06,
+      "loss": 0.3549,
+      "step": 872
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3974018692970276,
+      "learning_rate": 7.608447820845771e-06,
+      "loss": 0.2379,
+      "step": 873
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3716736435890198,
+      "learning_rate": 7.527226472074678e-06,
+      "loss": 0.2553,
+      "step": 874
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.45256009697914124,
+      "learning_rate": 7.446405675168938e-06,
+      "loss": 0.3242,
+      "step": 875
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3661549985408783,
+      "learning_rate": 7.365986192332625e-06,
+      "loss": 0.186,
+      "step": 876
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.4198785722255707,
+      "learning_rate": 7.285968781985092e-06,
+      "loss": 0.248,
+      "step": 877
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3356267213821411,
+      "learning_rate": 7.206354198753862e-06,
+      "loss": 0.1901,
+      "step": 878
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5052947402000427,
+      "learning_rate": 7.127143193467445e-06,
+      "loss": 0.3564,
+      "step": 879
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5338817238807678,
+      "learning_rate": 7.048336513148307e-06,
+      "loss": 0.3659,
+      "step": 880
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5839039087295532,
+      "learning_rate": 6.969934901005809e-06,
+      "loss": 0.419,
+      "step": 881
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.42304497957229614,
+      "learning_rate": 6.8919390964291895e-06,
+      "loss": 0.3328,
+      "step": 882
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.39831215143203735,
+      "learning_rate": 6.814349834980621e-06,
+      "loss": 0.2877,
+      "step": 883
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.46370935440063477,
+      "learning_rate": 6.7371678483882264e-06,
+      "loss": 0.394,
+      "step": 884
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3856734335422516,
+      "learning_rate": 6.660393864539222e-06,
+      "loss": 0.2537,
+      "step": 885
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3688035011291504,
+      "learning_rate": 6.58402860747302e-06,
+      "loss": 0.1973,
+      "step": 886
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4165421426296234,
+      "learning_rate": 6.508072797374454e-06,
+      "loss": 0.321,
+      "step": 887
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4613915979862213,
+      "learning_rate": 6.4325271505669025e-06,
+      "loss": 0.3161,
+      "step": 888
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.43267375230789185,
+      "learning_rate": 6.3573923795056e-06,
+      "loss": 0.3161,
+      "step": 889
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4351928234100342,
+      "learning_rate": 6.282669192770896e-06,
+      "loss": 0.3331,
+      "step": 890
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4788655638694763,
+      "learning_rate": 6.208358295061572e-06,
+      "loss": 0.3003,
+      "step": 891
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3836725950241089,
+      "learning_rate": 6.134460387188206e-06,
+      "loss": 0.252,
+      "step": 892
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.38111889362335205,
+      "learning_rate": 6.060976166066546e-06,
+      "loss": 0.258,
+      "step": 893
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.46063032746315,
+      "learning_rate": 5.9879063247109405e-06,
+      "loss": 0.3298,
+      "step": 894
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.4190021753311157,
+      "learning_rate": 5.915251552227829e-06,
+      "loss": 0.2766,
+      "step": 895
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5162159204483032,
+      "learning_rate": 5.8430125338092115e-06,
+      "loss": 0.3708,
+      "step": 896
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.400238573551178,
+      "learning_rate": 5.771189950726191e-06,
+      "loss": 0.2258,
+      "step": 897
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.45645174384117126,
+      "learning_rate": 5.699784480322568e-06,
+      "loss": 0.3665,
+      "step": 898
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.45055222511291504,
+      "learning_rate": 5.628796796008434e-06,
+      "loss": 0.3089,
+      "step": 899
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3893721401691437,
+      "learning_rate": 5.558227567253832e-06,
+      "loss": 0.2761,
+      "step": 900
+    },
+    {
+      "epoch": 0.85,
+      "eval_loss": 0.27538353204727173,
+      "eval_runtime": 357.8998,
+      "eval_samples_per_second": 0.953,
+      "eval_steps_per_second": 0.24,
+      "step": 900
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.48486000299453735,
+      "learning_rate": 5.488077459582425e-06,
+      "loss": 0.3343,
+      "step": 901
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.39481961727142334,
+      "learning_rate": 5.418347134565249e-06,
+      "loss": 0.2881,
+      "step": 902
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.4042988419532776,
+      "learning_rate": 5.349037249814442e-06,
+      "loss": 0.2922,
+      "step": 903
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.4478388726711273,
+      "learning_rate": 5.28014845897708e-06,
+      "loss": 0.2936,
+      "step": 904
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.4552001357078552,
+      "learning_rate": 5.211681411728969e-06,
+      "loss": 0.3182,
+      "step": 905
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3898690640926361,
+      "learning_rate": 5.143636753768549e-06,
+      "loss": 0.2386,
+      "step": 906
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3873755633831024,
+      "learning_rate": 5.076015126810784e-06,
+      "loss": 0.2986,
+      "step": 907
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.406667023897171,
+      "learning_rate": 5.008817168581137e-06,
+      "loss": 0.2655,
+      "step": 908
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.4281293749809265,
+      "learning_rate": 4.94204351280953e-06,
+      "loss": 0.3063,
+      "step": 909
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.4260796010494232,
+      "learning_rate": 4.8756947892243725e-06,
+      "loss": 0.3075,
+      "step": 910
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3630758225917816,
+      "learning_rate": 4.809771623546627e-06,
+      "loss": 0.2135,
+      "step": 911
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3312900960445404,
+      "learning_rate": 4.744274637483936e-06,
+      "loss": 0.1985,
+      "step": 912
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.36237314343452454,
+      "learning_rate": 4.6792044487247e-06,
+      "loss": 0.2839,
+      "step": 913
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.40924111008644104,
+      "learning_rate": 4.614561670932288e-06,
+      "loss": 0.2704,
+      "step": 914
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.38105857372283936,
+      "learning_rate": 4.550346913739256e-06,
+      "loss": 0.2213,
+      "step": 915
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.388007789850235,
+      "learning_rate": 4.486560782741578e-06,
+      "loss": 0.2179,
+      "step": 916
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.34383153915405273,
+      "learning_rate": 4.423203879492943e-06,
+      "loss": 0.2049,
+      "step": 917
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4638625383377075,
+      "learning_rate": 4.360276801499086e-06,
+      "loss": 0.338,
+      "step": 918
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4769915044307709,
+      "learning_rate": 4.29778014221216e-06,
+      "loss": 0.2742,
+      "step": 919
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4313514828681946,
+      "learning_rate": 4.2357144910251e-06,
+      "loss": 0.2994,
+      "step": 920
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.45999112725257874,
+      "learning_rate": 4.174080433266136e-06,
+      "loss": 0.3357,
+      "step": 921
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4573007822036743,
+      "learning_rate": 4.112878550193194e-06,
+      "loss": 0.3172,
+      "step": 922
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.36643993854522705,
+      "learning_rate": 4.05210941898847e-06,
+      "loss": 0.2128,
+      "step": 923
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.37925997376441956,
+      "learning_rate": 3.991773612752952e-06,
+      "loss": 0.2358,
+      "step": 924
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.41036054491996765,
+      "learning_rate": 3.9318717005010495e-06,
+      "loss": 0.3105,
+      "step": 925
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3971090316772461,
+      "learning_rate": 3.872404247155193e-06,
+      "loss": 0.3083,
+      "step": 926
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.43480557203292847,
+      "learning_rate": 3.8133718135405285e-06,
+      "loss": 0.3432,
+      "step": 927
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.4274615943431854,
+      "learning_rate": 3.754774956379614e-06,
+      "loss": 0.2778,
+      "step": 928
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.4346570670604706,
+      "learning_rate": 3.696614228287187e-06,
+      "loss": 0.3049,
+      "step": 929
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.49182531237602234,
+      "learning_rate": 3.6388901777649477e-06,
+      "loss": 0.3099,
+      "step": 930
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.4507730305194855,
+      "learning_rate": 3.581603349196372e-06,
+      "loss": 0.2935,
+      "step": 931
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.4616379141807556,
+      "learning_rate": 3.5247542828415747e-06,
+      "loss": 0.3775,
+      "step": 932
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.397449254989624,
+      "learning_rate": 3.468343514832251e-06,
+      "loss": 0.2384,
+      "step": 933
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.4080866277217865,
+      "learning_rate": 3.4123715771665786e-06,
+      "loss": 0.2142,
+      "step": 934
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.42029279470443726,
+      "learning_rate": 3.356838997704226e-06,
+      "loss": 0.319,
+      "step": 935
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.4343917965888977,
+      "learning_rate": 3.3017463001613626e-06,
+      "loss": 0.2456,
+      "step": 936
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3546516001224518,
+      "learning_rate": 3.2470940041057106e-06,
+      "loss": 0.2046,
+      "step": 937
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.37147656083106995,
+      "learning_rate": 3.1928826249516987e-06,
+      "loss": 0.2059,
+      "step": 938
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.4479697644710541,
+      "learning_rate": 3.139112673955513e-06,
+      "loss": 0.2751,
+      "step": 939
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.40827590227127075,
+      "learning_rate": 3.08578465821035e-06,
+      "loss": 0.2909,
+      "step": 940
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3929819166660309,
+      "learning_rate": 3.0328990806415934e-06,
+      "loss": 0.2879,
+      "step": 941
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5057595372200012,
+      "learning_rate": 2.9804564400020994e-06,
+      "loss": 0.3474,
+      "step": 942
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.42703625559806824,
+      "learning_rate": 2.9284572308674596e-06,
+      "loss": 0.3362,
+      "step": 943
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.42836809158325195,
+      "learning_rate": 2.8769019436313715e-06,
+      "loss": 0.3082,
+      "step": 944
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3881003260612488,
+      "learning_rate": 2.825791064500993e-06,
+      "loss": 0.2751,
+      "step": 945
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4846949577331543,
+      "learning_rate": 2.7751250754923573e-06,
+      "loss": 0.2891,
+      "step": 946
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.39968496561050415,
+      "learning_rate": 2.7249044544258363e-06,
+      "loss": 0.3095,
+      "step": 947
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4672655463218689,
+      "learning_rate": 2.675129674921639e-06,
+      "loss": 0.3978,
+      "step": 948
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5235310792922974,
+      "learning_rate": 2.625801206395312e-06,
+      "loss": 0.379,
+      "step": 949
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.41804197430610657,
+      "learning_rate": 2.576919514053355e-06,
+      "loss": 0.264,
+      "step": 950
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.526220977306366,
+      "learning_rate": 2.528485058888813e-06,
+      "loss": 0.3815,
+      "step": 951
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.41712749004364014,
+      "learning_rate": 2.48049829767692e-06,
+      "loss": 0.3206,
+      "step": 952
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5365576148033142,
+      "learning_rate": 2.4329596829708144e-06,
+      "loss": 0.392,
+      "step": 953
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.39783185720443726,
+      "learning_rate": 2.385869663097251e-06,
+      "loss": 0.2809,
+      "step": 954
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4556587040424347,
+      "learning_rate": 2.3392286821523724e-06,
+      "loss": 0.3399,
+      "step": 955
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.40517574548721313,
+      "learning_rate": 2.2930371799975594e-06,
+      "loss": 0.2141,
+      "step": 956
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.40251103043556213,
+      "learning_rate": 2.2472955922552163e-06,
+      "loss": 0.2466,
+      "step": 957
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.4396744668483734,
+      "learning_rate": 2.202004350304715e-06,
+      "loss": 0.3454,
+      "step": 958
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.4512024521827698,
+      "learning_rate": 2.157163881278312e-06,
+      "loss": 0.3142,
+      "step": 959
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.42345571517944336,
+      "learning_rate": 2.112774608057111e-06,
+      "loss": 0.2933,
+      "step": 960
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.47581082582473755,
+      "learning_rate": 2.068836949267089e-06,
+      "loss": 0.3281,
+      "step": 961
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.4577794075012207,
+      "learning_rate": 2.0253513192751373e-06,
+      "loss": 0.2667,
+      "step": 962
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.41427233815193176,
+      "learning_rate": 1.982318128185151e-06,
+      "loss": 0.2909,
+      "step": 963
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.45944905281066895,
+      "learning_rate": 1.9397377818341944e-06,
+      "loss": 0.2948,
+      "step": 964
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.45142343640327454,
+      "learning_rate": 1.8976106817886196e-06,
+      "loss": 0.3679,
+      "step": 965
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4514009654521942,
+      "learning_rate": 1.855937225340315e-06,
+      "loss": 0.3024,
+      "step": 966
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4211220145225525,
+      "learning_rate": 1.8147178055029579e-06,
+      "loss": 0.275,
+      "step": 967
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3520624041557312,
+      "learning_rate": 1.7739528110083004e-06,
+      "loss": 0.2447,
+      "step": 968
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.44592034816741943,
+      "learning_rate": 1.7336426263024897e-06,
+      "loss": 0.3103,
+      "step": 969
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3339546024799347,
+      "learning_rate": 1.6937876315424705e-06,
+      "loss": 0.1964,
+      "step": 970
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3771350085735321,
+      "learning_rate": 1.6543882025923886e-06,
+      "loss": 0.2632,
+      "step": 971
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4589972496032715,
+      "learning_rate": 1.6154447110200255e-06,
+      "loss": 0.339,
+      "step": 972
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4086770713329315,
+      "learning_rate": 1.5769575240933422e-06,
+      "loss": 0.2235,
+      "step": 973
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6194506287574768,
+      "learning_rate": 1.5389270047769578e-06,
+      "loss": 0.4712,
+      "step": 974
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4194900691509247,
+      "learning_rate": 1.5013535117287647e-06,
+      "loss": 0.2939,
+      "step": 975
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3534494638442993,
+      "learning_rate": 1.4642373992965364e-06,
+      "loss": 0.1908,
+      "step": 976
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3865540325641632,
+      "learning_rate": 1.4275790175145908e-06,
+      "loss": 0.218,
+      "step": 977
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.39693301916122437,
+      "learning_rate": 1.3913787121004719e-06,
+      "loss": 0.2441,
+      "step": 978
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.4401733875274658,
+      "learning_rate": 1.3556368244517114e-06,
+      "loss": 0.3099,
+      "step": 979
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.4274381697177887,
+      "learning_rate": 1.3203536916425841e-06,
+      "loss": 0.2667,
+      "step": 980
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.42912420630455017,
+      "learning_rate": 1.2855296464209688e-06,
+      "loss": 0.3108,
+      "step": 981
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.4918239712715149,
+      "learning_rate": 1.2511650172051637e-06,
+      "loss": 0.4166,
+      "step": 982
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.4177713096141815,
+      "learning_rate": 1.217260128080816e-06,
+      "loss": 0.2737,
+      "step": 983
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.4006113111972809,
+      "learning_rate": 1.1838152987978578e-06,
+      "loss": 0.2749,
+      "step": 984
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.4339311122894287,
+      "learning_rate": 1.1508308447674975e-06,
+      "loss": 0.2834,
+      "step": 985
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.4010700285434723,
+      "learning_rate": 1.1183070770592441e-06,
+      "loss": 0.285,
+      "step": 986
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3905528485774994,
+      "learning_rate": 1.0862443023979651e-06,
+      "loss": 0.2601,
+      "step": 987
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3383557200431824,
+      "learning_rate": 1.0546428231609895e-06,
+      "loss": 0.2232,
+      "step": 988
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5410901308059692,
+      "learning_rate": 1.0235029373752758e-06,
+      "loss": 0.3973,
+      "step": 989
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.42802780866622925,
+      "learning_rate": 9.928249387145982e-07,
+      "loss": 0.2575,
+      "step": 990
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.40100714564323425,
+      "learning_rate": 9.6260911649676e-07,
+      "loss": 0.2064,
+      "step": 991
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.36955365538597107,
+      "learning_rate": 9.32855755680867e-07,
+      "loss": 0.2152,
+      "step": 992
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.39849358797073364,
+      "learning_rate": 9.035651368646648e-07,
+      "loss": 0.2989,
+      "step": 993
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.504355251789093,
+      "learning_rate": 8.747375362818666e-07,
+      "loss": 0.4096,
+      "step": 994
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.45022833347320557,
+      "learning_rate": 8.463732257995571e-07,
+      "loss": 0.3108,
+      "step": 995
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.32840028405189514,
+      "learning_rate": 8.184724729156379e-07,
+      "loss": 0.2003,
+      "step": 996
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.43470728397369385,
+      "learning_rate": 7.910355407562742e-07,
+      "loss": 0.2808,
+      "step": 997
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3786812424659729,
+      "learning_rate": 7.640626880734581e-07,
+      "loss": 0.2073,
+      "step": 998
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.35407590866088867,
+      "learning_rate": 7.375541692425326e-07,
+      "loss": 0.2183,
+      "step": 999
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.46359875798225403,
+      "learning_rate": 7.115102342598101e-07,
+      "loss": 0.338,
+      "step": 1000
+    },
+    {
+      "epoch": 0.95,
+      "eval_loss": 0.2737712562084198,
+      "eval_runtime": 353.7642,
+      "eval_samples_per_second": 0.964,
+      "eval_steps_per_second": 0.243,
+      "step": 1000
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.41465336084365845,
+      "learning_rate": 6.859311287402081e-07,
+      "loss": 0.2373,
+      "step": 1001
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.36273032426834106,
+      "learning_rate": 6.608170939149283e-07,
+      "loss": 0.1952,
+      "step": 1002
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.4153614342212677,
+      "learning_rate": 6.361683666291973e-07,
+      "loss": 0.3195,
+      "step": 1003
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5194461345672607,
+      "learning_rate": 6.119851793400189e-07,
+      "loss": 0.3115,
+      "step": 1004
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.41203445196151733,
+      "learning_rate": 5.882677601139919e-07,
+      "loss": 0.2507,
+      "step": 1005
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.43692079186439514,
+      "learning_rate": 5.650163326251345e-07,
+      "loss": 0.255,
+      "step": 1006
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5047875642776489,
+      "learning_rate": 5.422311161528193e-07,
+      "loss": 0.3724,
+      "step": 1007
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.44066694378852844,
+      "learning_rate": 5.199123255796634e-07,
+      "loss": 0.2655,
+      "step": 1008
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3251340687274933,
+      "learning_rate": 4.980601713895305e-07,
+      "loss": 0.1632,
+      "step": 1009
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4146948456764221,
+      "learning_rate": 4.7667485966552685e-07,
+      "loss": 0.2316,
+      "step": 1010
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4053228795528412,
+      "learning_rate": 4.55756592088058e-07,
+      "loss": 0.3262,
+      "step": 1011
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4250980019569397,
+      "learning_rate": 4.353055659329419e-07,
+      "loss": 0.3648,
+      "step": 1012
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4724719226360321,
+      "learning_rate": 4.153219740695435e-07,
+      "loss": 0.3675,
+      "step": 1013
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4801291525363922,
+      "learning_rate": 3.958060049589485e-07,
+      "loss": 0.3561,
+      "step": 1014
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.38172417879104614,
+      "learning_rate": 3.767578426521923e-07,
+      "loss": 0.2341,
+      "step": 1015
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.37145552039146423,
+      "learning_rate": 3.5817766678850615e-07,
+      "loss": 0.2414,
+      "step": 1016
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.37331122159957886,
+      "learning_rate": 3.4006565259366295e-07,
+      "loss": 0.2482,
+      "step": 1017
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3793918192386627,
+      "learning_rate": 3.224219708782894e-07,
+      "loss": 0.2814,
+      "step": 1018
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.47230786085128784,
+      "learning_rate": 3.052467880362675e-07,
+      "loss": 0.3887,
+      "step": 1019
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4130839705467224,
+      "learning_rate": 2.88540266043158e-07,
+      "loss": 0.2995,
+      "step": 1020
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4309743642807007,
+      "learning_rate": 2.723025624546849e-07,
+      "loss": 0.3233,
+      "step": 1021
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4043530523777008,
+      "learning_rate": 2.5653383040524227e-07,
+      "loss": 0.3045,
+      "step": 1022
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.432908296585083,
+      "learning_rate": 2.412342186064565e-07,
+      "loss": 0.2307,
+      "step": 1023
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4279992878437042,
+      "learning_rate": 2.2640387134577058e-07,
+      "loss": 0.3555,
+      "step": 1024
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3497396409511566,
+      "learning_rate": 2.1204292848509555e-07,
+      "loss": 0.2301,
+      "step": 1025
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4449145197868347,
+      "learning_rate": 1.98151525459489e-07,
+      "loss": 0.2748,
+      "step": 1026
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.40879833698272705,
+      "learning_rate": 1.8472979327587292e-07,
+      "loss": 0.2129,
+      "step": 1027
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4308483600616455,
+      "learning_rate": 1.717778585118013e-07,
+      "loss": 0.3235,
+      "step": 1028
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.42937713861465454,
+      "learning_rate": 1.5929584331427216e-07,
+      "loss": 0.3039,
+      "step": 1029
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5298446416854858,
+      "learning_rate": 1.4728386539856754e-07,
+      "loss": 0.3446,
+      "step": 1030
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.457104355096817,
+      "learning_rate": 1.3574203804713747e-07,
+      "loss": 0.4001,
+      "step": 1031
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.39928048849105835,
+      "learning_rate": 1.246704701085566e-07,
+      "loss": 0.2695,
+      "step": 1032
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.4090886116027832,
+      "learning_rate": 1.1406926599646373e-07,
+      "loss": 0.273,
+      "step": 1033
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.4059845805168152,
+      "learning_rate": 1.0393852568860719e-07,
+      "loss": 0.2309,
+      "step": 1034
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.32376325130462646,
+      "learning_rate": 9.427834472588992e-08,
+      "loss": 0.1461,
+      "step": 1035
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.42355772852897644,
+      "learning_rate": 8.508881421145365e-08,
+      "loss": 0.2592,
+      "step": 1036
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.4212473928928375,
+      "learning_rate": 7.637002080985168e-08,
+      "loss": 0.2276,
+      "step": 1037
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.4643043279647827,
+      "learning_rate": 6.812204674619959e-08,
+      "loss": 0.3243,
+      "step": 1038
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.42566314339637756,
+      "learning_rate": 6.034496980542037e-08,
+      "loss": 0.2772,
+      "step": 1039
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.40134018659591675,
+      "learning_rate": 5.303886333151153e-08,
+      "loss": 0.2685,
+      "step": 1040
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.40590885281562805,
+      "learning_rate": 4.620379622682358e-08,
+      "loss": 0.2859,
+      "step": 1041
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.4037347435951233,
+      "learning_rate": 3.9839832951465986e-08,
+      "loss": 0.247,
+      "step": 1042
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3922388255596161,
+      "learning_rate": 3.394703352263551e-08,
+      "loss": 0.2599,
+      "step": 1043
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.4479919970035553,
+      "learning_rate": 2.8525453514099965e-08,
+      "loss": 0.3304,
+      "step": 1044
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.36609113216400146,
+      "learning_rate": 2.3575144055643094e-08,
+      "loss": 0.1947,
+      "step": 1045
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.4246656894683838,
+      "learning_rate": 1.9096151832609375e-08,
+      "loss": 0.2425,
+      "step": 1046
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.38966405391693115,
+      "learning_rate": 1.5088519085437734e-08,
+      "loss": 0.2894,
+      "step": 1047
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.39244353771209717,
+      "learning_rate": 1.1552283609272962e-08,
+      "loss": 0.223,
+      "step": 1048
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.4142623543739319,
+      "learning_rate": 8.487478753615997e-09,
+      "loss": 0.3221,
+      "step": 1049
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.4871939420700073,
+      "learning_rate": 5.894133422001957e-09,
+      "loss": 0.3634,
+      "step": 1050
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.35148271918296814,
+      "learning_rate": 3.772272071722593e-09,
+      "loss": 0.2149,
+      "step": 1051
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.4478323459625244,
+      "learning_rate": 2.1219147136264382e-09,
+      "loss": 0.27,
+      "step": 1052
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.4414227306842804,
+      "learning_rate": 9.43076911874563e-10,
+      "loss": 0.3492,
+      "step": 1053
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.4506877362728119,
+      "learning_rate": 2.3576978384065584e-10,
+      "loss": 0.3233,
+      "step": 1054
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.40581271052360535,
+      "learning_rate": 0.0,
+      "loss": 0.2496,
+      "step": 1055
+    },
+    {
+      "epoch": 1.0,
+      "step": 1055,
+      "total_flos": 1.1584683941193318e+17,
+      "train_loss": 0.31083106594910553,
+      "train_runtime": 54396.122,
+      "train_samples_per_second": 0.621,
+      "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
+}