diff --git a/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa-streaming_baseline_lora_agg_s512/README.md b/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa-streaming_baseline_lora_agg_s512/README.md
new file mode 100644
index 0000000000000000000000000000000000000000..d4576fe074287232d3836bf69c21d3f2593290d9
--- /dev/null
+++ b/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa-streaming_baseline_lora_agg_s512/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-streaming_baseline_lora_agg_s512/adapter_config.json b/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa-streaming_baseline_lora_agg_s512/adapter_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..6f3405393c29c080ebca3dfa161166fb5ee55547
--- /dev/null
+++ b/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa-streaming_baseline_lora_agg_s512/adapter_config.json
@@ -0,0 +1,26 @@
+{
+  "auto_mapping": null,
+  "base_model_name_or_path": "/data/wiedmann/hub/models--lmms-lab--llava-onevision-qwen2-7b-ov/snapshots/0b07bf7565e244cf4f39982249eafe8cd799d6dd",
+  "bias": "none",
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "lora_alpha": 256,
+  "lora_dropout": 0.05,
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "r": 128,
+  "revision": null,
+  "target_modules": [
+    "o_proj",
+    "down_proj",
+    "k_proj",
+    "gate_proj",
+    "v_proj",
+    "q_proj",
+    "up_proj"
+  ],
+  "task_type": "CAUSAL_LM"
+}
\ No newline at end of file
diff --git a/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa-streaming_baseline_lora_agg_s512/adapter_model.bin b/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa-streaming_baseline_lora_agg_s512/adapter_model.bin
new file mode 100644
index 0000000000000000000000000000000000000000..7dc1d5f101effd92b3296a2b92f2f6e1a281d645
--- /dev/null
+++ b/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa-streaming_baseline_lora_agg_s512/adapter_model.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:482a76f7bff433719bfe204e171b3361f0c36b7331b516ba8d1f718ff429713d
+size 692127130
diff --git a/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa-streaming_baseline_lora_agg_s512/config.json b/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa-streaming_baseline_lora_agg_s512/config.json
new file mode 100644
index 0000000000000000000000000000000000000000..e9f8e0e13da00f5eb71bf8a3460b68f319187baa
--- /dev/null
+++ b/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa-streaming_baseline_lora_agg_s512/config.json
@@ -0,0 +1,67 @@
+{
+  "_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": 1,
+  "mm_streaming_input_dim": 1152,
+  "mm_streaming_num_heads": 8,
+  "mm_streaming_num_layers": 4,
+  "mm_streaming_num_state_tokens": 512,
+  "mm_streaming_patches_per_frame": 729,
+  "mm_streaming_state_dim": 1152,
+  "mm_streaming_vision_chunk_size": 16,
+  "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-streaming_baseline_lora_agg_s512/generation_config.json b/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa-streaming_baseline_lora_agg_s512/generation_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..19a297221acb87418d4388a3decef2282c6d7316
--- /dev/null
+++ b/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa-streaming_baseline_lora_agg_s512/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-streaming_baseline_lora_agg_s512/non_lora_trainables.bin b/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa-streaming_baseline_lora_agg_s512/non_lora_trainables.bin
new file mode 100644
index 0000000000000000000000000000000000000000..e229f20225d783aeb0c762c5239ea1c91ff424e8
--- /dev/null
+++ b/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa-streaming_baseline_lora_agg_s512/non_lora_trainables.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d8e589d969f9ff8251c350ee9c1d3f9377f0d809b42724ba2b4bc5abbf7fa0f7
+size 207902844
diff --git a/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa-streaming_baseline_lora_agg_s512/trainer_state.json b/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa-streaming_baseline_lora_agg_s512/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..623902347b28884ecfa76f0cd8a12f089495fcdb
--- /dev/null
+++ b/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa-streaming_baseline_lora_agg_s512/trainer_state.json
@@ -0,0 +1,7492 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9999413730433253,
+  "eval_steps": 500,
+  "global_step": 1066,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 17.227703094482422,
+      "learning_rate": 3.125e-06,
+      "loss": 1.0838,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 11.005305290222168,
+      "learning_rate": 6.25e-06,
+      "loss": 1.1249,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 12.119623184204102,
+      "learning_rate": 9.375000000000001e-06,
+      "loss": 1.0595,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.532540321350098,
+      "learning_rate": 1.25e-05,
+      "loss": 0.9128,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.108962059020996,
+      "learning_rate": 1.5625e-05,
+      "loss": 0.8711,
+      "step": 5
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.5435802936553955,
+      "learning_rate": 1.8750000000000002e-05,
+      "loss": 0.6673,
+      "step": 6
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.0784573554992676,
+      "learning_rate": 2.1875e-05,
+      "loss": 0.5435,
+      "step": 7
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4409202337265015,
+      "learning_rate": 2.5e-05,
+      "loss": 0.4465,
+      "step": 8
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9887893199920654,
+      "learning_rate": 2.8125000000000003e-05,
+      "loss": 0.4325,
+      "step": 9
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9212615489959717,
+      "learning_rate": 3.125e-05,
+      "loss": 0.3307,
+      "step": 10
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0202654600143433,
+      "learning_rate": 3.4375e-05,
+      "loss": 0.2758,
+      "step": 11
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7716664671897888,
+      "learning_rate": 3.7500000000000003e-05,
+      "loss": 0.4347,
+      "step": 12
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8882746696472168,
+      "learning_rate": 4.0625000000000005e-05,
+      "loss": 0.5303,
+      "step": 13
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9287745356559753,
+      "learning_rate": 4.375e-05,
+      "loss": 0.3661,
+      "step": 14
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7302972674369812,
+      "learning_rate": 4.6875e-05,
+      "loss": 0.4894,
+      "step": 15
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7104817628860474,
+      "learning_rate": 5e-05,
+      "loss": 0.3584,
+      "step": 16
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5547098517417908,
+      "learning_rate": 5.3125000000000004e-05,
+      "loss": 0.2798,
+      "step": 17
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7443296909332275,
+      "learning_rate": 5.6250000000000005e-05,
+      "loss": 0.35,
+      "step": 18
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5101932287216187,
+      "learning_rate": 5.9375e-05,
+      "loss": 0.3115,
+      "step": 19
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6928537487983704,
+      "learning_rate": 6.25e-05,
+      "loss": 0.4309,
+      "step": 20
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5658664107322693,
+      "learning_rate": 6.562500000000001e-05,
+      "loss": 0.3284,
+      "step": 21
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.65761399269104,
+      "learning_rate": 6.875e-05,
+      "loss": 0.4236,
+      "step": 22
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5779470801353455,
+      "learning_rate": 7.1875e-05,
+      "loss": 0.3631,
+      "step": 23
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.570232629776001,
+      "learning_rate": 7.500000000000001e-05,
+      "loss": 0.3179,
+      "step": 24
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5996125936508179,
+      "learning_rate": 7.8125e-05,
+      "loss": 0.3948,
+      "step": 25
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6183748841285706,
+      "learning_rate": 8.125000000000001e-05,
+      "loss": 0.4139,
+      "step": 26
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.578999400138855,
+      "learning_rate": 8.4375e-05,
+      "loss": 0.398,
+      "step": 27
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7683339715003967,
+      "learning_rate": 8.75e-05,
+      "loss": 0.3236,
+      "step": 28
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5391294360160828,
+      "learning_rate": 9.062500000000001e-05,
+      "loss": 0.3597,
+      "step": 29
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6637127995491028,
+      "learning_rate": 9.375e-05,
+      "loss": 0.4444,
+      "step": 30
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.49874716997146606,
+      "learning_rate": 9.687500000000001e-05,
+      "loss": 0.3222,
+      "step": 31
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4726903736591339,
+      "learning_rate": 0.0001,
+      "loss": 0.2823,
+      "step": 32
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5357120633125305,
+      "learning_rate": 9.999976921990784e-05,
+      "loss": 0.3139,
+      "step": 33
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6340093016624451,
+      "learning_rate": 9.999907688176173e-05,
+      "loss": 0.4162,
+      "step": 34
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4693492650985718,
+      "learning_rate": 9.999792299195278e-05,
+      "loss": 0.2762,
+      "step": 35
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5481849312782288,
+      "learning_rate": 9.999630756113278e-05,
+      "loss": 0.3765,
+      "step": 36
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4791087806224823,
+      "learning_rate": 9.999423060421411e-05,
+      "loss": 0.3108,
+      "step": 37
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4711155295372009,
+      "learning_rate": 9.999169214036958e-05,
+      "loss": 0.3442,
+      "step": 38
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4753577709197998,
+      "learning_rate": 9.998869219303227e-05,
+      "loss": 0.3113,
+      "step": 39
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5480284094810486,
+      "learning_rate": 9.998523078989529e-05,
+      "loss": 0.3737,
+      "step": 40
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4719812870025635,
+      "learning_rate": 9.998130796291156e-05,
+      "loss": 0.3651,
+      "step": 41
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5594419836997986,
+      "learning_rate": 9.997692374829352e-05,
+      "loss": 0.3126,
+      "step": 42
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4275471568107605,
+      "learning_rate": 9.997207818651274e-05,
+      "loss": 0.2936,
+      "step": 43
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.43990057706832886,
+      "learning_rate": 9.996677132229957e-05,
+      "loss": 0.3226,
+      "step": 44
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.48392900824546814,
+      "learning_rate": 9.996100320464274e-05,
+      "loss": 0.3621,
+      "step": 45
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4566228687763214,
+      "learning_rate": 9.995477388678897e-05,
+      "loss": 0.2735,
+      "step": 46
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5327995419502258,
+      "learning_rate": 9.994808342624234e-05,
+      "loss": 0.3073,
+      "step": 47
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.40377333760261536,
+      "learning_rate": 9.994093188476382e-05,
+      "loss": 0.2833,
+      "step": 48
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4591901898384094,
+      "learning_rate": 9.993331932837079e-05,
+      "loss": 0.3515,
+      "step": 49
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4833613634109497,
+      "learning_rate": 9.992524582733629e-05,
+      "loss": 0.2718,
+      "step": 50
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4145682752132416,
+      "learning_rate": 9.991671145618846e-05,
+      "loss": 0.2361,
+      "step": 51
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.46532148122787476,
+      "learning_rate": 9.99077162937098e-05,
+      "loss": 0.3232,
+      "step": 52
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.49215883016586304,
+      "learning_rate": 9.989826042293652e-05,
+      "loss": 0.3338,
+      "step": 53
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4783603250980377,
+      "learning_rate": 9.988834393115767e-05,
+      "loss": 0.3179,
+      "step": 54
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5149618983268738,
+      "learning_rate": 9.987796690991439e-05,
+      "loss": 0.3614,
+      "step": 55
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4666965901851654,
+      "learning_rate": 9.98671294549991e-05,
+      "loss": 0.3194,
+      "step": 56
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.42810237407684326,
+      "learning_rate": 9.985583166645455e-05,
+      "loss": 0.2451,
+      "step": 57
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.419660747051239,
+      "learning_rate": 9.98440736485729e-05,
+      "loss": 0.2881,
+      "step": 58
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.40833503007888794,
+      "learning_rate": 9.983185550989487e-05,
+      "loss": 0.2195,
+      "step": 59
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4693005681037903,
+      "learning_rate": 9.981917736320851e-05,
+      "loss": 0.3611,
+      "step": 60
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.36646148562431335,
+      "learning_rate": 9.980603932554845e-05,
+      "loss": 0.2019,
+      "step": 61
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.39927858114242554,
+      "learning_rate": 9.979244151819453e-05,
+      "loss": 0.2467,
+      "step": 62
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4634983241558075,
+      "learning_rate": 9.97783840666709e-05,
+      "loss": 0.3597,
+      "step": 63
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4355047941207886,
+      "learning_rate": 9.976386710074478e-05,
+      "loss": 0.2993,
+      "step": 64
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4115259349346161,
+      "learning_rate": 9.974889075442521e-05,
+      "loss": 0.2293,
+      "step": 65
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4722943902015686,
+      "learning_rate": 9.97334551659619e-05,
+      "loss": 0.3061,
+      "step": 66
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4350064694881439,
+      "learning_rate": 9.971756047784393e-05,
+      "loss": 0.3101,
+      "step": 67
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.44171565771102905,
+      "learning_rate": 9.970120683679838e-05,
+      "loss": 0.3909,
+      "step": 68
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.45926347374916077,
+      "learning_rate": 9.968439439378905e-05,
+      "loss": 0.3899,
+      "step": 69
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.44425445795059204,
+      "learning_rate": 9.966712330401504e-05,
+      "loss": 0.3668,
+      "step": 70
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4987228810787201,
+      "learning_rate": 9.964939372690926e-05,
+      "loss": 0.4006,
+      "step": 71
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4817635715007782,
+      "learning_rate": 9.96312058261371e-05,
+      "loss": 0.3439,
+      "step": 72
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.43129855394363403,
+      "learning_rate": 9.961255976959473e-05,
+      "loss": 0.3223,
+      "step": 73
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4915347993373871,
+      "learning_rate": 9.959345572940771e-05,
+      "loss": 0.4157,
+      "step": 74
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.38367170095443726,
+      "learning_rate": 9.957389388192935e-05,
+      "loss": 0.2921,
+      "step": 75
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4222044050693512,
+      "learning_rate": 9.9553874407739e-05,
+      "loss": 0.267,
+      "step": 76
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.41161587834358215,
+      "learning_rate": 9.953339749164057e-05,
+      "loss": 0.3667,
+      "step": 77
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.44461458921432495,
+      "learning_rate": 9.951246332266057e-05,
+      "loss": 0.3618,
+      "step": 78
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.44937530159950256,
+      "learning_rate": 9.949107209404665e-05,
+      "loss": 0.4015,
+      "step": 79
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4108678698539734,
+      "learning_rate": 9.946922400326554e-05,
+      "loss": 0.374,
+      "step": 80
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4602833092212677,
+      "learning_rate": 9.944691925200145e-05,
+      "loss": 0.3695,
+      "step": 81
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4525197446346283,
+      "learning_rate": 9.942415804615406e-05,
+      "loss": 0.3845,
+      "step": 82
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4771604537963867,
+      "learning_rate": 9.940094059583671e-05,
+      "loss": 0.3861,
+      "step": 83
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4231971800327301,
+      "learning_rate": 9.937726711537442e-05,
+      "loss": 0.2658,
+      "step": 84
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4093407392501831,
+      "learning_rate": 9.93531378233019e-05,
+      "loss": 0.3377,
+      "step": 85
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.49191173911094666,
+      "learning_rate": 9.932855294236154e-05,
+      "loss": 0.3626,
+      "step": 86
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5030225515365601,
+      "learning_rate": 9.930351269950143e-05,
+      "loss": 0.4074,
+      "step": 87
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.47146856784820557,
+      "learning_rate": 9.927801732587312e-05,
+      "loss": 0.2998,
+      "step": 88
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3858545124530792,
+      "learning_rate": 9.925206705682962e-05,
+      "loss": 0.2586,
+      "step": 89
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5475620627403259,
+      "learning_rate": 9.92256621319231e-05,
+      "loss": 0.4691,
+      "step": 90
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.47144803404808044,
+      "learning_rate": 9.919880279490286e-05,
+      "loss": 0.3133,
+      "step": 91
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3688579201698303,
+      "learning_rate": 9.917148929371288e-05,
+      "loss": 0.1735,
+      "step": 92
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4857231080532074,
+      "learning_rate": 9.914372188048964e-05,
+      "loss": 0.3737,
+      "step": 93
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.38670048117637634,
+      "learning_rate": 9.911550081155983e-05,
+      "loss": 0.2712,
+      "step": 94
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4856467843055725,
+      "learning_rate": 9.908682634743784e-05,
+      "loss": 0.3925,
+      "step": 95
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.41404563188552856,
+      "learning_rate": 9.905769875282352e-05,
+      "loss": 0.2892,
+      "step": 96
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.42182257771492004,
+      "learning_rate": 9.902811829659961e-05,
+      "loss": 0.304,
+      "step": 97
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4935455620288849,
+      "learning_rate": 9.899808525182935e-05,
+      "loss": 0.3808,
+      "step": 98
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3772177994251251,
+      "learning_rate": 9.896759989575386e-05,
+      "loss": 0.2935,
+      "step": 99
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4526858925819397,
+      "learning_rate": 9.893666250978971e-05,
+      "loss": 0.3774,
+      "step": 100
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.32689258456230164,
+      "learning_rate": 9.890527337952617e-05,
+      "loss": 0.2034,
+      "step": 101
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4143233597278595,
+      "learning_rate": 9.887343279472272e-05,
+      "loss": 0.3017,
+      "step": 102
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4620525538921356,
+      "learning_rate": 9.884114104930628e-05,
+      "loss": 0.3201,
+      "step": 103
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3897378146648407,
+      "learning_rate": 9.880839844136854e-05,
+      "loss": 0.2629,
+      "step": 104
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.450862318277359,
+      "learning_rate": 9.877520527316317e-05,
+      "loss": 0.3982,
+      "step": 105
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.38423511385917664,
+      "learning_rate": 9.874156185110306e-05,
+      "loss": 0.3077,
+      "step": 106
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.47380977869033813,
+      "learning_rate": 9.870746848575751e-05,
+      "loss": 0.3188,
+      "step": 107
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.43851879239082336,
+      "learning_rate": 9.86729254918493e-05,
+      "loss": 0.2765,
+      "step": 108
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4031079113483429,
+      "learning_rate": 9.863793318825186e-05,
+      "loss": 0.2989,
+      "step": 109
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4153116047382355,
+      "learning_rate": 9.860249189798627e-05,
+      "loss": 0.4076,
+      "step": 110
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.35746559500694275,
+      "learning_rate": 9.856660194821829e-05,
+      "loss": 0.2245,
+      "step": 111
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4301232397556305,
+      "learning_rate": 9.853026367025535e-05,
+      "loss": 0.3171,
+      "step": 112
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.41360700130462646,
+      "learning_rate": 9.849347739954352e-05,
+      "loss": 0.2908,
+      "step": 113
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5251216292381287,
+      "learning_rate": 9.845624347566433e-05,
+      "loss": 0.466,
+      "step": 114
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3653389513492584,
+      "learning_rate": 9.841856224233174e-05,
+      "loss": 0.2221,
+      "step": 115
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.48610758781433105,
+      "learning_rate": 9.83804340473889e-05,
+      "loss": 0.4226,
+      "step": 116
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.41285499930381775,
+      "learning_rate": 9.83418592428049e-05,
+      "loss": 0.3015,
+      "step": 117
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4366799294948578,
+      "learning_rate": 9.830283818467163e-05,
+      "loss": 0.3452,
+      "step": 118
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.43274393677711487,
+      "learning_rate": 9.826337123320046e-05,
+      "loss": 0.3277,
+      "step": 119
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.43159425258636475,
+      "learning_rate": 9.822345875271883e-05,
+      "loss": 0.3384,
+      "step": 120
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4645746052265167,
+      "learning_rate": 9.818310111166699e-05,
+      "loss": 0.3445,
+      "step": 121
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.531746506690979,
+      "learning_rate": 9.814229868259452e-05,
+      "loss": 0.3606,
+      "step": 122
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.46642574667930603,
+      "learning_rate": 9.810105184215699e-05,
+      "loss": 0.4172,
+      "step": 123
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4294590353965759,
+      "learning_rate": 9.805936097111234e-05,
+      "loss": 0.2697,
+      "step": 124
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5032444596290588,
+      "learning_rate": 9.801722645431754e-05,
+      "loss": 0.4007,
+      "step": 125
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.42096102237701416,
+      "learning_rate": 9.797464868072488e-05,
+      "loss": 0.3039,
+      "step": 126
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.43327075242996216,
+      "learning_rate": 9.793162804337845e-05,
+      "loss": 0.321,
+      "step": 127
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4488172233104706,
+      "learning_rate": 9.788816493941051e-05,
+      "loss": 0.3466,
+      "step": 128
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4482726454734802,
+      "learning_rate": 9.784425977003784e-05,
+      "loss": 0.3556,
+      "step": 129
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.43109023571014404,
+      "learning_rate": 9.779991294055802e-05,
+      "loss": 0.3729,
+      "step": 130
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.36907973885536194,
+      "learning_rate": 9.775512486034563e-05,
+      "loss": 0.2535,
+      "step": 131
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5850874781608582,
+      "learning_rate": 9.770989594284857e-05,
+      "loss": 0.5767,
+      "step": 132
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4706707000732422,
+      "learning_rate": 9.766422660558421e-05,
+      "loss": 0.3897,
+      "step": 133
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.38747847080230713,
+      "learning_rate": 9.761811727013548e-05,
+      "loss": 0.2789,
+      "step": 134
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.37999868392944336,
+      "learning_rate": 9.757156836214706e-05,
+      "loss": 0.3114,
+      "step": 135
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4177742302417755,
+      "learning_rate": 9.752458031132141e-05,
+      "loss": 0.3577,
+      "step": 136
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.39297688007354736,
+      "learning_rate": 9.747715355141478e-05,
+      "loss": 0.2334,
+      "step": 137
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3654375672340393,
+      "learning_rate": 9.742928852023325e-05,
+      "loss": 0.3229,
+      "step": 138
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.432533323764801,
+      "learning_rate": 9.73809856596287e-05,
+      "loss": 0.3597,
+      "step": 139
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4045366644859314,
+      "learning_rate": 9.733224541549464e-05,
+      "loss": 0.3186,
+      "step": 140
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4264724552631378,
+      "learning_rate": 9.728306823776221e-05,
+      "loss": 0.329,
+      "step": 141
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3992307484149933,
+      "learning_rate": 9.723345458039594e-05,
+      "loss": 0.3513,
+      "step": 142
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.339410662651062,
+      "learning_rate": 9.718340490138965e-05,
+      "loss": 0.188,
+      "step": 143
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4419720470905304,
+      "learning_rate": 9.713291966276206e-05,
+      "loss": 0.3414,
+      "step": 144
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4582366645336151,
+      "learning_rate": 9.708199933055272e-05,
+      "loss": 0.4222,
+      "step": 145
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4727243483066559,
+      "learning_rate": 9.70306443748176e-05,
+      "loss": 0.3994,
+      "step": 146
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3670468330383301,
+      "learning_rate": 9.697885526962474e-05,
+      "loss": 0.2829,
+      "step": 147
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.40513771772384644,
+      "learning_rate": 9.692663249304992e-05,
+      "loss": 0.333,
+      "step": 148
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.48286768794059753,
+      "learning_rate": 9.687397652717223e-05,
+      "loss": 0.4656,
+      "step": 149
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.45031237602233887,
+      "learning_rate": 9.682088785806963e-05,
+      "loss": 0.3733,
+      "step": 150
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3925604224205017,
+      "learning_rate": 9.67673669758144e-05,
+      "loss": 0.2769,
+      "step": 151
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4268004894256592,
+      "learning_rate": 9.671341437446877e-05,
+      "loss": 0.3381,
+      "step": 152
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5197659730911255,
+      "learning_rate": 9.665903055208014e-05,
+      "loss": 0.4432,
+      "step": 153
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.44246628880500793,
+      "learning_rate": 9.660421601067666e-05,
+      "loss": 0.3358,
+      "step": 154
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.36518609523773193,
+      "learning_rate": 9.654897125626252e-05,
+      "loss": 0.2827,
+      "step": 155
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.37158283591270447,
+      "learning_rate": 9.649329679881334e-05,
+      "loss": 0.2822,
+      "step": 156
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.31241849064826965,
+      "learning_rate": 9.643719315227133e-05,
+      "loss": 0.1952,
+      "step": 157
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4282618761062622,
+      "learning_rate": 9.63806608345407e-05,
+      "loss": 0.3535,
+      "step": 158
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4605000913143158,
+      "learning_rate": 9.632370036748279e-05,
+      "loss": 0.3793,
+      "step": 159
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4260792136192322,
+      "learning_rate": 9.626631227691127e-05,
+      "loss": 0.2857,
+      "step": 160
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.39072418212890625,
+      "learning_rate": 9.62084970925873e-05,
+      "loss": 0.2469,
+      "step": 161
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.43903443217277527,
+      "learning_rate": 9.615025534821462e-05,
+      "loss": 0.2761,
+      "step": 162
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4896968603134155,
+      "learning_rate": 9.609158758143464e-05,
+      "loss": 0.3861,
+      "step": 163
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4410502314567566,
+      "learning_rate": 9.603249433382144e-05,
+      "loss": 0.2861,
+      "step": 164
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.45812466740608215,
+      "learning_rate": 9.597297615087685e-05,
+      "loss": 0.271,
+      "step": 165
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3951738476753235,
+      "learning_rate": 9.591303358202535e-05,
+      "loss": 0.2225,
+      "step": 166
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.46651744842529297,
+      "learning_rate": 9.585266718060897e-05,
+      "loss": 0.3533,
+      "step": 167
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4054509401321411,
+      "learning_rate": 9.579187750388227e-05,
+      "loss": 0.2154,
+      "step": 168
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.433390736579895,
+      "learning_rate": 9.573066511300714e-05,
+      "loss": 0.3388,
+      "step": 169
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.43540117144584656,
+      "learning_rate": 9.566903057304764e-05,
+      "loss": 0.3604,
+      "step": 170
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4319068193435669,
+      "learning_rate": 9.560697445296474e-05,
+      "loss": 0.3286,
+      "step": 171
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5118317604064941,
+      "learning_rate": 9.554449732561113e-05,
+      "loss": 0.3459,
+      "step": 172
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.45201918482780457,
+      "learning_rate": 9.548159976772592e-05,
+      "loss": 0.3469,
+      "step": 173
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.47841161489486694,
+      "learning_rate": 9.541828235992926e-05,
+      "loss": 0.356,
+      "step": 174
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3838687539100647,
+      "learning_rate": 9.535454568671704e-05,
+      "loss": 0.2688,
+      "step": 175
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4381600022315979,
+      "learning_rate": 9.529039033645548e-05,
+      "loss": 0.2392,
+      "step": 176
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.38000673055648804,
+      "learning_rate": 9.522581690137567e-05,
+      "loss": 0.3034,
+      "step": 177
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3562740981578827,
+      "learning_rate": 9.516082597756815e-05,
+      "loss": 0.232,
+      "step": 178
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4523466229438782,
+      "learning_rate": 9.509541816497737e-05,
+      "loss": 0.4132,
+      "step": 179
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4095892906188965,
+      "learning_rate": 9.50295940673962e-05,
+      "loss": 0.2558,
+      "step": 180
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4435361921787262,
+      "learning_rate": 9.496335429246026e-05,
+      "loss": 0.3918,
+      "step": 181
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.426711767911911,
+      "learning_rate": 9.489669945164242e-05,
+      "loss": 0.328,
+      "step": 182
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.35389336943626404,
+      "learning_rate": 9.482963016024709e-05,
+      "loss": 0.2219,
+      "step": 183
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.36992374062538147,
+      "learning_rate": 9.476214703740454e-05,
+      "loss": 0.2427,
+      "step": 184
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5090086460113525,
+      "learning_rate": 9.469425070606524e-05,
+      "loss": 0.5443,
+      "step": 185
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.42029502987861633,
+      "learning_rate": 9.462594179299406e-05,
+      "loss": 0.2334,
+      "step": 186
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.42248085141181946,
+      "learning_rate": 9.45572209287645e-05,
+      "loss": 0.3449,
+      "step": 187
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5659753084182739,
+      "learning_rate": 9.44880887477528e-05,
+      "loss": 0.381,
+      "step": 188
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.45616188645362854,
+      "learning_rate": 9.441854588813228e-05,
+      "loss": 0.29,
+      "step": 189
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.394564688205719,
+      "learning_rate": 9.43485929918672e-05,
+      "loss": 0.3241,
+      "step": 190
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.503383994102478,
+      "learning_rate": 9.427823070470699e-05,
+      "loss": 0.4226,
+      "step": 191
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3788263201713562,
+      "learning_rate": 9.420745967618026e-05,
+      "loss": 0.2751,
+      "step": 192
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.30289629101753235,
+      "learning_rate": 9.413628055958878e-05,
+      "loss": 0.188,
+      "step": 193
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3985527753829956,
+      "learning_rate": 9.406469401200151e-05,
+      "loss": 0.3384,
+      "step": 194
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4570313096046448,
+      "learning_rate": 9.399270069424842e-05,
+      "loss": 0.3379,
+      "step": 195
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.45788517594337463,
+      "learning_rate": 9.392030127091452e-05,
+      "loss": 0.2914,
+      "step": 196
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.39778071641921997,
+      "learning_rate": 9.384749641033359e-05,
+      "loss": 0.28,
+      "step": 197
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4381253719329834,
+      "learning_rate": 9.377428678458214e-05,
+      "loss": 0.356,
+      "step": 198
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4055528938770294,
+      "learning_rate": 9.370067306947316e-05,
+      "loss": 0.2767,
+      "step": 199
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.47014090418815613,
+      "learning_rate": 9.362665594454984e-05,
+      "loss": 0.4312,
+      "step": 200
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4234636425971985,
+      "learning_rate": 9.355223609307933e-05,
+      "loss": 0.2813,
+      "step": 201
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.47207698225975037,
+      "learning_rate": 9.347741420204643e-05,
+      "loss": 0.3601,
+      "step": 202
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4559366703033447,
+      "learning_rate": 9.340219096214727e-05,
+      "loss": 0.29,
+      "step": 203
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.46230751276016235,
+      "learning_rate": 9.33265670677829e-05,
+      "loss": 0.37,
+      "step": 204
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.43081924319267273,
+      "learning_rate": 9.325054321705289e-05,
+      "loss": 0.3043,
+      "step": 205
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.43205636739730835,
+      "learning_rate": 9.317412011174886e-05,
+      "loss": 0.3097,
+      "step": 206
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4734487235546112,
+      "learning_rate": 9.309729845734813e-05,
+      "loss": 0.4025,
+      "step": 207
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3640989661216736,
+      "learning_rate": 9.302007896300698e-05,
+      "loss": 0.2752,
+      "step": 208
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.41770240664482117,
+      "learning_rate": 9.29424623415543e-05,
+      "loss": 0.316,
+      "step": 209
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.48017314076423645,
+      "learning_rate": 9.286444930948496e-05,
+      "loss": 0.3579,
+      "step": 210
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.37423351407051086,
+      "learning_rate": 9.278604058695313e-05,
+      "loss": 0.2375,
+      "step": 211
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4382573962211609,
+      "learning_rate": 9.270723689776568e-05,
+      "loss": 0.3206,
+      "step": 212
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4625137150287628,
+      "learning_rate": 9.262803896937555e-05,
+      "loss": 0.404,
+      "step": 213
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4267883896827698,
+      "learning_rate": 9.254844753287493e-05,
+      "loss": 0.2745,
+      "step": 214
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.48072972893714905,
+      "learning_rate": 9.24684633229886e-05,
+      "loss": 0.3158,
+      "step": 215
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5718352794647217,
+      "learning_rate": 9.238808707806706e-05,
+      "loss": 0.357,
+      "step": 216
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4857891798019409,
+      "learning_rate": 9.230731954007983e-05,
+      "loss": 0.3711,
+      "step": 217
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5256755352020264,
+      "learning_rate": 9.222616145460849e-05,
+      "loss": 0.4199,
+      "step": 218
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.40874987840652466,
+      "learning_rate": 9.214461357083985e-05,
+      "loss": 0.2117,
+      "step": 219
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.36349964141845703,
+      "learning_rate": 9.206267664155907e-05,
+      "loss": 0.249,
+      "step": 220
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4281284511089325,
+      "learning_rate": 9.198035142314259e-05,
+      "loss": 0.3793,
+      "step": 221
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.45442864298820496,
+      "learning_rate": 9.189763867555129e-05,
+      "loss": 0.3745,
+      "step": 222
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.445965975522995,
+      "learning_rate": 9.181453916232339e-05,
+      "loss": 0.2544,
+      "step": 223
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4232853651046753,
+      "learning_rate": 9.173105365056742e-05,
+      "loss": 0.3489,
+      "step": 224
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4950341284275055,
+      "learning_rate": 9.164718291095515e-05,
+      "loss": 0.3743,
+      "step": 225
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4112965762615204,
+      "learning_rate": 9.156292771771446e-05,
+      "loss": 0.338,
+      "step": 226
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3986303508281708,
+      "learning_rate": 9.14782888486222e-05,
+      "loss": 0.3071,
+      "step": 227
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.38412225246429443,
+      "learning_rate": 9.1393267084997e-05,
+      "loss": 0.2964,
+      "step": 228
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4604569673538208,
+      "learning_rate": 9.130786321169209e-05,
+      "loss": 0.3754,
+      "step": 229
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4137207865715027,
+      "learning_rate": 9.122207801708802e-05,
+      "loss": 0.3311,
+      "step": 230
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.37709876894950867,
+      "learning_rate": 9.113591229308538e-05,
+      "loss": 0.2439,
+      "step": 231
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.42634347081184387,
+      "learning_rate": 9.104936683509755e-05,
+      "loss": 0.3066,
+      "step": 232
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.37276193499565125,
+      "learning_rate": 9.096244244204324e-05,
+      "loss": 0.3085,
+      "step": 233
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.42944371700286865,
+      "learning_rate": 9.087513991633924e-05,
+      "loss": 0.3185,
+      "step": 234
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.41260093450546265,
+      "learning_rate": 9.078746006389298e-05,
+      "loss": 0.3474,
+      "step": 235
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.32577750086784363,
+      "learning_rate": 9.069940369409499e-05,
+      "loss": 0.1917,
+      "step": 236
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.49313199520111084,
+      "learning_rate": 9.061097161981159e-05,
+      "loss": 0.4111,
+      "step": 237
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4716200828552246,
+      "learning_rate": 9.052216465737726e-05,
+      "loss": 0.3746,
+      "step": 238
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4609526991844177,
+      "learning_rate": 9.043298362658714e-05,
+      "loss": 0.3682,
+      "step": 239
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.37921881675720215,
+      "learning_rate": 9.034342935068952e-05,
+      "loss": 0.1989,
+      "step": 240
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.43589839339256287,
+      "learning_rate": 9.025350265637815e-05,
+      "loss": 0.3716,
+      "step": 241
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.45863112807273865,
+      "learning_rate": 9.016320437378466e-05,
+      "loss": 0.3152,
+      "step": 242
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4138740003108978,
+      "learning_rate": 9.007253533647089e-05,
+      "loss": 0.3435,
+      "step": 243
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4855858385562897,
+      "learning_rate": 8.998149638142119e-05,
+      "loss": 0.3306,
+      "step": 244
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.49547135829925537,
+      "learning_rate": 8.98900883490347e-05,
+      "loss": 0.4374,
+      "step": 245
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3813140094280243,
+      "learning_rate": 8.979831208311758e-05,
+      "loss": 0.2895,
+      "step": 246
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4578895568847656,
+      "learning_rate": 8.970616843087524e-05,
+      "loss": 0.3151,
+      "step": 247
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.36515703797340393,
+      "learning_rate": 8.96136582429045e-05,
+      "loss": 0.2167,
+      "step": 248
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.46578672528266907,
+      "learning_rate": 8.952078237318575e-05,
+      "loss": 0.3441,
+      "step": 249
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.42086178064346313,
+      "learning_rate": 8.942754167907507e-05,
+      "loss": 0.2692,
+      "step": 250
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5085225105285645,
+      "learning_rate": 8.933393702129628e-05,
+      "loss": 0.2555,
+      "step": 251
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.44771888852119446,
+      "learning_rate": 8.923996926393305e-05,
+      "loss": 0.3989,
+      "step": 252
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.42549923062324524,
+      "learning_rate": 8.91456392744209e-05,
+      "loss": 0.3679,
+      "step": 253
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.48210227489471436,
+      "learning_rate": 8.905094792353917e-05,
+      "loss": 0.3376,
+      "step": 254
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4970339238643646,
+      "learning_rate": 8.895589608540297e-05,
+      "loss": 0.3605,
+      "step": 255
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3592231869697571,
+      "learning_rate": 8.886048463745525e-05,
+      "loss": 0.1728,
+      "step": 256
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4326871633529663,
+      "learning_rate": 8.876471446045847e-05,
+      "loss": 0.2955,
+      "step": 257
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.42976316809654236,
+      "learning_rate": 8.866858643848665e-05,
+      "loss": 0.2432,
+      "step": 258
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.394156277179718,
+      "learning_rate": 8.857210145891715e-05,
+      "loss": 0.2881,
+      "step": 259
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.41744494438171387,
+      "learning_rate": 8.847526041242246e-05,
+      "loss": 0.1836,
+      "step": 260
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.40021783113479614,
+      "learning_rate": 8.8378064192962e-05,
+      "loss": 0.2429,
+      "step": 261
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4498775601387024,
+      "learning_rate": 8.82805136977739e-05,
+      "loss": 0.3673,
+      "step": 262
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.490108847618103,
+      "learning_rate": 8.818260982736661e-05,
+      "loss": 0.3611,
+      "step": 263
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.39929094910621643,
+      "learning_rate": 8.808435348551071e-05,
+      "loss": 0.2208,
+      "step": 264
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3677091896533966,
+      "learning_rate": 8.798574557923053e-05,
+      "loss": 0.2226,
+      "step": 265
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3725321888923645,
+      "learning_rate": 8.788678701879573e-05,
+      "loss": 0.3111,
+      "step": 266
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3611381947994232,
+      "learning_rate": 8.778747871771292e-05,
+      "loss": 0.237,
+      "step": 267
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4395051598548889,
+      "learning_rate": 8.768782159271727e-05,
+      "loss": 0.3296,
+      "step": 268
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.40998443961143494,
+      "learning_rate": 8.758781656376398e-05,
+      "loss": 0.249,
+      "step": 269
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.46020665764808655,
+      "learning_rate": 8.748746455401986e-05,
+      "loss": 0.3302,
+      "step": 270
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.46815627813339233,
+      "learning_rate": 8.738676648985476e-05,
+      "loss": 0.3139,
+      "step": 271
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4065473973751068,
+      "learning_rate": 8.7285723300833e-05,
+      "loss": 0.3707,
+      "step": 272
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3898315727710724,
+      "learning_rate": 8.718433591970485e-05,
+      "loss": 0.288,
+      "step": 273
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.478915810585022,
+      "learning_rate": 8.708260528239788e-05,
+      "loss": 0.3895,
+      "step": 274
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4072932004928589,
+      "learning_rate": 8.698053232800832e-05,
+      "loss": 0.2634,
+      "step": 275
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.41820770502090454,
+      "learning_rate": 8.68781179987924e-05,
+      "loss": 0.3607,
+      "step": 276
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3776008188724518,
+      "learning_rate": 8.677536324015765e-05,
+      "loss": 0.278,
+      "step": 277
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3983832001686096,
+      "learning_rate": 8.667226900065419e-05,
+      "loss": 0.3231,
+      "step": 278
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.472797691822052,
+      "learning_rate": 8.656883623196592e-05,
+      "loss": 0.3177,
+      "step": 279
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.44463932514190674,
+      "learning_rate": 8.646506588890183e-05,
+      "loss": 0.3068,
+      "step": 280
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4819587469100952,
+      "learning_rate": 8.636095892938707e-05,
+      "loss": 0.3476,
+      "step": 281
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.45399779081344604,
+      "learning_rate": 8.62565163144542e-05,
+      "loss": 0.4011,
+      "step": 282
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3941989243030548,
+      "learning_rate": 8.615173900823426e-05,
+      "loss": 0.2842,
+      "step": 283
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.39772456884384155,
+      "learning_rate": 8.60466279779479e-05,
+      "loss": 0.288,
+      "step": 284
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3934876322746277,
+      "learning_rate": 8.594118419389647e-05,
+      "loss": 0.3304,
+      "step": 285
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4301411807537079,
+      "learning_rate": 8.583540862945301e-05,
+      "loss": 0.307,
+      "step": 286
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3396373391151428,
+      "learning_rate": 8.57293022610533e-05,
+      "loss": 0.2122,
+      "step": 287
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4123905301094055,
+      "learning_rate": 8.562286606818684e-05,
+      "loss": 0.238,
+      "step": 288
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5258879661560059,
+      "learning_rate": 8.55161010333878e-05,
+      "loss": 0.5028,
+      "step": 289
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4868035316467285,
+      "learning_rate": 8.540900814222598e-05,
+      "loss": 0.3512,
+      "step": 290
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.49566733837127686,
+      "learning_rate": 8.530158838329765e-05,
+      "loss": 0.3262,
+      "step": 291
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5323141813278198,
+      "learning_rate": 8.519384274821649e-05,
+      "loss": 0.4359,
+      "step": 292
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.41219809651374817,
+      "learning_rate": 8.508577223160442e-05,
+      "loss": 0.259,
+      "step": 293
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.46348467469215393,
+      "learning_rate": 8.497737783108238e-05,
+      "loss": 0.3208,
+      "step": 294
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6380013227462769,
+      "learning_rate": 8.486866054726114e-05,
+      "loss": 0.3092,
+      "step": 295
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4216512441635132,
+      "learning_rate": 8.475962138373213e-05,
+      "loss": 0.2662,
+      "step": 296
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4462231397628784,
+      "learning_rate": 8.465026134705805e-05,
+      "loss": 0.3089,
+      "step": 297
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.43319839239120483,
+      "learning_rate": 8.454058144676366e-05,
+      "loss": 0.2952,
+      "step": 298
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3860713243484497,
+      "learning_rate": 8.443058269532651e-05,
+      "loss": 0.3128,
+      "step": 299
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4144044518470764,
+      "learning_rate": 8.432026610816745e-05,
+      "loss": 0.2749,
+      "step": 300
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.437470018863678,
+      "learning_rate": 8.420963270364137e-05,
+      "loss": 0.274,
+      "step": 301
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.464390367269516,
+      "learning_rate": 8.409868350302774e-05,
+      "loss": 0.3264,
+      "step": 302
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.45976755023002625,
+      "learning_rate": 8.398741953052127e-05,
+      "loss": 0.2807,
+      "step": 303
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.44180965423583984,
+      "learning_rate": 8.387584181322233e-05,
+      "loss": 0.3016,
+      "step": 304
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4997859597206116,
+      "learning_rate": 8.376395138112754e-05,
+      "loss": 0.3449,
+      "step": 305
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4143840968608856,
+      "learning_rate": 8.365174926712032e-05,
+      "loss": 0.3417,
+      "step": 306
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4447815418243408,
+      "learning_rate": 8.353923650696118e-05,
+      "loss": 0.3467,
+      "step": 307
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4479486644268036,
+      "learning_rate": 8.342641413927837e-05,
+      "loss": 0.3673,
+      "step": 308
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.44752955436706543,
+      "learning_rate": 8.331328320555812e-05,
+      "loss": 0.3458,
+      "step": 309
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5090377926826477,
+      "learning_rate": 8.319984475013512e-05,
+      "loss": 0.3991,
+      "step": 310
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.43693166971206665,
+      "learning_rate": 8.308609982018286e-05,
+      "loss": 0.3353,
+      "step": 311
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.38203293085098267,
+      "learning_rate": 8.297204946570398e-05,
+      "loss": 0.2731,
+      "step": 312
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4455333650112152,
+      "learning_rate": 8.285769473952052e-05,
+      "loss": 0.2632,
+      "step": 313
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4769796133041382,
+      "learning_rate": 8.274303669726426e-05,
+      "loss": 0.3187,
+      "step": 314
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.45285284519195557,
+      "learning_rate": 8.262807639736692e-05,
+      "loss": 0.4072,
+      "step": 315
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.47556719183921814,
+      "learning_rate": 8.251281490105045e-05,
+      "loss": 0.3732,
+      "step": 316
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.385691374540329,
+      "learning_rate": 8.239725327231721e-05,
+      "loss": 0.3075,
+      "step": 317
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.41788721084594727,
+      "learning_rate": 8.228139257794012e-05,
+      "loss": 0.3112,
+      "step": 318
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4231437146663666,
+      "learning_rate": 8.216523388745287e-05,
+      "loss": 0.3517,
+      "step": 319
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4116215407848358,
+      "learning_rate": 8.204877827313997e-05,
+      "loss": 0.3299,
+      "step": 320
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3918690085411072,
+      "learning_rate": 8.193202681002692e-05,
+      "loss": 0.337,
+      "step": 321
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.28665515780448914,
+      "learning_rate": 8.181498057587027e-05,
+      "loss": 0.1694,
+      "step": 322
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3561767637729645,
+      "learning_rate": 8.169764065114764e-05,
+      "loss": 0.2718,
+      "step": 323
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4175536632537842,
+      "learning_rate": 8.158000811904778e-05,
+      "loss": 0.2497,
+      "step": 324
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.41284066438674927,
+      "learning_rate": 8.146208406546053e-05,
+      "loss": 0.3359,
+      "step": 325
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4605599343776703,
+      "learning_rate": 8.134386957896688e-05,
+      "loss": 0.4093,
+      "step": 326
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.45733022689819336,
+      "learning_rate": 8.122536575082882e-05,
+      "loss": 0.4192,
+      "step": 327
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.46098488569259644,
+      "learning_rate": 8.110657367497933e-05,
+      "loss": 0.2819,
+      "step": 328
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3386194407939911,
+      "learning_rate": 8.098749444801224e-05,
+      "loss": 0.2341,
+      "step": 329
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.42128705978393555,
+      "learning_rate": 8.08681291691722e-05,
+      "loss": 0.3533,
+      "step": 330
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.43376579880714417,
+      "learning_rate": 8.074847894034434e-05,
+      "loss": 0.2797,
+      "step": 331
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.41651150584220886,
+      "learning_rate": 8.062854486604435e-05,
+      "loss": 0.256,
+      "step": 332
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.38273122906684875,
+      "learning_rate": 8.050832805340806e-05,
+      "loss": 0.2789,
+      "step": 333
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.38854745030403137,
+      "learning_rate": 8.038782961218136e-05,
+      "loss": 0.328,
+      "step": 334
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.49828365445137024,
+      "learning_rate": 8.026705065470996e-05,
+      "loss": 0.3411,
+      "step": 335
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.39919033646583557,
+      "learning_rate": 8.014599229592894e-05,
+      "loss": 0.3047,
+      "step": 336
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.440532922744751,
+      "learning_rate": 8.002465565335271e-05,
+      "loss": 0.2559,
+      "step": 337
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4270270764827728,
+      "learning_rate": 7.990304184706455e-05,
+      "loss": 0.2757,
+      "step": 338
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4185905158519745,
+      "learning_rate": 7.978115199970621e-05,
+      "loss": 0.3783,
+      "step": 339
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.38478323817253113,
+      "learning_rate": 7.965898723646776e-05,
+      "loss": 0.3041,
+      "step": 340
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5213392376899719,
+      "learning_rate": 7.953654868507699e-05,
+      "loss": 0.418,
+      "step": 341
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.42577335238456726,
+      "learning_rate": 7.941383747578912e-05,
+      "loss": 0.3395,
+      "step": 342
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4232735335826874,
+      "learning_rate": 7.929085474137629e-05,
+      "loss": 0.3455,
+      "step": 343
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.36636146903038025,
+      "learning_rate": 7.91676016171172e-05,
+      "loss": 0.3177,
+      "step": 344
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.46816015243530273,
+      "learning_rate": 7.904407924078654e-05,
+      "loss": 0.4576,
+      "step": 345
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.43102020025253296,
+      "learning_rate": 7.892028875264451e-05,
+      "loss": 0.3307,
+      "step": 346
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4004449248313904,
+      "learning_rate": 7.879623129542633e-05,
+      "loss": 0.2942,
+      "step": 347
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4232417941093445,
+      "learning_rate": 7.867190801433166e-05,
+      "loss": 0.3691,
+      "step": 348
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4015827775001526,
+      "learning_rate": 7.854732005701402e-05,
+      "loss": 0.345,
+      "step": 349
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3961988389492035,
+      "learning_rate": 7.842246857357023e-05,
+      "loss": 0.2919,
+      "step": 350
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3362734317779541,
+      "learning_rate": 7.829735471652978e-05,
+      "loss": 0.2247,
+      "step": 351
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.37961891293525696,
+      "learning_rate": 7.817197964084411e-05,
+      "loss": 0.2637,
+      "step": 352
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3325817883014679,
+      "learning_rate": 7.804634450387616e-05,
+      "loss": 0.1882,
+      "step": 353
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.40928661823272705,
+      "learning_rate": 7.792045046538941e-05,
+      "loss": 0.3098,
+      "step": 354
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4568662941455841,
+      "learning_rate": 7.77942986875374e-05,
+      "loss": 0.419,
+      "step": 355
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.395332008600235,
+      "learning_rate": 7.766789033485287e-05,
+      "loss": 0.3353,
+      "step": 356
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4348996579647064,
+      "learning_rate": 7.75412265742371e-05,
+      "loss": 0.3883,
+      "step": 357
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.432668536901474,
+      "learning_rate": 7.741430857494904e-05,
+      "loss": 0.3164,
+      "step": 358
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.43673163652420044,
+      "learning_rate": 7.728713750859458e-05,
+      "loss": 0.3445,
+      "step": 359
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.42592430114746094,
+      "learning_rate": 7.715971454911577e-05,
+      "loss": 0.3516,
+      "step": 360
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4502371549606323,
+      "learning_rate": 7.703204087277988e-05,
+      "loss": 0.3562,
+      "step": 361
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3586786091327667,
+      "learning_rate": 7.690411765816864e-05,
+      "loss": 0.2582,
+      "step": 362
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3953322470188141,
+      "learning_rate": 7.677594608616729e-05,
+      "loss": 0.3064,
+      "step": 363
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3941778838634491,
+      "learning_rate": 7.66475273399537e-05,
+      "loss": 0.2823,
+      "step": 364
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3739788234233856,
+      "learning_rate": 7.651886260498751e-05,
+      "loss": 0.267,
+      "step": 365
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.43802833557128906,
+      "learning_rate": 7.638995306899908e-05,
+      "loss": 0.3207,
+      "step": 366
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3720547556877136,
+      "learning_rate": 7.626079992197857e-05,
+      "loss": 0.2774,
+      "step": 367
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.42035406827926636,
+      "learning_rate": 7.613140435616503e-05,
+      "loss": 0.3196,
+      "step": 368
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.401703804731369,
+      "learning_rate": 7.600176756603525e-05,
+      "loss": 0.2521,
+      "step": 369
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3969535827636719,
+      "learning_rate": 7.587189074829284e-05,
+      "loss": 0.3049,
+      "step": 370
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.33402374386787415,
+      "learning_rate": 7.57417751018572e-05,
+      "loss": 0.2278,
+      "step": 371
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.37405410408973694,
+      "learning_rate": 7.561142182785233e-05,
+      "loss": 0.2392,
+      "step": 372
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.42111271619796753,
+      "learning_rate": 7.548083212959588e-05,
+      "loss": 0.308,
+      "step": 373
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.421460896730423,
+      "learning_rate": 7.535000721258791e-05,
+      "loss": 0.2638,
+      "step": 374
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.42545416951179504,
+      "learning_rate": 7.521894828449994e-05,
+      "loss": 0.3674,
+      "step": 375
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4488581418991089,
+      "learning_rate": 7.508765655516358e-05,
+      "loss": 0.3428,
+      "step": 376
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.390317440032959,
+      "learning_rate": 7.495613323655953e-05,
+      "loss": 0.3062,
+      "step": 377
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4021550714969635,
+      "learning_rate": 7.482437954280635e-05,
+      "loss": 0.2613,
+      "step": 378
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4348108768463135,
+      "learning_rate": 7.469239669014923e-05,
+      "loss": 0.3695,
+      "step": 379
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.39812782406806946,
+      "learning_rate": 7.456018589694873e-05,
+      "loss": 0.2887,
+      "step": 380
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5493481755256653,
+      "learning_rate": 7.442774838366965e-05,
+      "loss": 0.3927,
+      "step": 381
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2790215313434601,
+      "learning_rate": 7.429508537286963e-05,
+      "loss": 0.1896,
+      "step": 382
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.43105021119117737,
+      "learning_rate": 7.416219808918794e-05,
+      "loss": 0.351,
+      "step": 383
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4434210956096649,
+      "learning_rate": 7.402908775933419e-05,
+      "loss": 0.3709,
+      "step": 384
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4587249457836151,
+      "learning_rate": 7.389575561207692e-05,
+      "loss": 0.4602,
+      "step": 385
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4942215383052826,
+      "learning_rate": 7.376220287823236e-05,
+      "loss": 0.3745,
+      "step": 386
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.45109692215919495,
+      "learning_rate": 7.3628430790653e-05,
+      "loss": 0.3226,
+      "step": 387
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.39248013496398926,
+      "learning_rate": 7.349444058421619e-05,
+      "loss": 0.2886,
+      "step": 388
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4016898274421692,
+      "learning_rate": 7.336023349581287e-05,
+      "loss": 0.2429,
+      "step": 389
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3656561076641083,
+      "learning_rate": 7.322581076433596e-05,
+      "loss": 0.2889,
+      "step": 390
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.49339476227760315,
+      "learning_rate": 7.309117363066912e-05,
+      "loss": 0.4121,
+      "step": 391
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.41815677285194397,
+      "learning_rate": 7.295632333767513e-05,
+      "loss": 0.3564,
+      "step": 392
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.35798338055610657,
+      "learning_rate": 7.28212611301845e-05,
+      "loss": 0.234,
+      "step": 393
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4400104284286499,
+      "learning_rate": 7.2685988254984e-05,
+      "loss": 0.363,
+      "step": 394
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.43570441007614136,
+      "learning_rate": 7.255050596080509e-05,
+      "loss": 0.3474,
+      "step": 395
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2705797255039215,
+      "learning_rate": 7.241481549831243e-05,
+      "loss": 0.1613,
+      "step": 396
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4214705526828766,
+      "learning_rate": 7.22789181200923e-05,
+      "loss": 0.3396,
+      "step": 397
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4547910988330841,
+      "learning_rate": 7.214281508064107e-05,
+      "loss": 0.3565,
+      "step": 398
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3604981601238251,
+      "learning_rate": 7.200650763635366e-05,
+      "loss": 0.2394,
+      "step": 399
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4512694478034973,
+      "learning_rate": 7.186999704551181e-05,
+      "loss": 0.3133,
+      "step": 400
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.45708590745925903,
+      "learning_rate": 7.173328456827263e-05,
+      "loss": 0.319,
+      "step": 401
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3279622197151184,
+      "learning_rate": 7.15963714666568e-05,
+      "loss": 0.2018,
+      "step": 402
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4273524284362793,
+      "learning_rate": 7.145925900453709e-05,
+      "loss": 0.3532,
+      "step": 403
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3724408745765686,
+      "learning_rate": 7.132194844762654e-05,
+      "loss": 0.2762,
+      "step": 404
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.44304215908050537,
+      "learning_rate": 7.118444106346687e-05,
+      "loss": 0.3618,
+      "step": 405
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4273759424686432,
+      "learning_rate": 7.104673812141675e-05,
+      "loss": 0.3664,
+      "step": 406
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.40296387672424316,
+      "learning_rate": 7.090884089264011e-05,
+      "loss": 0.2687,
+      "step": 407
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3528042137622833,
+      "learning_rate": 7.077075065009433e-05,
+      "loss": 0.2294,
+      "step": 408
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4306895136833191,
+      "learning_rate": 7.063246866851858e-05,
+      "loss": 0.2743,
+      "step": 409
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5282109379768372,
+      "learning_rate": 7.049399622442198e-05,
+      "loss": 0.3679,
+      "step": 410
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.45601069927215576,
+      "learning_rate": 7.035533459607189e-05,
+      "loss": 0.3349,
+      "step": 411
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.39970582723617554,
+      "learning_rate": 7.021648506348204e-05,
+      "loss": 0.3156,
+      "step": 412
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4105047583580017,
+      "learning_rate": 7.007744890840073e-05,
+      "loss": 0.2743,
+      "step": 413
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.42897555232048035,
+      "learning_rate": 6.993822741429907e-05,
+      "loss": 0.3822,
+      "step": 414
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3480586111545563,
+      "learning_rate": 6.979882186635897e-05,
+      "loss": 0.2126,
+      "step": 415
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.43617862462997437,
+      "learning_rate": 6.965923355146147e-05,
+      "loss": 0.3459,
+      "step": 416
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3757456839084625,
+      "learning_rate": 6.951946375817474e-05,
+      "loss": 0.2828,
+      "step": 417
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.45733776688575745,
+      "learning_rate": 6.937951377674221e-05,
+      "loss": 0.2947,
+      "step": 418
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.44172319769859314,
+      "learning_rate": 6.923938489907066e-05,
+      "loss": 0.3259,
+      "step": 419
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4123852849006653,
+      "learning_rate": 6.909907841871829e-05,
+      "loss": 0.2887,
+      "step": 420
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4304922819137573,
+      "learning_rate": 6.895859563088283e-05,
+      "loss": 0.3682,
+      "step": 421
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.36229416728019714,
+      "learning_rate": 6.881793783238948e-05,
+      "loss": 0.2454,
+      "step": 422
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.46588000655174255,
+      "learning_rate": 6.867710632167903e-05,
+      "loss": 0.3546,
+      "step": 423
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3604878783226013,
+      "learning_rate": 6.853610239879586e-05,
+      "loss": 0.2702,
+      "step": 424
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.34859415888786316,
+      "learning_rate": 6.839492736537588e-05,
+      "loss": 0.1924,
+      "step": 425
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.462001234292984,
+      "learning_rate": 6.82535825246346e-05,
+      "loss": 0.2681,
+      "step": 426
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.46837642788887024,
+      "learning_rate": 6.811206918135502e-05,
+      "loss": 0.3998,
+      "step": 427
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4874202013015747,
+      "learning_rate": 6.797038864187564e-05,
+      "loss": 0.3459,
+      "step": 428
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3754540681838989,
+      "learning_rate": 6.782854221407838e-05,
+      "loss": 0.297,
+      "step": 429
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.43778693675994873,
+      "learning_rate": 6.768653120737652e-05,
+      "loss": 0.3855,
+      "step": 430
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4074949324131012,
+      "learning_rate": 6.754435693270258e-05,
+      "loss": 0.2835,
+      "step": 431
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4566798806190491,
+      "learning_rate": 6.740202070249623e-05,
+      "loss": 0.349,
+      "step": 432
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4901081919670105,
+      "learning_rate": 6.725952383069222e-05,
+      "loss": 0.4543,
+      "step": 433
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.43744221329689026,
+      "learning_rate": 6.711686763270818e-05,
+      "loss": 0.374,
+      "step": 434
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.41882264614105225,
+      "learning_rate": 6.697405342543258e-05,
+      "loss": 0.3751,
+      "step": 435
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.36602655053138733,
+      "learning_rate": 6.683108252721238e-05,
+      "loss": 0.3011,
+      "step": 436
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.47255802154541016,
+      "learning_rate": 6.668795625784113e-05,
+      "loss": 0.4474,
+      "step": 437
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3598809540271759,
+      "learning_rate": 6.654467593854657e-05,
+      "loss": 0.2321,
+      "step": 438
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4139677584171295,
+      "learning_rate": 6.640124289197845e-05,
+      "loss": 0.3092,
+      "step": 439
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4301883578300476,
+      "learning_rate": 6.625765844219652e-05,
+      "loss": 0.369,
+      "step": 440
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.39680930972099304,
+      "learning_rate": 6.611392391465802e-05,
+      "loss": 0.2968,
+      "step": 441
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4317682683467865,
+      "learning_rate": 6.597004063620567e-05,
+      "loss": 0.3328,
+      "step": 442
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.29227694869041443,
+      "learning_rate": 6.582600993505534e-05,
+      "loss": 0.1873,
+      "step": 443
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4805541932582855,
+      "learning_rate": 6.568183314078376e-05,
+      "loss": 0.4159,
+      "step": 444
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3404616117477417,
+      "learning_rate": 6.553751158431627e-05,
+      "loss": 0.2327,
+      "step": 445
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3709631860256195,
+      "learning_rate": 6.539304659791456e-05,
+      "loss": 0.2642,
+      "step": 446
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.38298705220222473,
+      "learning_rate": 6.524843951516434e-05,
+      "loss": 0.2727,
+      "step": 447
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.43321123719215393,
+      "learning_rate": 6.510369167096308e-05,
+      "loss": 0.292,
+      "step": 448
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3808426558971405,
+      "learning_rate": 6.495880440150756e-05,
+      "loss": 0.2678,
+      "step": 449
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3932049572467804,
+      "learning_rate": 6.481377904428171e-05,
+      "loss": 0.2624,
+      "step": 450
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4109935760498047,
+      "learning_rate": 6.466861693804413e-05,
+      "loss": 0.3203,
+      "step": 451
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.48460057377815247,
+      "learning_rate": 6.45233194228158e-05,
+      "loss": 0.4028,
+      "step": 452
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4326816201210022,
+      "learning_rate": 6.437788783986766e-05,
+      "loss": 0.2372,
+      "step": 453
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.46684136986732483,
+      "learning_rate": 6.42323235317083e-05,
+      "loss": 0.3116,
+      "step": 454
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4408712685108185,
+      "learning_rate": 6.408662784207149e-05,
+      "loss": 0.293,
+      "step": 455
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.43929120898246765,
+      "learning_rate": 6.394080211590381e-05,
+      "loss": 0.2862,
+      "step": 456
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.33374032378196716,
+      "learning_rate": 6.379484769935223e-05,
+      "loss": 0.2255,
+      "step": 457
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4163099527359009,
+      "learning_rate": 6.364876593975173e-05,
+      "loss": 0.2681,
+      "step": 458
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.34640195965766907,
+      "learning_rate": 6.350255818561277e-05,
+      "loss": 0.2094,
+      "step": 459
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4374372363090515,
+      "learning_rate": 6.335622578660889e-05,
+      "loss": 0.3127,
+      "step": 460
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3736974895000458,
+      "learning_rate": 6.320977009356431e-05,
+      "loss": 0.2477,
+      "step": 461
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.49957776069641113,
+      "learning_rate": 6.306319245844133e-05,
+      "loss": 0.3696,
+      "step": 462
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.41424816846847534,
+      "learning_rate": 6.291649423432799e-05,
+      "loss": 0.2518,
+      "step": 463
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.37798556685447693,
+      "learning_rate": 6.276967677542542e-05,
+      "loss": 0.2697,
+      "step": 464
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.403611958026886,
+      "learning_rate": 6.262274143703554e-05,
+      "loss": 0.3054,
+      "step": 465
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4166502356529236,
+      "learning_rate": 6.24756895755484e-05,
+      "loss": 0.2946,
+      "step": 466
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3577952980995178,
+      "learning_rate": 6.232852254842962e-05,
+      "loss": 0.2579,
+      "step": 467
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4579060673713684,
+      "learning_rate": 6.218124171420806e-05,
+      "loss": 0.3533,
+      "step": 468
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.404363751411438,
+      "learning_rate": 6.203384843246307e-05,
+      "loss": 0.3408,
+      "step": 469
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3781503736972809,
+      "learning_rate": 6.188634406381207e-05,
+      "loss": 0.3318,
+      "step": 470
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.47970715165138245,
+      "learning_rate": 6.173872996989793e-05,
+      "loss": 0.3375,
+      "step": 471
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3317432701587677,
+      "learning_rate": 6.159100751337642e-05,
+      "loss": 0.2029,
+      "step": 472
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4206933081150055,
+      "learning_rate": 6.144317805790361e-05,
+      "loss": 0.2868,
+      "step": 473
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.38284099102020264,
+      "learning_rate": 6.129524296812335e-05,
+      "loss": 0.2567,
+      "step": 474
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.33891499042510986,
+      "learning_rate": 6.114720360965453e-05,
+      "loss": 0.1857,
+      "step": 475
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.41084223985671997,
+      "learning_rate": 6.099906134907868e-05,
+      "loss": 0.2805,
+      "step": 476
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.43213552236557007,
+      "learning_rate": 6.085081755392714e-05,
+      "loss": 0.3463,
+      "step": 477
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.45232778787612915,
+      "learning_rate": 6.07024735926686e-05,
+      "loss": 0.3279,
+      "step": 478
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4519384205341339,
+      "learning_rate": 6.055403083469637e-05,
+      "loss": 0.3846,
+      "step": 479
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.48919233679771423,
+      "learning_rate": 6.04054906503158e-05,
+      "loss": 0.4084,
+      "step": 480
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.45906567573547363,
+      "learning_rate": 6.0256854410731565e-05,
+      "loss": 0.3372,
+      "step": 481
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4604879319667816,
+      "learning_rate": 6.010812348803509e-05,
+      "loss": 0.376,
+      "step": 482
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.44598522782325745,
+      "learning_rate": 5.99592992551918e-05,
+      "loss": 0.4184,
+      "step": 483
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.41852816939353943,
+      "learning_rate": 5.9810383086028535e-05,
+      "loss": 0.377,
+      "step": 484
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3676040768623352,
+      "learning_rate": 5.9661376355220734e-05,
+      "loss": 0.2941,
+      "step": 485
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.48775678873062134,
+      "learning_rate": 5.9512280438279914e-05,
+      "loss": 0.4142,
+      "step": 486
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.38730305433273315,
+      "learning_rate": 5.936309671154084e-05,
+      "loss": 0.3353,
+      "step": 487
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.43073874711990356,
+      "learning_rate": 5.9213826552148886e-05,
+      "loss": 0.343,
+      "step": 488
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3993862569332123,
+      "learning_rate": 5.906447133804731e-05,
+      "loss": 0.3534,
+      "step": 489
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.37199288606643677,
+      "learning_rate": 5.891503244796448e-05,
+      "loss": 0.2556,
+      "step": 490
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3543163239955902,
+      "learning_rate": 5.8765511261401254e-05,
+      "loss": 0.2985,
+      "step": 491
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.35919004678726196,
+      "learning_rate": 5.861590915861817e-05,
+      "loss": 0.2354,
+      "step": 492
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.37146124243736267,
+      "learning_rate": 5.846622752062268e-05,
+      "loss": 0.2635,
+      "step": 493
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4142633378505707,
+      "learning_rate": 5.831646772915651e-05,
+      "loss": 0.3471,
+      "step": 494
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4175507128238678,
+      "learning_rate": 5.816663116668276e-05,
+      "loss": 0.3043,
+      "step": 495
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4025022089481354,
+      "learning_rate": 5.801671921637328e-05,
+      "loss": 0.3176,
+      "step": 496
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5054760575294495,
+      "learning_rate": 5.786673326209584e-05,
+      "loss": 0.4023,
+      "step": 497
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3766479194164276,
+      "learning_rate": 5.7716674688401286e-05,
+      "loss": 0.265,
+      "step": 498
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.482490599155426,
+      "learning_rate": 5.756654488051091e-05,
+      "loss": 0.3873,
+      "step": 499
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3957984149456024,
+      "learning_rate": 5.7416345224303524e-05,
+      "loss": 0.2965,
+      "step": 500
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3828335702419281,
+      "learning_rate": 5.7266077106302785e-05,
+      "loss": 0.2825,
+      "step": 501
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.34701448678970337,
+      "learning_rate": 5.7115741913664264e-05,
+      "loss": 0.2666,
+      "step": 502
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3745706081390381,
+      "learning_rate": 5.696534103416276e-05,
+      "loss": 0.2752,
+      "step": 503
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4428030848503113,
+      "learning_rate": 5.6814875856179414e-05,
+      "loss": 0.3607,
+      "step": 504
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4908398985862732,
+      "learning_rate": 5.666434776868895e-05,
+      "loss": 0.3612,
+      "step": 505
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.47323817014694214,
+      "learning_rate": 5.651375816124679e-05,
+      "loss": 0.3226,
+      "step": 506
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4364282786846161,
+      "learning_rate": 5.636310842397629e-05,
+      "loss": 0.3154,
+      "step": 507
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.44495850801467896,
+      "learning_rate": 5.621239994755583e-05,
+      "loss": 0.3132,
+      "step": 508
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.41207343339920044,
+      "learning_rate": 5.606163412320608e-05,
+      "loss": 0.2979,
+      "step": 509
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.408840537071228,
+      "learning_rate": 5.5910812342677065e-05,
+      "loss": 0.2689,
+      "step": 510
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.36593517661094666,
+      "learning_rate": 5.575993599823536e-05,
+      "loss": 0.2513,
+      "step": 511
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.37152597308158875,
+      "learning_rate": 5.560900648265124e-05,
+      "loss": 0.1832,
+      "step": 512
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.48371148109436035,
+      "learning_rate": 5.545802518918579e-05,
+      "loss": 0.3981,
+      "step": 513
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5247592329978943,
+      "learning_rate": 5.5306993511578096e-05,
+      "loss": 0.4218,
+      "step": 514
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4305899441242218,
+      "learning_rate": 5.515591284403234e-05,
+      "loss": 0.3061,
+      "step": 515
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.40413033962249756,
+      "learning_rate": 5.5004784581204927e-05,
+      "loss": 0.3007,
+      "step": 516
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.36091166734695435,
+      "learning_rate": 5.485361011819164e-05,
+      "loss": 0.2439,
+      "step": 517
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.39443257451057434,
+      "learning_rate": 5.4702390850514726e-05,
+      "loss": 0.2617,
+      "step": 518
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.40204355120658875,
+      "learning_rate": 5.455112817411006e-05,
+      "loss": 0.3121,
+      "step": 519
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.41540104150772095,
+      "learning_rate": 5.4399823485314226e-05,
+      "loss": 0.2636,
+      "step": 520
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.44149285554885864,
+      "learning_rate": 5.4248478180851604e-05,
+      "loss": 0.387,
+      "step": 521
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.36926665902137756,
+      "learning_rate": 5.409709365782154e-05,
+      "loss": 0.254,
+      "step": 522
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4551641047000885,
+      "learning_rate": 5.3945671313685386e-05,
+      "loss": 0.3855,
+      "step": 523
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.37731683254241943,
+      "learning_rate": 5.379421254625366e-05,
+      "loss": 0.2773,
+      "step": 524
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.40048646926879883,
+      "learning_rate": 5.364271875367311e-05,
+      "loss": 0.294,
+      "step": 525
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4262080192565918,
+      "learning_rate": 5.3491191334413746e-05,
+      "loss": 0.3774,
+      "step": 526
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.38077598810195923,
+      "learning_rate": 5.3339631687256084e-05,
+      "loss": 0.2447,
+      "step": 527
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.39433252811431885,
+      "learning_rate": 5.318804121127807e-05,
+      "loss": 0.2666,
+      "step": 528
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.45546483993530273,
+      "learning_rate": 5.3036421305842276e-05,
+      "loss": 0.3913,
+      "step": 529
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3501354455947876,
+      "learning_rate": 5.288477337058293e-05,
+      "loss": 0.2395,
+      "step": 530
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4665021002292633,
+      "learning_rate": 5.273309880539301e-05,
+      "loss": 0.3696,
+      "step": 531
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5128461718559265,
+      "learning_rate": 5.258139901041132e-05,
+      "loss": 0.2984,
+      "step": 532
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3114360272884369,
+      "learning_rate": 5.242967538600957e-05,
+      "loss": 0.1855,
+      "step": 533
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4211101531982422,
+      "learning_rate": 5.227792933277943e-05,
+      "loss": 0.3319,
+      "step": 534
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3312818706035614,
+      "learning_rate": 5.212616225151965e-05,
+      "loss": 0.2227,
+      "step": 535
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.43743807077407837,
+      "learning_rate": 5.197437554322304e-05,
+      "loss": 0.4077,
+      "step": 536
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.415711373090744,
+      "learning_rate": 5.182257060906365e-05,
+      "loss": 0.2987,
+      "step": 537
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.46717381477355957,
+      "learning_rate": 5.167074885038373e-05,
+      "loss": 0.33,
+      "step": 538
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.36937227845191956,
+      "learning_rate": 5.151891166868086e-05,
+      "loss": 0.2459,
+      "step": 539
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5149711966514587,
+      "learning_rate": 5.1367060465595006e-05,
+      "loss": 0.462,
+      "step": 540
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.384420782327652,
+      "learning_rate": 5.121519664289553e-05,
+      "loss": 0.2958,
+      "step": 541
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.46390560269355774,
+      "learning_rate": 5.106332160246834e-05,
+      "loss": 0.4127,
+      "step": 542
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.37238529324531555,
+      "learning_rate": 5.0911436746302834e-05,
+      "loss": 0.2882,
+      "step": 543
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.37248557806015015,
+      "learning_rate": 5.075954347647909e-05,
+      "loss": 0.2447,
+      "step": 544
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3653389811515808,
+      "learning_rate": 5.0607643195154796e-05,
+      "loss": 0.2843,
+      "step": 545
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3526271879673004,
+      "learning_rate": 5.045573730455241e-05,
+      "loss": 0.285,
+      "step": 546
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3367946445941925,
+      "learning_rate": 5.030382720694612e-05,
+      "loss": 0.2531,
+      "step": 547
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4197223484516144,
+      "learning_rate": 5.0151914304649015e-05,
+      "loss": 0.3536,
+      "step": 548
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.33859139680862427,
+      "learning_rate": 5e-05,
+      "loss": 0.2301,
+      "step": 549
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.43413838744163513,
+      "learning_rate": 4.984808569535101e-05,
+      "loss": 0.3349,
+      "step": 550
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4138838052749634,
+      "learning_rate": 4.969617279305388e-05,
+      "loss": 0.3194,
+      "step": 551
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.37030553817749023,
+      "learning_rate": 4.954426269544761e-05,
+      "loss": 0.2617,
+      "step": 552
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.39585164189338684,
+      "learning_rate": 4.939235680484522e-05,
+      "loss": 0.3128,
+      "step": 553
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.35292455554008484,
+      "learning_rate": 4.924045652352092e-05,
+      "loss": 0.2415,
+      "step": 554
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.383457213640213,
+      "learning_rate": 4.908856325369718e-05,
+      "loss": 0.3018,
+      "step": 555
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3327413499355316,
+      "learning_rate": 4.893667839753168e-05,
+      "loss": 0.1913,
+      "step": 556
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4065321087837219,
+      "learning_rate": 4.878480335710448e-05,
+      "loss": 0.3515,
+      "step": 557
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.2919169068336487,
+      "learning_rate": 4.8632939534405006e-05,
+      "loss": 0.1718,
+      "step": 558
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3973611295223236,
+      "learning_rate": 4.8481088331319146e-05,
+      "loss": 0.32,
+      "step": 559
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3425373435020447,
+      "learning_rate": 4.832925114961629e-05,
+      "loss": 0.2478,
+      "step": 560
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.472787082195282,
+      "learning_rate": 4.817742939093635e-05,
+      "loss": 0.4112,
+      "step": 561
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.42442649602890015,
+      "learning_rate": 4.8025624456776966e-05,
+      "loss": 0.2729,
+      "step": 562
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.36462414264678955,
+      "learning_rate": 4.787383774848037e-05,
+      "loss": 0.3008,
+      "step": 563
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4791617691516876,
+      "learning_rate": 4.772207066722057e-05,
+      "loss": 0.3457,
+      "step": 564
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.46410971879959106,
+      "learning_rate": 4.757032461399044e-05,
+      "loss": 0.3273,
+      "step": 565
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4102344214916229,
+      "learning_rate": 4.7418600989588694e-05,
+      "loss": 0.235,
+      "step": 566
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3302304148674011,
+      "learning_rate": 4.726690119460701e-05,
+      "loss": 0.1944,
+      "step": 567
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.39677003026008606,
+      "learning_rate": 4.7115226629417075e-05,
+      "loss": 0.3123,
+      "step": 568
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.432862251996994,
+      "learning_rate": 4.6963578694157736e-05,
+      "loss": 0.2621,
+      "step": 569
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.44008541107177734,
+      "learning_rate": 4.681195878872194e-05,
+      "loss": 0.3357,
+      "step": 570
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3981214463710785,
+      "learning_rate": 4.666036831274392e-05,
+      "loss": 0.2633,
+      "step": 571
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4169214963912964,
+      "learning_rate": 4.6508808665586265e-05,
+      "loss": 0.3009,
+      "step": 572
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.33570775389671326,
+      "learning_rate": 4.635728124632692e-05,
+      "loss": 0.2555,
+      "step": 573
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4198802709579468,
+      "learning_rate": 4.6205787453746336e-05,
+      "loss": 0.2845,
+      "step": 574
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.34704431891441345,
+      "learning_rate": 4.605432868631462e-05,
+      "loss": 0.233,
+      "step": 575
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.48909562826156616,
+      "learning_rate": 4.590290634217848e-05,
+      "loss": 0.3706,
+      "step": 576
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.43792209029197693,
+      "learning_rate": 4.57515218191484e-05,
+      "loss": 0.3334,
+      "step": 577
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3764708936214447,
+      "learning_rate": 4.5600176514685786e-05,
+      "loss": 0.2736,
+      "step": 578
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.40787622332572937,
+      "learning_rate": 4.5448871825889946e-05,
+      "loss": 0.3078,
+      "step": 579
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.41174522042274475,
+      "learning_rate": 4.52976091494853e-05,
+      "loss": 0.2572,
+      "step": 580
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3787226378917694,
+      "learning_rate": 4.514638988180837e-05,
+      "loss": 0.2752,
+      "step": 581
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.47147271037101746,
+      "learning_rate": 4.4995215418795085e-05,
+      "loss": 0.3972,
+      "step": 582
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3915804922580719,
+      "learning_rate": 4.484408715596768e-05,
+      "loss": 0.293,
+      "step": 583
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.413932204246521,
+      "learning_rate": 4.4693006488421915e-05,
+      "loss": 0.2767,
+      "step": 584
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.43163156509399414,
+      "learning_rate": 4.454197481081422e-05,
+      "loss": 0.2971,
+      "step": 585
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3995768427848816,
+      "learning_rate": 4.439099351734878e-05,
+      "loss": 0.2823,
+      "step": 586
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4052314758300781,
+      "learning_rate": 4.4240064001764646e-05,
+      "loss": 0.2714,
+      "step": 587
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.39411696791648865,
+      "learning_rate": 4.4089187657322953e-05,
+      "loss": 0.2645,
+      "step": 588
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3517491817474365,
+      "learning_rate": 4.393836587679394e-05,
+      "loss": 0.2443,
+      "step": 589
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.46535736322402954,
+      "learning_rate": 4.3787600052444174e-05,
+      "loss": 0.3524,
+      "step": 590
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.47955119609832764,
+      "learning_rate": 4.363689157602373e-05,
+      "loss": 0.4257,
+      "step": 591
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.37164661288261414,
+      "learning_rate": 4.348624183875322e-05,
+      "loss": 0.2939,
+      "step": 592
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3547576367855072,
+      "learning_rate": 4.333565223131107e-05,
+      "loss": 0.2171,
+      "step": 593
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4252079427242279,
+      "learning_rate": 4.318512414382059e-05,
+      "loss": 0.3381,
+      "step": 594
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4120088517665863,
+      "learning_rate": 4.3034658965837255e-05,
+      "loss": 0.1831,
+      "step": 595
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4044457972049713,
+      "learning_rate": 4.288425808633575e-05,
+      "loss": 0.3326,
+      "step": 596
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3945488929748535,
+      "learning_rate": 4.273392289369723e-05,
+      "loss": 0.2759,
+      "step": 597
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3605932593345642,
+      "learning_rate": 4.258365477569648e-05,
+      "loss": 0.2268,
+      "step": 598
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4327079653739929,
+      "learning_rate": 4.2433455119489105e-05,
+      "loss": 0.3237,
+      "step": 599
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4783308207988739,
+      "learning_rate": 4.228332531159871e-05,
+      "loss": 0.3675,
+      "step": 600
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4235597550868988,
+      "learning_rate": 4.2133266737904176e-05,
+      "loss": 0.2976,
+      "step": 601
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4399956464767456,
+      "learning_rate": 4.1983280783626724e-05,
+      "loss": 0.3662,
+      "step": 602
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.42588597536087036,
+      "learning_rate": 4.183336883331723e-05,
+      "loss": 0.3611,
+      "step": 603
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.43737921118736267,
+      "learning_rate": 4.1683532270843504e-05,
+      "loss": 0.3477,
+      "step": 604
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.40459975600242615,
+      "learning_rate": 4.153377247937732e-05,
+      "loss": 0.3054,
+      "step": 605
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3984995484352112,
+      "learning_rate": 4.138409084138185e-05,
+      "loss": 0.3077,
+      "step": 606
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4512704312801361,
+      "learning_rate": 4.1234488738598744e-05,
+      "loss": 0.2709,
+      "step": 607
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.33205512166023254,
+      "learning_rate": 4.108496755203553e-05,
+      "loss": 0.2092,
+      "step": 608
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.40625834465026855,
+      "learning_rate": 4.0935528661952716e-05,
+      "loss": 0.3054,
+      "step": 609
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.41261643171310425,
+      "learning_rate": 4.0786173447851126e-05,
+      "loss": 0.3156,
+      "step": 610
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3921923041343689,
+      "learning_rate": 4.063690328845916e-05,
+      "loss": 0.2664,
+      "step": 611
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.34725549817085266,
+      "learning_rate": 4.04877195617201e-05,
+      "loss": 0.2566,
+      "step": 612
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4461452066898346,
+      "learning_rate": 4.033862364477927e-05,
+      "loss": 0.3574,
+      "step": 613
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.34110385179519653,
+      "learning_rate": 4.0189616913971484e-05,
+      "loss": 0.2254,
+      "step": 614
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.41009894013404846,
+      "learning_rate": 4.0040700744808204e-05,
+      "loss": 0.3814,
+      "step": 615
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3577955663204193,
+      "learning_rate": 3.989187651196493e-05,
+      "loss": 0.2292,
+      "step": 616
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4174287021160126,
+      "learning_rate": 3.974314558926844e-05,
+      "loss": 0.3083,
+      "step": 617
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3687293529510498,
+      "learning_rate": 3.9594509349684216e-05,
+      "loss": 0.3006,
+      "step": 618
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3968195617198944,
+      "learning_rate": 3.9445969165303647e-05,
+      "loss": 0.2811,
+      "step": 619
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.43467170000076294,
+      "learning_rate": 3.929752640733141e-05,
+      "loss": 0.3129,
+      "step": 620
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4192885160446167,
+      "learning_rate": 3.914918244607287e-05,
+      "loss": 0.2833,
+      "step": 621
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3982818126678467,
+      "learning_rate": 3.900093865092134e-05,
+      "loss": 0.2464,
+      "step": 622
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.37380045652389526,
+      "learning_rate": 3.885279639034546e-05,
+      "loss": 0.2454,
+      "step": 623
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.47077518701553345,
+      "learning_rate": 3.870475703187667e-05,
+      "loss": 0.265,
+      "step": 624
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.47373396158218384,
+      "learning_rate": 3.855682194209639e-05,
+      "loss": 0.2975,
+      "step": 625
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.30882200598716736,
+      "learning_rate": 3.840899248662358e-05,
+      "loss": 0.2194,
+      "step": 626
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.401368111371994,
+      "learning_rate": 3.8261270030102084e-05,
+      "loss": 0.2833,
+      "step": 627
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3969556391239166,
+      "learning_rate": 3.8113655936187947e-05,
+      "loss": 0.3015,
+      "step": 628
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3934876024723053,
+      "learning_rate": 3.796615156753696e-05,
+      "loss": 0.3399,
+      "step": 629
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4074617028236389,
+      "learning_rate": 3.7818758285791955e-05,
+      "loss": 0.3068,
+      "step": 630
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3954191207885742,
+      "learning_rate": 3.767147745157039e-05,
+      "loss": 0.3337,
+      "step": 631
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.47569915652275085,
+      "learning_rate": 3.7524310424451635e-05,
+      "loss": 0.3639,
+      "step": 632
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.39985620975494385,
+      "learning_rate": 3.7377258562964454e-05,
+      "loss": 0.3208,
+      "step": 633
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4269794821739197,
+      "learning_rate": 3.723032322457458e-05,
+      "loss": 0.3463,
+      "step": 634
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4925236105918884,
+      "learning_rate": 3.708350576567204e-05,
+      "loss": 0.41,
+      "step": 635
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3888967037200928,
+      "learning_rate": 3.6936807541558674e-05,
+      "loss": 0.2667,
+      "step": 636
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4810844659805298,
+      "learning_rate": 3.6790229906435705e-05,
+      "loss": 0.4566,
+      "step": 637
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3134874403476715,
+      "learning_rate": 3.664377421339111e-05,
+      "loss": 0.2074,
+      "step": 638
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4054643213748932,
+      "learning_rate": 3.6497441814387247e-05,
+      "loss": 0.2916,
+      "step": 639
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3976866900920868,
+      "learning_rate": 3.6351234060248286e-05,
+      "loss": 0.2898,
+      "step": 640
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.44661563634872437,
+      "learning_rate": 3.6205152300647784e-05,
+      "loss": 0.3798,
+      "step": 641
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.37204062938690186,
+      "learning_rate": 3.605919788409621e-05,
+      "loss": 0.3046,
+      "step": 642
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.40015891194343567,
+      "learning_rate": 3.591337215792852e-05,
+      "loss": 0.3225,
+      "step": 643
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.39118486642837524,
+      "learning_rate": 3.5767676468291713e-05,
+      "loss": 0.2303,
+      "step": 644
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4793816804885864,
+      "learning_rate": 3.562211216013235e-05,
+      "loss": 0.4683,
+      "step": 645
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3733760714530945,
+      "learning_rate": 3.5476680577184206e-05,
+      "loss": 0.3253,
+      "step": 646
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.339495986700058,
+      "learning_rate": 3.533138306195588e-05,
+      "loss": 0.2335,
+      "step": 647
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.43037405610084534,
+      "learning_rate": 3.5186220955718306e-05,
+      "loss": 0.2914,
+      "step": 648
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3812549114227295,
+      "learning_rate": 3.5041195598492446e-05,
+      "loss": 0.3047,
+      "step": 649
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3607620298862457,
+      "learning_rate": 3.489630832903694e-05,
+      "loss": 0.2662,
+      "step": 650
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.47065141797065735,
+      "learning_rate": 3.475156048483567e-05,
+      "loss": 0.3258,
+      "step": 651
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4050087034702301,
+      "learning_rate": 3.460695340208546e-05,
+      "loss": 0.3524,
+      "step": 652
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.39014732837677,
+      "learning_rate": 3.446248841568375e-05,
+      "loss": 0.2748,
+      "step": 653
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4207385182380676,
+      "learning_rate": 3.431816685921625e-05,
+      "loss": 0.3261,
+      "step": 654
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3538001775741577,
+      "learning_rate": 3.417399006494466e-05,
+      "loss": 0.233,
+      "step": 655
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3425646722316742,
+      "learning_rate": 3.402995936379433e-05,
+      "loss": 0.245,
+      "step": 656
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3696725368499756,
+      "learning_rate": 3.3886076085341986e-05,
+      "loss": 0.2547,
+      "step": 657
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3415139615535736,
+      "learning_rate": 3.37423415578035e-05,
+      "loss": 0.2382,
+      "step": 658
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3608959913253784,
+      "learning_rate": 3.3598757108021546e-05,
+      "loss": 0.2394,
+      "step": 659
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3897416889667511,
+      "learning_rate": 3.345532406145345e-05,
+      "loss": 0.2854,
+      "step": 660
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.34789225459098816,
+      "learning_rate": 3.331204374215888e-05,
+      "loss": 0.2357,
+      "step": 661
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.37091711163520813,
+      "learning_rate": 3.316891747278761e-05,
+      "loss": 0.2424,
+      "step": 662
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.47823840379714966,
+      "learning_rate": 3.302594657456744e-05,
+      "loss": 0.3736,
+      "step": 663
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.45948362350463867,
+      "learning_rate": 3.288313236729183e-05,
+      "loss": 0.3395,
+      "step": 664
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.43615004420280457,
+      "learning_rate": 3.274047616930781e-05,
+      "loss": 0.343,
+      "step": 665
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.4098581373691559,
+      "learning_rate": 3.259797929750378e-05,
+      "loss": 0.2629,
+      "step": 666
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.38520053029060364,
+      "learning_rate": 3.245564306729744e-05,
+      "loss": 0.2948,
+      "step": 667
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3814550042152405,
+      "learning_rate": 3.231346879262349e-05,
+      "loss": 0.2802,
+      "step": 668
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.38907739520072937,
+      "learning_rate": 3.217145778592162e-05,
+      "loss": 0.3131,
+      "step": 669
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4280015230178833,
+      "learning_rate": 3.202961135812437e-05,
+      "loss": 0.3395,
+      "step": 670
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4221903681755066,
+      "learning_rate": 3.1887930818644996e-05,
+      "loss": 0.3522,
+      "step": 671
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4077076315879822,
+      "learning_rate": 3.1746417475365405e-05,
+      "loss": 0.2843,
+      "step": 672
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4645012319087982,
+      "learning_rate": 3.1605072634624125e-05,
+      "loss": 0.4123,
+      "step": 673
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4186570644378662,
+      "learning_rate": 3.146389760120416e-05,
+      "loss": 0.2707,
+      "step": 674
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.40125572681427,
+      "learning_rate": 3.132289367832097e-05,
+      "loss": 0.2917,
+      "step": 675
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4461178183555603,
+      "learning_rate": 3.118206216761053e-05,
+      "loss": 0.3186,
+      "step": 676
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3696655035018921,
+      "learning_rate": 3.104140436911719e-05,
+      "loss": 0.1924,
+      "step": 677
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3892281949520111,
+      "learning_rate": 3.0900921581281725e-05,
+      "loss": 0.3101,
+      "step": 678
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4376410245895386,
+      "learning_rate": 3.076061510092935e-05,
+      "loss": 0.3385,
+      "step": 679
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3946727216243744,
+      "learning_rate": 3.062048622325779e-05,
+      "loss": 0.3105,
+      "step": 680
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4012112319469452,
+      "learning_rate": 3.0480536241825263e-05,
+      "loss": 0.2442,
+      "step": 681
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.45755696296691895,
+      "learning_rate": 3.034076644853853e-05,
+      "loss": 0.2932,
+      "step": 682
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4443381428718567,
+      "learning_rate": 3.0201178133641038e-05,
+      "loss": 0.3799,
+      "step": 683
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3474355638027191,
+      "learning_rate": 3.0061772585700953e-05,
+      "loss": 0.2334,
+      "step": 684
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.38603806495666504,
+      "learning_rate": 2.992255109159926e-05,
+      "loss": 0.2574,
+      "step": 685
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.38428762555122375,
+      "learning_rate": 2.9783514936517965e-05,
+      "loss": 0.3081,
+      "step": 686
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.41110917925834656,
+      "learning_rate": 2.9644665403928117e-05,
+      "loss": 0.2946,
+      "step": 687
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4486719071865082,
+      "learning_rate": 2.950600377557804e-05,
+      "loss": 0.4257,
+      "step": 688
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4164861738681793,
+      "learning_rate": 2.9367531331481436e-05,
+      "loss": 0.2622,
+      "step": 689
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.40194183588027954,
+      "learning_rate": 2.9229249349905684e-05,
+      "loss": 0.3232,
+      "step": 690
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.38287442922592163,
+      "learning_rate": 2.909115910735991e-05,
+      "loss": 0.2665,
+      "step": 691
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3484087288379669,
+      "learning_rate": 2.895326187858326e-05,
+      "loss": 0.2387,
+      "step": 692
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.48390090465545654,
+      "learning_rate": 2.881555893653314e-05,
+      "loss": 0.478,
+      "step": 693
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3917997181415558,
+      "learning_rate": 2.8678051552373487e-05,
+      "loss": 0.2926,
+      "step": 694
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3424125611782074,
+      "learning_rate": 2.854074099546291e-05,
+      "loss": 0.2481,
+      "step": 695
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.38915279507637024,
+      "learning_rate": 2.8403628533343206e-05,
+      "loss": 0.2039,
+      "step": 696
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.39113256335258484,
+      "learning_rate": 2.826671543172738e-05,
+      "loss": 0.3135,
+      "step": 697
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3531236946582794,
+      "learning_rate": 2.8130002954488183e-05,
+      "loss": 0.2567,
+      "step": 698
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4020720422267914,
+      "learning_rate": 2.799349236364634e-05,
+      "loss": 0.3434,
+      "step": 699
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.41817712783813477,
+      "learning_rate": 2.7857184919358937e-05,
+      "loss": 0.3242,
+      "step": 700
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3442646861076355,
+      "learning_rate": 2.7721081879907718e-05,
+      "loss": 0.2424,
+      "step": 701
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.49456846714019775,
+      "learning_rate": 2.7585184501687577e-05,
+      "loss": 0.356,
+      "step": 702
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4054080545902252,
+      "learning_rate": 2.74494940391949e-05,
+      "loss": 0.3587,
+      "step": 703
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4222458302974701,
+      "learning_rate": 2.731401174501601e-05,
+      "loss": 0.317,
+      "step": 704
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.36371931433677673,
+      "learning_rate": 2.7178738869815506e-05,
+      "loss": 0.2638,
+      "step": 705
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.34860721230506897,
+      "learning_rate": 2.7043676662324878e-05,
+      "loss": 0.2298,
+      "step": 706
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.42635902762413025,
+      "learning_rate": 2.69088263693309e-05,
+      "loss": 0.3104,
+      "step": 707
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.411121129989624,
+      "learning_rate": 2.6774189235664026e-05,
+      "loss": 0.2629,
+      "step": 708
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.46932339668273926,
+      "learning_rate": 2.663976650418715e-05,
+      "loss": 0.4047,
+      "step": 709
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.46861177682876587,
+      "learning_rate": 2.650555941578381e-05,
+      "loss": 0.4143,
+      "step": 710
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3969976603984833,
+      "learning_rate": 2.6371569209347014e-05,
+      "loss": 0.292,
+      "step": 711
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.41862115263938904,
+      "learning_rate": 2.6237797121767634e-05,
+      "loss": 0.3706,
+      "step": 712
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3939150869846344,
+      "learning_rate": 2.6104244387923082e-05,
+      "loss": 0.2726,
+      "step": 713
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.33569246530532837,
+      "learning_rate": 2.5970912240665813e-05,
+      "loss": 0.2001,
+      "step": 714
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3980376422405243,
+      "learning_rate": 2.5837801910812053e-05,
+      "loss": 0.3019,
+      "step": 715
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3491917550563812,
+      "learning_rate": 2.5704914627130374e-05,
+      "loss": 0.2664,
+      "step": 716
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.43504101037979126,
+      "learning_rate": 2.5572251616330373e-05,
+      "loss": 0.3871,
+      "step": 717
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.41263502836227417,
+      "learning_rate": 2.5439814103051284e-05,
+      "loss": 0.2307,
+      "step": 718
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.31717926263809204,
+      "learning_rate": 2.530760330985079e-05,
+      "loss": 0.1846,
+      "step": 719
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.451673686504364,
+      "learning_rate": 2.517562045719367e-05,
+      "loss": 0.403,
+      "step": 720
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3923144042491913,
+      "learning_rate": 2.504386676344047e-05,
+      "loss": 0.3311,
+      "step": 721
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.41828757524490356,
+      "learning_rate": 2.4912343444836445e-05,
+      "loss": 0.3399,
+      "step": 722
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.44174039363861084,
+      "learning_rate": 2.4781051715500076e-05,
+      "loss": 0.3893,
+      "step": 723
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.34206709265708923,
+      "learning_rate": 2.46499927874121e-05,
+      "loss": 0.2145,
+      "step": 724
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3767687678337097,
+      "learning_rate": 2.4519167870404125e-05,
+      "loss": 0.2524,
+      "step": 725
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4103870987892151,
+      "learning_rate": 2.4388578172147675e-05,
+      "loss": 0.3165,
+      "step": 726
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.37169212102890015,
+      "learning_rate": 2.4258224898142807e-05,
+      "loss": 0.2359,
+      "step": 727
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.42850708961486816,
+      "learning_rate": 2.4128109251707155e-05,
+      "loss": 0.3953,
+      "step": 728
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3800533711910248,
+      "learning_rate": 2.399823243396476e-05,
+      "loss": 0.2891,
+      "step": 729
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3955475091934204,
+      "learning_rate": 2.3868595643834995e-05,
+      "loss": 0.2531,
+      "step": 730
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.42857804894447327,
+      "learning_rate": 2.373920007802144e-05,
+      "loss": 0.322,
+      "step": 731
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3623318672180176,
+      "learning_rate": 2.361004693100094e-05,
+      "loss": 0.2604,
+      "step": 732
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.40041908621788025,
+      "learning_rate": 2.3481137395012513e-05,
+      "loss": 0.3004,
+      "step": 733
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3790295124053955,
+      "learning_rate": 2.3352472660046295e-05,
+      "loss": 0.2634,
+      "step": 734
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3655712902545929,
+      "learning_rate": 2.322405391383273e-05,
+      "loss": 0.3042,
+      "step": 735
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.42415592074394226,
+      "learning_rate": 2.3095882341831372e-05,
+      "loss": 0.4077,
+      "step": 736
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.37483200430870056,
+      "learning_rate": 2.296795912722014e-05,
+      "loss": 0.3072,
+      "step": 737
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.35293281078338623,
+      "learning_rate": 2.284028545088423e-05,
+      "loss": 0.259,
+      "step": 738
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.34485095739364624,
+      "learning_rate": 2.2712862491405436e-05,
+      "loss": 0.266,
+      "step": 739
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.344819039106369,
+      "learning_rate": 2.258569142505098e-05,
+      "loss": 0.2453,
+      "step": 740
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4387754797935486,
+      "learning_rate": 2.2458773425762912e-05,
+      "loss": 0.2779,
+      "step": 741
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4539426267147064,
+      "learning_rate": 2.2332109665147127e-05,
+      "loss": 0.3791,
+      "step": 742
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.41004621982574463,
+      "learning_rate": 2.2205701312462617e-05,
+      "loss": 0.3598,
+      "step": 743
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3861682415008545,
+      "learning_rate": 2.2079549534610606e-05,
+      "loss": 0.2982,
+      "step": 744
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.40412721037864685,
+      "learning_rate": 2.1953655496123853e-05,
+      "loss": 0.2879,
+      "step": 745
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.41516509652137756,
+      "learning_rate": 2.1828020359155905e-05,
+      "loss": 0.2866,
+      "step": 746
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.35956189036369324,
+      "learning_rate": 2.1702645283470236e-05,
+      "loss": 0.2449,
+      "step": 747
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.45450302958488464,
+      "learning_rate": 2.1577531426429782e-05,
+      "loss": 0.3618,
+      "step": 748
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.32821208238601685,
+      "learning_rate": 2.1452679942985993e-05,
+      "loss": 0.2288,
+      "step": 749
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4255446195602417,
+      "learning_rate": 2.132809198566837e-05,
+      "loss": 0.3431,
+      "step": 750
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4676855206489563,
+      "learning_rate": 2.1203768704573672e-05,
+      "loss": 0.3602,
+      "step": 751
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.37337955832481384,
+      "learning_rate": 2.1079711247355505e-05,
+      "loss": 0.2727,
+      "step": 752
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.422907292842865,
+      "learning_rate": 2.095592075921347e-05,
+      "loss": 0.2386,
+      "step": 753
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3308090269565582,
+      "learning_rate": 2.08323983828828e-05,
+      "loss": 0.191,
+      "step": 754
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.42989543080329895,
+      "learning_rate": 2.0709145258623704e-05,
+      "loss": 0.3802,
+      "step": 755
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.414399117231369,
+      "learning_rate": 2.0586162524210895e-05,
+      "loss": 0.2893,
+      "step": 756
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4370255172252655,
+      "learning_rate": 2.0463451314923015e-05,
+      "loss": 0.3331,
+      "step": 757
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.42983582615852356,
+      "learning_rate": 2.0341012763532243e-05,
+      "loss": 0.3425,
+      "step": 758
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.36104878783226013,
+      "learning_rate": 2.021884800029379e-05,
+      "loss": 0.2644,
+      "step": 759
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.38013267517089844,
+      "learning_rate": 2.009695815293548e-05,
+      "loss": 0.2912,
+      "step": 760
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4188981056213379,
+      "learning_rate": 1.9975344346647297e-05,
+      "loss": 0.3136,
+      "step": 761
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.38251587748527527,
+      "learning_rate": 1.9854007704071064e-05,
+      "loss": 0.2962,
+      "step": 762
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.32296597957611084,
+      "learning_rate": 1.973294934529007e-05,
+      "loss": 0.2121,
+      "step": 763
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3597976565361023,
+      "learning_rate": 1.961217038781863e-05,
+      "loss": 0.2609,
+      "step": 764
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.44841957092285156,
+      "learning_rate": 1.9491671946591962e-05,
+      "loss": 0.3221,
+      "step": 765
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.4504457414150238,
+      "learning_rate": 1.9371455133955675e-05,
+      "loss": 0.3516,
+      "step": 766
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.33190175890922546,
+      "learning_rate": 1.925152105965567e-05,
+      "loss": 0.2645,
+      "step": 767
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.33879125118255615,
+      "learning_rate": 1.9131870830827818e-05,
+      "loss": 0.2267,
+      "step": 768
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.42365795373916626,
+      "learning_rate": 1.9012505551987765e-05,
+      "loss": 0.2736,
+      "step": 769
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.37389981746673584,
+      "learning_rate": 1.8893426325020686e-05,
+      "loss": 0.2634,
+      "step": 770
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.41957971453666687,
+      "learning_rate": 1.8774634249171185e-05,
+      "loss": 0.4077,
+      "step": 771
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3793964982032776,
+      "learning_rate": 1.8656130421033123e-05,
+      "loss": 0.2957,
+      "step": 772
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.34412625432014465,
+      "learning_rate": 1.8537915934539486e-05,
+      "loss": 0.1677,
+      "step": 773
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3888733685016632,
+      "learning_rate": 1.841999188095224e-05,
+      "loss": 0.3111,
+      "step": 774
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.4144531786441803,
+      "learning_rate": 1.830235934885237e-05,
+      "loss": 0.3651,
+      "step": 775
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3423333168029785,
+      "learning_rate": 1.818501942412975e-05,
+      "loss": 0.2843,
+      "step": 776
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3905755877494812,
+      "learning_rate": 1.8067973189973075e-05,
+      "loss": 0.3341,
+      "step": 777
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.4139542579650879,
+      "learning_rate": 1.7951221726860045e-05,
+      "loss": 0.3393,
+      "step": 778
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.4078083634376526,
+      "learning_rate": 1.7834766112547142e-05,
+      "loss": 0.3124,
+      "step": 779
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3850538730621338,
+      "learning_rate": 1.771860742205988e-05,
+      "loss": 0.3099,
+      "step": 780
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.4191945493221283,
+      "learning_rate": 1.7602746727682796e-05,
+      "loss": 0.3639,
+      "step": 781
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3619956076145172,
+      "learning_rate": 1.7487185098949565e-05,
+      "loss": 0.2781,
+      "step": 782
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.42559903860092163,
+      "learning_rate": 1.7371923602633078e-05,
+      "loss": 0.3629,
+      "step": 783
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.43343913555145264,
+      "learning_rate": 1.725696330273575e-05,
+      "loss": 0.3785,
+      "step": 784
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.39102593064308167,
+      "learning_rate": 1.7142305260479474e-05,
+      "loss": 0.3651,
+      "step": 785
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3920254111289978,
+      "learning_rate": 1.7027950534296027e-05,
+      "loss": 0.378,
+      "step": 786
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3507162928581238,
+      "learning_rate": 1.6913900179817144e-05,
+      "loss": 0.2284,
+      "step": 787
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.385535329580307,
+      "learning_rate": 1.6800155249864896e-05,
+      "loss": 0.3308,
+      "step": 788
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.37612423300743103,
+      "learning_rate": 1.668671679444192e-05,
+      "loss": 0.2848,
+      "step": 789
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.4130227267742157,
+      "learning_rate": 1.6573585860721646e-05,
+      "loss": 0.3532,
+      "step": 790
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.37785717844963074,
+      "learning_rate": 1.646076349303884e-05,
+      "loss": 0.2867,
+      "step": 791
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3971453309059143,
+      "learning_rate": 1.63482507328797e-05,
+      "loss": 0.2969,
+      "step": 792
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.301319420337677,
+      "learning_rate": 1.6236048618872456e-05,
+      "loss": 0.2183,
+      "step": 793
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.486827552318573,
+      "learning_rate": 1.6124158186777676e-05,
+      "loss": 0.3165,
+      "step": 794
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4306073784828186,
+      "learning_rate": 1.6012580469478743e-05,
+      "loss": 0.3385,
+      "step": 795
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3236197829246521,
+      "learning_rate": 1.5901316496972262e-05,
+      "loss": 0.1959,
+      "step": 796
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.383569598197937,
+      "learning_rate": 1.5790367296358644e-05,
+      "loss": 0.2979,
+      "step": 797
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.31395941972732544,
+      "learning_rate": 1.5679733891832556e-05,
+      "loss": 0.2087,
+      "step": 798
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3919505178928375,
+      "learning_rate": 1.55694173046735e-05,
+      "loss": 0.2757,
+      "step": 799
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3555279076099396,
+      "learning_rate": 1.5459418553236343e-05,
+      "loss": 0.2804,
+      "step": 800
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3989741802215576,
+      "learning_rate": 1.5349738652941968e-05,
+      "loss": 0.3274,
+      "step": 801
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.38423246145248413,
+      "learning_rate": 1.5240378616267886e-05,
+      "loss": 0.2666,
+      "step": 802
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3812728524208069,
+      "learning_rate": 1.5131339452738863e-05,
+      "loss": 0.3723,
+      "step": 803
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.37258511781692505,
+      "learning_rate": 1.5022622168917649e-05,
+      "loss": 0.2737,
+      "step": 804
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.35156306624412537,
+      "learning_rate": 1.4914227768395595e-05,
+      "loss": 0.2617,
+      "step": 805
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.37645992636680603,
+      "learning_rate": 1.4806157251783515e-05,
+      "loss": 0.303,
+      "step": 806
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3626248836517334,
+      "learning_rate": 1.4698411616702356e-05,
+      "loss": 0.253,
+      "step": 807
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.37690359354019165,
+      "learning_rate": 1.4590991857774038e-05,
+      "loss": 0.3221,
+      "step": 808
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.416437029838562,
+      "learning_rate": 1.4483898966612209e-05,
+      "loss": 0.3789,
+      "step": 809
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3635142743587494,
+      "learning_rate": 1.437713393181317e-05,
+      "loss": 0.2786,
+      "step": 810
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.4015759527683258,
+      "learning_rate": 1.4270697738946704e-05,
+      "loss": 0.2582,
+      "step": 811
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3174329698085785,
+      "learning_rate": 1.4164591370547004e-05,
+      "loss": 0.2301,
+      "step": 812
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.38481318950653076,
+      "learning_rate": 1.4058815806103542e-05,
+      "loss": 0.2948,
+      "step": 813
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.4067227840423584,
+      "learning_rate": 1.3953372022052107e-05,
+      "loss": 0.3354,
+      "step": 814
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3462645411491394,
+      "learning_rate": 1.3848260991765755e-05,
+      "loss": 0.294,
+      "step": 815
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.4777897000312805,
+      "learning_rate": 1.3743483685545811e-05,
+      "loss": 0.4079,
+      "step": 816
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.4179569184780121,
+      "learning_rate": 1.363904107061294e-05,
+      "loss": 0.3217,
+      "step": 817
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.4066832661628723,
+      "learning_rate": 1.3534934111098179e-05,
+      "loss": 0.337,
+      "step": 818
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3682103455066681,
+      "learning_rate": 1.3431163768034077e-05,
+      "loss": 0.3227,
+      "step": 819
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3473730683326721,
+      "learning_rate": 1.3327730999345817e-05,
+      "loss": 0.2303,
+      "step": 820
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.35099923610687256,
+      "learning_rate": 1.3224636759842363e-05,
+      "loss": 0.2924,
+      "step": 821
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3207169473171234,
+      "learning_rate": 1.3121882001207614e-05,
+      "loss": 0.2208,
+      "step": 822
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3512991666793823,
+      "learning_rate": 1.3019467671991692e-05,
+      "loss": 0.2392,
+      "step": 823
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.4191690981388092,
+      "learning_rate": 1.2917394717602121e-05,
+      "loss": 0.3623,
+      "step": 824
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.34676387906074524,
+      "learning_rate": 1.2815664080295159e-05,
+      "loss": 0.252,
+      "step": 825
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.34901660680770874,
+      "learning_rate": 1.2714276699166994e-05,
+      "loss": 0.2367,
+      "step": 826
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3845536708831787,
+      "learning_rate": 1.261323351014525e-05,
+      "loss": 0.2552,
+      "step": 827
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3202722370624542,
+      "learning_rate": 1.251253544598014e-05,
+      "loss": 0.2054,
+      "step": 828
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.35958996415138245,
+      "learning_rate": 1.241218343623602e-05,
+      "loss": 0.3064,
+      "step": 829
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.37509864568710327,
+      "learning_rate": 1.2312178407282749e-05,
+      "loss": 0.3027,
+      "step": 830
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3806156814098358,
+      "learning_rate": 1.2212521282287092e-05,
+      "loss": 0.305,
+      "step": 831
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.43754497170448303,
+      "learning_rate": 1.2113212981204292e-05,
+      "loss": 0.3643,
+      "step": 832
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.41764360666275024,
+      "learning_rate": 1.2014254420769466e-05,
+      "loss": 0.3289,
+      "step": 833
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3798913061618805,
+      "learning_rate": 1.1915646514489292e-05,
+      "loss": 0.2902,
+      "step": 834
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.38450661301612854,
+      "learning_rate": 1.1817390172633403e-05,
+      "loss": 0.3193,
+      "step": 835
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3482416570186615,
+      "learning_rate": 1.1719486302226118e-05,
+      "loss": 0.243,
+      "step": 836
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5255814790725708,
+      "learning_rate": 1.1621935807038003e-05,
+      "loss": 0.4848,
+      "step": 837
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.41135770082473755,
+      "learning_rate": 1.152473958757756e-05,
+      "loss": 0.3362,
+      "step": 838
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3493039309978485,
+      "learning_rate": 1.1427898541082855e-05,
+      "loss": 0.2379,
+      "step": 839
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.4206957519054413,
+      "learning_rate": 1.133141356151336e-05,
+      "loss": 0.3234,
+      "step": 840
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.4049489498138428,
+      "learning_rate": 1.123528553954154e-05,
+      "loss": 0.3105,
+      "step": 841
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3023333251476288,
+      "learning_rate": 1.1139515362544755e-05,
+      "loss": 0.2024,
+      "step": 842
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.4536070227622986,
+      "learning_rate": 1.1044103914597031e-05,
+      "loss": 0.3822,
+      "step": 843
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5311452150344849,
+      "learning_rate": 1.0949052076460853e-05,
+      "loss": 0.4121,
+      "step": 844
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.4352911412715912,
+      "learning_rate": 1.085436072557911e-05,
+      "loss": 0.3284,
+      "step": 845
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3684670031070709,
+      "learning_rate": 1.0760030736066951e-05,
+      "loss": 0.2503,
+      "step": 846
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.38302332162857056,
+      "learning_rate": 1.0666062978703733e-05,
+      "loss": 0.2825,
+      "step": 847
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.36409202218055725,
+      "learning_rate": 1.0572458320924943e-05,
+      "loss": 0.2636,
+      "step": 848
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.44230782985687256,
+      "learning_rate": 1.0479217626814253e-05,
+      "loss": 0.3134,
+      "step": 849
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.32951366901397705,
+      "learning_rate": 1.0386341757095502e-05,
+      "loss": 0.198,
+      "step": 850
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.39737051725387573,
+      "learning_rate": 1.0293831569124774e-05,
+      "loss": 0.3287,
+      "step": 851
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.40820056200027466,
+      "learning_rate": 1.0201687916882418e-05,
+      "loss": 0.349,
+      "step": 852
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3584843575954437,
+      "learning_rate": 1.0109911650965314e-05,
+      "loss": 0.2142,
+      "step": 853
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5511006116867065,
+      "learning_rate": 1.0018503618578818e-05,
+      "loss": 0.5335,
+      "step": 854
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.40527984499931335,
+      "learning_rate": 9.927464663529118e-06,
+      "loss": 0.2826,
+      "step": 855
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4436098337173462,
+      "learning_rate": 9.836795626215356e-06,
+      "loss": 0.3708,
+      "step": 856
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3601125180721283,
+      "learning_rate": 9.746497343621857e-06,
+      "loss": 0.2946,
+      "step": 857
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3351758122444153,
+      "learning_rate": 9.656570649310481e-06,
+      "loss": 0.2355,
+      "step": 858
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3661154806613922,
+      "learning_rate": 9.567016373412857e-06,
+      "loss": 0.3093,
+      "step": 859
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.4545595347881317,
+      "learning_rate": 9.477835342622759e-06,
+      "loss": 0.3889,
+      "step": 860
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.4591468274593353,
+      "learning_rate": 9.389028380188419e-06,
+      "loss": 0.3956,
+      "step": 861
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3811791241168976,
+      "learning_rate": 9.300596305905013e-06,
+      "loss": 0.2942,
+      "step": 862
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3912218511104584,
+      "learning_rate": 9.212539936107029e-06,
+      "loss": 0.2768,
+      "step": 863
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3407474458217621,
+      "learning_rate": 9.124860083660769e-06,
+      "loss": 0.2694,
+      "step": 864
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3362795412540436,
+      "learning_rate": 9.037557557956766e-06,
+      "loss": 0.2052,
+      "step": 865
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.4200963079929352,
+      "learning_rate": 8.950633164902467e-06,
+      "loss": 0.3292,
+      "step": 866
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.40156418085098267,
+      "learning_rate": 8.86408770691462e-06,
+      "loss": 0.2826,
+      "step": 867
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3317236304283142,
+      "learning_rate": 8.777921982911996e-06,
+      "loss": 0.24,
+      "step": 868
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.37346014380455017,
+      "learning_rate": 8.692136788307903e-06,
+      "loss": 0.2693,
+      "step": 869
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.32652783393859863,
+      "learning_rate": 8.606732915003002e-06,
+      "loss": 0.2489,
+      "step": 870
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.36683058738708496,
+      "learning_rate": 8.521711151377803e-06,
+      "loss": 0.2614,
+      "step": 871
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.39762526750564575,
+      "learning_rate": 8.437072282285535e-06,
+      "loss": 0.3236,
+      "step": 872
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3494615852832794,
+      "learning_rate": 8.35281708904485e-06,
+      "loss": 0.2425,
+      "step": 873
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.38728398084640503,
+      "learning_rate": 8.268946349432582e-06,
+      "loss": 0.3143,
+      "step": 874
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.41258835792541504,
+      "learning_rate": 8.185460837676612e-06,
+      "loss": 0.2985,
+      "step": 875
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.4381546974182129,
+      "learning_rate": 8.102361324448715e-06,
+      "loss": 0.2868,
+      "step": 876
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.41193243861198425,
+      "learning_rate": 8.019648576857425e-06,
+      "loss": 0.3096,
+      "step": 877
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.4166334271430969,
+      "learning_rate": 7.937323358440935e-06,
+      "loss": 0.3443,
+      "step": 878
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3498817980289459,
+      "learning_rate": 7.85538642916015e-06,
+      "loss": 0.2727,
+      "step": 879
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.33879581093788147,
+      "learning_rate": 7.773838545391515e-06,
+      "loss": 0.2048,
+      "step": 880
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3829089403152466,
+      "learning_rate": 7.692680459920188e-06,
+      "loss": 0.2927,
+      "step": 881
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3402465879917145,
+      "learning_rate": 7.6119129219329395e-06,
+      "loss": 0.2253,
+      "step": 882
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.424730122089386,
+      "learning_rate": 7.5315366770114195e-06,
+      "loss": 0.335,
+      "step": 883
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.4533472955226898,
+      "learning_rate": 7.4515524671250725e-06,
+      "loss": 0.2951,
+      "step": 884
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.38933825492858887,
+      "learning_rate": 7.371961030624452e-06,
+      "loss": 0.257,
+      "step": 885
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.33994266390800476,
+      "learning_rate": 7.292763102234329e-06,
+      "loss": 0.2736,
+      "step": 886
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.356049507856369,
+      "learning_rate": 7.213959413046894e-06,
+      "loss": 0.2118,
+      "step": 887
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.35656169056892395,
+      "learning_rate": 7.135550690515052e-06,
+      "loss": 0.2757,
+      "step": 888
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3579034209251404,
+      "learning_rate": 7.057537658445701e-06,
+      "loss": 0.2367,
+      "step": 889
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.2979848086833954,
+      "learning_rate": 6.979921036993042e-06,
+      "loss": 0.1987,
+      "step": 890
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.42152541875839233,
+      "learning_rate": 6.902701542651874e-06,
+      "loss": 0.2744,
+      "step": 891
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3124920725822449,
+      "learning_rate": 6.825879888251135e-06,
+      "loss": 0.1845,
+      "step": 892
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3792358934879303,
+      "learning_rate": 6.749456782947122e-06,
+      "loss": 0.2557,
+      "step": 893
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.39623990654945374,
+      "learning_rate": 6.6734329322171165e-06,
+      "loss": 0.3049,
+      "step": 894
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3868322968482971,
+      "learning_rate": 6.597809037852726e-06,
+      "loss": 0.3433,
+      "step": 895
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3803116977214813,
+      "learning_rate": 6.522585797953579e-06,
+      "loss": 0.2398,
+      "step": 896
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4614750146865845,
+      "learning_rate": 6.447763906920679e-06,
+      "loss": 0.4015,
+      "step": 897
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4475559592247009,
+      "learning_rate": 6.373344055450165e-06,
+      "loss": 0.336,
+      "step": 898
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.366828054189682,
+      "learning_rate": 6.2993269305268495e-06,
+      "loss": 0.2346,
+      "step": 899
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.41806674003601074,
+      "learning_rate": 6.2257132154178665e-06,
+      "loss": 0.3423,
+      "step": 900
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.4186936020851135,
+      "learning_rate": 6.152503589666425e-06,
+      "loss": 0.3321,
+      "step": 901
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.4075973331928253,
+      "learning_rate": 6.079698729085498e-06,
+      "loss": 0.306,
+      "step": 902
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3074958920478821,
+      "learning_rate": 6.007299305751585e-06,
+      "loss": 0.185,
+      "step": 903
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3523412048816681,
+      "learning_rate": 5.935305987998496e-06,
+      "loss": 0.2477,
+      "step": 904
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3624049425125122,
+      "learning_rate": 5.863719440411214e-06,
+      "loss": 0.2392,
+      "step": 905
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3795306086540222,
+      "learning_rate": 5.792540323819751e-06,
+      "loss": 0.3398,
+      "step": 906
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3727322220802307,
+      "learning_rate": 5.721769295293034e-06,
+      "loss": 0.2522,
+      "step": 907
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.44352471828460693,
+      "learning_rate": 5.651407008132809e-06,
+      "loss": 0.3619,
+      "step": 908
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3567368984222412,
+      "learning_rate": 5.5814541118677284e-06,
+      "loss": 0.2106,
+      "step": 909
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.4105832874774933,
+      "learning_rate": 5.5119112522471924e-06,
+      "loss": 0.3556,
+      "step": 910
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.37798091769218445,
+      "learning_rate": 5.442779071235516e-06,
+      "loss": 0.2559,
+      "step": 911
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3410490155220032,
+      "learning_rate": 5.374058207005944e-06,
+      "loss": 0.2455,
+      "step": 912
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.378336101770401,
+      "learning_rate": 5.305749293934764e-06,
+      "loss": 0.2541,
+      "step": 913
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.4319799840450287,
+      "learning_rate": 5.237852962595469e-06,
+      "loss": 0.3012,
+      "step": 914
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.33922573924064636,
+      "learning_rate": 5.170369839752925e-06,
+      "loss": 0.2038,
+      "step": 915
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3704172372817993,
+      "learning_rate": 5.1033005483575925e-06,
+      "loss": 0.2578,
+      "step": 916
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.41262710094451904,
+      "learning_rate": 5.036645707539745e-06,
+      "loss": 0.2624,
+      "step": 917
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.40159058570861816,
+      "learning_rate": 4.9704059326038055e-06,
+      "loss": 0.2647,
+      "step": 918
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.38587838411331177,
+      "learning_rate": 4.90458183502262e-06,
+      "loss": 0.2894,
+      "step": 919
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.4216020107269287,
+      "learning_rate": 4.839174022431858e-06,
+      "loss": 0.3768,
+      "step": 920
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.41159510612487793,
+      "learning_rate": 4.7741830986243356e-06,
+      "loss": 0.3103,
+      "step": 921
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3718778192996979,
+      "learning_rate": 4.709609663544534e-06,
+      "loss": 0.2877,
+      "step": 922
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3785461187362671,
+      "learning_rate": 4.645454313282965e-06,
+      "loss": 0.2953,
+      "step": 923
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4794873893260956,
+      "learning_rate": 4.581717640070743e-06,
+      "loss": 0.3446,
+      "step": 924
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.40924200415611267,
+      "learning_rate": 4.5184002322740785e-06,
+      "loss": 0.3418,
+      "step": 925
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3224068880081177,
+      "learning_rate": 4.455502674388873e-06,
+      "loss": 0.2382,
+      "step": 926
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.46128109097480774,
+      "learning_rate": 4.3930255470352736e-06,
+      "loss": 0.35,
+      "step": 927
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4323473572731018,
+      "learning_rate": 4.330969426952375e-06,
+      "loss": 0.3307,
+      "step": 928
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.40308627486228943,
+      "learning_rate": 4.269334886992876e-06,
+      "loss": 0.2882,
+      "step": 929
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5716599822044373,
+      "learning_rate": 4.208122496117744e-06,
+      "loss": 0.2227,
+      "step": 930
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.38822227716445923,
+      "learning_rate": 4.147332819391048e-06,
+      "loss": 0.2759,
+      "step": 931
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.40629640221595764,
+      "learning_rate": 4.0869664179746694e-06,
+      "loss": 0.2665,
+      "step": 932
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.35057899355888367,
+      "learning_rate": 4.027023849123157e-06,
+      "loss": 0.2333,
+      "step": 933
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.393621563911438,
+      "learning_rate": 3.967505666178556e-06,
+      "loss": 0.2696,
+      "step": 934
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.36038023233413696,
+      "learning_rate": 3.908412418565371e-06,
+      "loss": 0.2546,
+      "step": 935
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.34687674045562744,
+      "learning_rate": 3.849744651785381e-06,
+      "loss": 0.2528,
+      "step": 936
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3856310546398163,
+      "learning_rate": 3.7915029074126974e-06,
+      "loss": 0.3603,
+      "step": 937
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5023623704910278,
+      "learning_rate": 3.7336877230887246e-06,
+      "loss": 0.4284,
+      "step": 938
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.4204338788986206,
+      "learning_rate": 3.676299632517216e-06,
+      "loss": 0.4169,
+      "step": 939
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5023021101951599,
+      "learning_rate": 3.619339165459307e-06,
+      "loss": 0.3397,
+      "step": 940
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.341836154460907,
+      "learning_rate": 3.562806847728678e-06,
+      "loss": 0.2697,
+      "step": 941
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.40847527980804443,
+      "learning_rate": 3.5067032011866783e-06,
+      "loss": 0.3658,
+      "step": 942
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3342381417751312,
+      "learning_rate": 3.4510287437374835e-06,
+      "loss": 0.1656,
+      "step": 943
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3953072428703308,
+      "learning_rate": 3.3957839893233536e-06,
+      "loss": 0.2481,
+      "step": 944
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3248569667339325,
+      "learning_rate": 3.340969447919873e-06,
+      "loss": 0.2469,
+      "step": 945
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3305179178714752,
+      "learning_rate": 3.286585625531241e-06,
+      "loss": 0.2154,
+      "step": 946
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.33991092443466187,
+      "learning_rate": 3.232633024185583e-06,
+      "loss": 0.2226,
+      "step": 947
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.30957362055778503,
+      "learning_rate": 3.1791121419303794e-06,
+      "loss": 0.2185,
+      "step": 948
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.4127846360206604,
+      "learning_rate": 3.1260234728277717e-06,
+      "loss": 0.2835,
+      "step": 949
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.39826324582099915,
+      "learning_rate": 3.0733675069500865e-06,
+      "loss": 0.2953,
+      "step": 950
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.4400089979171753,
+      "learning_rate": 3.0211447303752695e-06,
+      "loss": 0.296,
+      "step": 951
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.38311436772346497,
+      "learning_rate": 2.9693556251824185e-06,
+      "loss": 0.29,
+      "step": 952
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.46116960048675537,
+      "learning_rate": 2.9180006694472906e-06,
+      "loss": 0.4249,
+      "step": 953
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.40703484416007996,
+      "learning_rate": 2.867080337237954e-06,
+      "loss": 0.3635,
+      "step": 954
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.42896768450737,
+      "learning_rate": 2.8165950986103805e-06,
+      "loss": 0.3774,
+      "step": 955
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4287061393260956,
+      "learning_rate": 2.7665454196040664e-06,
+      "loss": 0.335,
+      "step": 956
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4173295199871063,
+      "learning_rate": 2.716931762237801e-06,
+      "loss": 0.2793,
+      "step": 957
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4245130121707916,
+      "learning_rate": 2.667754584505372e-06,
+      "loss": 0.3427,
+      "step": 958
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4839573800563812,
+      "learning_rate": 2.6190143403713174e-06,
+      "loss": 0.3515,
+      "step": 959
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.434766948223114,
+      "learning_rate": 2.5707114797667465e-06,
+      "loss": 0.2847,
+      "step": 960
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4157213568687439,
+      "learning_rate": 2.522846448585231e-06,
+      "loss": 0.352,
+      "step": 961
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.40966856479644775,
+      "learning_rate": 2.4754196886785986e-06,
+      "loss": 0.27,
+      "step": 962
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3235345184803009,
+      "learning_rate": 2.4284316378529404e-06,
+      "loss": 0.2657,
+      "step": 963
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3809812366962433,
+      "learning_rate": 2.3818827298645207e-06,
+      "loss": 0.2526,
+      "step": 964
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.33908405900001526,
+      "learning_rate": 2.335773394415802e-06,
+      "loss": 0.2192,
+      "step": 965
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3907183110713959,
+      "learning_rate": 2.2901040571514322e-06,
+      "loss": 0.2976,
+      "step": 966
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.4197416305541992,
+      "learning_rate": 2.2448751396543787e-06,
+      "loss": 0.2952,
+      "step": 967
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3461250364780426,
+      "learning_rate": 2.2000870594419908e-06,
+      "loss": 0.1856,
+      "step": 968
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.36310264468193054,
+      "learning_rate": 2.155740229962161e-06,
+      "loss": 0.261,
+      "step": 969
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.349942147731781,
+      "learning_rate": 2.1118350605894955e-06,
+      "loss": 0.3009,
+      "step": 970
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.35959210991859436,
+      "learning_rate": 2.068371956621562e-06,
+      "loss": 0.2599,
+      "step": 971
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.39764103293418884,
+      "learning_rate": 2.0253513192751373e-06,
+      "loss": 0.3175,
+      "step": 972
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3876802623271942,
+      "learning_rate": 1.982773545682459e-06,
+      "loss": 0.298,
+      "step": 973
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.34669044613838196,
+      "learning_rate": 1.9406390288876586e-06,
+      "loss": 0.2254,
+      "step": 974
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3759954273700714,
+      "learning_rate": 1.8989481578430223e-06,
+      "loss": 0.2477,
+      "step": 975
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.42282500863075256,
+      "learning_rate": 1.8577013174054857e-06,
+      "loss": 0.3294,
+      "step": 976
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.40045222640037537,
+      "learning_rate": 1.8168988883330185e-06,
+      "loss": 0.3472,
+      "step": 977
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4552788734436035,
+      "learning_rate": 1.7765412472811771e-06,
+      "loss": 0.4431,
+      "step": 978
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3732775151729584,
+      "learning_rate": 1.7366287667995417e-06,
+      "loss": 0.2226,
+      "step": 979
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3581148386001587,
+      "learning_rate": 1.697161815328363e-06,
+      "loss": 0.2576,
+      "step": 980
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3295496106147766,
+      "learning_rate": 1.6581407571951092e-06,
+      "loss": 0.2751,
+      "step": 981
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3627816438674927,
+      "learning_rate": 1.6195659526111185e-06,
+      "loss": 0.2633,
+      "step": 982
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.42608195543289185,
+      "learning_rate": 1.5814377576682527e-06,
+      "loss": 0.3034,
+      "step": 983
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3478807508945465,
+      "learning_rate": 1.5437565243356656e-06,
+      "loss": 0.1997,
+      "step": 984
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.41831809282302856,
+      "learning_rate": 1.5065226004564893e-06,
+      "loss": 0.3793,
+      "step": 985
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3832615911960602,
+      "learning_rate": 1.4697363297446477e-06,
+      "loss": 0.2761,
+      "step": 986
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.4452967941761017,
+      "learning_rate": 1.4333980517817203e-06,
+      "loss": 0.3798,
+      "step": 987
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3327552080154419,
+      "learning_rate": 1.3975081020137392e-06,
+      "loss": 0.204,
+      "step": 988
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5041511058807373,
+      "learning_rate": 1.3620668117481472e-06,
+      "loss": 0.3522,
+      "step": 989
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.37212684750556946,
+      "learning_rate": 1.3270745081506997e-06,
+      "loss": 0.3307,
+      "step": 990
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5050026774406433,
+      "learning_rate": 1.292531514242501e-06,
+      "loss": 0.4735,
+      "step": 991
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.4048091471195221,
+      "learning_rate": 1.2584381488969454e-06,
+      "loss": 0.2996,
+      "step": 992
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3142896890640259,
+      "learning_rate": 1.2247947268368364e-06,
+      "loss": 0.2256,
+      "step": 993
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.37308236956596375,
+      "learning_rate": 1.191601558631461e-06,
+      "loss": 0.26,
+      "step": 994
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.35677245259284973,
+      "learning_rate": 1.1588589506937198e-06,
+      "loss": 0.2756,
+      "step": 995
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5449343919754028,
+      "learning_rate": 1.126567205277279e-06,
+      "loss": 0.3556,
+      "step": 996
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3877151310443878,
+      "learning_rate": 1.094726620473835e-06,
+      "loss": 0.2792,
+      "step": 997
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3979206085205078,
+      "learning_rate": 1.0633374902103088e-06,
+      "loss": 0.3125,
+      "step": 998
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.41529783606529236,
+      "learning_rate": 1.0324001042461395e-06,
+      "loss": 0.2613,
+      "step": 999
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.4048457145690918,
+      "learning_rate": 1.0019147481706625e-06,
+      "loss": 0.3312,
+      "step": 1000
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.4872767925262451,
+      "learning_rate": 9.718817034003901e-07,
+      "loss": 0.4226,
+      "step": 1001
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.32919400930404663,
+      "learning_rate": 9.423012471764914e-07,
+      "loss": 0.1977,
+      "step": 1002
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.342472106218338,
+      "learning_rate": 9.131736525621603e-07,
+      "loss": 0.2218,
+      "step": 1003
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.4028159976005554,
+      "learning_rate": 8.844991884401854e-07,
+      "loss": 0.286,
+      "step": 1004
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3890497386455536,
+      "learning_rate": 8.56278119510362e-07,
+      "loss": 0.274,
+      "step": 1005
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.4593218266963959,
+      "learning_rate": 8.285107062871333e-07,
+      "loss": 0.3539,
+      "step": 1006
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.35354185104370117,
+      "learning_rate": 8.011972050971483e-07,
+      "loss": 0.272,
+      "step": 1007
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.40247949957847595,
+      "learning_rate": 7.74337868076902e-07,
+      "loss": 0.3272,
+      "step": 1008
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.4523579478263855,
+      "learning_rate": 7.479329431703985e-07,
+      "loss": 0.4128,
+      "step": 1009
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.47494736313819885,
+      "learning_rate": 7.21982674126881e-07,
+      "loss": 0.3574,
+      "step": 1010
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.35004284977912903,
+      "learning_rate": 6.964873004985717e-07,
+      "loss": 0.2341,
+      "step": 1011
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.47898706793785095,
+      "learning_rate": 6.714470576384579e-07,
+      "loss": 0.3763,
+      "step": 1012
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3801863491535187,
+      "learning_rate": 6.468621766981154e-07,
+      "loss": 0.2777,
+      "step": 1013
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3755449056625366,
+      "learning_rate": 6.227328846255931e-07,
+      "loss": 0.2736,
+      "step": 1014
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3443692922592163,
+      "learning_rate": 5.990594041632991e-07,
+      "loss": 0.2544,
+      "step": 1015
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.36282068490982056,
+      "learning_rate": 5.758419538459459e-07,
+      "loss": 0.2778,
+      "step": 1016
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.45862850546836853,
+      "learning_rate": 5.530807479985633e-07,
+      "loss": 0.3454,
+      "step": 1017
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3214671313762665,
+      "learning_rate": 5.307759967344672e-07,
+      "loss": 0.2648,
+      "step": 1018
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4228048622608185,
+      "learning_rate": 5.089279059533658e-07,
+      "loss": 0.3266,
+      "step": 1019
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.43997371196746826,
+      "learning_rate": 4.87536677339423e-07,
+      "loss": 0.3662,
+      "step": 1020
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3612721264362335,
+      "learning_rate": 4.666025083594483e-07,
+      "loss": 0.2639,
+      "step": 1021
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4024774432182312,
+      "learning_rate": 4.461255922609986e-07,
+      "loss": 0.3015,
+      "step": 1022
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4298073649406433,
+      "learning_rate": 4.261061180706627e-07,
+      "loss": 0.3631,
+      "step": 1023
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3973303735256195,
+      "learning_rate": 4.065442705922906e-07,
+      "loss": 0.3697,
+      "step": 1024
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4075637757778168,
+      "learning_rate": 3.8744023040528374e-07,
+      "loss": 0.3019,
+      "step": 1025
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4226757884025574,
+      "learning_rate": 3.687941738629186e-07,
+      "loss": 0.3541,
+      "step": 1026
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3980466425418854,
+      "learning_rate": 3.5060627309074224e-07,
+      "loss": 0.2771,
+      "step": 1027
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3921937048435211,
+      "learning_rate": 3.3287669598497383e-07,
+      "loss": 0.2716,
+      "step": 1028
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.43594250082969666,
+      "learning_rate": 3.156056062109503e-07,
+      "loss": 0.3511,
+      "step": 1029
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.33162230253219604,
+      "learning_rate": 2.987931632016272e-07,
+      "loss": 0.2191,
+      "step": 1030
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4063693881034851,
+      "learning_rate": 2.824395221560805e-07,
+      "loss": 0.2922,
+      "step": 1031
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3888567090034485,
+      "learning_rate": 2.665448340381016e-07,
+      "loss": 0.2674,
+      "step": 1032
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3596007525920868,
+      "learning_rate": 2.511092455747932e-07,
+      "loss": 0.2949,
+      "step": 1033
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3734399676322937,
+      "learning_rate": 2.361328992552314e-07,
+      "loss": 0.2538,
+      "step": 1034
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.48918384313583374,
+      "learning_rate": 2.2161593332910013e-07,
+      "loss": 0.3673,
+      "step": 1035
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.43824249505996704,
+      "learning_rate": 2.0755848180547543e-07,
+      "loss": 0.3915,
+      "step": 1036
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3258342146873474,
+      "learning_rate": 1.9396067445155986e-07,
+      "loss": 0.2233,
+      "step": 1037
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4354231357574463,
+      "learning_rate": 1.8082263679148337e-07,
+      "loss": 0.3945,
+      "step": 1038
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.41941532492637634,
+      "learning_rate": 1.681444901051432e-07,
+      "loss": 0.3513,
+      "step": 1039
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.42694929242134094,
+      "learning_rate": 1.5592635142709367e-07,
+      "loss": 0.2924,
+      "step": 1040
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.377742201089859,
+      "learning_rate": 1.4416833354546356e-07,
+      "loss": 0.2663,
+      "step": 1041
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.39117175340652466,
+      "learning_rate": 1.328705450009071e-07,
+      "loss": 0.2678,
+      "step": 1042
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.44528332352638245,
+      "learning_rate": 1.2203309008561592e-07,
+      "loss": 0.3609,
+      "step": 1043
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.464968740940094,
+      "learning_rate": 1.1165606884234181e-07,
+      "loss": 0.3227,
+      "step": 1044
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.4331037700176239,
+      "learning_rate": 1.0173957706348659e-07,
+      "loss": 0.3268,
+      "step": 1045
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.40670907497406006,
+      "learning_rate": 9.228370629019711e-08,
+      "loss": 0.2878,
+      "step": 1046
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.4464401602745056,
+      "learning_rate": 8.328854381154938e-08,
+      "loss": 0.3512,
+      "step": 1047
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.36872437596321106,
+      "learning_rate": 7.475417266371576e-08,
+      "loss": 0.2733,
+      "step": 1048
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.35684168338775635,
+      "learning_rate": 6.668067162921566e-08,
+      "loss": 0.263,
+      "step": 1049
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.41713425517082214,
+      "learning_rate": 5.906811523618272e-08,
+      "loss": 0.2968,
+      "step": 1050
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.4490773379802704,
+      "learning_rate": 5.191657375767656e-08,
+      "loss": 0.3842,
+      "step": 1051
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.39667680859565735,
+      "learning_rate": 4.522611321103321e-08,
+      "loss": 0.2798,
+      "step": 1052
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.34880930185317993,
+      "learning_rate": 3.8996795357254535e-08,
+      "loss": 0.2241,
+      "step": 1053
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5244704484939575,
+      "learning_rate": 3.322867770044202e-08,
+      "loss": 0.378,
+      "step": 1054
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.4561891555786133,
+      "learning_rate": 2.792181348726941e-08,
+      "loss": 0.3333,
+      "step": 1055
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.422680526971817,
+      "learning_rate": 2.3076251706477536e-08,
+      "loss": 0.3322,
+      "step": 1056
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.45002683997154236,
+      "learning_rate": 1.869203708843581e-08,
+      "loss": 0.3724,
+      "step": 1057
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3882070779800415,
+      "learning_rate": 1.476921010471477e-08,
+      "loss": 0.3165,
+      "step": 1058
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.42735832929611206,
+      "learning_rate": 1.1307806967741919e-08,
+      "loss": 0.2957,
+      "step": 1059
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3215475380420685,
+      "learning_rate": 8.307859630429793e-09,
+      "loss": 0.2327,
+      "step": 1060
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.4432857632637024,
+      "learning_rate": 5.7693957858984125e-09,
+      "loss": 0.3602,
+      "step": 1061
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.41549310088157654,
+      "learning_rate": 3.6924388672254785e-09,
+      "loss": 0.3299,
+      "step": 1062
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.3883119523525238,
+      "learning_rate": 2.0770080472298783e-09,
+      "loss": 0.2913,
+      "step": 1063
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5132927298545837,
+      "learning_rate": 9.231182382773984e-10,
+      "loss": 0.3904,
+      "step": 1064
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.3842669427394867,
+      "learning_rate": 2.307800921641512e-10,
+      "loss": 0.2883,
+      "step": 1065
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.3865910768508911,
+      "learning_rate": 0.0,
+      "loss": 0.276,
+      "step": 1066
+    },
+    {
+      "epoch": 1.0,
+      "step": 1066,
+      "total_flos": 1.171159722670162e+17,
+      "train_loss": 0.31279155516937573,
+      "train_runtime": 54620.1987,
+      "train_samples_per_second": 0.625,
+      "train_steps_per_second": 0.02
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 1066,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 100,
+  "total_flos": 1.171159722670162e+17,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}