diff --git a/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa_shuffled-streaming_baseline_lora_corr_agg_fc4_s2048/README.md b/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa_shuffled-streaming_baseline_lora_corr_agg_fc4_s2048/README.md
new file mode 100644
index 0000000000000000000000000000000000000000..d4576fe074287232d3836bf69c21d3f2593290d9
--- /dev/null
+++ b/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa_shuffled-streaming_baseline_lora_corr_agg_fc4_s2048/README.md
@@ -0,0 +1,9 @@
+---
+library_name: peft
+---
+## Training procedure
+
+### Framework versions
+
+
+- PEFT 0.4.0
diff --git a/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa_shuffled-streaming_baseline_lora_corr_agg_fc4_s2048/adapter_config.json b/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa_shuffled-streaming_baseline_lora_corr_agg_fc4_s2048/adapter_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..4af632a4c7fa1e034e8dbb9c8ee7569985bdb54b
--- /dev/null
+++ b/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa_shuffled-streaming_baseline_lora_corr_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": 32,
+  "lora_dropout": 0.05,
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "r": 64,
+  "revision": null,
+  "target_modules": [
+    "gate_proj",
+    "q_proj",
+    "o_proj",
+    "v_proj",
+    "up_proj",
+    "k_proj",
+    "down_proj"
+  ],
+  "task_type": "CAUSAL_LM"
+}
\ No newline at end of file
diff --git a/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa_shuffled-streaming_baseline_lora_corr_agg_fc4_s2048/adapter_model.bin b/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa_shuffled-streaming_baseline_lora_corr_agg_fc4_s2048/adapter_model.bin
new file mode 100644
index 0000000000000000000000000000000000000000..3f0a1e758d46700da8e2116fd1c2e6458bd89fc8
--- /dev/null
+++ b/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa_shuffled-streaming_baseline_lora_corr_agg_fc4_s2048/adapter_model.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dae8934b14f78f93271b7d4f86516e3c0fa37acd2e77fae2f298e6ec2bc5d882
+size 346162458
diff --git a/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa_shuffled-streaming_baseline_lora_corr_agg_fc4_s2048/config.json b/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa_shuffled-streaming_baseline_lora_corr_agg_fc4_s2048/config.json
new file mode 100644
index 0000000000000000000000000000000000000000..a48d68518cce4598c62f52102a6c0a9e1adc9faa
--- /dev/null
+++ b/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa_shuffled-streaming_baseline_lora_corr_agg_fc4_s2048/config.json
@@ -0,0 +1,68 @@
+{
+  "_name_or_path": "/data/wiedmann/hub/models--lmms-lab--llava-onevision-qwen2-7b-ov/snapshots/0b07bf7565e244cf4f39982249eafe8cd799d6dd",
+  "add_faster_video": false,
+  "add_time_instruction": false,
+  "architectures": [
+    "LlavaQwenForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 151643,
+  "eos_token_id": 151645,
+  "faster_token_stride": 10,
+  "force_sample": false,
+  "hidden_act": "silu",
+  "hidden_size": 3584,
+  "image_aspect_ratio": "square",
+  "image_crop_resolution": null,
+  "image_grid_pinpoints": null,
+  "image_split_resolution": null,
+  "image_token_index": 151646,
+  "initializer_range": 0.02,
+  "intermediate_size": 18944,
+  "max_position_embeddings": 32768,
+  "max_window_layers": 28,
+  "mm_hidden_size": 1152,
+  "mm_newline_position": "no_token",
+  "mm_patch_merge_type": "spatial_unpad",
+  "mm_projector_lr": null,
+  "mm_projector_type": "mlp2x_gelu",
+  "mm_resampler_type": "streaming_agg",
+  "mm_spatial_pool_mode": "bilinear",
+  "mm_spatial_pool_stride": null,
+  "mm_streaming_frames_per_chunk": 4,
+  "mm_streaming_input_dim": 1152,
+  "mm_streaming_mlp_ratio": 4.0,
+  "mm_streaming_num_heads": 8,
+  "mm_streaming_num_layers": 4,
+  "mm_streaming_num_state_tokens": 2048,
+  "mm_streaming_patches_per_frame": 729,
+  "mm_streaming_state_dim": 1152,
+  "mm_streaming_vision_chunk_size": 8,
+  "mm_tunable_parts": "mm_mlp_adapter,mm_vision_resampler",
+  "mm_use_im_patch_token": false,
+  "mm_use_im_start_end": false,
+  "mm_vision_select_feature": "patch",
+  "mm_vision_select_layer": -2,
+  "mm_vision_tower": "google/siglip-so400m-patch14-384",
+  "mm_vision_tower_lr": null,
+  "model_type": "llava_qwen",
+  "num_attention_heads": 28,
+  "num_hidden_layers": 28,
+  "num_key_value_heads": 4,
+  "pos_skipping_range": 4096,
+  "rms_norm_eps": 1e-06,
+  "rope_scaling": null,
+  "rope_theta": 1000000.0,
+  "sliding_window": 131072,
+  "tie_word_embeddings": false,
+  "tokenizer_model_max_length": 8192,
+  "tokenizer_padding_side": "right",
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.40.0.dev0",
+  "use_cache": true,
+  "use_mm_proj": true,
+  "use_pos_skipping": false,
+  "use_sliding_window": false,
+  "vision_tower_pretrained": null,
+  "vocab_size": 152064
+}
diff --git a/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa_shuffled-streaming_baseline_lora_corr_agg_fc4_s2048/generation_config.json b/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa_shuffled-streaming_baseline_lora_corr_agg_fc4_s2048/generation_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..19a297221acb87418d4388a3decef2282c6d7316
--- /dev/null
+++ b/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa_shuffled-streaming_baseline_lora_corr_agg_fc4_s2048/generation_config.json
@@ -0,0 +1,14 @@
+{
+  "bos_token_id": 151643,
+  "do_sample": true,
+  "eos_token_id": [
+    151645,
+    151643
+  ],
+  "pad_token_id": 151643,
+  "repetition_penalty": 1.05,
+  "temperature": 0.7,
+  "top_k": 20,
+  "top_p": 0.8,
+  "transformers_version": "4.40.0.dev0"
+}
diff --git a/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa_shuffled-streaming_baseline_lora_corr_agg_fc4_s2048/non_lora_trainables.bin b/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa_shuffled-streaming_baseline_lora_corr_agg_fc4_s2048/non_lora_trainables.bin
new file mode 100644
index 0000000000000000000000000000000000000000..4bd792fc725b9fe38dded0314e1f80f108d5fcc6
--- /dev/null
+++ b/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa_shuffled-streaming_baseline_lora_corr_agg_fc4_s2048/non_lora_trainables.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:83df14521784f0469cb8816cfa0481571642869eba79790c45ec99df93989734
+size 211441788
diff --git a/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa_shuffled-streaming_baseline_lora_corr_agg_fc4_s2048/trainer_state.json b/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa_shuffled-streaming_baseline_lora_corr_agg_fc4_s2048/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..06236f1dc48c1bab877465287b3a3e5e98d068a5
--- /dev/null
+++ b/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa_shuffled-streaming_baseline_lora_corr_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": 17.0417537689209,
+      "learning_rate": 3.125e-06,
+      "loss": 1.1519,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 15.427328109741211,
+      "learning_rate": 6.25e-06,
+      "loss": 1.0785,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 14.720137596130371,
+      "learning_rate": 9.375000000000001e-06,
+      "loss": 1.164,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 10.265087127685547,
+      "learning_rate": 1.25e-05,
+      "loss": 1.1291,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.201363563537598,
+      "learning_rate": 1.5625e-05,
+      "loss": 0.9853,
+      "step": 5
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.612859725952148,
+      "learning_rate": 1.8750000000000002e-05,
+      "loss": 0.9617,
+      "step": 6
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.7551286220550537,
+      "learning_rate": 2.1875e-05,
+      "loss": 0.8711,
+      "step": 7
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.353327989578247,
+      "learning_rate": 2.5e-05,
+      "loss": 0.7447,
+      "step": 8
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2674058675765991,
+      "learning_rate": 2.8125000000000003e-05,
+      "loss": 0.5473,
+      "step": 9
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6474534273147583,
+      "learning_rate": 3.125e-05,
+      "loss": 0.5326,
+      "step": 10
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2107363939285278,
+      "learning_rate": 3.4375e-05,
+      "loss": 0.4587,
+      "step": 11
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8501068353652954,
+      "learning_rate": 3.7500000000000003e-05,
+      "loss": 0.3679,
+      "step": 12
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0735009908676147,
+      "learning_rate": 4.0625000000000005e-05,
+      "loss": 0.458,
+      "step": 13
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.4377739429473877,
+      "learning_rate": 4.375e-05,
+      "loss": 0.4188,
+      "step": 14
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5296626687049866,
+      "learning_rate": 4.6875e-05,
+      "loss": 0.62,
+      "step": 15
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.39675450325012207,
+      "learning_rate": 5e-05,
+      "loss": 0.4481,
+      "step": 16
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.40650835633277893,
+      "learning_rate": 5.3125000000000004e-05,
+      "loss": 0.4877,
+      "step": 17
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.33760541677474976,
+      "learning_rate": 5.6250000000000005e-05,
+      "loss": 0.3327,
+      "step": 18
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.45227017998695374,
+      "learning_rate": 5.9375e-05,
+      "loss": 0.4035,
+      "step": 19
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.32842957973480225,
+      "learning_rate": 6.25e-05,
+      "loss": 0.3088,
+      "step": 20
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.246881365776062,
+      "learning_rate": 6.562500000000001e-05,
+      "loss": 0.4249,
+      "step": 21
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.1798517107963562,
+      "learning_rate": 6.875e-05,
+      "loss": 0.191,
+      "step": 22
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.22029730677604675,
+      "learning_rate": 7.1875e-05,
+      "loss": 0.318,
+      "step": 23
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.23356413841247559,
+      "learning_rate": 7.500000000000001e-05,
+      "loss": 0.4912,
+      "step": 24
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.19552457332611084,
+      "learning_rate": 7.8125e-05,
+      "loss": 0.2547,
+      "step": 25
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.19979548454284668,
+      "learning_rate": 8.125000000000001e-05,
+      "loss": 0.3621,
+      "step": 26
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.19841191172599792,
+      "learning_rate": 8.4375e-05,
+      "loss": 0.3666,
+      "step": 27
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.19183498620986938,
+      "learning_rate": 8.75e-05,
+      "loss": 0.3491,
+      "step": 28
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.18138259649276733,
+      "learning_rate": 9.062500000000001e-05,
+      "loss": 0.352,
+      "step": 29
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2047562599182129,
+      "learning_rate": 9.375e-05,
+      "loss": 0.3212,
+      "step": 30
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.18068565428256989,
+      "learning_rate": 9.687500000000001e-05,
+      "loss": 0.32,
+      "step": 31
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.31302621960639954,
+      "learning_rate": 0.0001,
+      "loss": 0.3398,
+      "step": 32
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.22792330384254456,
+      "learning_rate": 9.999976423021617e-05,
+      "loss": 0.3693,
+      "step": 33
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.17064064741134644,
+      "learning_rate": 9.999905692308813e-05,
+      "loss": 0.3797,
+      "step": 34
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2116643637418747,
+      "learning_rate": 9.999787808528638e-05,
+      "loss": 0.4705,
+      "step": 35
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.16474707424640656,
+      "learning_rate": 9.999622772792828e-05,
+      "loss": 0.3065,
+      "step": 36
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.17513303458690643,
+      "learning_rate": 9.9994105866578e-05,
+      "loss": 0.3853,
+      "step": 37
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.20958589017391205,
+      "learning_rate": 9.999151252124639e-05,
+      "loss": 0.3858,
+      "step": 38
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.2161528766155243,
+      "learning_rate": 9.998844771639073e-05,
+      "loss": 0.3145,
+      "step": 39
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.16423137485980988,
+      "learning_rate": 9.998491148091457e-05,
+      "loss": 0.3188,
+      "step": 40
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.2106671780347824,
+      "learning_rate": 9.99809038481674e-05,
+      "loss": 0.45,
+      "step": 41
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1796143800020218,
+      "learning_rate": 9.997642485594436e-05,
+      "loss": 0.3425,
+      "step": 42
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.17102830111980438,
+      "learning_rate": 9.99714745464859e-05,
+      "loss": 0.3739,
+      "step": 43
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.18903662264347076,
+      "learning_rate": 9.996605296647737e-05,
+      "loss": 0.3039,
+      "step": 44
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1893555074930191,
+      "learning_rate": 9.996016016704854e-05,
+      "loss": 0.4995,
+      "step": 45
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.18810036778450012,
+      "learning_rate": 9.995379620377318e-05,
+      "loss": 0.2729,
+      "step": 46
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.17215709388256073,
+      "learning_rate": 9.99469611366685e-05,
+      "loss": 0.257,
+      "step": 47
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.17270013689994812,
+      "learning_rate": 9.993965503019457e-05,
+      "loss": 0.4027,
+      "step": 48
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.17746330797672272,
+      "learning_rate": 9.99318779532538e-05,
+      "loss": 0.3545,
+      "step": 49
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.17929814755916595,
+      "learning_rate": 9.992362997919016e-05,
+      "loss": 0.3208,
+      "step": 50
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.19912242889404297,
+      "learning_rate": 9.991491118578856e-05,
+      "loss": 0.3807,
+      "step": 51
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.1734088957309723,
+      "learning_rate": 9.990572165527413e-05,
+      "loss": 0.3583,
+      "step": 52
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.22232988476753235,
+      "learning_rate": 9.98960614743114e-05,
+      "loss": 0.4393,
+      "step": 53
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.20737454295158386,
+      "learning_rate": 9.988593073400354e-05,
+      "loss": 0.4323,
+      "step": 54
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.18521550297737122,
+      "learning_rate": 9.987532952989145e-05,
+      "loss": 0.3836,
+      "step": 55
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.17583943903446198,
+      "learning_rate": 9.986425796195287e-05,
+      "loss": 0.2777,
+      "step": 56
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.19251686334609985,
+      "learning_rate": 9.985271613460144e-05,
+      "loss": 0.408,
+      "step": 57
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.16808868944644928,
+      "learning_rate": 9.984070415668574e-05,
+      "loss": 0.2523,
+      "step": 58
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.20617179572582245,
+      "learning_rate": 9.982822214148821e-05,
+      "loss": 0.4156,
+      "step": 59
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.19919346272945404,
+      "learning_rate": 9.981527020672413e-05,
+      "loss": 0.3182,
+      "step": 60
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.20069947838783264,
+      "learning_rate": 9.980184847454052e-05,
+      "loss": 0.3954,
+      "step": 61
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.17549441754817963,
+      "learning_rate": 9.978795707151492e-05,
+      "loss": 0.3074,
+      "step": 62
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.1613418161869049,
+      "learning_rate": 9.977359612865423e-05,
+      "loss": 0.3025,
+      "step": 63
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.17346088588237762,
+      "learning_rate": 9.975876578139355e-05,
+      "loss": 0.343,
+      "step": 64
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.18235662579536438,
+      "learning_rate": 9.974346616959476e-05,
+      "loss": 0.4583,
+      "step": 65
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.15650813281536102,
+      "learning_rate": 9.972769743754531e-05,
+      "loss": 0.3005,
+      "step": 66
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.13963809609413147,
+      "learning_rate": 9.971145973395684e-05,
+      "loss": 0.2472,
+      "step": 67
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.13272036612033844,
+      "learning_rate": 9.969475321196373e-05,
+      "loss": 0.2145,
+      "step": 68
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.14247752726078033,
+      "learning_rate": 9.96775780291217e-05,
+      "loss": 0.3014,
+      "step": 69
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.1368098258972168,
+      "learning_rate": 9.965993434740633e-05,
+      "loss": 0.2801,
+      "step": 70
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.16953808069229126,
+      "learning_rate": 9.96418223332115e-05,
+      "loss": 0.3398,
+      "step": 71
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.13653875887393951,
+      "learning_rate": 9.962324215734782e-05,
+      "loss": 0.2038,
+      "step": 72
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.1879628598690033,
+      "learning_rate": 9.960419399504106e-05,
+      "loss": 0.3986,
+      "step": 73
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.13999062776565552,
+      "learning_rate": 9.958467802593046e-05,
+      "loss": 0.2681,
+      "step": 74
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.14363595843315125,
+      "learning_rate": 9.956469443406706e-05,
+      "loss": 0.3083,
+      "step": 75
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.12918664515018463,
+      "learning_rate": 9.954424340791196e-05,
+      "loss": 0.2932,
+      "step": 76
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.18551352620124817,
+      "learning_rate": 9.952332514033447e-05,
+      "loss": 0.3485,
+      "step": 77
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.15228018164634705,
+      "learning_rate": 9.950193982861047e-05,
+      "loss": 0.3871,
+      "step": 78
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.17414754629135132,
+      "learning_rate": 9.948008767442034e-05,
+      "loss": 0.331,
+      "step": 79
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.1374177485704422,
+      "learning_rate": 9.94577688838472e-05,
+      "loss": 0.291,
+      "step": 80
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.16502645611763,
+      "learning_rate": 9.943498366737487e-05,
+      "loss": 0.3643,
+      "step": 81
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.15110145509243011,
+      "learning_rate": 9.941173223988602e-05,
+      "loss": 0.3078,
+      "step": 82
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.19556356966495514,
+      "learning_rate": 9.938801482065998e-05,
+      "loss": 0.3402,
+      "step": 83
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.17671266198158264,
+      "learning_rate": 9.93638316333708e-05,
+      "loss": 0.39,
+      "step": 84
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.15854957699775696,
+      "learning_rate": 9.933918290608509e-05,
+      "loss": 0.3322,
+      "step": 85
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.161972314119339,
+      "learning_rate": 9.93140688712598e-05,
+      "loss": 0.3564,
+      "step": 86
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.18441465497016907,
+      "learning_rate": 9.928848976574019e-05,
+      "loss": 0.4031,
+      "step": 87
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.13918839395046234,
+      "learning_rate": 9.926244583075748e-05,
+      "loss": 0.3174,
+      "step": 88
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.1949048787355423,
+      "learning_rate": 9.923593731192655e-05,
+      "loss": 0.3629,
+      "step": 89
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.15310326218605042,
+      "learning_rate": 9.920896445924373e-05,
+      "loss": 0.3054,
+      "step": 90
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.1382163166999817,
+      "learning_rate": 9.918152752708438e-05,
+      "loss": 0.1912,
+      "step": 91
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.16869135200977325,
+      "learning_rate": 9.915362677420044e-05,
+      "loss": 0.3724,
+      "step": 92
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.17062540352344513,
+      "learning_rate": 9.912526246371814e-05,
+      "loss": 0.4316,
+      "step": 93
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.17472337186336517,
+      "learning_rate": 9.909643486313533e-05,
+      "loss": 0.3693,
+      "step": 94
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.1831342577934265,
+      "learning_rate": 9.906714424431913e-05,
+      "loss": 0.3863,
+      "step": 95
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.16841727495193481,
+      "learning_rate": 9.903739088350325e-05,
+      "loss": 0.2671,
+      "step": 96
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.1663845330476761,
+      "learning_rate": 9.900717506128541e-05,
+      "loss": 0.3058,
+      "step": 97
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.15532377362251282,
+      "learning_rate": 9.897649706262473e-05,
+      "loss": 0.31,
+      "step": 98
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.19599847495555878,
+      "learning_rate": 9.894535717683902e-05,
+      "loss": 0.3777,
+      "step": 99
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.16897916793823242,
+      "learning_rate": 9.891375569760204e-05,
+      "loss": 0.3445,
+      "step": 100
+    },
+    {
+      "epoch": 0.09,
+      "eval_loss": 0.31261664628982544,
+      "eval_runtime": 363.2139,
+      "eval_samples_per_second": 0.939,
+      "eval_steps_per_second": 0.237,
+      "step": 100
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.17293377220630646,
+      "learning_rate": 9.888169292294076e-05,
+      "loss": 0.2886,
+      "step": 101
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.17156994342803955,
+      "learning_rate": 9.884916915523251e-05,
+      "loss": 0.2793,
+      "step": 102
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.23175673186779022,
+      "learning_rate": 9.881618470120215e-05,
+      "loss": 0.3556,
+      "step": 103
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.18181762099266052,
+      "learning_rate": 9.87827398719192e-05,
+      "loss": 0.3706,
+      "step": 104
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.17873138189315796,
+      "learning_rate": 9.874883498279484e-05,
+      "loss": 0.2635,
+      "step": 105
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.1691795438528061,
+      "learning_rate": 9.871447035357903e-05,
+      "loss": 0.3097,
+      "step": 106
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.16361039876937866,
+      "learning_rate": 9.867964630835743e-05,
+      "loss": 0.2754,
+      "step": 107
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.15922842919826508,
+      "learning_rate": 9.864436317554831e-05,
+      "loss": 0.2901,
+      "step": 108
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.16168159246444702,
+      "learning_rate": 9.860862128789953e-05,
+      "loss": 0.2782,
+      "step": 109
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.16325147449970245,
+      "learning_rate": 9.857242098248542e-05,
+      "loss": 0.2287,
+      "step": 110
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.18361088633537292,
+      "learning_rate": 9.853576260070347e-05,
+      "loss": 0.3341,
+      "step": 111
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.1589210331439972,
+      "learning_rate": 9.849864648827125e-05,
+      "loss": 0.2797,
+      "step": 112
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.18811821937561035,
+      "learning_rate": 9.846107299522304e-05,
+      "loss": 0.3781,
+      "step": 113
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.20984318852424622,
+      "learning_rate": 9.842304247590667e-05,
+      "loss": 0.4218,
+      "step": 114
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.16862577199935913,
+      "learning_rate": 9.838455528897998e-05,
+      "loss": 0.2605,
+      "step": 115
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.15139126777648926,
+      "learning_rate": 9.834561179740762e-05,
+      "loss": 0.2669,
+      "step": 116
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.18524199724197388,
+      "learning_rate": 9.830621236845754e-05,
+      "loss": 0.2994,
+      "step": 117
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.17054533958435059,
+      "learning_rate": 9.826635737369751e-05,
+      "loss": 0.3163,
+      "step": 118
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.1616528034210205,
+      "learning_rate": 9.82260471889917e-05,
+      "loss": 0.2628,
+      "step": 119
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.16226598620414734,
+      "learning_rate": 9.818528219449705e-05,
+      "loss": 0.2578,
+      "step": 120
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.1630934476852417,
+      "learning_rate": 9.814406277465968e-05,
+      "loss": 0.2654,
+      "step": 121
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.1647336632013321,
+      "learning_rate": 9.810238931821138e-05,
+      "loss": 0.3335,
+      "step": 122
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.18229910731315613,
+      "learning_rate": 9.806026221816581e-05,
+      "loss": 0.405,
+      "step": 123
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.166782945394516,
+      "learning_rate": 9.801768187181486e-05,
+      "loss": 0.3577,
+      "step": 124
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.15582101047039032,
+      "learning_rate": 9.797464868072488e-05,
+      "loss": 0.2554,
+      "step": 125
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.17277398705482483,
+      "learning_rate": 9.793116305073292e-05,
+      "loss": 0.3231,
+      "step": 126
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.18093496561050415,
+      "learning_rate": 9.788722539194291e-05,
+      "loss": 0.3783,
+      "step": 127
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.18339870870113373,
+      "learning_rate": 9.784283611872169e-05,
+      "loss": 0.311,
+      "step": 128
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.16993552446365356,
+      "learning_rate": 9.77979956496953e-05,
+      "loss": 0.3498,
+      "step": 129
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.1930403858423233,
+      "learning_rate": 9.77527044077448e-05,
+      "loss": 0.4549,
+      "step": 130
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.17571842670440674,
+      "learning_rate": 9.770696282000244e-05,
+      "loss": 0.3548,
+      "step": 131
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.18186169862747192,
+      "learning_rate": 9.766077131784763e-05,
+      "loss": 0.4074,
+      "step": 132
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.15717683732509613,
+      "learning_rate": 9.761413033690276e-05,
+      "loss": 0.3097,
+      "step": 133
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.16943234205245972,
+      "learning_rate": 9.756704031702918e-05,
+      "loss": 0.3589,
+      "step": 134
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.15487098693847656,
+      "learning_rate": 9.751950170232309e-05,
+      "loss": 0.2785,
+      "step": 135
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.1598738133907318,
+      "learning_rate": 9.74715149411112e-05,
+      "loss": 0.3288,
+      "step": 136
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.1726771891117096,
+      "learning_rate": 9.742308048594665e-05,
+      "loss": 0.3682,
+      "step": 137
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.11256855726242065,
+      "learning_rate": 9.73741987936047e-05,
+      "loss": 0.163,
+      "step": 138
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.12890499830245972,
+      "learning_rate": 9.732487032507837e-05,
+      "loss": 0.2296,
+      "step": 139
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.18539537489414215,
+      "learning_rate": 9.727509554557417e-05,
+      "loss": 0.3892,
+      "step": 140
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.16778677701950073,
+      "learning_rate": 9.722487492450764e-05,
+      "loss": 0.277,
+      "step": 141
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.18062999844551086,
+      "learning_rate": 9.717420893549902e-05,
+      "loss": 0.3701,
+      "step": 142
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.1723717451095581,
+      "learning_rate": 9.712309805636863e-05,
+      "loss": 0.3018,
+      "step": 143
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.17957985401153564,
+      "learning_rate": 9.707154276913256e-05,
+      "loss": 0.3784,
+      "step": 144
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.17122071981430054,
+      "learning_rate": 9.701954355999792e-05,
+      "loss": 0.3666,
+      "step": 145
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.1683095395565033,
+      "learning_rate": 9.696710091935841e-05,
+      "loss": 0.3231,
+      "step": 146
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.15086056292057037,
+      "learning_rate": 9.691421534178966e-05,
+      "loss": 0.2831,
+      "step": 147
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.1477445662021637,
+      "learning_rate": 9.686088732604449e-05,
+      "loss": 0.2993,
+      "step": 148
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.1293020248413086,
+      "learning_rate": 9.68071173750483e-05,
+      "loss": 0.2241,
+      "step": 149
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.16265176236629486,
+      "learning_rate": 9.675290599589429e-05,
+      "loss": 0.3636,
+      "step": 150
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.1667158454656601,
+      "learning_rate": 9.669825369983865e-05,
+      "loss": 0.273,
+      "step": 151
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.14929796755313873,
+      "learning_rate": 9.664316100229578e-05,
+      "loss": 0.3321,
+      "step": 152
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.1772039234638214,
+      "learning_rate": 9.658762842283343e-05,
+      "loss": 0.3396,
+      "step": 153
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.14490586519241333,
+      "learning_rate": 9.653165648516776e-05,
+      "loss": 0.2598,
+      "step": 154
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.19450931251049042,
+      "learning_rate": 9.647524571715843e-05,
+      "loss": 0.374,
+      "step": 155
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.1437874436378479,
+      "learning_rate": 9.641839665080363e-05,
+      "loss": 0.2915,
+      "step": 156
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.1743856966495514,
+      "learning_rate": 9.636110982223505e-05,
+      "loss": 0.2875,
+      "step": 157
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.15949201583862305,
+      "learning_rate": 9.630338577171282e-05,
+      "loss": 0.317,
+      "step": 158
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.15031678974628448,
+      "learning_rate": 9.624522504362039e-05,
+      "loss": 0.2868,
+      "step": 159
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.16286946833133698,
+      "learning_rate": 9.618662818645948e-05,
+      "loss": 0.3244,
+      "step": 160
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.1371241956949234,
+      "learning_rate": 9.612759575284483e-05,
+      "loss": 0.2099,
+      "step": 161
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.1905253380537033,
+      "learning_rate": 9.606812829949896e-05,
+      "loss": 0.308,
+      "step": 162
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.1650477796792984,
+      "learning_rate": 9.600822638724705e-05,
+      "loss": 0.2766,
+      "step": 163
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.1493254005908966,
+      "learning_rate": 9.594789058101153e-05,
+      "loss": 0.2525,
+      "step": 164
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.15554015338420868,
+      "learning_rate": 9.588712144980681e-05,
+      "loss": 0.3232,
+      "step": 165
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.17346766591072083,
+      "learning_rate": 9.582591956673387e-05,
+      "loss": 0.3218,
+      "step": 166
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.17289134860038757,
+      "learning_rate": 9.576428550897489e-05,
+      "loss": 0.2463,
+      "step": 167
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.16859036684036255,
+      "learning_rate": 9.570221985778785e-05,
+      "loss": 0.3326,
+      "step": 168
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.16982051730155945,
+      "learning_rate": 9.563972319850092e-05,
+      "loss": 0.2987,
+      "step": 169
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.12437541782855988,
+      "learning_rate": 9.557679612050708e-05,
+      "loss": 0.2082,
+      "step": 170
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.1824512481689453,
+      "learning_rate": 9.551343921725843e-05,
+      "loss": 0.3735,
+      "step": 171
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.21117272973060608,
+      "learning_rate": 9.544965308626074e-05,
+      "loss": 0.4558,
+      "step": 172
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.14839278161525726,
+      "learning_rate": 9.538543832906773e-05,
+      "loss": 0.2664,
+      "step": 173
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.16749627888202667,
+      "learning_rate": 9.53207955512753e-05,
+      "loss": 0.3023,
+      "step": 174
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.18060199916362762,
+      "learning_rate": 9.525572536251607e-05,
+      "loss": 0.4061,
+      "step": 175
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.1660703420639038,
+      "learning_rate": 9.519022837645337e-05,
+      "loss": 0.3268,
+      "step": 176
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.16949373483657837,
+      "learning_rate": 9.512430521077565e-05,
+      "loss": 0.2895,
+      "step": 177
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.1692509651184082,
+      "learning_rate": 9.505795648719048e-05,
+      "loss": 0.3315,
+      "step": 178
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.18185889720916748,
+      "learning_rate": 9.499118283141887e-05,
+      "loss": 0.3509,
+      "step": 179
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.14933206140995026,
+      "learning_rate": 9.492398487318921e-05,
+      "loss": 0.3104,
+      "step": 180
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.15998147428035736,
+      "learning_rate": 9.485636324623146e-05,
+      "loss": 0.2588,
+      "step": 181
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.17629262804985046,
+      "learning_rate": 9.478831858827104e-05,
+      "loss": 0.3526,
+      "step": 182
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.17032553255558014,
+      "learning_rate": 9.471985154102292e-05,
+      "loss": 0.3241,
+      "step": 183
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.22255390882492065,
+      "learning_rate": 9.465096275018556e-05,
+      "loss": 0.4323,
+      "step": 184
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.2063310295343399,
+      "learning_rate": 9.458165286543476e-05,
+      "loss": 0.3328,
+      "step": 185
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.1893690526485443,
+      "learning_rate": 9.451192254041758e-05,
+      "loss": 0.3388,
+      "step": 186
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.19564178586006165,
+      "learning_rate": 9.444177243274618e-05,
+      "loss": 0.4384,
+      "step": 187
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.18728984892368317,
+      "learning_rate": 9.437120320399158e-05,
+      "loss": 0.3444,
+      "step": 188
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.16391535103321075,
+      "learning_rate": 9.430021551967744e-05,
+      "loss": 0.2808,
+      "step": 189
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.16459453105926514,
+      "learning_rate": 9.422881004927381e-05,
+      "loss": 0.2735,
+      "step": 190
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.17424438893795013,
+      "learning_rate": 9.415698746619079e-05,
+      "loss": 0.3394,
+      "step": 191
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.17163540422916412,
+      "learning_rate": 9.408474844777217e-05,
+      "loss": 0.2553,
+      "step": 192
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.20092585682868958,
+      "learning_rate": 9.401209367528908e-05,
+      "loss": 0.4694,
+      "step": 193
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.17325465381145477,
+      "learning_rate": 9.393902383393347e-05,
+      "loss": 0.3004,
+      "step": 194
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.23620380461215973,
+      "learning_rate": 9.386553961281179e-05,
+      "loss": 0.4225,
+      "step": 195
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.17715439200401306,
+      "learning_rate": 9.379164170493844e-05,
+      "loss": 0.3612,
+      "step": 196
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.19171035289764404,
+      "learning_rate": 9.371733080722911e-05,
+      "loss": 0.3455,
+      "step": 197
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.14322207868099213,
+      "learning_rate": 9.36426076204944e-05,
+      "loss": 0.2158,
+      "step": 198
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.18233761191368103,
+      "learning_rate": 9.35674728494331e-05,
+      "loss": 0.3489,
+      "step": 199
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.1562356948852539,
+      "learning_rate": 9.349192720262555e-05,
+      "loss": 0.2929,
+      "step": 200
+    },
+    {
+      "epoch": 0.19,
+      "eval_loss": 0.30594682693481445,
+      "eval_runtime": 359.1561,
+      "eval_samples_per_second": 0.949,
+      "eval_steps_per_second": 0.239,
+      "step": 200
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.16633079946041107,
+      "learning_rate": 9.341597139252698e-05,
+      "loss": 0.3186,
+      "step": 201
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.16485023498535156,
+      "learning_rate": 9.333960613546079e-05,
+      "loss": 0.2947,
+      "step": 202
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.16471706330776215,
+      "learning_rate": 9.326283215161178e-05,
+      "loss": 0.2635,
+      "step": 203
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.13085497915744781,
+      "learning_rate": 9.31856501650194e-05,
+      "loss": 0.1591,
+      "step": 204
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.1691998988389969,
+      "learning_rate": 9.310806090357082e-05,
+      "loss": 0.3602,
+      "step": 205
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.1322520673274994,
+      "learning_rate": 9.303006509899418e-05,
+      "loss": 0.1798,
+      "step": 206
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.17757368087768555,
+      "learning_rate": 9.29516634868517e-05,
+      "loss": 0.3225,
+      "step": 207
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.1936919093132019,
+      "learning_rate": 9.287285680653254e-05,
+      "loss": 0.357,
+      "step": 208
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.15838877856731415,
+      "learning_rate": 9.279364580124614e-05,
+      "loss": 0.3097,
+      "step": 209
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.1514427661895752,
+      "learning_rate": 9.27140312180149e-05,
+      "loss": 0.2339,
+      "step": 210
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.20150162279605865,
+      "learning_rate": 9.263401380766739e-05,
+      "loss": 0.3936,
+      "step": 211
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.18278218805789948,
+      "learning_rate": 9.255359432483105e-05,
+      "loss": 0.3434,
+      "step": 212
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.17064377665519714,
+      "learning_rate": 9.247277352792533e-05,
+      "loss": 0.261,
+      "step": 213
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.17509956657886505,
+      "learning_rate": 9.239155217915422e-05,
+      "loss": 0.3588,
+      "step": 214
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.14489436149597168,
+      "learning_rate": 9.230993104449939e-05,
+      "loss": 0.2207,
+      "step": 215
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.14888685941696167,
+      "learning_rate": 9.222791089371265e-05,
+      "loss": 0.2994,
+      "step": 216
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.15331536531448364,
+      "learning_rate": 9.214549250030899e-05,
+      "loss": 0.2647,
+      "step": 217
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.17798304557800293,
+      "learning_rate": 9.206267664155907e-05,
+      "loss": 0.3989,
+      "step": 218
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.13959749042987823,
+      "learning_rate": 9.197946409848194e-05,
+      "loss": 0.2351,
+      "step": 219
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.14366888999938965,
+      "learning_rate": 9.189585565583778e-05,
+      "loss": 0.2906,
+      "step": 220
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.15726351737976074,
+      "learning_rate": 9.181185210212034e-05,
+      "loss": 0.3184,
+      "step": 221
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.15729475021362305,
+      "learning_rate": 9.172745422954961e-05,
+      "loss": 0.2635,
+      "step": 222
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.16062325239181519,
+      "learning_rate": 9.164266283406433e-05,
+      "loss": 0.3381,
+      "step": 223
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.16578790545463562,
+      "learning_rate": 9.155747871531444e-05,
+      "loss": 0.3235,
+      "step": 224
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.18538223206996918,
+      "learning_rate": 9.14719026766536e-05,
+      "loss": 0.3675,
+      "step": 225
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.16532543301582336,
+      "learning_rate": 9.13859355251316e-05,
+      "loss": 0.2875,
+      "step": 226
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.1482064127922058,
+      "learning_rate": 9.129957807148666e-05,
+      "loss": 0.2316,
+      "step": 227
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.17890916764736176,
+      "learning_rate": 9.121283113013795e-05,
+      "loss": 0.3895,
+      "step": 228
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.13888315856456757,
+      "learning_rate": 9.112569551917774e-05,
+      "loss": 0.2447,
+      "step": 229
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.17399254441261292,
+      "learning_rate": 9.103817206036382e-05,
+      "loss": 0.3667,
+      "step": 230
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.1870443969964981,
+      "learning_rate": 9.095026157911166e-05,
+      "loss": 0.3447,
+      "step": 231
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.17662768065929413,
+      "learning_rate": 9.086196490448668e-05,
+      "loss": 0.2666,
+      "step": 232
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.1861211210489273,
+      "learning_rate": 9.077328286919638e-05,
+      "loss": 0.3676,
+      "step": 233
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.18413136899471283,
+      "learning_rate": 9.068421630958254e-05,
+      "loss": 0.3718,
+      "step": 234
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.1329324096441269,
+      "learning_rate": 9.059476606561328e-05,
+      "loss": 0.226,
+      "step": 235
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.16132499277591705,
+      "learning_rate": 9.050493298087523e-05,
+      "loss": 0.3458,
+      "step": 236
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.1686651110649109,
+      "learning_rate": 9.041471790256542e-05,
+      "loss": 0.3162,
+      "step": 237
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.1615574061870575,
+      "learning_rate": 9.032412168148345e-05,
+      "loss": 0.3162,
+      "step": 238
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.18377281725406647,
+      "learning_rate": 9.02331451720234e-05,
+      "loss": 0.3404,
+      "step": 239
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.13473980128765106,
+      "learning_rate": 9.014178923216572e-05,
+      "loss": 0.2299,
+      "step": 240
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.15975143015384674,
+      "learning_rate": 9.005005472346924e-05,
+      "loss": 0.259,
+      "step": 241
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.17034611105918884,
+      "learning_rate": 8.995794251106295e-05,
+      "loss": 0.3683,
+      "step": 242
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.16486890614032745,
+      "learning_rate": 8.986545346363792e-05,
+      "loss": 0.3102,
+      "step": 243
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.1553836464881897,
+      "learning_rate": 8.977258845343905e-05,
+      "loss": 0.3179,
+      "step": 244
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.15077827870845795,
+      "learning_rate": 8.967934835625689e-05,
+      "loss": 0.2789,
+      "step": 245
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.19002769887447357,
+      "learning_rate": 8.958573405141932e-05,
+      "loss": 0.3379,
+      "step": 246
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.18748913705348969,
+      "learning_rate": 8.949174642178332e-05,
+      "loss": 0.3834,
+      "step": 247
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.17668350040912628,
+      "learning_rate": 8.939738635372664e-05,
+      "loss": 0.3183,
+      "step": 248
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.19625459611415863,
+      "learning_rate": 8.930265473713938e-05,
+      "loss": 0.3684,
+      "step": 249
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.17532126605510712,
+      "learning_rate": 8.920755246541563e-05,
+      "loss": 0.3553,
+      "step": 250
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.16110257804393768,
+      "learning_rate": 8.911208043544513e-05,
+      "loss": 0.3126,
+      "step": 251
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.15233579277992249,
+      "learning_rate": 8.90162395476046e-05,
+      "loss": 0.2948,
+      "step": 252
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.19746264815330505,
+      "learning_rate": 8.89200307057495e-05,
+      "loss": 0.463,
+      "step": 253
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.1695135235786438,
+      "learning_rate": 8.882345481720533e-05,
+      "loss": 0.3773,
+      "step": 254
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.16552560031414032,
+      "learning_rate": 8.872651279275916e-05,
+      "loss": 0.3222,
+      "step": 255
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.17647133767604828,
+      "learning_rate": 8.862920554665098e-05,
+      "loss": 0.3102,
+      "step": 256
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.15897108614444733,
+      "learning_rate": 8.853153399656513e-05,
+      "loss": 0.3019,
+      "step": 257
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.17882820963859558,
+      "learning_rate": 8.843349906362163e-05,
+      "loss": 0.3806,
+      "step": 258
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.19276972115039825,
+      "learning_rate": 8.833510167236747e-05,
+      "loss": 0.4031,
+      "step": 259
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.1630733758211136,
+      "learning_rate": 8.823634275076791e-05,
+      "loss": 0.3392,
+      "step": 260
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.16516664624214172,
+      "learning_rate": 8.813722323019773e-05,
+      "loss": 0.3851,
+      "step": 261
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.15051980316638947,
+      "learning_rate": 8.803774404543247e-05,
+      "loss": 0.2463,
+      "step": 262
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.15737298130989075,
+      "learning_rate": 8.793790613463955e-05,
+      "loss": 0.3288,
+      "step": 263
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.12310407310724258,
+      "learning_rate": 8.783771043936948e-05,
+      "loss": 0.2211,
+      "step": 264
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.16717009246349335,
+      "learning_rate": 8.773715790454695e-05,
+      "loss": 0.2936,
+      "step": 265
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.14910729229450226,
+      "learning_rate": 8.763624947846195e-05,
+      "loss": 0.2257,
+      "step": 266
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.17754556238651276,
+      "learning_rate": 8.753498611276079e-05,
+      "loss": 0.3535,
+      "step": 267
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.18893520534038544,
+      "learning_rate": 8.743336876243712e-05,
+      "loss": 0.3989,
+      "step": 268
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.20323459804058075,
+      "learning_rate": 8.7331398385823e-05,
+      "loss": 0.4341,
+      "step": 269
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.19121398031711578,
+      "learning_rate": 8.722907594457974e-05,
+      "loss": 0.3873,
+      "step": 270
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.1796387881040573,
+      "learning_rate": 8.712640240368899e-05,
+      "loss": 0.3972,
+      "step": 271
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.18387527763843536,
+      "learning_rate": 8.702337873144343e-05,
+      "loss": 0.2989,
+      "step": 272
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.151773139834404,
+      "learning_rate": 8.692000589943785e-05,
+      "loss": 0.2631,
+      "step": 273
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.18373416364192963,
+      "learning_rate": 8.681628488255986e-05,
+      "loss": 0.3892,
+      "step": 274
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.14465570449829102,
+      "learning_rate": 8.671221665898073e-05,
+      "loss": 0.2657,
+      "step": 275
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.1640452742576599,
+      "learning_rate": 8.660780221014616e-05,
+      "loss": 0.3306,
+      "step": 276
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.1678406149148941,
+      "learning_rate": 8.650304252076703e-05,
+      "loss": 0.3193,
+      "step": 277
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.14313830435276031,
+      "learning_rate": 8.639793857881011e-05,
+      "loss": 0.2583,
+      "step": 278
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.18983900547027588,
+      "learning_rate": 8.629249137548874e-05,
+      "loss": 0.3715,
+      "step": 279
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.15354441106319427,
+      "learning_rate": 8.618670190525352e-05,
+      "loss": 0.2644,
+      "step": 280
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.16019406914710999,
+      "learning_rate": 8.608057116578283e-05,
+      "loss": 0.2849,
+      "step": 281
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.17929953336715698,
+      "learning_rate": 8.597410015797358e-05,
+      "loss": 0.397,
+      "step": 282
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.13921594619750977,
+      "learning_rate": 8.586728988593158e-05,
+      "loss": 0.2252,
+      "step": 283
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.1677287220954895,
+      "learning_rate": 8.576014135696226e-05,
+      "loss": 0.3246,
+      "step": 284
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.15673357248306274,
+      "learning_rate": 8.565265558156101e-05,
+      "loss": 0.2795,
+      "step": 285
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.1804131716489792,
+      "learning_rate": 8.554483357340379e-05,
+      "loss": 0.347,
+      "step": 286
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.13836736977100372,
+      "learning_rate": 8.543667634933742e-05,
+      "loss": 0.2477,
+      "step": 287
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.15196296572685242,
+      "learning_rate": 8.532818492937014e-05,
+      "loss": 0.2784,
+      "step": 288
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.15890322625637054,
+      "learning_rate": 8.521936033666187e-05,
+      "loss": 0.2866,
+      "step": 289
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.17324712872505188,
+      "learning_rate": 8.511020359751466e-05,
+      "loss": 0.3181,
+      "step": 290
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.12436319142580032,
+      "learning_rate": 8.500071574136295e-05,
+      "loss": 0.2089,
+      "step": 291
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.16586050391197205,
+      "learning_rate": 8.489089780076386e-05,
+      "loss": 0.285,
+      "step": 292
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.16333432495594025,
+      "learning_rate": 8.478075081138745e-05,
+      "loss": 0.2924,
+      "step": 293
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.194902241230011,
+      "learning_rate": 8.467027581200702e-05,
+      "loss": 0.3929,
+      "step": 294
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.17525450885295868,
+      "learning_rate": 8.455947384448924e-05,
+      "loss": 0.3437,
+      "step": 295
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.1408596783876419,
+      "learning_rate": 8.444834595378434e-05,
+      "loss": 0.2465,
+      "step": 296
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.16492432355880737,
+      "learning_rate": 8.433689318791626e-05,
+      "loss": 0.3292,
+      "step": 297
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.1690223217010498,
+      "learning_rate": 8.422511659797279e-05,
+      "loss": 0.3061,
+      "step": 298
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.15156665444374084,
+      "learning_rate": 8.411301723809563e-05,
+      "loss": 0.2755,
+      "step": 299
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.1627221703529358,
+      "learning_rate": 8.400059616547047e-05,
+      "loss": 0.3216,
+      "step": 300
+    },
+    {
+      "epoch": 0.28,
+      "eval_loss": 0.3006282448768616,
+      "eval_runtime": 354.7293,
+      "eval_samples_per_second": 0.961,
+      "eval_steps_per_second": 0.242,
+      "step": 300
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.1748882681131363,
+      "learning_rate": 8.388785444031695e-05,
+      "loss": 0.372,
+      "step": 301
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.16050395369529724,
+      "learning_rate": 8.377479312587879e-05,
+      "loss": 0.268,
+      "step": 302
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.15824644267559052,
+      "learning_rate": 8.366141328841366e-05,
+      "loss": 0.3367,
+      "step": 303
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.14442604780197144,
+      "learning_rate": 8.354771599718312e-05,
+      "loss": 0.2645,
+      "step": 304
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.17349877953529358,
+      "learning_rate": 8.343370232444261e-05,
+      "loss": 0.3432,
+      "step": 305
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.15489909052848816,
+      "learning_rate": 8.331937334543132e-05,
+      "loss": 0.2906,
+      "step": 306
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.20159928500652313,
+      "learning_rate": 8.320473013836196e-05,
+      "loss": 0.4224,
+      "step": 307
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.17371705174446106,
+      "learning_rate": 8.308977378441072e-05,
+      "loss": 0.4127,
+      "step": 308
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.15670941770076752,
+      "learning_rate": 8.297450536770697e-05,
+      "loss": 0.2912,
+      "step": 309
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.16518990695476532,
+      "learning_rate": 8.28589259753231e-05,
+      "loss": 0.2871,
+      "step": 310
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.17717592418193817,
+      "learning_rate": 8.274303669726426e-05,
+      "loss": 0.3291,
+      "step": 311
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.15787926316261292,
+      "learning_rate": 8.262683862645804e-05,
+      "loss": 0.3047,
+      "step": 312
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.1727159172296524,
+      "learning_rate": 8.251033285874421e-05,
+      "loss": 0.3429,
+      "step": 313
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.17631886899471283,
+      "learning_rate": 8.239352049286435e-05,
+      "loss": 0.3178,
+      "step": 314
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.17020989954471588,
+      "learning_rate": 8.22764026304515e-05,
+      "loss": 0.3241,
+      "step": 315
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.17571833729743958,
+      "learning_rate": 8.21589803760198e-05,
+      "loss": 0.3343,
+      "step": 316
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.1662856936454773,
+      "learning_rate": 8.204125483695403e-05,
+      "loss": 0.3218,
+      "step": 317
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.15965956449508667,
+      "learning_rate": 8.192322712349917e-05,
+      "loss": 0.2543,
+      "step": 318
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.16337652504444122,
+      "learning_rate": 8.180489834875e-05,
+      "loss": 0.3056,
+      "step": 319
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.15989547967910767,
+      "learning_rate": 8.168626962864045e-05,
+      "loss": 0.2647,
+      "step": 320
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.18423987925052643,
+      "learning_rate": 8.156734208193327e-05,
+      "loss": 0.3181,
+      "step": 321
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1749279499053955,
+      "learning_rate": 8.144811683020932e-05,
+      "loss": 0.2939,
+      "step": 322
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.16700512170791626,
+      "learning_rate": 8.132859499785707e-05,
+      "loss": 0.3293,
+      "step": 323
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.15745440125465393,
+      "learning_rate": 8.120877771206201e-05,
+      "loss": 0.3074,
+      "step": 324
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.19622637331485748,
+      "learning_rate": 8.108866610279593e-05,
+      "loss": 0.3966,
+      "step": 325
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1764141470193863,
+      "learning_rate": 8.096826130280639e-05,
+      "loss": 0.3099,
+      "step": 326
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.16999737918376923,
+      "learning_rate": 8.08475644476059e-05,
+      "loss": 0.3663,
+      "step": 327
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.16933573782444,
+      "learning_rate": 8.072657667546136e-05,
+      "loss": 0.3762,
+      "step": 328
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.17103281617164612,
+      "learning_rate": 8.060529912738315e-05,
+      "loss": 0.37,
+      "step": 329
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1775260865688324,
+      "learning_rate": 8.048373294711455e-05,
+      "loss": 0.3327,
+      "step": 330
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.16094951331615448,
+      "learning_rate": 8.036187928112087e-05,
+      "loss": 0.3096,
+      "step": 331
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.17832496762275696,
+      "learning_rate": 8.023973927857857e-05,
+      "loss": 0.3979,
+      "step": 332
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.20816633105278015,
+      "learning_rate": 8.011731409136454e-05,
+      "loss": 0.3812,
+      "step": 333
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.16785018146038055,
+      "learning_rate": 7.99946048740452e-05,
+      "loss": 0.2534,
+      "step": 334
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.1813991665840149,
+      "learning_rate": 7.987161278386554e-05,
+      "loss": 0.3782,
+      "step": 335
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.14834709465503693,
+      "learning_rate": 7.974833898073832e-05,
+      "loss": 0.3042,
+      "step": 336
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.15628859400749207,
+      "learning_rate": 7.962478462723307e-05,
+      "loss": 0.3429,
+      "step": 337
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.16038833558559418,
+      "learning_rate": 7.950095088856508e-05,
+      "loss": 0.3417,
+      "step": 338
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.1356244832277298,
+      "learning_rate": 7.937683893258454e-05,
+      "loss": 0.2192,
+      "step": 339
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.18800334632396698,
+      "learning_rate": 7.925244992976538e-05,
+      "loss": 0.3934,
+      "step": 340
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.14651575684547424,
+      "learning_rate": 7.912778505319436e-05,
+      "loss": 0.322,
+      "step": 341
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.12999705970287323,
+      "learning_rate": 7.900284547855991e-05,
+      "loss": 0.1922,
+      "step": 342
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.18193797767162323,
+      "learning_rate": 7.88776323841411e-05,
+      "loss": 0.342,
+      "step": 343
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.21355745196342468,
+      "learning_rate": 7.875214695079647e-05,
+      "loss": 0.3832,
+      "step": 344
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.1443042755126953,
+      "learning_rate": 7.862639036195298e-05,
+      "loss": 0.2474,
+      "step": 345
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.18863625824451447,
+      "learning_rate": 7.850036380359478e-05,
+      "loss": 0.3491,
+      "step": 346
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.17483064532279968,
+      "learning_rate": 7.837406846425204e-05,
+      "loss": 0.3693,
+      "step": 347
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.19009007513523102,
+      "learning_rate": 7.824750553498976e-05,
+      "loss": 0.3906,
+      "step": 348
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.20057331025600433,
+      "learning_rate": 7.812067620939652e-05,
+      "loss": 0.3399,
+      "step": 349
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.16113601624965668,
+      "learning_rate": 7.799358168357323e-05,
+      "loss": 0.2819,
+      "step": 350
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.19135306775569916,
+      "learning_rate": 7.786622315612183e-05,
+      "loss": 0.366,
+      "step": 351
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.15706925094127655,
+      "learning_rate": 7.773860182813403e-05,
+      "loss": 0.2585,
+      "step": 352
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.17346857488155365,
+      "learning_rate": 7.761071890317994e-05,
+      "loss": 0.3225,
+      "step": 353
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.15920041501522064,
+      "learning_rate": 7.748257558729677e-05,
+      "loss": 0.3121,
+      "step": 354
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.15750929713249207,
+      "learning_rate": 7.735417308897736e-05,
+      "loss": 0.3587,
+      "step": 355
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.16349010169506073,
+      "learning_rate": 7.72255126191589e-05,
+      "loss": 0.3122,
+      "step": 356
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.19034729897975922,
+      "learning_rate": 7.709659539121144e-05,
+      "loss": 0.3066,
+      "step": 357
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.1531616896390915,
+      "learning_rate": 7.696742262092644e-05,
+      "loss": 0.3019,
+      "step": 358
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.14113670587539673,
+      "learning_rate": 7.683799552650534e-05,
+      "loss": 0.2285,
+      "step": 359
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.13064084947109222,
+      "learning_rate": 7.67083153285481e-05,
+      "loss": 0.2327,
+      "step": 360
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.15126869082450867,
+      "learning_rate": 7.657838325004159e-05,
+      "loss": 0.2563,
+      "step": 361
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.16526973247528076,
+      "learning_rate": 7.644820051634812e-05,
+      "loss": 0.3232,
+      "step": 362
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.15968774259090424,
+      "learning_rate": 7.63177683551939e-05,
+      "loss": 0.2859,
+      "step": 363
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.16833890974521637,
+      "learning_rate": 7.618708799665744e-05,
+      "loss": 0.2766,
+      "step": 364
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.16992513835430145,
+      "learning_rate": 7.605616067315793e-05,
+      "loss": 0.351,
+      "step": 365
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.17396047711372375,
+      "learning_rate": 7.592498761944362e-05,
+      "loss": 0.2247,
+      "step": 366
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.11733554303646088,
+      "learning_rate": 7.579357007258022e-05,
+      "loss": 0.1722,
+      "step": 367
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.1603459119796753,
+      "learning_rate": 7.56619092719392e-05,
+      "loss": 0.2589,
+      "step": 368
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.16132675111293793,
+      "learning_rate": 7.553000645918611e-05,
+      "loss": 0.1977,
+      "step": 369
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.14037099480628967,
+      "learning_rate": 7.539786287826885e-05,
+      "loss": 0.2178,
+      "step": 370
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.1363111436367035,
+      "learning_rate": 7.526547977540592e-05,
+      "loss": 0.216,
+      "step": 371
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.15175656974315643,
+      "learning_rate": 7.51328583990748e-05,
+      "loss": 0.2548,
+      "step": 372
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.168841153383255,
+      "learning_rate": 7.500000000000001e-05,
+      "loss": 0.2916,
+      "step": 373
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.1467706561088562,
+      "learning_rate": 7.486690583114136e-05,
+      "loss": 0.2048,
+      "step": 374
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.19271841645240784,
+      "learning_rate": 7.473357714768222e-05,
+      "loss": 0.3504,
+      "step": 375
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.19736838340759277,
+      "learning_rate": 7.460001520701755e-05,
+      "loss": 0.4486,
+      "step": 376
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.17552991211414337,
+      "learning_rate": 7.446622126874218e-05,
+      "loss": 0.337,
+      "step": 377
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.18846909701824188,
+      "learning_rate": 7.433219659463881e-05,
+      "loss": 0.3549,
+      "step": 378
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.17685943841934204,
+      "learning_rate": 7.41979424486662e-05,
+      "loss": 0.3594,
+      "step": 379
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.1731223464012146,
+      "learning_rate": 7.406346009694713e-05,
+      "loss": 0.2771,
+      "step": 380
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.13795405626296997,
+      "learning_rate": 7.392875080775665e-05,
+      "loss": 0.2764,
+      "step": 381
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.16157139837741852,
+      "learning_rate": 7.379381585150996e-05,
+      "loss": 0.3048,
+      "step": 382
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.15121662616729736,
+      "learning_rate": 7.365865650075046e-05,
+      "loss": 0.2874,
+      "step": 383
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.12418895214796066,
+      "learning_rate": 7.35232740301378e-05,
+      "loss": 0.2273,
+      "step": 384
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.15809480845928192,
+      "learning_rate": 7.338766971643579e-05,
+      "loss": 0.3168,
+      "step": 385
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.16992981731891632,
+      "learning_rate": 7.325184483850042e-05,
+      "loss": 0.3104,
+      "step": 386
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.13675522804260254,
+      "learning_rate": 7.311580067726783e-05,
+      "loss": 0.2231,
+      "step": 387
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.15234754979610443,
+      "learning_rate": 7.297953851574206e-05,
+      "loss": 0.2851,
+      "step": 388
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.19200845062732697,
+      "learning_rate": 7.284305963898314e-05,
+      "loss": 0.4363,
+      "step": 389
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.17089393734931946,
+      "learning_rate": 7.27063653340949e-05,
+      "loss": 0.3317,
+      "step": 390
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.16668805480003357,
+      "learning_rate": 7.25694568902128e-05,
+      "loss": 0.3262,
+      "step": 391
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.11876419931650162,
+      "learning_rate": 7.243233559849179e-05,
+      "loss": 0.2037,
+      "step": 392
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.14661411941051483,
+      "learning_rate": 7.229500275209418e-05,
+      "loss": 0.2554,
+      "step": 393
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.15959692001342773,
+      "learning_rate": 7.215745964617737e-05,
+      "loss": 0.2678,
+      "step": 394
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.16589178144931793,
+      "learning_rate": 7.201970757788172e-05,
+      "loss": 0.3295,
+      "step": 395
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.149629145860672,
+      "learning_rate": 7.188174784631824e-05,
+      "loss": 0.2354,
+      "step": 396
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.17092368006706238,
+      "learning_rate": 7.174358175255636e-05,
+      "loss": 0.2992,
+      "step": 397
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.16202566027641296,
+      "learning_rate": 7.160521059961169e-05,
+      "loss": 0.2839,
+      "step": 398
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.12684136629104614,
+      "learning_rate": 7.146663569243371e-05,
+      "loss": 0.2192,
+      "step": 399
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.18242771923542023,
+      "learning_rate": 7.132785833789344e-05,
+      "loss": 0.3784,
+      "step": 400
+    },
+    {
+      "epoch": 0.38,
+      "eval_loss": 0.29874736070632935,
+      "eval_runtime": 356.6774,
+      "eval_samples_per_second": 0.956,
+      "eval_steps_per_second": 0.241,
+      "step": 400
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.1540268063545227,
+      "learning_rate": 7.118887984477116e-05,
+      "loss": 0.2856,
+      "step": 401
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.14653688669204712,
+      "learning_rate": 7.104970152374405e-05,
+      "loss": 0.3003,
+      "step": 402
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.14971095323562622,
+      "learning_rate": 7.091032468737381e-05,
+      "loss": 0.2886,
+      "step": 403
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.17707420885562897,
+      "learning_rate": 7.077075065009433e-05,
+      "loss": 0.2881,
+      "step": 404
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.1450321227312088,
+      "learning_rate": 7.063098072819919e-05,
+      "loss": 0.2964,
+      "step": 405
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.13553856313228607,
+      "learning_rate": 7.049101623982937e-05,
+      "loss": 0.2256,
+      "step": 406
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.1460329294204712,
+      "learning_rate": 7.03508585049608e-05,
+      "loss": 0.2502,
+      "step": 407
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.16372466087341309,
+      "learning_rate": 7.021050884539177e-05,
+      "loss": 0.3391,
+      "step": 408
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.14569127559661865,
+      "learning_rate": 7.006996858473069e-05,
+      "loss": 0.2621,
+      "step": 409
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.15142247080802917,
+      "learning_rate": 6.99292390483834e-05,
+      "loss": 0.2422,
+      "step": 410
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.14533023536205292,
+      "learning_rate": 6.97883215635408e-05,
+      "loss": 0.2589,
+      "step": 411
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.1775951385498047,
+      "learning_rate": 6.964721745916631e-05,
+      "loss": 0.3276,
+      "step": 412
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.15809454023838043,
+      "learning_rate": 6.950592806598327e-05,
+      "loss": 0.2783,
+      "step": 413
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.14700832962989807,
+      "learning_rate": 6.936445471646249e-05,
+      "loss": 0.2648,
+      "step": 414
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.1538705825805664,
+      "learning_rate": 6.922279874480959e-05,
+      "loss": 0.336,
+      "step": 415
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.16012847423553467,
+      "learning_rate": 6.908096148695251e-05,
+      "loss": 0.3275,
+      "step": 416
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.1486840695142746,
+      "learning_rate": 6.89389442805288e-05,
+      "loss": 0.2706,
+      "step": 417
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.16760365664958954,
+      "learning_rate": 6.879674846487314e-05,
+      "loss": 0.3073,
+      "step": 418
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.16847726702690125,
+      "learning_rate": 6.865437538100457e-05,
+      "loss": 0.3278,
+      "step": 419
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.14246656000614166,
+      "learning_rate": 6.851182637161396e-05,
+      "loss": 0.2357,
+      "step": 420
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.15778616070747375,
+      "learning_rate": 6.836910278105124e-05,
+      "loss": 0.2936,
+      "step": 421
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.16941332817077637,
+      "learning_rate": 6.822620595531286e-05,
+      "loss": 0.3449,
+      "step": 422
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.14975254237651825,
+      "learning_rate": 6.808313724202894e-05,
+      "loss": 0.2815,
+      "step": 423
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.1396080106496811,
+      "learning_rate": 6.793989799045066e-05,
+      "loss": 0.2325,
+      "step": 424
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.1809430569410324,
+      "learning_rate": 6.779648955143754e-05,
+      "loss": 0.3065,
+      "step": 425
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.17733648419380188,
+      "learning_rate": 6.765291327744463e-05,
+      "loss": 0.328,
+      "step": 426
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.1549312025308609,
+      "learning_rate": 6.750917052250982e-05,
+      "loss": 0.2911,
+      "step": 427
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.16912095248699188,
+      "learning_rate": 6.736526264224101e-05,
+      "loss": 0.309,
+      "step": 428
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.1673160344362259,
+      "learning_rate": 6.722119099380345e-05,
+      "loss": 0.3327,
+      "step": 429
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.14743894338607788,
+      "learning_rate": 6.707695693590676e-05,
+      "loss": 0.2115,
+      "step": 430
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.14556078612804413,
+      "learning_rate": 6.693256182879225e-05,
+      "loss": 0.2715,
+      "step": 431
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.17107875645160675,
+      "learning_rate": 6.678800703422003e-05,
+      "loss": 0.347,
+      "step": 432
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.1924813836812973,
+      "learning_rate": 6.664329391545626e-05,
+      "loss": 0.3873,
+      "step": 433
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.1918533742427826,
+      "learning_rate": 6.64984238372601e-05,
+      "loss": 0.4215,
+      "step": 434
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.15770386159420013,
+      "learning_rate": 6.635339816587109e-05,
+      "loss": 0.2942,
+      "step": 435
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.20068824291229248,
+      "learning_rate": 6.620821826899606e-05,
+      "loss": 0.388,
+      "step": 436
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.18431293964385986,
+      "learning_rate": 6.606288551579629e-05,
+      "loss": 0.3363,
+      "step": 437
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.1866261065006256,
+      "learning_rate": 6.59174012768747e-05,
+      "loss": 0.3931,
+      "step": 438
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.1555357724428177,
+      "learning_rate": 6.577176692426279e-05,
+      "loss": 0.3402,
+      "step": 439
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.17006400227546692,
+      "learning_rate": 6.562598383140772e-05,
+      "loss": 0.3341,
+      "step": 440
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.16117194294929504,
+      "learning_rate": 6.548005337315944e-05,
+      "loss": 0.2928,
+      "step": 441
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.12271099537611008,
+      "learning_rate": 6.533397692575766e-05,
+      "loss": 0.1775,
+      "step": 442
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.20110560953617096,
+      "learning_rate": 6.518775586681887e-05,
+      "loss": 0.432,
+      "step": 443
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.14445729553699493,
+      "learning_rate": 6.504139157532337e-05,
+      "loss": 0.2885,
+      "step": 444
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.1538679301738739,
+      "learning_rate": 6.489488543160224e-05,
+      "loss": 0.2286,
+      "step": 445
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.13938476145267487,
+      "learning_rate": 6.47482388173244e-05,
+      "loss": 0.238,
+      "step": 446
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.17891260981559753,
+      "learning_rate": 6.460145311548341e-05,
+      "loss": 0.3711,
+      "step": 447
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.16656829416751862,
+      "learning_rate": 6.445452971038464e-05,
+      "loss": 0.3239,
+      "step": 448
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.1773829311132431,
+      "learning_rate": 6.430746998763204e-05,
+      "loss": 0.3784,
+      "step": 449
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.16422712802886963,
+      "learning_rate": 6.416027533411519e-05,
+      "loss": 0.3766,
+      "step": 450
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.20201419293880463,
+      "learning_rate": 6.401294713799618e-05,
+      "loss": 0.4286,
+      "step": 451
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.15801064670085907,
+      "learning_rate": 6.386548678869644e-05,
+      "loss": 0.2996,
+      "step": 452
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.14631867408752441,
+      "learning_rate": 6.37178956768838e-05,
+      "loss": 0.2659,
+      "step": 453
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.16085556149482727,
+      "learning_rate": 6.35701751944592e-05,
+      "loss": 0.3103,
+      "step": 454
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.15359799563884735,
+      "learning_rate": 6.342232673454371e-05,
+      "loss": 0.3311,
+      "step": 455
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.12510977685451508,
+      "learning_rate": 6.32743516914653e-05,
+      "loss": 0.204,
+      "step": 456
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.17648649215698242,
+      "learning_rate": 6.312625146074575e-05,
+      "loss": 0.354,
+      "step": 457
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.14568190276622772,
+      "learning_rate": 6.297802743908741e-05,
+      "loss": 0.2626,
+      "step": 458
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.1486312448978424,
+      "learning_rate": 6.282968102436009e-05,
+      "loss": 0.2973,
+      "step": 459
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.13261553645133972,
+      "learning_rate": 6.268121361558792e-05,
+      "loss": 0.2255,
+      "step": 460
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.1478547602891922,
+      "learning_rate": 6.253262661293604e-05,
+      "loss": 0.3431,
+      "step": 461
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.15811577439308167,
+      "learning_rate": 6.238392141769743e-05,
+      "loss": 0.2496,
+      "step": 462
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.1656562089920044,
+      "learning_rate": 6.22350994322798e-05,
+      "loss": 0.3869,
+      "step": 463
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.16209283471107483,
+      "learning_rate": 6.208616206019224e-05,
+      "loss": 0.2512,
+      "step": 464
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.14161382615566254,
+      "learning_rate": 6.193711070603202e-05,
+      "loss": 0.2572,
+      "step": 465
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.15951459109783173,
+      "learning_rate": 6.178794677547137e-05,
+      "loss": 0.2672,
+      "step": 466
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.159026637673378,
+      "learning_rate": 6.163867167524419e-05,
+      "loss": 0.3189,
+      "step": 467
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.1612398326396942,
+      "learning_rate": 6.14892868131328e-05,
+      "loss": 0.2857,
+      "step": 468
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.1422705203294754,
+      "learning_rate": 6.133979359795468e-05,
+      "loss": 0.2676,
+      "step": 469
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.16060587763786316,
+      "learning_rate": 6.119019343954914e-05,
+      "loss": 0.2783,
+      "step": 470
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.14463238418102264,
+      "learning_rate": 6.104048774876406e-05,
+      "loss": 0.3008,
+      "step": 471
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.1663757562637329,
+      "learning_rate": 6.0890677937442574e-05,
+      "loss": 0.3604,
+      "step": 472
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.210937038064003,
+      "learning_rate": 6.074076541840977e-05,
+      "loss": 0.487,
+      "step": 473
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.1401451826095581,
+      "learning_rate": 6.059075160545933e-05,
+      "loss": 0.2273,
+      "step": 474
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.1536555141210556,
+      "learning_rate": 6.0440637913340224e-05,
+      "loss": 0.2376,
+      "step": 475
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.1482958197593689,
+      "learning_rate": 6.029042575774334e-05,
+      "loss": 0.2931,
+      "step": 476
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.1802617907524109,
+      "learning_rate": 6.0140116555288204e-05,
+      "loss": 0.3287,
+      "step": 477
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.1453169733285904,
+      "learning_rate": 5.998971172350953e-05,
+      "loss": 0.314,
+      "step": 478
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.1587706059217453,
+      "learning_rate": 5.9839212680843925e-05,
+      "loss": 0.2609,
+      "step": 479
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.1905590146780014,
+      "learning_rate": 5.968862084661643e-05,
+      "loss": 0.3676,
+      "step": 480
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.14204907417297363,
+      "learning_rate": 5.953793764102722e-05,
+      "loss": 0.2973,
+      "step": 481
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.16098394989967346,
+      "learning_rate": 5.938716448513818e-05,
+      "loss": 0.2798,
+      "step": 482
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.20171226561069489,
+      "learning_rate": 5.923630280085948e-05,
+      "loss": 0.4642,
+      "step": 483
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.1643485724925995,
+      "learning_rate": 5.908535401093618e-05,
+      "loss": 0.3596,
+      "step": 484
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.1599271446466446,
+      "learning_rate": 5.8934319538934824e-05,
+      "loss": 0.2906,
+      "step": 485
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.1449955701828003,
+      "learning_rate": 5.878320080923001e-05,
+      "loss": 0.2689,
+      "step": 486
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.200435072183609,
+      "learning_rate": 5.863199924699095e-05,
+      "loss": 0.2931,
+      "step": 487
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.21762442588806152,
+      "learning_rate": 5.848071627816803e-05,
+      "loss": 0.4407,
+      "step": 488
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.18104234337806702,
+      "learning_rate": 5.8329353329479366e-05,
+      "loss": 0.3422,
+      "step": 489
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.15091653168201447,
+      "learning_rate": 5.8177911828397336e-05,
+      "loss": 0.2591,
+      "step": 490
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.17679543793201447,
+      "learning_rate": 5.802639320313514e-05,
+      "loss": 0.3619,
+      "step": 491
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.18017727136611938,
+      "learning_rate": 5.787479888263333e-05,
+      "loss": 0.3283,
+      "step": 492
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.1720423400402069,
+      "learning_rate": 5.77231302965463e-05,
+      "loss": 0.2893,
+      "step": 493
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.17164617776870728,
+      "learning_rate": 5.757138887522884e-05,
+      "loss": 0.2864,
+      "step": 494
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.14751997590065002,
+      "learning_rate": 5.7419576049722646e-05,
+      "loss": 0.2521,
+      "step": 495
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.15324720740318298,
+      "learning_rate": 5.7267693251742794e-05,
+      "loss": 0.2874,
+      "step": 496
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.1822265237569809,
+      "learning_rate": 5.7115741913664264e-05,
+      "loss": 0.2915,
+      "step": 497
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.18057996034622192,
+      "learning_rate": 5.696372346850841e-05,
+      "loss": 0.3805,
+      "step": 498
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.1679065078496933,
+      "learning_rate": 5.68116393499295e-05,
+      "loss": 0.3261,
+      "step": 499
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.14962001144886017,
+      "learning_rate": 5.6659490992201094e-05,
+      "loss": 0.2526,
+      "step": 500
+    },
+    {
+      "epoch": 0.47,
+      "eval_loss": 0.29161664843559265,
+      "eval_runtime": 355.6352,
+      "eval_samples_per_second": 0.959,
+      "eval_steps_per_second": 0.242,
+      "step": 500
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.20530852675437927,
+      "learning_rate": 5.650727983020262e-05,
+      "loss": 0.4471,
+      "step": 501
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.14180253446102142,
+      "learning_rate": 5.635500729940577e-05,
+      "loss": 0.2105,
+      "step": 502
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.2044840157032013,
+      "learning_rate": 5.620267483586105e-05,
+      "loss": 0.3889,
+      "step": 503
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.1825052797794342,
+      "learning_rate": 5.6050283876184116e-05,
+      "loss": 0.3661,
+      "step": 504
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.20767174661159515,
+      "learning_rate": 5.5897835857542317e-05,
+      "loss": 0.3642,
+      "step": 505
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.16215144097805023,
+      "learning_rate": 5.574533221764109e-05,
+      "loss": 0.2642,
+      "step": 506
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.16340447962284088,
+      "learning_rate": 5.5592774394710465e-05,
+      "loss": 0.278,
+      "step": 507
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.17503058910369873,
+      "learning_rate": 5.5440163827491454e-05,
+      "loss": 0.2906,
+      "step": 508
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.159480020403862,
+      "learning_rate": 5.5287501955222444e-05,
+      "loss": 0.2665,
+      "step": 509
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.15985313057899475,
+      "learning_rate": 5.513479021762572e-05,
+      "loss": 0.293,
+      "step": 510
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.15223486721515656,
+      "learning_rate": 5.4982030054893785e-05,
+      "loss": 0.2242,
+      "step": 511
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.16948065161705017,
+      "learning_rate": 5.482922290767589e-05,
+      "loss": 0.2513,
+      "step": 512
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.1636878401041031,
+      "learning_rate": 5.467637021706438e-05,
+      "loss": 0.2841,
+      "step": 513
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.21989518404006958,
+      "learning_rate": 5.452347342458104e-05,
+      "loss": 0.3895,
+      "step": 514
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.18184396624565125,
+      "learning_rate": 5.437053397216364e-05,
+      "loss": 0.312,
+      "step": 515
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.17834916710853577,
+      "learning_rate": 5.4217553302152237e-05,
+      "loss": 0.3103,
+      "step": 516
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.20788073539733887,
+      "learning_rate": 5.406453285727564e-05,
+      "loss": 0.3476,
+      "step": 517
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.17035901546478271,
+      "learning_rate": 5.39114740806377e-05,
+      "loss": 0.2784,
+      "step": 518
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.14679253101348877,
+      "learning_rate": 5.3758378415703825e-05,
+      "loss": 0.2407,
+      "step": 519
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.20116353034973145,
+      "learning_rate": 5.360524730628728e-05,
+      "loss": 0.4063,
+      "step": 520
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.17486828565597534,
+      "learning_rate": 5.345208219653561e-05,
+      "loss": 0.3035,
+      "step": 521
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.16789518296718597,
+      "learning_rate": 5.3298884530917014e-05,
+      "loss": 0.3335,
+      "step": 522
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.15769682824611664,
+      "learning_rate": 5.3145655754206714e-05,
+      "loss": 0.2577,
+      "step": 523
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.14827242493629456,
+      "learning_rate": 5.2992397311473316e-05,
+      "loss": 0.2084,
+      "step": 524
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.18390865623950958,
+      "learning_rate": 5.2839110648065214e-05,
+      "loss": 0.33,
+      "step": 525
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.18774594366550446,
+      "learning_rate": 5.2685797209596976e-05,
+      "loss": 0.3179,
+      "step": 526
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.18476159870624542,
+      "learning_rate": 5.2532458441935636e-05,
+      "loss": 0.3182,
+      "step": 527
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2046561986207962,
+      "learning_rate": 5.2379095791187124e-05,
+      "loss": 0.3421,
+      "step": 528
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.17056624591350555,
+      "learning_rate": 5.222571070368257e-05,
+      "loss": 0.2906,
+      "step": 529
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.18982867896556854,
+      "learning_rate": 5.2072304625964785e-05,
+      "loss": 0.271,
+      "step": 530
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.17283333837985992,
+      "learning_rate": 5.191887900477443e-05,
+      "loss": 0.3419,
+      "step": 531
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.17433173954486847,
+      "learning_rate": 5.176543528703657e-05,
+      "loss": 0.3086,
+      "step": 532
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.15211868286132812,
+      "learning_rate": 5.161197491984684e-05,
+      "loss": 0.274,
+      "step": 533
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.1734624058008194,
+      "learning_rate": 5.1458499350458e-05,
+      "loss": 0.2258,
+      "step": 534
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2026003748178482,
+      "learning_rate": 5.130501002626609e-05,
+      "loss": 0.4652,
+      "step": 535
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.17500601708889008,
+      "learning_rate": 5.11515083947969e-05,
+      "loss": 0.2742,
+      "step": 536
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.17626890540122986,
+      "learning_rate": 5.099799590369231e-05,
+      "loss": 0.3188,
+      "step": 537
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.19223853945732117,
+      "learning_rate": 5.084447400069655e-05,
+      "loss": 0.3395,
+      "step": 538
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.17484283447265625,
+      "learning_rate": 5.069094413364272e-05,
+      "loss": 0.2892,
+      "step": 539
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2019236534833908,
+      "learning_rate": 5.053740775043892e-05,
+      "loss": 0.3382,
+      "step": 540
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2101663500070572,
+      "learning_rate": 5.038386629905475e-05,
+      "loss": 0.3229,
+      "step": 541
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.18816357851028442,
+      "learning_rate": 5.023032122750759e-05,
+      "loss": 0.3378,
+      "step": 542
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.17948800325393677,
+      "learning_rate": 5.007677398384901e-05,
+      "loss": 0.2664,
+      "step": 543
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.19135427474975586,
+      "learning_rate": 4.9923226016151e-05,
+      "loss": 0.3223,
+      "step": 544
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.1658935844898224,
+      "learning_rate": 4.976967877249242e-05,
+      "loss": 0.3339,
+      "step": 545
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.1513405591249466,
+      "learning_rate": 4.961613370094527e-05,
+      "loss": 0.2547,
+      "step": 546
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.1806260049343109,
+      "learning_rate": 4.9462592249561095e-05,
+      "loss": 0.351,
+      "step": 547
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.16001088917255402,
+      "learning_rate": 4.9309055866357295e-05,
+      "loss": 0.2371,
+      "step": 548
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.1843252331018448,
+      "learning_rate": 4.915552599930345e-05,
+      "loss": 0.3452,
+      "step": 549
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.18163682520389557,
+      "learning_rate": 4.900200409630771e-05,
+      "loss": 0.2775,
+      "step": 550
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.17243829369544983,
+      "learning_rate": 4.884849160520311e-05,
+      "loss": 0.2437,
+      "step": 551
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.17966440320014954,
+      "learning_rate": 4.869498997373393e-05,
+      "loss": 0.2797,
+      "step": 552
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.20064568519592285,
+      "learning_rate": 4.8541500649542014e-05,
+      "loss": 0.3909,
+      "step": 553
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.17336438596248627,
+      "learning_rate": 4.838802508015316e-05,
+      "loss": 0.2968,
+      "step": 554
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.20279474556446075,
+      "learning_rate": 4.823456471296344e-05,
+      "loss": 0.3045,
+      "step": 555
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.16955551505088806,
+      "learning_rate": 4.8081120995225574e-05,
+      "loss": 0.2838,
+      "step": 556
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.20965775847434998,
+      "learning_rate": 4.792769537403523e-05,
+      "loss": 0.4009,
+      "step": 557
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.22277423739433289,
+      "learning_rate": 4.777428929631743e-05,
+      "loss": 0.3693,
+      "step": 558
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.21148189902305603,
+      "learning_rate": 4.762090420881289e-05,
+      "loss": 0.3701,
+      "step": 559
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.22442913055419922,
+      "learning_rate": 4.746754155806437e-05,
+      "loss": 0.4355,
+      "step": 560
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.17229177057743073,
+      "learning_rate": 4.731420279040303e-05,
+      "loss": 0.2967,
+      "step": 561
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.137137308716774,
+      "learning_rate": 4.716088935193479e-05,
+      "loss": 0.1845,
+      "step": 562
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2031993567943573,
+      "learning_rate": 4.7007602688526695e-05,
+      "loss": 0.3388,
+      "step": 563
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.1469893455505371,
+      "learning_rate": 4.6854344245793305e-05,
+      "loss": 0.2142,
+      "step": 564
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.19145740568637848,
+      "learning_rate": 4.6701115469083e-05,
+      "loss": 0.2655,
+      "step": 565
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.1763337254524231,
+      "learning_rate": 4.65479178034644e-05,
+      "loss": 0.1885,
+      "step": 566
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.17556172609329224,
+      "learning_rate": 4.6394752693712734e-05,
+      "loss": 0.2945,
+      "step": 567
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.1767197549343109,
+      "learning_rate": 4.624162158429619e-05,
+      "loss": 0.2572,
+      "step": 568
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.21139641106128693,
+      "learning_rate": 4.608852591936231e-05,
+      "loss": 0.3731,
+      "step": 569
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.20019468665122986,
+      "learning_rate": 4.593546714272438e-05,
+      "loss": 0.3419,
+      "step": 570
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.20723475515842438,
+      "learning_rate": 4.578244669784777e-05,
+      "loss": 0.3456,
+      "step": 571
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.13808375597000122,
+      "learning_rate": 4.562946602783636e-05,
+      "loss": 0.1996,
+      "step": 572
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.39727845788002014,
+      "learning_rate": 4.547652657541897e-05,
+      "loss": 0.4441,
+      "step": 573
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.18654167652130127,
+      "learning_rate": 4.532362978293564e-05,
+      "loss": 0.3799,
+      "step": 574
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.16589882969856262,
+      "learning_rate": 4.517077709232411e-05,
+      "loss": 0.2352,
+      "step": 575
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.21066059172153473,
+      "learning_rate": 4.501796994510622e-05,
+      "loss": 0.3851,
+      "step": 576
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.1282036453485489,
+      "learning_rate": 4.4865209782374304e-05,
+      "loss": 0.1942,
+      "step": 577
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.15674398839473724,
+      "learning_rate": 4.471249804477758e-05,
+      "loss": 0.2139,
+      "step": 578
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.17126213014125824,
+      "learning_rate": 4.4559836172508565e-05,
+      "loss": 0.3144,
+      "step": 579
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.20061375200748444,
+      "learning_rate": 4.4407225605289547e-05,
+      "loss": 0.2262,
+      "step": 580
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.15191729366779327,
+      "learning_rate": 4.4254667782358924e-05,
+      "loss": 0.2553,
+      "step": 581
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.18683789670467377,
+      "learning_rate": 4.410216414245771e-05,
+      "loss": 0.3661,
+      "step": 582
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.17604923248291016,
+      "learning_rate": 4.394971612381591e-05,
+      "loss": 0.2922,
+      "step": 583
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.1844872087240219,
+      "learning_rate": 4.379732516413897e-05,
+      "loss": 0.267,
+      "step": 584
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.17229197919368744,
+      "learning_rate": 4.364499270059424e-05,
+      "loss": 0.2994,
+      "step": 585
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.18299497663974762,
+      "learning_rate": 4.34927201697974e-05,
+      "loss": 0.3174,
+      "step": 586
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.1788688451051712,
+      "learning_rate": 4.334050900779893e-05,
+      "loss": 0.2826,
+      "step": 587
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.2060079276561737,
+      "learning_rate": 4.3188360650070525e-05,
+      "loss": 0.3407,
+      "step": 588
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.15371231734752655,
+      "learning_rate": 4.303627653149159e-05,
+      "loss": 0.2168,
+      "step": 589
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.16864553093910217,
+      "learning_rate": 4.288425808633575e-05,
+      "loss": 0.2415,
+      "step": 590
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.17563757300376892,
+      "learning_rate": 4.2732306748257224e-05,
+      "loss": 0.2815,
+      "step": 591
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.16491085290908813,
+      "learning_rate": 4.258042395027738e-05,
+      "loss": 0.2288,
+      "step": 592
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.19919921457767487,
+      "learning_rate": 4.2428611124771184e-05,
+      "loss": 0.2786,
+      "step": 593
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.16612334549427032,
+      "learning_rate": 4.2276869703453724e-05,
+      "loss": 0.2952,
+      "step": 594
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.16550858318805695,
+      "learning_rate": 4.21252011173667e-05,
+      "loss": 0.2522,
+      "step": 595
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.184719979763031,
+      "learning_rate": 4.1973606796864884e-05,
+      "loss": 0.2768,
+      "step": 596
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.18673363327980042,
+      "learning_rate": 4.182208817160269e-05,
+      "loss": 0.3728,
+      "step": 597
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.16201287508010864,
+      "learning_rate": 4.167064667052065e-05,
+      "loss": 0.2425,
+      "step": 598
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.18673589825630188,
+      "learning_rate": 4.151928372183198e-05,
+      "loss": 0.3056,
+      "step": 599
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.19868288934230804,
+      "learning_rate": 4.136800075300906e-05,
+      "loss": 0.3117,
+      "step": 600
+    },
+    {
+      "epoch": 0.57,
+      "eval_loss": 0.2891244888305664,
+      "eval_runtime": 366.4873,
+      "eval_samples_per_second": 0.93,
+      "eval_steps_per_second": 0.235,
+      "step": 600
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.20592208206653595,
+      "learning_rate": 4.1216799190770006e-05,
+      "loss": 0.4301,
+      "step": 601
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.1483517736196518,
+      "learning_rate": 4.1065680461065194e-05,
+      "loss": 0.2446,
+      "step": 602
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.16961555182933807,
+      "learning_rate": 4.0914645989063845e-05,
+      "loss": 0.2783,
+      "step": 603
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.18571127951145172,
+      "learning_rate": 4.076369719914055e-05,
+      "loss": 0.2804,
+      "step": 604
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.221944198012352,
+      "learning_rate": 4.061283551486185e-05,
+      "loss": 0.4535,
+      "step": 605
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.19506922364234924,
+      "learning_rate": 4.0462062358972796e-05,
+      "loss": 0.3896,
+      "step": 606
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.21058465540409088,
+      "learning_rate": 4.03113791533836e-05,
+      "loss": 0.4188,
+      "step": 607
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.17764399945735931,
+      "learning_rate": 4.016078731915608e-05,
+      "loss": 0.2872,
+      "step": 608
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.1822466403245926,
+      "learning_rate": 4.001028827649046e-05,
+      "loss": 0.3104,
+      "step": 609
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.1656925082206726,
+      "learning_rate": 3.9859883444711794e-05,
+      "loss": 0.2686,
+      "step": 610
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.15843911468982697,
+      "learning_rate": 3.970957424225666e-05,
+      "loss": 0.2582,
+      "step": 611
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.16249459981918335,
+      "learning_rate": 3.955936208665979e-05,
+      "loss": 0.2439,
+      "step": 612
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.171095609664917,
+      "learning_rate": 3.940924839454066e-05,
+      "loss": 0.2781,
+      "step": 613
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.20573212206363678,
+      "learning_rate": 3.925923458159023e-05,
+      "loss": 0.3921,
+      "step": 614
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.17610518634319305,
+      "learning_rate": 3.9109322062557424e-05,
+      "loss": 0.2876,
+      "step": 615
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.1425466537475586,
+      "learning_rate": 3.895951225123595e-05,
+      "loss": 0.1929,
+      "step": 616
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.1828867495059967,
+      "learning_rate": 3.880980656045087e-05,
+      "loss": 0.298,
+      "step": 617
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.1653434783220291,
+      "learning_rate": 3.866020640204533e-05,
+      "loss": 0.2681,
+      "step": 618
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.16614679992198944,
+      "learning_rate": 3.851071318686721e-05,
+      "loss": 0.2504,
+      "step": 619
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.14689838886260986,
+      "learning_rate": 3.8361328324755825e-05,
+      "loss": 0.2324,
+      "step": 620
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.17827895283699036,
+      "learning_rate": 3.821205322452863e-05,
+      "loss": 0.3296,
+      "step": 621
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.19474762678146362,
+      "learning_rate": 3.8062889293967976e-05,
+      "loss": 0.4104,
+      "step": 622
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.1724865436553955,
+      "learning_rate": 3.791383793980776e-05,
+      "loss": 0.3454,
+      "step": 623
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.14381413161754608,
+      "learning_rate": 3.7764900567720196e-05,
+      "loss": 0.2138,
+      "step": 624
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.19949211180210114,
+      "learning_rate": 3.761607858230257e-05,
+      "loss": 0.4009,
+      "step": 625
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.16575078666210175,
+      "learning_rate": 3.746737338706397e-05,
+      "loss": 0.25,
+      "step": 626
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.18713515996932983,
+      "learning_rate": 3.7318786384412074e-05,
+      "loss": 0.3878,
+      "step": 627
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.17751425504684448,
+      "learning_rate": 3.71703189756399e-05,
+      "loss": 0.3235,
+      "step": 628
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.15603938698768616,
+      "learning_rate": 3.70219725609126e-05,
+      "loss": 0.2682,
+      "step": 629
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.16794337332248688,
+      "learning_rate": 3.6873748539254246e-05,
+      "loss": 0.2561,
+      "step": 630
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.18881405889987946,
+      "learning_rate": 3.6725648308534696e-05,
+      "loss": 0.3884,
+      "step": 631
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.17776301503181458,
+      "learning_rate": 3.65776732654563e-05,
+      "loss": 0.3409,
+      "step": 632
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.19741865992546082,
+      "learning_rate": 3.6429824805540815e-05,
+      "loss": 0.3526,
+      "step": 633
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.14986495673656464,
+      "learning_rate": 3.628210432311621e-05,
+      "loss": 0.2145,
+      "step": 634
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.20826153457164764,
+      "learning_rate": 3.6134513211303556e-05,
+      "loss": 0.3786,
+      "step": 635
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.20555941760540009,
+      "learning_rate": 3.598705286200382e-05,
+      "loss": 0.3032,
+      "step": 636
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.1856733113527298,
+      "learning_rate": 3.58397246658848e-05,
+      "loss": 0.374,
+      "step": 637
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.20580483973026276,
+      "learning_rate": 3.5692530012367955e-05,
+      "loss": 0.3444,
+      "step": 638
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.19250930845737457,
+      "learning_rate": 3.554547028961537e-05,
+      "loss": 0.2934,
+      "step": 639
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.16029620170593262,
+      "learning_rate": 3.5398546884516604e-05,
+      "loss": 0.2181,
+      "step": 640
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.18877767026424408,
+      "learning_rate": 3.5251761182675625e-05,
+      "loss": 0.3525,
+      "step": 641
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.19860802590847015,
+      "learning_rate": 3.510511456839777e-05,
+      "loss": 0.3048,
+      "step": 642
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.19587402045726776,
+      "learning_rate": 3.495860842467664e-05,
+      "loss": 0.3364,
+      "step": 643
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.18687398731708527,
+      "learning_rate": 3.481224413318114e-05,
+      "loss": 0.328,
+      "step": 644
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.19980676472187042,
+      "learning_rate": 3.4666023074242356e-05,
+      "loss": 0.3375,
+      "step": 645
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2091117948293686,
+      "learning_rate": 3.4519946626840573e-05,
+      "loss": 0.3351,
+      "step": 646
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.20576684176921844,
+      "learning_rate": 3.4374016168592296e-05,
+      "loss": 0.3126,
+      "step": 647
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2082446664571762,
+      "learning_rate": 3.422823307573722e-05,
+      "loss": 0.3528,
+      "step": 648
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2016439139842987,
+      "learning_rate": 3.40825987231253e-05,
+      "loss": 0.3788,
+      "step": 649
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.21504685282707214,
+      "learning_rate": 3.393711448420372e-05,
+      "loss": 0.2901,
+      "step": 650
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.1966167837381363,
+      "learning_rate": 3.379178173100396e-05,
+      "loss": 0.29,
+      "step": 651
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.15545183420181274,
+      "learning_rate": 3.364660183412892e-05,
+      "loss": 0.1902,
+      "step": 652
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.20119062066078186,
+      "learning_rate": 3.3501576162739904e-05,
+      "loss": 0.285,
+      "step": 653
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.1794392317533493,
+      "learning_rate": 3.335670608454376e-05,
+      "loss": 0.3512,
+      "step": 654
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.16064248979091644,
+      "learning_rate": 3.321199296577998e-05,
+      "loss": 0.2127,
+      "step": 655
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.1904190182685852,
+      "learning_rate": 3.3067438171207766e-05,
+      "loss": 0.3257,
+      "step": 656
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.18056736886501312,
+      "learning_rate": 3.292304306409325e-05,
+      "loss": 0.3166,
+      "step": 657
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.18243494629859924,
+      "learning_rate": 3.2778809006196564e-05,
+      "loss": 0.3098,
+      "step": 658
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.15375906229019165,
+      "learning_rate": 3.263473735775899e-05,
+      "loss": 0.1964,
+      "step": 659
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.16546106338500977,
+      "learning_rate": 3.249082947749019e-05,
+      "loss": 0.2491,
+      "step": 660
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.19515836238861084,
+      "learning_rate": 3.234708672255538e-05,
+      "loss": 0.2869,
+      "step": 661
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.16240686178207397,
+      "learning_rate": 3.220351044856247e-05,
+      "loss": 0.2522,
+      "step": 662
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.17125046253204346,
+      "learning_rate": 3.206010200954935e-05,
+      "loss": 0.2671,
+      "step": 663
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.20373588800430298,
+      "learning_rate": 3.191686275797107e-05,
+      "loss": 0.3702,
+      "step": 664
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.1734769642353058,
+      "learning_rate": 3.177379404468715e-05,
+      "loss": 0.2941,
+      "step": 665
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.13766516745090485,
+      "learning_rate": 3.163089721894876e-05,
+      "loss": 0.1891,
+      "step": 666
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.17288701236248016,
+      "learning_rate": 3.148817362838606e-05,
+      "loss": 0.3149,
+      "step": 667
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.17593587934970856,
+      "learning_rate": 3.134562461899545e-05,
+      "loss": 0.315,
+      "step": 668
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.1737486720085144,
+      "learning_rate": 3.120325153512687e-05,
+      "loss": 0.2567,
+      "step": 669
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.17551515996456146,
+      "learning_rate": 3.10610557194712e-05,
+      "loss": 0.2747,
+      "step": 670
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.206053227186203,
+      "learning_rate": 3.091903851304751e-05,
+      "loss": 0.4144,
+      "step": 671
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.15633688867092133,
+      "learning_rate": 3.077720125519042e-05,
+      "loss": 0.2517,
+      "step": 672
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.19218087196350098,
+      "learning_rate": 3.063554528353752e-05,
+      "loss": 0.3255,
+      "step": 673
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.16911399364471436,
+      "learning_rate": 3.0494071934016737e-05,
+      "loss": 0.2474,
+      "step": 674
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.18369029462337494,
+      "learning_rate": 3.03527825408337e-05,
+      "loss": 0.3381,
+      "step": 675
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.16539664566516876,
+      "learning_rate": 3.0211678436459212e-05,
+      "loss": 0.2616,
+      "step": 676
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.19227956235408783,
+      "learning_rate": 3.0070760951616618e-05,
+      "loss": 0.3163,
+      "step": 677
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.1829083412885666,
+      "learning_rate": 2.993003141526933e-05,
+      "loss": 0.3135,
+      "step": 678
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.16197767853736877,
+      "learning_rate": 2.978949115460824e-05,
+      "loss": 0.2139,
+      "step": 679
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.17858240008354187,
+      "learning_rate": 2.9649141495039223e-05,
+      "loss": 0.2648,
+      "step": 680
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.16623885929584503,
+      "learning_rate": 2.950898376017064e-05,
+      "loss": 0.2742,
+      "step": 681
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.16351516544818878,
+      "learning_rate": 2.9369019271800823e-05,
+      "loss": 0.2351,
+      "step": 682
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.18213540315628052,
+      "learning_rate": 2.9229249349905684e-05,
+      "loss": 0.355,
+      "step": 683
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.17309141159057617,
+      "learning_rate": 2.908967531262618e-05,
+      "loss": 0.2252,
+      "step": 684
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.17787668108940125,
+      "learning_rate": 2.895029847625595e-05,
+      "loss": 0.2907,
+      "step": 685
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2267909198999405,
+      "learning_rate": 2.8811120155228844e-05,
+      "loss": 0.4296,
+      "step": 686
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.19319239258766174,
+      "learning_rate": 2.8672141662106578e-05,
+      "loss": 0.3285,
+      "step": 687
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.1958734095096588,
+      "learning_rate": 2.853336430756631e-05,
+      "loss": 0.3381,
+      "step": 688
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.22564010322093964,
+      "learning_rate": 2.8394789400388328e-05,
+      "loss": 0.3791,
+      "step": 689
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.23828552663326263,
+      "learning_rate": 2.8256418247443662e-05,
+      "loss": 0.3,
+      "step": 690
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.163003072142601,
+      "learning_rate": 2.811825215368179e-05,
+      "loss": 0.2569,
+      "step": 691
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2118333876132965,
+      "learning_rate": 2.798029242211828e-05,
+      "loss": 0.375,
+      "step": 692
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.18281732499599457,
+      "learning_rate": 2.7842540353822632e-05,
+      "loss": 0.2896,
+      "step": 693
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.17668594419956207,
+      "learning_rate": 2.770499724790584e-05,
+      "loss": 0.2656,
+      "step": 694
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.19043265283107758,
+      "learning_rate": 2.7567664401508225e-05,
+      "loss": 0.3223,
+      "step": 695
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.21557258069515228,
+      "learning_rate": 2.7430543109787222e-05,
+      "loss": 0.3892,
+      "step": 696
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.1652493178844452,
+      "learning_rate": 2.729363466590511e-05,
+      "loss": 0.2218,
+      "step": 697
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.17171581089496613,
+      "learning_rate": 2.7156940361016864e-05,
+      "loss": 0.2707,
+      "step": 698
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.18436965346336365,
+      "learning_rate": 2.702046148425795e-05,
+      "loss": 0.2667,
+      "step": 699
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.1945919543504715,
+      "learning_rate": 2.688419932273219e-05,
+      "loss": 0.3658,
+      "step": 700
+    },
+    {
+      "epoch": 0.66,
+      "eval_loss": 0.2868694067001343,
+      "eval_runtime": 372.3804,
+      "eval_samples_per_second": 0.916,
+      "eval_steps_per_second": 0.231,
+      "step": 700
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.20651818811893463,
+      "learning_rate": 2.6748155161499567e-05,
+      "loss": 0.3567,
+      "step": 701
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.1465105265378952,
+      "learning_rate": 2.6612330283564223e-05,
+      "loss": 0.2049,
+      "step": 702
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.17382213473320007,
+      "learning_rate": 2.6476725969862227e-05,
+      "loss": 0.1899,
+      "step": 703
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.1563204675912857,
+      "learning_rate": 2.634134349924956e-05,
+      "loss": 0.2319,
+      "step": 704
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.17278626561164856,
+      "learning_rate": 2.6206184148490066e-05,
+      "loss": 0.3044,
+      "step": 705
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.17541703581809998,
+      "learning_rate": 2.6071249192243363e-05,
+      "loss": 0.3149,
+      "step": 706
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.20070253312587738,
+      "learning_rate": 2.5936539903052892e-05,
+      "loss": 0.3409,
+      "step": 707
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.19257397949695587,
+      "learning_rate": 2.580205755133384e-05,
+      "loss": 0.3917,
+      "step": 708
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.19976547360420227,
+      "learning_rate": 2.5667803405361213e-05,
+      "loss": 0.3921,
+      "step": 709
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.204472154378891,
+      "learning_rate": 2.5533778731257824e-05,
+      "loss": 0.3973,
+      "step": 710
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.17846406996250153,
+      "learning_rate": 2.5399984792982457e-05,
+      "loss": 0.2811,
+      "step": 711
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.18965056538581848,
+      "learning_rate": 2.5266422852317795e-05,
+      "loss": 0.2984,
+      "step": 712
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.20487138628959656,
+      "learning_rate": 2.513309416885865e-05,
+      "loss": 0.358,
+      "step": 713
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.19671356678009033,
+      "learning_rate": 2.500000000000001e-05,
+      "loss": 0.2543,
+      "step": 714
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.1853562593460083,
+      "learning_rate": 2.4867141600925214e-05,
+      "loss": 0.2873,
+      "step": 715
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.18355299532413483,
+      "learning_rate": 2.4734520224594093e-05,
+      "loss": 0.3451,
+      "step": 716
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.1895030289888382,
+      "learning_rate": 2.4602137121731195e-05,
+      "loss": 0.3195,
+      "step": 717
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.17159874737262726,
+      "learning_rate": 2.44699935408139e-05,
+      "loss": 0.3058,
+      "step": 718
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.13373179733753204,
+      "learning_rate": 2.433809072806081e-05,
+      "loss": 0.1542,
+      "step": 719
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.1691519320011139,
+      "learning_rate": 2.4206429927419794e-05,
+      "loss": 0.2622,
+      "step": 720
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.18795622885227203,
+      "learning_rate": 2.40750123805564e-05,
+      "loss": 0.2757,
+      "step": 721
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.19379259645938873,
+      "learning_rate": 2.3943839326842092e-05,
+      "loss": 0.3125,
+      "step": 722
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.17294247448444366,
+      "learning_rate": 2.3812912003342568e-05,
+      "loss": 0.253,
+      "step": 723
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.20651890337467194,
+      "learning_rate": 2.3682231644806108e-05,
+      "loss": 0.3294,
+      "step": 724
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.18134303390979767,
+      "learning_rate": 2.3551799483651894e-05,
+      "loss": 0.2887,
+      "step": 725
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.21852369606494904,
+      "learning_rate": 2.342161674995843e-05,
+      "loss": 0.3801,
+      "step": 726
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.17804594337940216,
+      "learning_rate": 2.3291684671451903e-05,
+      "loss": 0.3067,
+      "step": 727
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.17285257577896118,
+      "learning_rate": 2.3162004473494657e-05,
+      "loss": 0.255,
+      "step": 728
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.16711926460266113,
+      "learning_rate": 2.3032577379073577e-05,
+      "loss": 0.3151,
+      "step": 729
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.16523627936840057,
+      "learning_rate": 2.2903404608788582e-05,
+      "loss": 0.2419,
+      "step": 730
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.17762525379657745,
+      "learning_rate": 2.2774487380841115e-05,
+      "loss": 0.312,
+      "step": 731
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.20041225850582123,
+      "learning_rate": 2.2645826911022655e-05,
+      "loss": 0.3041,
+      "step": 732
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.18052154779434204,
+      "learning_rate": 2.2517424412703252e-05,
+      "loss": 0.2665,
+      "step": 733
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.1542147845029831,
+      "learning_rate": 2.2389281096820075e-05,
+      "loss": 0.249,
+      "step": 734
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.17372718453407288,
+      "learning_rate": 2.2261398171865978e-05,
+      "loss": 0.3133,
+      "step": 735
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.17182888090610504,
+      "learning_rate": 2.2133776843878186e-05,
+      "loss": 0.2619,
+      "step": 736
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.19321058690547943,
+      "learning_rate": 2.2006418316426775e-05,
+      "loss": 0.2751,
+      "step": 737
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.1866815835237503,
+      "learning_rate": 2.187932379060348e-05,
+      "loss": 0.3245,
+      "step": 738
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.1817731410264969,
+      "learning_rate": 2.1752494465010242e-05,
+      "loss": 0.2273,
+      "step": 739
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.1846882700920105,
+      "learning_rate": 2.1625931535747964e-05,
+      "loss": 0.3202,
+      "step": 740
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.1985662877559662,
+      "learning_rate": 2.1499636196405228e-05,
+      "loss": 0.3466,
+      "step": 741
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.19262108206748962,
+      "learning_rate": 2.1373609638047032e-05,
+      "loss": 0.2712,
+      "step": 742
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.1885438710451126,
+      "learning_rate": 2.1247853049203543e-05,
+      "loss": 0.2686,
+      "step": 743
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.19670408964157104,
+      "learning_rate": 2.112236761585892e-05,
+      "loss": 0.3338,
+      "step": 744
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.20213522017002106,
+      "learning_rate": 2.09971545214401e-05,
+      "loss": 0.3043,
+      "step": 745
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.18916374444961548,
+      "learning_rate": 2.087221494680563e-05,
+      "loss": 0.3369,
+      "step": 746
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.14264284074306488,
+      "learning_rate": 2.074755007023461e-05,
+      "loss": 0.2282,
+      "step": 747
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.14444948732852936,
+      "learning_rate": 2.0623161067415463e-05,
+      "loss": 0.1499,
+      "step": 748
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.20152431726455688,
+      "learning_rate": 2.049904911143492e-05,
+      "loss": 0.3041,
+      "step": 749
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.19557496905326843,
+      "learning_rate": 2.0375215372766944e-05,
+      "loss": 0.308,
+      "step": 750
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.20917536318302155,
+      "learning_rate": 2.025166101926168e-05,
+      "loss": 0.4191,
+      "step": 751
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.16045525670051575,
+      "learning_rate": 2.012838721613447e-05,
+      "loss": 0.212,
+      "step": 752
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.23069824278354645,
+      "learning_rate": 2.0005395125954812e-05,
+      "loss": 0.3855,
+      "step": 753
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.1648103892803192,
+      "learning_rate": 1.9882685908635462e-05,
+      "loss": 0.2716,
+      "step": 754
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.1953677386045456,
+      "learning_rate": 1.9760260721421426e-05,
+      "loss": 0.3168,
+      "step": 755
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.212935671210289,
+      "learning_rate": 1.963812071887913e-05,
+      "loss": 0.3683,
+      "step": 756
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.1765817254781723,
+      "learning_rate": 1.951626705288544e-05,
+      "loss": 0.2998,
+      "step": 757
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.18489320576190948,
+      "learning_rate": 1.9394700872616855e-05,
+      "loss": 0.2328,
+      "step": 758
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2066463828086853,
+      "learning_rate": 1.9273423324538658e-05,
+      "loss": 0.3681,
+      "step": 759
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.17519253492355347,
+      "learning_rate": 1.9152435552394104e-05,
+      "loss": 0.2621,
+      "step": 760
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.17596815526485443,
+      "learning_rate": 1.903173869719362e-05,
+      "loss": 0.2313,
+      "step": 761
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.18167275190353394,
+      "learning_rate": 1.891133389720407e-05,
+      "loss": 0.2698,
+      "step": 762
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.1822601705789566,
+      "learning_rate": 1.8791222287937982e-05,
+      "loss": 0.2911,
+      "step": 763
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.18289314210414886,
+      "learning_rate": 1.8671405002142918e-05,
+      "loss": 0.2086,
+      "step": 764
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.1928934007883072,
+      "learning_rate": 1.855188316979068e-05,
+      "loss": 0.3029,
+      "step": 765
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.1944347769021988,
+      "learning_rate": 1.8432657918066732e-05,
+      "loss": 0.3018,
+      "step": 766
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.21636322140693665,
+      "learning_rate": 1.831373037135955e-05,
+      "loss": 0.3807,
+      "step": 767
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.1668556034564972,
+      "learning_rate": 1.8195101651250017e-05,
+      "loss": 0.207,
+      "step": 768
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.19037510454654694,
+      "learning_rate": 1.807677287650083e-05,
+      "loss": 0.2871,
+      "step": 769
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.1802353858947754,
+      "learning_rate": 1.7958745163045986e-05,
+      "loss": 0.2862,
+      "step": 770
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.23819811642169952,
+      "learning_rate": 1.7841019623980217e-05,
+      "loss": 0.3169,
+      "step": 771
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.16739079356193542,
+      "learning_rate": 1.77235973695485e-05,
+      "loss": 0.2206,
+      "step": 772
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.20899198949337006,
+      "learning_rate": 1.760647950713566e-05,
+      "loss": 0.3186,
+      "step": 773
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.1680791974067688,
+      "learning_rate": 1.74896671412558e-05,
+      "loss": 0.2524,
+      "step": 774
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.22058913111686707,
+      "learning_rate": 1.737316137354197e-05,
+      "loss": 0.3846,
+      "step": 775
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.17867128551006317,
+      "learning_rate": 1.725696330273575e-05,
+      "loss": 0.3146,
+      "step": 776
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.18511047959327698,
+      "learning_rate": 1.7141074024676912e-05,
+      "loss": 0.2953,
+      "step": 777
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.17963749170303345,
+      "learning_rate": 1.7025494632293048e-05,
+      "loss": 0.1958,
+      "step": 778
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.20299947261810303,
+      "learning_rate": 1.6910226215589303e-05,
+      "loss": 0.3565,
+      "step": 779
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.208424910902977,
+      "learning_rate": 1.6795269861638042e-05,
+      "loss": 0.3485,
+      "step": 780
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.19364427030086517,
+      "learning_rate": 1.6680626654568686e-05,
+      "loss": 0.2683,
+      "step": 781
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.21714358031749725,
+      "learning_rate": 1.656629767555739e-05,
+      "loss": 0.3886,
+      "step": 782
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.19313904643058777,
+      "learning_rate": 1.645228400281689e-05,
+      "loss": 0.2884,
+      "step": 783
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.1811634600162506,
+      "learning_rate": 1.6338586711586356e-05,
+      "loss": 0.2505,
+      "step": 784
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.18979164958000183,
+      "learning_rate": 1.6225206874121218e-05,
+      "loss": 0.3217,
+      "step": 785
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.18322841823101044,
+      "learning_rate": 1.6112145559683055e-05,
+      "loss": 0.2848,
+      "step": 786
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2041218876838684,
+      "learning_rate": 1.5999403834529548e-05,
+      "loss": 0.4002,
+      "step": 787
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.17978321015834808,
+      "learning_rate": 1.5886982761904377e-05,
+      "loss": 0.2538,
+      "step": 788
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2092827707529068,
+      "learning_rate": 1.5774883402027208e-05,
+      "loss": 0.3359,
+      "step": 789
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.20518158376216888,
+      "learning_rate": 1.5663106812083743e-05,
+      "loss": 0.3143,
+      "step": 790
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.1681492030620575,
+      "learning_rate": 1.555165404621567e-05,
+      "loss": 0.2597,
+      "step": 791
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.201442688703537,
+      "learning_rate": 1.5440526155510763e-05,
+      "loss": 0.3707,
+      "step": 792
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.20597267150878906,
+      "learning_rate": 1.532972418799298e-05,
+      "loss": 0.3683,
+      "step": 793
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.19196146726608276,
+      "learning_rate": 1.5219249188612556e-05,
+      "loss": 0.2935,
+      "step": 794
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.1902860403060913,
+      "learning_rate": 1.5109102199236152e-05,
+      "loss": 0.3359,
+      "step": 795
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.22097539901733398,
+      "learning_rate": 1.4999284258637053e-05,
+      "loss": 0.3503,
+      "step": 796
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.18898749351501465,
+      "learning_rate": 1.488979640248534e-05,
+      "loss": 0.2899,
+      "step": 797
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.1775866448879242,
+      "learning_rate": 1.4780639663338125e-05,
+      "loss": 0.2566,
+      "step": 798
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.19556714594364166,
+      "learning_rate": 1.4671815070629868e-05,
+      "loss": 0.2388,
+      "step": 799
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.15021736919879913,
+      "learning_rate": 1.4563323650662586e-05,
+      "loss": 0.1983,
+      "step": 800
+    },
+    {
+      "epoch": 0.76,
+      "eval_loss": 0.28555864095687866,
+      "eval_runtime": 356.3129,
+      "eval_samples_per_second": 0.957,
+      "eval_steps_per_second": 0.241,
+      "step": 800
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.17953263223171234,
+      "learning_rate": 1.4455166426596223e-05,
+      "loss": 0.2164,
+      "step": 801
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.24168966710567474,
+      "learning_rate": 1.434734441843899e-05,
+      "loss": 0.4737,
+      "step": 802
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.17398898303508759,
+      "learning_rate": 1.4239858643037751e-05,
+      "loss": 0.2849,
+      "step": 803
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.16854502260684967,
+      "learning_rate": 1.4132710114068426e-05,
+      "loss": 0.2591,
+      "step": 804
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.1947992742061615,
+      "learning_rate": 1.4025899842026442e-05,
+      "loss": 0.191,
+      "step": 805
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.18380802869796753,
+      "learning_rate": 1.3919428834217163e-05,
+      "loss": 0.2148,
+      "step": 806
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.18306519091129303,
+      "learning_rate": 1.3813298094746491e-05,
+      "loss": 0.2384,
+      "step": 807
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.17643406987190247,
+      "learning_rate": 1.3707508624511262e-05,
+      "loss": 0.2051,
+      "step": 808
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2269081324338913,
+      "learning_rate": 1.36020614211899e-05,
+      "loss": 0.3664,
+      "step": 809
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.1772211641073227,
+      "learning_rate": 1.349695747923298e-05,
+      "loss": 0.2479,
+      "step": 810
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.19540081918239594,
+      "learning_rate": 1.339219778985385e-05,
+      "loss": 0.2828,
+      "step": 811
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.19767455756664276,
+      "learning_rate": 1.3287783341019278e-05,
+      "loss": 0.2818,
+      "step": 812
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.19986826181411743,
+      "learning_rate": 1.3183715117440142e-05,
+      "loss": 0.3318,
+      "step": 813
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.20941559970378876,
+      "learning_rate": 1.3079994100562159e-05,
+      "loss": 0.3224,
+      "step": 814
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.19123049080371857,
+      "learning_rate": 1.2976621268556571e-05,
+      "loss": 0.3185,
+      "step": 815
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.18940964341163635,
+      "learning_rate": 1.2873597596311027e-05,
+      "loss": 0.2734,
+      "step": 816
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.23779705166816711,
+      "learning_rate": 1.2770924055420258e-05,
+      "loss": 0.3136,
+      "step": 817
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.20298561453819275,
+      "learning_rate": 1.2668601614177017e-05,
+      "loss": 0.3548,
+      "step": 818
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.18004286289215088,
+      "learning_rate": 1.2566631237562893e-05,
+      "loss": 0.2498,
+      "step": 819
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.18635989725589752,
+      "learning_rate": 1.246501388723923e-05,
+      "loss": 0.2534,
+      "step": 820
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.18110227584838867,
+      "learning_rate": 1.2363750521538064e-05,
+      "loss": 0.2658,
+      "step": 821
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.19258196651935577,
+      "learning_rate": 1.2262842095453065e-05,
+      "loss": 0.3771,
+      "step": 822
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.18437983095645905,
+      "learning_rate": 1.2162289560630524e-05,
+      "loss": 0.2936,
+      "step": 823
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.21418513357639313,
+      "learning_rate": 1.2062093865360458e-05,
+      "loss": 0.3452,
+      "step": 824
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.17180278897285461,
+      "learning_rate": 1.1962255954567537e-05,
+      "loss": 0.2426,
+      "step": 825
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.18029405176639557,
+      "learning_rate": 1.1862776769802276e-05,
+      "loss": 0.2719,
+      "step": 826
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.23371167480945587,
+      "learning_rate": 1.1763657249232107e-05,
+      "loss": 0.3428,
+      "step": 827
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.16515472531318665,
+      "learning_rate": 1.1664898327632551e-05,
+      "loss": 0.2415,
+      "step": 828
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.20264777541160583,
+      "learning_rate": 1.1566500936378388e-05,
+      "loss": 0.3379,
+      "step": 829
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.24041464924812317,
+      "learning_rate": 1.146846600343488e-05,
+      "loss": 0.3508,
+      "step": 830
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.22278086841106415,
+      "learning_rate": 1.1370794453349037e-05,
+      "loss": 0.4055,
+      "step": 831
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.23650290071964264,
+      "learning_rate": 1.1273487207240847e-05,
+      "loss": 0.2911,
+      "step": 832
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.19014495611190796,
+      "learning_rate": 1.1176545182794674e-05,
+      "loss": 0.1874,
+      "step": 833
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.1839723289012909,
+      "learning_rate": 1.1079969294250514e-05,
+      "loss": 0.2684,
+      "step": 834
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.1958753913640976,
+      "learning_rate": 1.0983760452395413e-05,
+      "loss": 0.3203,
+      "step": 835
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.20766787230968475,
+      "learning_rate": 1.0887919564554894e-05,
+      "loss": 0.33,
+      "step": 836
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.20762580633163452,
+      "learning_rate": 1.079244753458437e-05,
+      "loss": 0.3401,
+      "step": 837
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.1961548924446106,
+      "learning_rate": 1.0697345262860636e-05,
+      "loss": 0.3373,
+      "step": 838
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.18646018207073212,
+      "learning_rate": 1.0602613646273374e-05,
+      "loss": 0.355,
+      "step": 839
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.24027031660079956,
+      "learning_rate": 1.0508253578216693e-05,
+      "loss": 0.3792,
+      "step": 840
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.20653152465820312,
+      "learning_rate": 1.0414265948580692e-05,
+      "loss": 0.3173,
+      "step": 841
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.19750098884105682,
+      "learning_rate": 1.032065164374313e-05,
+      "loss": 0.2582,
+      "step": 842
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.18974408507347107,
+      "learning_rate": 1.0227411546560962e-05,
+      "loss": 0.2545,
+      "step": 843
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.17909228801727295,
+      "learning_rate": 1.01345465363621e-05,
+      "loss": 0.2938,
+      "step": 844
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.20879067480564117,
+      "learning_rate": 1.0042057488937067e-05,
+      "loss": 0.3014,
+      "step": 845
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.20528557896614075,
+      "learning_rate": 9.949945276530781e-06,
+      "loss": 0.3875,
+      "step": 846
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.167794868350029,
+      "learning_rate": 9.858210767834292e-06,
+      "loss": 0.2457,
+      "step": 847
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2016701102256775,
+      "learning_rate": 9.766854827976617e-06,
+      "loss": 0.2984,
+      "step": 848
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.18046480417251587,
+      "learning_rate": 9.675878318516546e-06,
+      "loss": 0.2217,
+      "step": 849
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.24706487357616425,
+      "learning_rate": 9.58528209743459e-06,
+      "loss": 0.4065,
+      "step": 850
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.207551047205925,
+      "learning_rate": 9.495067019124792e-06,
+      "loss": 0.3255,
+      "step": 851
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2020636349916458,
+      "learning_rate": 9.405233934386726e-06,
+      "loss": 0.2841,
+      "step": 852
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.17741268873214722,
+      "learning_rate": 9.315783690417479e-06,
+      "loss": 0.2441,
+      "step": 853
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.1885654777288437,
+      "learning_rate": 9.226717130803636e-06,
+      "loss": 0.2709,
+      "step": 854
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.1923096776008606,
+      "learning_rate": 9.138035095513336e-06,
+      "loss": 0.3319,
+      "step": 855
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.15192262828350067,
+      "learning_rate": 9.049738420888349e-06,
+      "loss": 0.2523,
+      "step": 856
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.1906556636095047,
+      "learning_rate": 8.961827939636196e-06,
+      "loss": 0.2693,
+      "step": 857
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.162446066737175,
+      "learning_rate": 8.87430448082227e-06,
+      "loss": 0.148,
+      "step": 858
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.16881592571735382,
+      "learning_rate": 8.787168869862066e-06,
+      "loss": 0.23,
+      "step": 859
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2027132511138916,
+      "learning_rate": 8.700421928513352e-06,
+      "loss": 0.3928,
+      "step": 860
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.20462428033351898,
+      "learning_rate": 8.614064474868421e-06,
+      "loss": 0.2782,
+      "step": 861
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2035517543554306,
+      "learning_rate": 8.528097323346407e-06,
+      "loss": 0.3295,
+      "step": 862
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.151922345161438,
+      "learning_rate": 8.442521284685573e-06,
+      "loss": 0.1953,
+      "step": 863
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.18760909140110016,
+      "learning_rate": 8.357337165935675e-06,
+      "loss": 0.2629,
+      "step": 864
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.22610962390899658,
+      "learning_rate": 8.27254577045039e-06,
+      "loss": 0.3479,
+      "step": 865
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.17555096745491028,
+      "learning_rate": 8.188147897879667e-06,
+      "loss": 0.2233,
+      "step": 866
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.1815774291753769,
+      "learning_rate": 8.104144344162229e-06,
+      "loss": 0.2461,
+      "step": 867
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.16811814904212952,
+      "learning_rate": 8.02053590151805e-06,
+      "loss": 0.2648,
+      "step": 868
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2018882781267166,
+      "learning_rate": 7.937323358440935e-06,
+      "loss": 0.2738,
+      "step": 869
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.1594676375389099,
+      "learning_rate": 7.854507499691005e-06,
+      "loss": 0.1658,
+      "step": 870
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.19084495306015015,
+      "learning_rate": 7.772089106287344e-06,
+      "loss": 0.3109,
+      "step": 871
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.21609549224376678,
+      "learning_rate": 7.690068955500624e-06,
+      "loss": 0.3769,
+      "step": 872
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.18996381759643555,
+      "learning_rate": 7.608447820845771e-06,
+      "loss": 0.2392,
+      "step": 873
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.21013544499874115,
+      "learning_rate": 7.527226472074678e-06,
+      "loss": 0.2753,
+      "step": 874
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.19846756756305695,
+      "learning_rate": 7.446405675168938e-06,
+      "loss": 0.3243,
+      "step": 875
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.18228833377361298,
+      "learning_rate": 7.365986192332625e-06,
+      "loss": 0.1948,
+      "step": 876
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.1821378618478775,
+      "learning_rate": 7.285968781985092e-06,
+      "loss": 0.2561,
+      "step": 877
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.15298676490783691,
+      "learning_rate": 7.206354198753862e-06,
+      "loss": 0.1948,
+      "step": 878
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.20400376617908478,
+      "learning_rate": 7.127143193467445e-06,
+      "loss": 0.3491,
+      "step": 879
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.23939305543899536,
+      "learning_rate": 7.048336513148307e-06,
+      "loss": 0.3556,
+      "step": 880
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.2803425192832947,
+      "learning_rate": 6.969934901005809e-06,
+      "loss": 0.418,
+      "step": 881
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.19635199010372162,
+      "learning_rate": 6.8919390964291895e-06,
+      "loss": 0.3403,
+      "step": 882
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.18559308350086212,
+      "learning_rate": 6.814349834980621e-06,
+      "loss": 0.2961,
+      "step": 883
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2659151554107666,
+      "learning_rate": 6.7371678483882264e-06,
+      "loss": 0.4128,
+      "step": 884
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.1881849616765976,
+      "learning_rate": 6.660393864539222e-06,
+      "loss": 0.2645,
+      "step": 885
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.1653018444776535,
+      "learning_rate": 6.58402860747302e-06,
+      "loss": 0.2017,
+      "step": 886
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.20394209027290344,
+      "learning_rate": 6.508072797374454e-06,
+      "loss": 0.3366,
+      "step": 887
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.1957215815782547,
+      "learning_rate": 6.4325271505669025e-06,
+      "loss": 0.3073,
+      "step": 888
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.19627298414707184,
+      "learning_rate": 6.3573923795056e-06,
+      "loss": 0.316,
+      "step": 889
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.19443422555923462,
+      "learning_rate": 6.282669192770896e-06,
+      "loss": 0.3242,
+      "step": 890
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.20445725321769714,
+      "learning_rate": 6.208358295061572e-06,
+      "loss": 0.2948,
+      "step": 891
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.1769063025712967,
+      "learning_rate": 6.134460387188206e-06,
+      "loss": 0.2592,
+      "step": 892
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.17925342917442322,
+      "learning_rate": 6.060976166066546e-06,
+      "loss": 0.2637,
+      "step": 893
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.206807941198349,
+      "learning_rate": 5.9879063247109405e-06,
+      "loss": 0.3309,
+      "step": 894
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2015414834022522,
+      "learning_rate": 5.915251552227829e-06,
+      "loss": 0.2982,
+      "step": 895
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.23987385630607605,
+      "learning_rate": 5.8430125338092115e-06,
+      "loss": 0.382,
+      "step": 896
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.17709748446941376,
+      "learning_rate": 5.771189950726191e-06,
+      "loss": 0.2363,
+      "step": 897
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.22804932296276093,
+      "learning_rate": 5.699784480322568e-06,
+      "loss": 0.3689,
+      "step": 898
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2100164145231247,
+      "learning_rate": 5.628796796008434e-06,
+      "loss": 0.3248,
+      "step": 899
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.18560853600502014,
+      "learning_rate": 5.558227567253832e-06,
+      "loss": 0.2851,
+      "step": 900
+    },
+    {
+      "epoch": 0.85,
+      "eval_loss": 0.2828112542629242,
+      "eval_runtime": 358.155,
+      "eval_samples_per_second": 0.952,
+      "eval_steps_per_second": 0.24,
+      "step": 900
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.20550298690795898,
+      "learning_rate": 5.488077459582425e-06,
+      "loss": 0.3295,
+      "step": 901
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.18917301297187805,
+      "learning_rate": 5.418347134565249e-06,
+      "loss": 0.299,
+      "step": 902
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.20705218613147736,
+      "learning_rate": 5.349037249814442e-06,
+      "loss": 0.2797,
+      "step": 903
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.22389914095401764,
+      "learning_rate": 5.28014845897708e-06,
+      "loss": 0.3068,
+      "step": 904
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.22722771763801575,
+      "learning_rate": 5.211681411728969e-06,
+      "loss": 0.3244,
+      "step": 905
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.18286827206611633,
+      "learning_rate": 5.143636753768549e-06,
+      "loss": 0.248,
+      "step": 906
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.1883745789527893,
+      "learning_rate": 5.076015126810784e-06,
+      "loss": 0.299,
+      "step": 907
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.18702787160873413,
+      "learning_rate": 5.008817168581137e-06,
+      "loss": 0.2696,
+      "step": 908
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2003723680973053,
+      "learning_rate": 4.94204351280953e-06,
+      "loss": 0.3124,
+      "step": 909
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.20706987380981445,
+      "learning_rate": 4.8756947892243725e-06,
+      "loss": 0.3214,
+      "step": 910
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2090759575366974,
+      "learning_rate": 4.809771623546627e-06,
+      "loss": 0.221,
+      "step": 911
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.16334488987922668,
+      "learning_rate": 4.744274637483936e-06,
+      "loss": 0.2097,
+      "step": 912
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.17283600568771362,
+      "learning_rate": 4.6792044487247e-06,
+      "loss": 0.2883,
+      "step": 913
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.1947123408317566,
+      "learning_rate": 4.614561670932288e-06,
+      "loss": 0.2777,
+      "step": 914
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.18705716729164124,
+      "learning_rate": 4.550346913739256e-06,
+      "loss": 0.2286,
+      "step": 915
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.18334023654460907,
+      "learning_rate": 4.486560782741578e-06,
+      "loss": 0.2215,
+      "step": 916
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.1559637188911438,
+      "learning_rate": 4.423203879492943e-06,
+      "loss": 0.2158,
+      "step": 917
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2183571457862854,
+      "learning_rate": 4.360276801499086e-06,
+      "loss": 0.3675,
+      "step": 918
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.1875716894865036,
+      "learning_rate": 4.29778014221216e-06,
+      "loss": 0.278,
+      "step": 919
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.21742533147335052,
+      "learning_rate": 4.2357144910251e-06,
+      "loss": 0.3144,
+      "step": 920
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.23520231246948242,
+      "learning_rate": 4.174080433266136e-06,
+      "loss": 0.3296,
+      "step": 921
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.20687828958034515,
+      "learning_rate": 4.112878550193194e-06,
+      "loss": 0.3105,
+      "step": 922
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2107243537902832,
+      "learning_rate": 4.05210941898847e-06,
+      "loss": 0.2241,
+      "step": 923
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2212332934141159,
+      "learning_rate": 3.991773612752952e-06,
+      "loss": 0.2429,
+      "step": 924
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2082129418849945,
+      "learning_rate": 3.9318717005010495e-06,
+      "loss": 0.3107,
+      "step": 925
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.20215757191181183,
+      "learning_rate": 3.872404247155193e-06,
+      "loss": 0.3241,
+      "step": 926
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.20882193744182587,
+      "learning_rate": 3.8133718135405285e-06,
+      "loss": 0.3569,
+      "step": 927
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.19976763427257538,
+      "learning_rate": 3.754774956379614e-06,
+      "loss": 0.2933,
+      "step": 928
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.21227557957172394,
+      "learning_rate": 3.696614228287187e-06,
+      "loss": 0.3195,
+      "step": 929
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2234586477279663,
+      "learning_rate": 3.6388901777649477e-06,
+      "loss": 0.319,
+      "step": 930
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.19070518016815186,
+      "learning_rate": 3.581603349196372e-06,
+      "loss": 0.2924,
+      "step": 931
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.23035258054733276,
+      "learning_rate": 3.5247542828415747e-06,
+      "loss": 0.3953,
+      "step": 932
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.1796349585056305,
+      "learning_rate": 3.468343514832251e-06,
+      "loss": 0.2388,
+      "step": 933
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.1925991177558899,
+      "learning_rate": 3.4123715771665786e-06,
+      "loss": 0.2178,
+      "step": 934
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.19209635257720947,
+      "learning_rate": 3.356838997704226e-06,
+      "loss": 0.3275,
+      "step": 935
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.19841033220291138,
+      "learning_rate": 3.3017463001613626e-06,
+      "loss": 0.2642,
+      "step": 936
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.18666499853134155,
+      "learning_rate": 3.2470940041057106e-06,
+      "loss": 0.2103,
+      "step": 937
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.23942196369171143,
+      "learning_rate": 3.1928826249516987e-06,
+      "loss": 0.2211,
+      "step": 938
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2525459825992584,
+      "learning_rate": 3.139112673955513e-06,
+      "loss": 0.2952,
+      "step": 939
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2017432302236557,
+      "learning_rate": 3.08578465821035e-06,
+      "loss": 0.3171,
+      "step": 940
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.18431301414966583,
+      "learning_rate": 3.0328990806415934e-06,
+      "loss": 0.2791,
+      "step": 941
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2199854552745819,
+      "learning_rate": 2.9804564400020994e-06,
+      "loss": 0.3402,
+      "step": 942
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.20603200793266296,
+      "learning_rate": 2.9284572308674596e-06,
+      "loss": 0.3483,
+      "step": 943
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.20404697954654694,
+      "learning_rate": 2.8769019436313715e-06,
+      "loss": 0.3204,
+      "step": 944
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.17376191914081573,
+      "learning_rate": 2.825791064500993e-06,
+      "loss": 0.2846,
+      "step": 945
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.22974978387355804,
+      "learning_rate": 2.7751250754923573e-06,
+      "loss": 0.3129,
+      "step": 946
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.18384508788585663,
+      "learning_rate": 2.7249044544258363e-06,
+      "loss": 0.3149,
+      "step": 947
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2168148010969162,
+      "learning_rate": 2.675129674921639e-06,
+      "loss": 0.4084,
+      "step": 948
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.24825100600719452,
+      "learning_rate": 2.625801206395312e-06,
+      "loss": 0.3891,
+      "step": 949
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.19692133367061615,
+      "learning_rate": 2.576919514053355e-06,
+      "loss": 0.269,
+      "step": 950
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2479449063539505,
+      "learning_rate": 2.528485058888813e-06,
+      "loss": 0.3801,
+      "step": 951
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.21182508766651154,
+      "learning_rate": 2.48049829767692e-06,
+      "loss": 0.3184,
+      "step": 952
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.24480792880058289,
+      "learning_rate": 2.4329596829708144e-06,
+      "loss": 0.407,
+      "step": 953
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.20409199595451355,
+      "learning_rate": 2.385869663097251e-06,
+      "loss": 0.2914,
+      "step": 954
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.20292706787586212,
+      "learning_rate": 2.3392286821523724e-06,
+      "loss": 0.3473,
+      "step": 955
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.19676654040813446,
+      "learning_rate": 2.2930371799975594e-06,
+      "loss": 0.2262,
+      "step": 956
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.1959904581308365,
+      "learning_rate": 2.2472955922552163e-06,
+      "loss": 0.2625,
+      "step": 957
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.21012650430202484,
+      "learning_rate": 2.202004350304715e-06,
+      "loss": 0.3484,
+      "step": 958
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2135014683008194,
+      "learning_rate": 2.157163881278312e-06,
+      "loss": 0.324,
+      "step": 959
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.20510633289813995,
+      "learning_rate": 2.112774608057111e-06,
+      "loss": 0.3069,
+      "step": 960
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.20598305761814117,
+      "learning_rate": 2.068836949267089e-06,
+      "loss": 0.3406,
+      "step": 961
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.23669765889644623,
+      "learning_rate": 2.0253513192751373e-06,
+      "loss": 0.2707,
+      "step": 962
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2071131020784378,
+      "learning_rate": 1.982318128185151e-06,
+      "loss": 0.3058,
+      "step": 963
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2210494577884674,
+      "learning_rate": 1.9397377818341944e-06,
+      "loss": 0.3139,
+      "step": 964
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2400975525379181,
+      "learning_rate": 1.8976106817886196e-06,
+      "loss": 0.3755,
+      "step": 965
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.21791237592697144,
+      "learning_rate": 1.855937225340315e-06,
+      "loss": 0.2989,
+      "step": 966
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.18651287257671356,
+      "learning_rate": 1.8147178055029579e-06,
+      "loss": 0.277,
+      "step": 967
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.1654125452041626,
+      "learning_rate": 1.7739528110083004e-06,
+      "loss": 0.254,
+      "step": 968
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.20894034206867218,
+      "learning_rate": 1.7336426263024897e-06,
+      "loss": 0.3052,
+      "step": 969
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.16818682849407196,
+      "learning_rate": 1.6937876315424705e-06,
+      "loss": 0.21,
+      "step": 970
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.17327508330345154,
+      "learning_rate": 1.6543882025923886e-06,
+      "loss": 0.2678,
+      "step": 971
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2144099920988083,
+      "learning_rate": 1.6154447110200255e-06,
+      "loss": 0.3411,
+      "step": 972
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.18619346618652344,
+      "learning_rate": 1.5769575240933422e-06,
+      "loss": 0.2253,
+      "step": 973
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.27366897463798523,
+      "learning_rate": 1.5389270047769578e-06,
+      "loss": 0.4898,
+      "step": 974
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.19534170627593994,
+      "learning_rate": 1.5013535117287647e-06,
+      "loss": 0.3085,
+      "step": 975
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.1718689352273941,
+      "learning_rate": 1.4642373992965364e-06,
+      "loss": 0.2004,
+      "step": 976
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.17628838121891022,
+      "learning_rate": 1.4275790175145908e-06,
+      "loss": 0.2296,
+      "step": 977
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.18485216796398163,
+      "learning_rate": 1.3913787121004719e-06,
+      "loss": 0.2718,
+      "step": 978
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.21213793754577637,
+      "learning_rate": 1.3556368244517114e-06,
+      "loss": 0.322,
+      "step": 979
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2154124677181244,
+      "learning_rate": 1.3203536916425841e-06,
+      "loss": 0.2837,
+      "step": 980
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.20204487442970276,
+      "learning_rate": 1.2855296464209688e-06,
+      "loss": 0.3087,
+      "step": 981
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2351517528295517,
+      "learning_rate": 1.2511650172051637e-06,
+      "loss": 0.4293,
+      "step": 982
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.22765348851680756,
+      "learning_rate": 1.217260128080816e-06,
+      "loss": 0.2933,
+      "step": 983
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.18321679532527924,
+      "learning_rate": 1.1838152987978578e-06,
+      "loss": 0.2817,
+      "step": 984
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.22217869758605957,
+      "learning_rate": 1.1508308447674975e-06,
+      "loss": 0.3046,
+      "step": 985
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.1905667781829834,
+      "learning_rate": 1.1183070770592441e-06,
+      "loss": 0.292,
+      "step": 986
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.17098748683929443,
+      "learning_rate": 1.0862443023979651e-06,
+      "loss": 0.2638,
+      "step": 987
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.20674315094947815,
+      "learning_rate": 1.0546428231609895e-06,
+      "loss": 0.2364,
+      "step": 988
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.23127688467502594,
+      "learning_rate": 1.0235029373752758e-06,
+      "loss": 0.3839,
+      "step": 989
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.19653750956058502,
+      "learning_rate": 9.928249387145982e-07,
+      "loss": 0.2689,
+      "step": 990
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.1779293417930603,
+      "learning_rate": 9.6260911649676e-07,
+      "loss": 0.2148,
+      "step": 991
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.1845840960741043,
+      "learning_rate": 9.32855755680867e-07,
+      "loss": 0.2274,
+      "step": 992
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.19880704581737518,
+      "learning_rate": 9.035651368646648e-07,
+      "loss": 0.3143,
+      "step": 993
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.23152576386928558,
+      "learning_rate": 8.747375362818666e-07,
+      "loss": 0.4096,
+      "step": 994
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.21079550683498383,
+      "learning_rate": 8.463732257995571e-07,
+      "loss": 0.327,
+      "step": 995
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.1455056518316269,
+      "learning_rate": 8.184724729156379e-07,
+      "loss": 0.2065,
+      "step": 996
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.23099635541439056,
+      "learning_rate": 7.910355407562742e-07,
+      "loss": 0.3063,
+      "step": 997
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2487201690673828,
+      "learning_rate": 7.640626880734581e-07,
+      "loss": 0.2217,
+      "step": 998
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.15706954896450043,
+      "learning_rate": 7.375541692425326e-07,
+      "loss": 0.2226,
+      "step": 999
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.26502522826194763,
+      "learning_rate": 7.115102342598101e-07,
+      "loss": 0.3473,
+      "step": 1000
+    },
+    {
+      "epoch": 0.95,
+      "eval_loss": 0.2819537818431854,
+      "eval_runtime": 356.0452,
+      "eval_samples_per_second": 0.958,
+      "eval_steps_per_second": 0.242,
+      "step": 1000
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.19536234438419342,
+      "learning_rate": 6.859311287402081e-07,
+      "loss": 0.2545,
+      "step": 1001
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.162049800157547,
+      "learning_rate": 6.608170939149283e-07,
+      "loss": 0.1906,
+      "step": 1002
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.20793919265270233,
+      "learning_rate": 6.361683666291973e-07,
+      "loss": 0.327,
+      "step": 1003
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2493748515844345,
+      "learning_rate": 6.119851793400189e-07,
+      "loss": 0.3237,
+      "step": 1004
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.22001808881759644,
+      "learning_rate": 5.882677601139919e-07,
+      "loss": 0.2669,
+      "step": 1005
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2077091932296753,
+      "learning_rate": 5.650163326251345e-07,
+      "loss": 0.2648,
+      "step": 1006
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.22986365854740143,
+      "learning_rate": 5.422311161528193e-07,
+      "loss": 0.3849,
+      "step": 1007
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.20212186872959137,
+      "learning_rate": 5.199123255796634e-07,
+      "loss": 0.275,
+      "step": 1008
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.15428291261196136,
+      "learning_rate": 4.980601713895305e-07,
+      "loss": 0.1793,
+      "step": 1009
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2178327590227127,
+      "learning_rate": 4.7667485966552685e-07,
+      "loss": 0.2409,
+      "step": 1010
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.19400638341903687,
+      "learning_rate": 4.55756592088058e-07,
+      "loss": 0.3414,
+      "step": 1011
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.23983286321163177,
+      "learning_rate": 4.353055659329419e-07,
+      "loss": 0.3887,
+      "step": 1012
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2387809157371521,
+      "learning_rate": 4.153219740695435e-07,
+      "loss": 0.3768,
+      "step": 1013
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.25704923272132874,
+      "learning_rate": 3.958060049589485e-07,
+      "loss": 0.3731,
+      "step": 1014
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.16717763245105743,
+      "learning_rate": 3.767578426521923e-07,
+      "loss": 0.2427,
+      "step": 1015
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.17705966532230377,
+      "learning_rate": 3.5817766678850615e-07,
+      "loss": 0.2551,
+      "step": 1016
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.1640429049730301,
+      "learning_rate": 3.4006565259366295e-07,
+      "loss": 0.253,
+      "step": 1017
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.19166415929794312,
+      "learning_rate": 3.224219708782894e-07,
+      "loss": 0.2827,
+      "step": 1018
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.20592908561229706,
+      "learning_rate": 3.052467880362675e-07,
+      "loss": 0.395,
+      "step": 1019
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.1967919021844864,
+      "learning_rate": 2.88540266043158e-07,
+      "loss": 0.3101,
+      "step": 1020
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.20888453722000122,
+      "learning_rate": 2.723025624546849e-07,
+      "loss": 0.3328,
+      "step": 1021
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.1885051280260086,
+      "learning_rate": 2.5653383040524227e-07,
+      "loss": 0.323,
+      "step": 1022
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.20738689601421356,
+      "learning_rate": 2.412342186064565e-07,
+      "loss": 0.2479,
+      "step": 1023
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.2259705662727356,
+      "learning_rate": 2.2640387134577058e-07,
+      "loss": 0.3722,
+      "step": 1024
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.1661883443593979,
+      "learning_rate": 2.1204292848509555e-07,
+      "loss": 0.2364,
+      "step": 1025
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.22001373767852783,
+      "learning_rate": 1.98151525459489e-07,
+      "loss": 0.2734,
+      "step": 1026
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.20108221471309662,
+      "learning_rate": 1.8472979327587292e-07,
+      "loss": 0.2135,
+      "step": 1027
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.20387016236782074,
+      "learning_rate": 1.717778585118013e-07,
+      "loss": 0.335,
+      "step": 1028
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2062751203775406,
+      "learning_rate": 1.5929584331427216e-07,
+      "loss": 0.3188,
+      "step": 1029
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.24346837401390076,
+      "learning_rate": 1.4728386539856754e-07,
+      "loss": 0.3692,
+      "step": 1030
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.21749813854694366,
+      "learning_rate": 1.3574203804713747e-07,
+      "loss": 0.4199,
+      "step": 1031
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.20605501532554626,
+      "learning_rate": 1.246704701085566e-07,
+      "loss": 0.2773,
+      "step": 1032
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.19620800018310547,
+      "learning_rate": 1.1406926599646373e-07,
+      "loss": 0.2831,
+      "step": 1033
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.21038460731506348,
+      "learning_rate": 1.0393852568860719e-07,
+      "loss": 0.243,
+      "step": 1034
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.17226473987102509,
+      "learning_rate": 9.427834472588992e-08,
+      "loss": 0.1471,
+      "step": 1035
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.21196429431438446,
+      "learning_rate": 8.508881421145365e-08,
+      "loss": 0.2825,
+      "step": 1036
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.19933149218559265,
+      "learning_rate": 7.637002080985168e-08,
+      "loss": 0.2474,
+      "step": 1037
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2043781280517578,
+      "learning_rate": 6.812204674619959e-08,
+      "loss": 0.3206,
+      "step": 1038
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.1937236487865448,
+      "learning_rate": 6.034496980542037e-08,
+      "loss": 0.2809,
+      "step": 1039
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2013004571199417,
+      "learning_rate": 5.303886333151153e-08,
+      "loss": 0.2785,
+      "step": 1040
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.19055452942848206,
+      "learning_rate": 4.620379622682358e-08,
+      "loss": 0.2804,
+      "step": 1041
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.23184803128242493,
+      "learning_rate": 3.9839832951465986e-08,
+      "loss": 0.2658,
+      "step": 1042
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.19381286203861237,
+      "learning_rate": 3.394703352263551e-08,
+      "loss": 0.2755,
+      "step": 1043
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2092810422182083,
+      "learning_rate": 2.8525453514099965e-08,
+      "loss": 0.3388,
+      "step": 1044
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.22127728164196014,
+      "learning_rate": 2.3575144055643094e-08,
+      "loss": 0.1961,
+      "step": 1045
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.20576807856559753,
+      "learning_rate": 1.9096151832609375e-08,
+      "loss": 0.2466,
+      "step": 1046
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.18368744850158691,
+      "learning_rate": 1.5088519085437734e-08,
+      "loss": 0.3002,
+      "step": 1047
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.1739194244146347,
+      "learning_rate": 1.1552283609272962e-08,
+      "loss": 0.2331,
+      "step": 1048
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.19566425681114197,
+      "learning_rate": 8.487478753615997e-09,
+      "loss": 0.3364,
+      "step": 1049
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.24177640676498413,
+      "learning_rate": 5.894133422001957e-09,
+      "loss": 0.3909,
+      "step": 1050
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.1735665202140808,
+      "learning_rate": 3.772272071722593e-09,
+      "loss": 0.2306,
+      "step": 1051
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.20549267530441284,
+      "learning_rate": 2.1219147136264382e-09,
+      "loss": 0.265,
+      "step": 1052
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.2152637243270874,
+      "learning_rate": 9.43076911874563e-10,
+      "loss": 0.3566,
+      "step": 1053
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.2164101004600525,
+      "learning_rate": 2.3576978384065584e-10,
+      "loss": 0.323,
+      "step": 1054
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.2054571509361267,
+      "learning_rate": 0.0,
+      "loss": 0.2629,
+      "step": 1055
+    },
+    {
+      "epoch": 1.0,
+      "step": 1055,
+      "total_flos": 1.133155957300265e+17,
+      "train_loss": 0.3136779415098977,
+      "train_runtime": 53935.6019,
+      "train_samples_per_second": 0.626,
+      "train_steps_per_second": 0.02
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 1055,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 100,
+  "total_flos": 1.133155957300265e+17,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}