diff --git a/llavaqwen2-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split00_all_mm_tune_olora256_512_llm/README.md b/llavaqwen2-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split00_all_mm_tune_olora256_512_llm/README.md
new file mode 100644
index 0000000000000000000000000000000000000000..d4576fe074287232d3836bf69c21d3f2593290d9
--- /dev/null
+++ b/llavaqwen2-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split00_all_mm_tune_olora256_512_llm/README.md
@@ -0,0 +1,9 @@
+---
+library_name: peft
+---
+## Training procedure
+
+### Framework versions
+
+
+- PEFT 0.4.0
diff --git a/llavaqwen2-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split00_all_mm_tune_olora256_512_llm/adapter_config.json b/llavaqwen2-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split00_all_mm_tune_olora256_512_llm/adapter_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..e1cb137dcd67002024395b7d0ec5f27ba099057b
--- /dev/null
+++ b/llavaqwen2-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split00_all_mm_tune_olora256_512_llm/adapter_config.json
@@ -0,0 +1,26 @@
+{
+  "auto_mapping": null,
+  "base_model_name_or_path": "lmms-lab/LLaVA-Video-7B-Qwen2",
+  "bias": "none",
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": "olora",
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "lora_alpha": 512,
+  "lora_dropout": 0.05,
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "r": 256,
+  "revision": null,
+  "target_modules": [
+    "v_proj",
+    "down_proj",
+    "k_proj",
+    "o_proj",
+    "q_proj",
+    "up_proj",
+    "gate_proj"
+  ],
+  "task_type": "CAUSAL_LM"
+}
\ No newline at end of file
diff --git a/llavaqwen2-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split00_all_mm_tune_olora256_512_llm/adapter_model.bin b/llavaqwen2-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split00_all_mm_tune_olora256_512_llm/adapter_model.bin
new file mode 100644
index 0000000000000000000000000000000000000000..a819e9f89f8f640aca9dae9b62ccaec1c53d904c
--- /dev/null
+++ b/llavaqwen2-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split00_all_mm_tune_olora256_512_llm/adapter_model.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c9ed0c732db686c9cc96998774e6c637420d6e91dd9fa322165ad0fb7df98a35
+size 1384057050
diff --git a/llavaqwen2-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split00_all_mm_tune_olora256_512_llm/config.json b/llavaqwen2-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split00_all_mm_tune_olora256_512_llm/config.json
new file mode 100644
index 0000000000000000000000000000000000000000..1a67f02d0063c3de7740207b9ab2a3eb7be1cbe3
--- /dev/null
+++ b/llavaqwen2-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split00_all_mm_tune_olora256_512_llm/config.json
@@ -0,0 +1,221 @@
+{
+  "_name_or_path": "lmms-lab/LLaVA-Video-7B-Qwen2",
+  "add_faster_video": false,
+  "add_time_instruction": true,
+  "architectures": [
+    "LlavaQwenForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 151643,
+  "eos_token_id": 151645,
+  "faster_token_stride": 10,
+  "force_sample": true,
+  "hidden_act": "silu",
+  "hidden_size": 3584,
+  "ignore_index": -100,
+  "image_aspect_ratio": "anyres_max_9",
+  "image_crop_resolution": null,
+  "image_grid_pinpoints": [
+    [
+      384,
+      384
+    ],
+    [
+      384,
+      768
+    ],
+    [
+      384,
+      1152
+    ],
+    [
+      384,
+      1536
+    ],
+    [
+      384,
+      1920
+    ],
+    [
+      384,
+      2304
+    ],
+    [
+      768,
+      384
+    ],
+    [
+      768,
+      768
+    ],
+    [
+      768,
+      1152
+    ],
+    [
+      768,
+      1536
+    ],
+    [
+      768,
+      1920
+    ],
+    [
+      768,
+      2304
+    ],
+    [
+      1152,
+      384
+    ],
+    [
+      1152,
+      768
+    ],
+    [
+      1152,
+      1152
+    ],
+    [
+      1152,
+      1536
+    ],
+    [
+      1152,
+      1920
+    ],
+    [
+      1152,
+      2304
+    ],
+    [
+      1536,
+      384
+    ],
+    [
+      1536,
+      768
+    ],
+    [
+      1536,
+      1152
+    ],
+    [
+      1536,
+      1536
+    ],
+    [
+      1536,
+      1920
+    ],
+    [
+      1536,
+      2304
+    ],
+    [
+      1920,
+      384
+    ],
+    [
+      1920,
+      768
+    ],
+    [
+      1920,
+      1152
+    ],
+    [
+      1920,
+      1536
+    ],
+    [
+      1920,
+      1920
+    ],
+    [
+      1920,
+      2304
+    ],
+    [
+      2304,
+      384
+    ],
+    [
+      2304,
+      768
+    ],
+    [
+      2304,
+      1152
+    ],
+    [
+      2304,
+      1536
+    ],
+    [
+      2304,
+      1920
+    ],
+    [
+      2304,
+      2304
+    ]
+  ],
+  "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": "grid",
+  "mm_patch_merge_type": "spatial_unpad",
+  "mm_projector_lr": 2e-05,
+  "mm_projector_type": "mlp2x_gelu",
+  "mm_resampler_type": null,
+  "mm_spatial_pool_mode": "bilinear",
+  "mm_spatial_pool_stride": 2,
+  "mm_tunable_parts": "mm_vision_tower,mm_mlp_adapter,mm_language_model",
+  "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",
+  "num_attention_heads": 28,
+  "num_hidden_layers": 28,
+  "num_key_value_heads": 4,
+  "pos_skipping_range": 4096,
+  "projector_hidden_act": "gelu",
+  "rms_norm_eps": 1e-06,
+  "rope_scaling": null,
+  "rope_theta": 1000000.0,
+  "sliding_window": 131072,
+  "text_config": {
+    "model_type": "llama"
+  },
+  "tie_word_embeddings": false,
+  "tokenizer_model_max_length": 32768,
+  "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_config": {
+    "hidden_size": 1024,
+    "image_size": 336,
+    "intermediate_size": 4096,
+    "model_type": "clip_vision_model",
+    "num_attention_heads": 16,
+    "num_hidden_layers": 24,
+    "patch_size": 14,
+    "projection_dim": 768,
+    "vocab_size": 32000
+  },
+  "vision_feature_layer": -2,
+  "vision_feature_select_strategy": "default",
+  "vision_tower_pretrained": null
+}
diff --git a/llavaqwen2-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split00_all_mm_tune_olora256_512_llm/generation_config.json b/llavaqwen2-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split00_all_mm_tune_olora256_512_llm/generation_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..19a297221acb87418d4388a3decef2282c6d7316
--- /dev/null
+++ b/llavaqwen2-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split00_all_mm_tune_olora256_512_llm/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/llavaqwen2-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split00_all_mm_tune_olora256_512_llm/non_lora_trainables.bin b/llavaqwen2-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split00_all_mm_tune_olora256_512_llm/non_lora_trainables.bin
new file mode 100644
index 0000000000000000000000000000000000000000..7a032beeac98ae79c66c8b1418b282f65f052797
--- /dev/null
+++ b/llavaqwen2-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split00_all_mm_tune_olora256_512_llm/non_lora_trainables.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0afc643d6c17a6e89d0a2528bebd61a99d7f47b7518d57a6787f55cfb1cc67fb
+size 33964208
diff --git a/llavaqwen2-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split00_all_mm_tune_olora256_512_llm/trainer_state.json b/llavaqwen2-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split00_all_mm_tune_olora256_512_llm/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..aa11d31093310a3cd11eb7634f7927e53c8a4e18
--- /dev/null
+++ b/llavaqwen2-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split00_all_mm_tune_olora256_512_llm/trainer_state.json
@@ -0,0 +1,3530 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 500,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.928880050484433,
+      "learning_rate": 6.666666666666667e-07,
+      "loss": 0.4907,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.588300236671549,
+      "learning_rate": 1.3333333333333334e-06,
+      "loss": 0.4211,
+      "step": 2
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.339991984294469,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 0.7389,
+      "step": 3
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.970866801586073,
+      "learning_rate": 2.666666666666667e-06,
+      "loss": 1.5158,
+      "step": 4
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.847088910892997,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 0.5582,
+      "step": 5
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.902184023132924,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 0.5615,
+      "step": 6
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1103083869089794,
+      "learning_rate": 4.666666666666667e-06,
+      "loss": 0.5533,
+      "step": 7
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3753084815529804,
+      "learning_rate": 5.333333333333334e-06,
+      "loss": 0.3678,
+      "step": 8
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6008535027371964,
+      "learning_rate": 6e-06,
+      "loss": 0.4722,
+      "step": 9
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4125509496168551,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 0.4045,
+      "step": 10
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.7161654242431323,
+      "learning_rate": 7.333333333333333e-06,
+      "loss": 0.7498,
+      "step": 11
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.402522894370017,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 0.4185,
+      "step": 12
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.513671206638328,
+      "learning_rate": 8.666666666666668e-06,
+      "loss": 0.309,
+      "step": 13
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9819893353895838,
+      "learning_rate": 9.333333333333334e-06,
+      "loss": 0.2349,
+      "step": 14
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3027073556008346,
+      "learning_rate": 1e-05,
+      "loss": 0.3234,
+      "step": 15
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4381709175193402,
+      "learning_rate": 9.999895105006995e-06,
+      "loss": 0.2945,
+      "step": 16
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.052815018764843,
+      "learning_rate": 9.99958042442916e-06,
+      "loss": 0.3493,
+      "step": 17
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6196920005213762,
+      "learning_rate": 9.999055971469864e-06,
+      "loss": 0.2761,
+      "step": 18
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4712226066536305,
+      "learning_rate": 9.998321768134101e-06,
+      "loss": 0.2012,
+      "step": 19
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8568837598037423,
+      "learning_rate": 9.997377845227577e-06,
+      "loss": 0.1712,
+      "step": 20
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9693693191969218,
+      "learning_rate": 9.9962242423554e-06,
+      "loss": 0.2447,
+      "step": 21
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9895575629794001,
+      "learning_rate": 9.99486100792044e-06,
+      "loss": 0.1854,
+      "step": 22
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7594217480912717,
+      "learning_rate": 9.993288199121283e-06,
+      "loss": 0.1996,
+      "step": 23
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0192540791922693,
+      "learning_rate": 9.991505881949837e-06,
+      "loss": 0.3051,
+      "step": 24
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3919378979755264,
+      "learning_rate": 9.98951413118856e-06,
+      "loss": 0.1417,
+      "step": 25
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7508302307462176,
+      "learning_rate": 9.987313030407325e-06,
+      "loss": 0.1757,
+      "step": 26
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0172169036173535,
+      "learning_rate": 9.984902671959911e-06,
+      "loss": 0.2013,
+      "step": 27
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3428007732278717,
+      "learning_rate": 9.982283156980133e-06,
+      "loss": 0.1983,
+      "step": 28
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7539923799727971,
+      "learning_rate": 9.979454595377594e-06,
+      "loss": 0.1219,
+      "step": 29
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8659715135984464,
+      "learning_rate": 9.97641710583307e-06,
+      "loss": 0.1899,
+      "step": 30
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7710723460098894,
+      "learning_rate": 9.973170815793543e-06,
+      "loss": 0.134,
+      "step": 31
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7935213054318757,
+      "learning_rate": 9.969715861466839e-06,
+      "loss": 0.1215,
+      "step": 32
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0478419132907202,
+      "learning_rate": 9.966052387815923e-06,
+      "loss": 0.1752,
+      "step": 33
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.056628710250906,
+      "learning_rate": 9.962180548552812e-06,
+      "loss": 0.1691,
+      "step": 34
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6599623594758851,
+      "learning_rate": 9.958100506132127e-06,
+      "loss": 0.11,
+      "step": 35
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8740425578465124,
+      "learning_rate": 9.953812431744274e-06,
+      "loss": 0.1382,
+      "step": 36
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9871589222528848,
+      "learning_rate": 9.94931650530827e-06,
+      "loss": 0.1378,
+      "step": 37
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0965428723520145,
+      "learning_rate": 9.944612915464183e-06,
+      "loss": 0.1911,
+      "step": 38
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8288408100645241,
+      "learning_rate": 9.93970185956522e-06,
+      "loss": 0.1526,
+      "step": 39
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8489303067819424,
+      "learning_rate": 9.934583543669454e-06,
+      "loss": 0.1406,
+      "step": 40
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4514902977171549,
+      "learning_rate": 9.929258182531167e-06,
+      "loss": 0.1848,
+      "step": 41
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8064037496605132,
+      "learning_rate": 9.923725999591846e-06,
+      "loss": 0.108,
+      "step": 42
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7308253552869113,
+      "learning_rate": 9.917987226970811e-06,
+      "loss": 0.13,
+      "step": 43
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.641530457475072,
+      "learning_rate": 9.912042105455462e-06,
+      "loss": 0.1276,
+      "step": 44
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8300848785648857,
+      "learning_rate": 9.905890884491196e-06,
+      "loss": 0.125,
+      "step": 45
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6144652207379732,
+      "learning_rate": 9.899533822170922e-06,
+      "loss": 0.117,
+      "step": 46
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8180138707863049,
+      "learning_rate": 9.892971185224244e-06,
+      "loss": 0.1293,
+      "step": 47
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9351068141307641,
+      "learning_rate": 9.886203249006265e-06,
+      "loss": 0.1322,
+      "step": 48
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7711243311834907,
+      "learning_rate": 9.879230297486034e-06,
+      "loss": 0.1266,
+      "step": 49
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2380701876377125,
+      "learning_rate": 9.872052623234632e-06,
+      "loss": 0.1255,
+      "step": 50
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.203808913144882,
+      "learning_rate": 9.864670527412891e-06,
+      "loss": 0.1921,
+      "step": 51
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2896863465751387,
+      "learning_rate": 9.857084319758772e-06,
+      "loss": 0.1581,
+      "step": 52
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7548162165103374,
+      "learning_rate": 9.849294318574353e-06,
+      "loss": 0.1463,
+      "step": 53
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0334367656162968,
+      "learning_rate": 9.841300850712479e-06,
+      "loss": 0.1332,
+      "step": 54
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6592483676317302,
+      "learning_rate": 9.833104251563058e-06,
+      "loss": 0.1155,
+      "step": 55
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7856988079219412,
+      "learning_rate": 9.824704865038967e-06,
+      "loss": 0.1223,
+      "step": 56
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9841576867137853,
+      "learning_rate": 9.816103043561648e-06,
+      "loss": 0.1457,
+      "step": 57
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6806196475936801,
+      "learning_rate": 9.807299148046301e-06,
+      "loss": 0.1189,
+      "step": 58
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8763554375141681,
+      "learning_rate": 9.798293547886748e-06,
+      "loss": 0.0842,
+      "step": 59
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4191117216965232,
+      "learning_rate": 9.789086620939936e-06,
+      "loss": 0.1708,
+      "step": 60
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0979618364115356,
+      "learning_rate": 9.779678753510082e-06,
+      "loss": 0.1723,
+      "step": 61
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7699797592772706,
+      "learning_rate": 9.770070340332457e-06,
+      "loss": 0.1287,
+      "step": 62
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8877764874268894,
+      "learning_rate": 9.76026178455684e-06,
+      "loss": 0.1876,
+      "step": 63
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6994853026271824,
+      "learning_rate": 9.75025349773058e-06,
+      "loss": 0.136,
+      "step": 64
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9216021155237354,
+      "learning_rate": 9.740045899781353e-06,
+      "loss": 0.1292,
+      "step": 65
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7453943118920977,
+      "learning_rate": 9.729639418999524e-06,
+      "loss": 0.155,
+      "step": 66
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.79773477245724,
+      "learning_rate": 9.719034492020183e-06,
+      "loss": 0.1769,
+      "step": 67
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6847427502446188,
+      "learning_rate": 9.708231563804828e-06,
+      "loss": 0.0942,
+      "step": 68
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2119073354718934,
+      "learning_rate": 9.697231087622691e-06,
+      "loss": 0.1372,
+      "step": 69
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5694650059673866,
+      "learning_rate": 9.68603352503172e-06,
+      "loss": 0.0993,
+      "step": 70
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.39819093816684803,
+      "learning_rate": 9.674639345859213e-06,
+      "loss": 0.0916,
+      "step": 71
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7362844200692608,
+      "learning_rate": 9.663049028182112e-06,
+      "loss": 0.1288,
+      "step": 72
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.48385101980537876,
+      "learning_rate": 9.651263058306932e-06,
+      "loss": 0.1037,
+      "step": 73
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7838597210641812,
+      "learning_rate": 9.639281930749363e-06,
+      "loss": 0.1471,
+      "step": 74
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9663173755107086,
+      "learning_rate": 9.627106148213521e-06,
+      "loss": 0.1437,
+      "step": 75
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4796822883173506,
+      "learning_rate": 9.61473622157086e-06,
+      "loss": 0.0926,
+      "step": 76
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.155197024481699,
+      "learning_rate": 9.602172669838721e-06,
+      "loss": 0.2146,
+      "step": 77
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1639055084652334,
+      "learning_rate": 9.589416020158577e-06,
+      "loss": 0.1724,
+      "step": 78
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.547966633302887,
+      "learning_rate": 9.5764668077739e-06,
+      "loss": 0.1084,
+      "step": 79
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8814626046410395,
+      "learning_rate": 9.563325576007702e-06,
+      "loss": 0.1372,
+      "step": 80
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6210765018596596,
+      "learning_rate": 9.549992876239753e-06,
+      "loss": 0.0694,
+      "step": 81
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.110500549491553,
+      "learning_rate": 9.536469267883432e-06,
+      "loss": 0.1839,
+      "step": 82
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2471118804751122,
+      "learning_rate": 9.52275531836226e-06,
+      "loss": 0.1467,
+      "step": 83
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9123055530527371,
+      "learning_rate": 9.508851603086094e-06,
+      "loss": 0.1337,
+      "step": 84
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9805374180275667,
+      "learning_rate": 9.494758705426978e-06,
+      "loss": 0.1071,
+      "step": 85
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5789811708800969,
+      "learning_rate": 9.480477216694674e-06,
+      "loss": 0.1159,
+      "step": 86
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7790018124634032,
+      "learning_rate": 9.466007736111846e-06,
+      "loss": 0.1393,
+      "step": 87
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7230480461407474,
+      "learning_rate": 9.451350870788922e-06,
+      "loss": 0.1321,
+      "step": 88
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.82321440490818,
+      "learning_rate": 9.436507235698613e-06,
+      "loss": 0.1687,
+      "step": 89
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1456530878281221,
+      "learning_rate": 9.421477453650118e-06,
+      "loss": 0.1707,
+      "step": 90
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.49974163825682083,
+      "learning_rate": 9.406262155262995e-06,
+      "loss": 0.0922,
+      "step": 91
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0660136843265906,
+      "learning_rate": 9.390861978940687e-06,
+      "loss": 0.1935,
+      "step": 92
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7010294818727274,
+      "learning_rate": 9.37527757084375e-06,
+      "loss": 0.1286,
+      "step": 93
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8687260383951586,
+      "learning_rate": 9.359509584862735e-06,
+      "loss": 0.1232,
+      "step": 94
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7431869719716694,
+      "learning_rate": 9.343558682590757e-06,
+      "loss": 0.1313,
+      "step": 95
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7125664703001464,
+      "learning_rate": 9.327425533295725e-06,
+      "loss": 0.1075,
+      "step": 96
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1341306934510087,
+      "learning_rate": 9.31111081389227e-06,
+      "loss": 0.1856,
+      "step": 97
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5920305603497104,
+      "learning_rate": 9.29461520891335e-06,
+      "loss": 0.0908,
+      "step": 98
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.115166027446719,
+      "learning_rate": 9.277939410481507e-06,
+      "loss": 0.1228,
+      "step": 99
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9285199941811209,
+      "learning_rate": 9.261084118279846e-06,
+      "loss": 0.1757,
+      "step": 100
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3218391768689233,
+      "learning_rate": 9.244050039522673e-06,
+      "loss": 0.1563,
+      "step": 101
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8936925872658409,
+      "learning_rate": 9.226837888925813e-06,
+      "loss": 0.1381,
+      "step": 102
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.616864481202159,
+      "learning_rate": 9.209448388676636e-06,
+      "loss": 0.0965,
+      "step": 103
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5886766076497468,
+      "learning_rate": 9.191882268403743e-06,
+      "loss": 0.0995,
+      "step": 104
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7418873901004167,
+      "learning_rate": 9.174140265146356e-06,
+      "loss": 0.1268,
+      "step": 105
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6683463268452202,
+      "learning_rate": 9.156223123323405e-06,
+      "loss": 0.0945,
+      "step": 106
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8100037014458216,
+      "learning_rate": 9.13813159470227e-06,
+      "loss": 0.0965,
+      "step": 107
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5813603939289472,
+      "learning_rate": 9.119866438367263e-06,
+      "loss": 0.0625,
+      "step": 108
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6823981550941903,
+      "learning_rate": 9.101428420687759e-06,
+      "loss": 0.1,
+      "step": 109
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0685791474330528,
+      "learning_rate": 9.082818315286054e-06,
+      "loss": 0.1305,
+      "step": 110
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.423832608761244,
+      "learning_rate": 9.0640369030049e-06,
+      "loss": 0.2443,
+      "step": 111
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7077886634021346,
+      "learning_rate": 9.045084971874738e-06,
+      "loss": 0.1361,
+      "step": 112
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8833141546052482,
+      "learning_rate": 9.025963317080641e-06,
+      "loss": 0.1054,
+      "step": 113
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7609054854017556,
+      "learning_rate": 9.006672740928952e-06,
+      "loss": 0.0669,
+      "step": 114
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8351944085373199,
+      "learning_rate": 8.987214052813605e-06,
+      "loss": 0.1025,
+      "step": 115
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5319790825694221,
+      "learning_rate": 8.967588069182184e-06,
+      "loss": 0.1023,
+      "step": 116
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6636541086727228,
+      "learning_rate": 8.947795613501658e-06,
+      "loss": 0.1215,
+      "step": 117
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5052715765394535,
+      "learning_rate": 8.927837516223824e-06,
+      "loss": 0.1127,
+      "step": 118
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8072207817927785,
+      "learning_rate": 8.907714614750473e-06,
+      "loss": 0.1361,
+      "step": 119
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9690886908335553,
+      "learning_rate": 8.887427753398249e-06,
+      "loss": 0.1899,
+      "step": 120
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7903552707748308,
+      "learning_rate": 8.866977783363219e-06,
+      "loss": 0.1321,
+      "step": 121
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0268300546563336,
+      "learning_rate": 8.846365562685178e-06,
+      "loss": 0.1555,
+      "step": 122
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6839614945715745,
+      "learning_rate": 8.825591956211614e-06,
+      "loss": 0.1055,
+      "step": 123
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8393856778492443,
+      "learning_rate": 8.804657835561456e-06,
+      "loss": 0.1667,
+      "step": 124
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7391743261975082,
+      "learning_rate": 8.783564079088478e-06,
+      "loss": 0.1233,
+      "step": 125
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8375549159216054,
+      "learning_rate": 8.762311571844453e-06,
+      "loss": 0.0841,
+      "step": 126
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.578947985021812,
+      "learning_rate": 8.74090120554202e-06,
+      "loss": 0.0695,
+      "step": 127
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5941668469144197,
+      "learning_rate": 8.719333878517274e-06,
+      "loss": 0.181,
+      "step": 128
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5770276591153891,
+      "learning_rate": 8.697610495692055e-06,
+      "loss": 0.1127,
+      "step": 129
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6171453740689381,
+      "learning_rate": 8.675731968536004e-06,
+      "loss": 0.1386,
+      "step": 130
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.138825520481436,
+      "learning_rate": 8.653699215028298e-06,
+      "loss": 0.15,
+      "step": 131
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0865950831257416,
+      "learning_rate": 8.63151315961915e-06,
+      "loss": 0.1578,
+      "step": 132
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9354551880186306,
+      "learning_rate": 8.609174733191012e-06,
+      "loss": 0.1444,
+      "step": 133
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.521471359027822,
+      "learning_rate": 8.586684873019513e-06,
+      "loss": 0.0994,
+      "step": 134
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6629304093571784,
+      "learning_rate": 8.564044522734147e-06,
+      "loss": 0.0978,
+      "step": 135
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6831151476298498,
+      "learning_rate": 8.541254632278667e-06,
+      "loss": 0.0779,
+      "step": 136
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7367583952734872,
+      "learning_rate": 8.518316157871232e-06,
+      "loss": 0.0911,
+      "step": 137
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6590736711583746,
+      "learning_rate": 8.495230061964289e-06,
+      "loss": 0.1356,
+      "step": 138
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.153818794835716,
+      "learning_rate": 8.471997313204183e-06,
+      "loss": 0.146,
+      "step": 139
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6008053964463435,
+      "learning_rate": 8.448618886390523e-06,
+      "loss": 0.0967,
+      "step": 140
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7825428576530492,
+      "learning_rate": 8.425095762435274e-06,
+      "loss": 0.1575,
+      "step": 141
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4077617053377632,
+      "learning_rate": 8.401428928321607e-06,
+      "loss": 0.082,
+      "step": 142
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6497568265720193,
+      "learning_rate": 8.377619377062483e-06,
+      "loss": 0.1463,
+      "step": 143
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6962679958043572,
+      "learning_rate": 8.353668107658984e-06,
+      "loss": 0.1304,
+      "step": 144
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5834749099786907,
+      "learning_rate": 8.329576125058406e-06,
+      "loss": 0.107,
+      "step": 145
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5503783852935425,
+      "learning_rate": 8.305344440112089e-06,
+      "loss": 0.124,
+      "step": 146
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7365506208146866,
+      "learning_rate": 8.280974069532999e-06,
+      "loss": 0.147,
+      "step": 147
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6978628022776826,
+      "learning_rate": 8.256466035853077e-06,
+      "loss": 0.102,
+      "step": 148
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7758804192060227,
+      "learning_rate": 8.231821367380335e-06,
+      "loss": 0.0845,
+      "step": 149
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5477123756526534,
+      "learning_rate": 8.207041098155701e-06,
+      "loss": 0.1259,
+      "step": 150
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6017501035266978,
+      "learning_rate": 8.182126267909642e-06,
+      "loss": 0.0587,
+      "step": 151
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7687312103332277,
+      "learning_rate": 8.157077922018537e-06,
+      "loss": 0.1451,
+      "step": 152
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0955298515155794,
+      "learning_rate": 8.13189711146081e-06,
+      "loss": 0.1656,
+      "step": 153
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5866736794773539,
+      "learning_rate": 8.106584892772844e-06,
+      "loss": 0.0905,
+      "step": 154
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7155438066725858,
+      "learning_rate": 8.081142328004638e-06,
+      "loss": 0.1528,
+      "step": 155
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6565395168851769,
+      "learning_rate": 8.055570484675252e-06,
+      "loss": 0.1251,
+      "step": 156
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2875502388618136,
+      "learning_rate": 8.029870435728018e-06,
+      "loss": 0.0913,
+      "step": 157
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5982546707253796,
+      "learning_rate": 8.004043259485519e-06,
+      "loss": 0.1115,
+      "step": 158
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.46008235561169636,
+      "learning_rate": 7.978090039604342e-06,
+      "loss": 0.087,
+      "step": 159
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.43711300108945894,
+      "learning_rate": 7.952011865029614e-06,
+      "loss": 0.1184,
+      "step": 160
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.918267884953489,
+      "learning_rate": 7.925809829949312e-06,
+      "loss": 0.1293,
+      "step": 161
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7265832047624626,
+      "learning_rate": 7.89948503374835e-06,
+      "loss": 0.1569,
+      "step": 162
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5409486213897078,
+      "learning_rate": 7.873038580962453e-06,
+      "loss": 0.0966,
+      "step": 163
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.41989818162281206,
+      "learning_rate": 7.846471581231814e-06,
+      "loss": 0.0985,
+      "step": 164
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8624237182282338,
+      "learning_rate": 7.819785149254534e-06,
+      "loss": 0.1495,
+      "step": 165
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3427024418979756,
+      "learning_rate": 7.792980404739849e-06,
+      "loss": 0.1465,
+      "step": 166
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6747523799681433,
+      "learning_rate": 7.766058472361154e-06,
+      "loss": 0.1482,
+      "step": 167
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.583861344914313,
+      "learning_rate": 7.739020481708816e-06,
+      "loss": 0.1241,
+      "step": 168
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.39792159088639656,
+      "learning_rate": 7.711867567242769e-06,
+      "loss": 0.078,
+      "step": 169
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.988133766491529,
+      "learning_rate": 7.68460086824492e-06,
+      "loss": 0.144,
+      "step": 170
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5438192691474706,
+      "learning_rate": 7.657221528771352e-06,
+      "loss": 0.1085,
+      "step": 171
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7255348589603168,
+      "learning_rate": 7.629730697604314e-06,
+      "loss": 0.095,
+      "step": 172
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6567578642812082,
+      "learning_rate": 7.602129528204023e-06,
+      "loss": 0.1258,
+      "step": 173
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6610744479736924,
+      "learning_rate": 7.574419178660269e-06,
+      "loss": 0.1004,
+      "step": 174
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.069879470565798,
+      "learning_rate": 7.546600811643816e-06,
+      "loss": 0.1755,
+      "step": 175
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9032056389882546,
+      "learning_rate": 7.5186755943576324e-06,
+      "loss": 0.1846,
+      "step": 176
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8292236391024451,
+      "learning_rate": 7.490644698487909e-06,
+      "loss": 0.1637,
+      "step": 177
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.513027858704895,
+      "learning_rate": 7.462509300154892e-06,
+      "loss": 0.1069,
+      "step": 178
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4333926139924782,
+      "learning_rate": 7.434270579863549e-06,
+      "loss": 0.1035,
+      "step": 179
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0347242731404998,
+      "learning_rate": 7.405929722454026e-06,
+      "loss": 0.145,
+      "step": 180
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5121541327494573,
+      "learning_rate": 7.3774879170519386e-06,
+      "loss": 0.119,
+      "step": 181
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4107383137309751,
+      "learning_rate": 7.348946357018479e-06,
+      "loss": 0.103,
+      "step": 182
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6738373632141648,
+      "learning_rate": 7.320306239900343e-06,
+      "loss": 0.1263,
+      "step": 183
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.47232430178292983,
+      "learning_rate": 7.291568767379484e-06,
+      "loss": 0.0802,
+      "step": 184
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6445442067197152,
+      "learning_rate": 7.262735145222696e-06,
+      "loss": 0.0718,
+      "step": 185
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.925502100538713,
+      "learning_rate": 7.233806583231012e-06,
+      "loss": 0.1209,
+      "step": 186
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.45358994344115217,
+      "learning_rate": 7.204784295188959e-06,
+      "loss": 0.11,
+      "step": 187
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.592074593684059,
+      "learning_rate": 7.1756694988136165e-06,
+      "loss": 0.1587,
+      "step": 188
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1993290449955363,
+      "learning_rate": 7.14646341570353e-06,
+      "loss": 0.1702,
+      "step": 189
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4325803651617202,
+      "learning_rate": 7.117167271287453e-06,
+      "loss": 0.079,
+      "step": 190
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8384022277203674,
+      "learning_rate": 7.0877822947729265e-06,
+      "loss": 0.1362,
+      "step": 191
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7303306509391421,
+      "learning_rate": 7.05830971909472e-06,
+      "loss": 0.1186,
+      "step": 192
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7516368588709544,
+      "learning_rate": 7.028750780863078e-06,
+      "loss": 0.0887,
+      "step": 193
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.938542188821057,
+      "learning_rate": 6.999106720311846e-06,
+      "loss": 0.1444,
+      "step": 194
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8717472174419715,
+      "learning_rate": 6.969378781246436e-06,
+      "loss": 0.1435,
+      "step": 195
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.67495945265085,
+      "learning_rate": 6.939568210991633e-06,
+      "loss": 0.1626,
+      "step": 196
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.342119658716737,
+      "learning_rate": 6.9096762603392595e-06,
+      "loss": 0.1953,
+      "step": 197
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6263126368661642,
+      "learning_rate": 6.8797041834956955e-06,
+      "loss": 0.1336,
+      "step": 198
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5941737014361337,
+      "learning_rate": 6.849653238029261e-06,
+      "loss": 0.153,
+      "step": 199
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7886180708514274,
+      "learning_rate": 6.819524684817439e-06,
+      "loss": 0.1223,
+      "step": 200
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7039612451739138,
+      "learning_rate": 6.78931978799398e-06,
+      "loss": 0.1335,
+      "step": 201
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8084406855671586,
+      "learning_rate": 6.7590398148958625e-06,
+      "loss": 0.1678,
+      "step": 202
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9860973166784854,
+      "learning_rate": 6.728686036010115e-06,
+      "loss": 0.153,
+      "step": 203
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5916914892453895,
+      "learning_rate": 6.698259724920503e-06,
+      "loss": 0.1385,
+      "step": 204
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6092574184981456,
+      "learning_rate": 6.667762158254104e-06,
+      "loss": 0.11,
+      "step": 205
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7140207170863097,
+      "learning_rate": 6.637194615627733e-06,
+      "loss": 0.1186,
+      "step": 206
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5017767013539293,
+      "learning_rate": 6.6065583795942625e-06,
+      "loss": 0.1103,
+      "step": 207
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7894584955176154,
+      "learning_rate": 6.5758547355887944e-06,
+      "loss": 0.1462,
+      "step": 208
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8876268352844191,
+      "learning_rate": 6.545084971874738e-06,
+      "loss": 0.1354,
+      "step": 209
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5673653854310852,
+      "learning_rate": 6.514250379489754e-06,
+      "loss": 0.0658,
+      "step": 210
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.36727540507660567,
+      "learning_rate": 6.483352252191585e-06,
+      "loss": 0.0771,
+      "step": 211
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8054289615117477,
+      "learning_rate": 6.452391886403767e-06,
+      "loss": 0.169,
+      "step": 212
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5441797169347325,
+      "learning_rate": 6.421370581161244e-06,
+      "loss": 0.1048,
+      "step": 213
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.89874238353916,
+      "learning_rate": 6.390289638055851e-06,
+      "loss": 0.1357,
+      "step": 214
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6949933563598905,
+      "learning_rate": 6.3591503611817155e-06,
+      "loss": 0.1054,
+      "step": 215
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3751201335684607,
+      "learning_rate": 6.3279540570805265e-06,
+      "loss": 0.1042,
+      "step": 216
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.47528910333657093,
+      "learning_rate": 6.296702034686726e-06,
+      "loss": 0.0769,
+      "step": 217
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8443699711520065,
+      "learning_rate": 6.265395605272581e-06,
+      "loss": 0.0964,
+      "step": 218
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3965451318034568,
+      "learning_rate": 6.234036082393171e-06,
+      "loss": 0.0664,
+      "step": 219
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8268942968030871,
+      "learning_rate": 6.202624781831269e-06,
+      "loss": 0.119,
+      "step": 220
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.41911373347383746,
+      "learning_rate": 6.171163021542134e-06,
+      "loss": 0.0799,
+      "step": 221
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9614476837400521,
+      "learning_rate": 6.139652121598219e-06,
+      "loss": 0.1709,
+      "step": 222
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9056929475746165,
+      "learning_rate": 6.108093404133772e-06,
+      "loss": 0.1738,
+      "step": 223
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.44336861675183054,
+      "learning_rate": 6.076488193289375e-06,
+      "loss": 0.0897,
+      "step": 224
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.49980277000157225,
+      "learning_rate": 6.044837815156377e-06,
+      "loss": 0.1065,
+      "step": 225
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8501092910803394,
+      "learning_rate": 6.013143597721252e-06,
+      "loss": 0.1185,
+      "step": 226
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0820255027100942,
+      "learning_rate": 5.981406870809889e-06,
+      "loss": 0.1042,
+      "step": 227
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5020359740609432,
+      "learning_rate": 5.949628966031785e-06,
+      "loss": 0.1051,
+      "step": 228
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5128920826197394,
+      "learning_rate": 5.9178112167241805e-06,
+      "loss": 0.0827,
+      "step": 229
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.45253579722167236,
+      "learning_rate": 5.885954957896115e-06,
+      "loss": 0.1069,
+      "step": 230
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.872506236560932,
+      "learning_rate": 5.854061526172402e-06,
+      "loss": 0.16,
+      "step": 231
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.42871376495620556,
+      "learning_rate": 5.822132259737565e-06,
+      "loss": 0.0795,
+      "step": 232
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.524971430598943,
+      "learning_rate": 5.7901684982796716e-06,
+      "loss": 0.0714,
+      "step": 233
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5362145472199313,
+      "learning_rate": 5.75817158293414e-06,
+      "loss": 0.1626,
+      "step": 234
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6267996709551327,
+      "learning_rate": 5.726142856227453e-06,
+      "loss": 0.1052,
+      "step": 235
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.661716060278592,
+      "learning_rate": 5.694083662020835e-06,
+      "loss": 0.1013,
+      "step": 236
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6582684822727328,
+      "learning_rate": 5.661995345453867e-06,
+      "loss": 0.1363,
+      "step": 237
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6930382608343246,
+      "learning_rate": 5.629879252888046e-06,
+      "loss": 0.1155,
+      "step": 238
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9097902075987594,
+      "learning_rate": 5.597736731850295e-06,
+      "loss": 0.1544,
+      "step": 239
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4492863280580798,
+      "learning_rate": 5.5655691309764225e-06,
+      "loss": 0.0981,
+      "step": 240
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.36242889591820937,
+      "learning_rate": 5.533377799954532e-06,
+      "loss": 0.0699,
+      "step": 241
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8398730038170705,
+      "learning_rate": 5.501164089468406e-06,
+      "loss": 0.133,
+      "step": 242
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9695085003069862,
+      "learning_rate": 5.4689293511408155e-06,
+      "loss": 0.1703,
+      "step": 243
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.838488275294773,
+      "learning_rate": 5.43667493747682e-06,
+      "loss": 0.1294,
+      "step": 244
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.45164214007362696,
+      "learning_rate": 5.404402201807022e-06,
+      "loss": 0.0822,
+      "step": 245
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7344032699831387,
+      "learning_rate": 5.372112498230771e-06,
+      "loss": 0.135,
+      "step": 246
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7068287460456814,
+      "learning_rate": 5.339807181559359e-06,
+      "loss": 0.1137,
+      "step": 247
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.43221336760669393,
+      "learning_rate": 5.307487607259175e-06,
+      "loss": 0.0663,
+      "step": 248
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6101399921737543,
+      "learning_rate": 5.275155131394825e-06,
+      "loss": 0.1045,
+      "step": 249
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6482460605334966,
+      "learning_rate": 5.242811110572243e-06,
+      "loss": 0.1245,
+      "step": 250
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5398676773671225,
+      "learning_rate": 5.210456901881761e-06,
+      "loss": 0.111,
+      "step": 251
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7645673370818892,
+      "learning_rate": 5.1780938628411795e-06,
+      "loss": 0.1363,
+      "step": 252
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5169589806284677,
+      "learning_rate": 5.145723351338799e-06,
+      "loss": 0.0993,
+      "step": 253
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7053804089435339,
+      "learning_rate": 5.11334672557645e-06,
+      "loss": 0.1194,
+      "step": 254
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5975589404773662,
+      "learning_rate": 5.080965344012509e-06,
+      "loss": 0.0934,
+      "step": 255
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.41552303091522763,
+      "learning_rate": 5.048580565304887e-06,
+      "loss": 0.0972,
+      "step": 256
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5489918448360888,
+      "learning_rate": 5.016193748254045e-06,
+      "loss": 0.0751,
+      "step": 257
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.22308617581959,
+      "learning_rate": 4.983806251745958e-06,
+      "loss": 0.1535,
+      "step": 258
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.43444777406659896,
+      "learning_rate": 4.951419434695115e-06,
+      "loss": 0.0886,
+      "step": 259
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.660673409335142,
+      "learning_rate": 4.919034655987493e-06,
+      "loss": 0.1149,
+      "step": 260
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3283375001173057,
+      "learning_rate": 4.886653274423551e-06,
+      "loss": 0.093,
+      "step": 261
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.280151105288537,
+      "learning_rate": 4.8542766486612035e-06,
+      "loss": 0.1663,
+      "step": 262
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3123335199204402,
+      "learning_rate": 4.821906137158822e-06,
+      "loss": 0.0724,
+      "step": 263
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.42192359918134575,
+      "learning_rate": 4.7895430981182415e-06,
+      "loss": 0.0708,
+      "step": 264
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5967015598287172,
+      "learning_rate": 4.757188889427761e-06,
+      "loss": 0.1091,
+      "step": 265
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8761618680748017,
+      "learning_rate": 4.724844868605176e-06,
+      "loss": 0.0922,
+      "step": 266
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5960199337090709,
+      "learning_rate": 4.6925123927408265e-06,
+      "loss": 0.119,
+      "step": 267
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8037891258649469,
+      "learning_rate": 4.660192818440642e-06,
+      "loss": 0.1221,
+      "step": 268
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5556173800063137,
+      "learning_rate": 4.627887501769231e-06,
+      "loss": 0.1177,
+      "step": 269
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9049550817206606,
+      "learning_rate": 4.59559779819298e-06,
+      "loss": 0.1591,
+      "step": 270
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1702651051641604,
+      "learning_rate": 4.5633250625231806e-06,
+      "loss": 0.182,
+      "step": 271
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8483471329393831,
+      "learning_rate": 4.531070648859186e-06,
+      "loss": 0.1109,
+      "step": 272
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.32161415640173774,
+      "learning_rate": 4.498835910531595e-06,
+      "loss": 0.0721,
+      "step": 273
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.43434061341999136,
+      "learning_rate": 4.4666222000454685e-06,
+      "loss": 0.1053,
+      "step": 274
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8637865606289452,
+      "learning_rate": 4.434430869023579e-06,
+      "loss": 0.0861,
+      "step": 275
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.628646931876877,
+      "learning_rate": 4.402263268149707e-06,
+      "loss": 0.1135,
+      "step": 276
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.34970147815123154,
+      "learning_rate": 4.370120747111956e-06,
+      "loss": 0.0723,
+      "step": 277
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4765585692957405,
+      "learning_rate": 4.338004654546136e-06,
+      "loss": 0.1145,
+      "step": 278
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7803990928349243,
+      "learning_rate": 4.3059163379791676e-06,
+      "loss": 0.1156,
+      "step": 279
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7671333748360498,
+      "learning_rate": 4.27385714377255e-06,
+      "loss": 0.1053,
+      "step": 280
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7121038442085383,
+      "learning_rate": 4.24182841706586e-06,
+      "loss": 0.1289,
+      "step": 281
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8800209303666127,
+      "learning_rate": 4.209831501720328e-06,
+      "loss": 0.1778,
+      "step": 282
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.665862686003406,
+      "learning_rate": 4.177867740262437e-06,
+      "loss": 0.1501,
+      "step": 283
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.121777185384995,
+      "learning_rate": 4.145938473827598e-06,
+      "loss": 0.1149,
+      "step": 284
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.40686853519037686,
+      "learning_rate": 4.1140450421038865e-06,
+      "loss": 0.085,
+      "step": 285
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4216094226014865,
+      "learning_rate": 4.08218878327582e-06,
+      "loss": 0.1087,
+      "step": 286
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8226728277647426,
+      "learning_rate": 4.050371033968216e-06,
+      "loss": 0.17,
+      "step": 287
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.48187584984948256,
+      "learning_rate": 4.018593129190113e-06,
+      "loss": 0.0978,
+      "step": 288
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4526743619684988,
+      "learning_rate": 3.98685640227875e-06,
+      "loss": 0.0858,
+      "step": 289
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2046001007142648,
+      "learning_rate": 3.955162184843625e-06,
+      "loss": 0.1467,
+      "step": 290
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5177528518475938,
+      "learning_rate": 3.9235118067106255e-06,
+      "loss": 0.0904,
+      "step": 291
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5460427300809364,
+      "learning_rate": 3.89190659586623e-06,
+      "loss": 0.0746,
+      "step": 292
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5241932527164748,
+      "learning_rate": 3.8603478784017845e-06,
+      "loss": 0.1297,
+      "step": 293
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.39990316884451654,
+      "learning_rate": 3.828836978457868e-06,
+      "loss": 0.0857,
+      "step": 294
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5332681583136303,
+      "learning_rate": 3.7973752181687336e-06,
+      "loss": 0.1067,
+      "step": 295
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7158724785835472,
+      "learning_rate": 3.7659639176068287e-06,
+      "loss": 0.1198,
+      "step": 296
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7926331004862835,
+      "learning_rate": 3.734604394727419e-06,
+      "loss": 0.1302,
+      "step": 297
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4726616087570671,
+      "learning_rate": 3.703297965313275e-06,
+      "loss": 0.0943,
+      "step": 298
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3365001206743048,
+      "learning_rate": 3.6720459429194743e-06,
+      "loss": 0.0761,
+      "step": 299
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5041980032450573,
+      "learning_rate": 3.6408496388182857e-06,
+      "loss": 0.1348,
+      "step": 300
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3980953006694637,
+      "learning_rate": 3.6097103619441505e-06,
+      "loss": 0.0745,
+      "step": 301
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.38002146091571526,
+      "learning_rate": 3.578629418838757e-06,
+      "loss": 0.0705,
+      "step": 302
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6150028421766434,
+      "learning_rate": 3.5476081135962335e-06,
+      "loss": 0.1014,
+      "step": 303
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.47726586681317523,
+      "learning_rate": 3.516647747808417e-06,
+      "loss": 0.0873,
+      "step": 304
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7303160554132148,
+      "learning_rate": 3.4857496205102475e-06,
+      "loss": 0.1218,
+      "step": 305
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5063263339961966,
+      "learning_rate": 3.4549150281252635e-06,
+      "loss": 0.1003,
+      "step": 306
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5173752740699328,
+      "learning_rate": 3.4241452644112085e-06,
+      "loss": 0.1441,
+      "step": 307
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.578168860033146,
+      "learning_rate": 3.3934416204057396e-06,
+      "loss": 0.1384,
+      "step": 308
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.256424858150544,
+      "learning_rate": 3.3628053843722674e-06,
+      "loss": 0.0444,
+      "step": 309
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.4025835046318359,
+      "learning_rate": 3.3322378417458985e-06,
+      "loss": 0.0658,
+      "step": 310
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6353664899766872,
+      "learning_rate": 3.3017402750794976e-06,
+      "loss": 0.1037,
+      "step": 311
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.065012599456067,
+      "learning_rate": 3.271313963989886e-06,
+      "loss": 0.1453,
+      "step": 312
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8933738678604561,
+      "learning_rate": 3.240960185104137e-06,
+      "loss": 0.1155,
+      "step": 313
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6295023999749918,
+      "learning_rate": 3.2106802120060197e-06,
+      "loss": 0.0817,
+      "step": 314
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6825569076251282,
+      "learning_rate": 3.180475315182563e-06,
+      "loss": 0.1074,
+      "step": 315
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.003002066232109,
+      "learning_rate": 3.1503467619707407e-06,
+      "loss": 0.1333,
+      "step": 316
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5539936219856008,
+      "learning_rate": 3.1202958165043053e-06,
+      "loss": 0.0892,
+      "step": 317
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5499033259681437,
+      "learning_rate": 3.090323739660742e-06,
+      "loss": 0.1171,
+      "step": 318
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7295234710326629,
+      "learning_rate": 3.060431789008368e-06,
+      "loss": 0.1339,
+      "step": 319
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7039705596052677,
+      "learning_rate": 3.0306212187535653e-06,
+      "loss": 0.1735,
+      "step": 320
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7237951210549803,
+      "learning_rate": 3.000893279688155e-06,
+      "loss": 0.1272,
+      "step": 321
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5977750840343024,
+      "learning_rate": 2.9712492191369245e-06,
+      "loss": 0.1161,
+      "step": 322
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9137575856242299,
+      "learning_rate": 2.9416902809052817e-06,
+      "loss": 0.1631,
+      "step": 323
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5604895386564465,
+      "learning_rate": 2.912217705227075e-06,
+      "loss": 0.0935,
+      "step": 324
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.312594001816466,
+      "learning_rate": 2.882832728712551e-06,
+      "loss": 0.2357,
+      "step": 325
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8125476078354773,
+      "learning_rate": 2.8535365842964713e-06,
+      "loss": 0.0856,
+      "step": 326
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6050490227647147,
+      "learning_rate": 2.8243305011863843e-06,
+      "loss": 0.0858,
+      "step": 327
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5686503856565632,
+      "learning_rate": 2.7952157048110406e-06,
+      "loss": 0.0786,
+      "step": 328
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3341595489320209,
+      "learning_rate": 2.7661934167689887e-06,
+      "loss": 0.0765,
+      "step": 329
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.48523257788152796,
+      "learning_rate": 2.7372648547773063e-06,
+      "loss": 0.0818,
+      "step": 330
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4353651802388437,
+      "learning_rate": 2.7084312326205164e-06,
+      "loss": 0.1027,
+      "step": 331
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5804653250200166,
+      "learning_rate": 2.6796937600996587e-06,
+      "loss": 0.1287,
+      "step": 332
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7258700154480898,
+      "learning_rate": 2.6510536429815224e-06,
+      "loss": 0.1477,
+      "step": 333
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5569984089795129,
+      "learning_rate": 2.622512082948063e-06,
+      "loss": 0.1006,
+      "step": 334
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5618985392420127,
+      "learning_rate": 2.594070277545975e-06,
+      "loss": 0.1022,
+      "step": 335
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.4057866106940638,
+      "learning_rate": 2.5657294201364526e-06,
+      "loss": 0.1026,
+      "step": 336
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3949073034443962,
+      "learning_rate": 2.5374906998451094e-06,
+      "loss": 0.0701,
+      "step": 337
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5821479003733443,
+      "learning_rate": 2.5093553015120937e-06,
+      "loss": 0.1141,
+      "step": 338
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.37657783683758206,
+      "learning_rate": 2.4813244056423692e-06,
+      "loss": 0.0807,
+      "step": 339
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8996794097563738,
+      "learning_rate": 2.4533991883561868e-06,
+      "loss": 0.1392,
+      "step": 340
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7862777875907287,
+      "learning_rate": 2.425580821339733e-06,
+      "loss": 0.1427,
+      "step": 341
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4158720785594425,
+      "learning_rate": 2.3978704717959777e-06,
+      "loss": 0.057,
+      "step": 342
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.445461113749062,
+      "learning_rate": 2.3702693023956853e-06,
+      "loss": 0.0982,
+      "step": 343
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3380111458364013,
+      "learning_rate": 2.342778471228648e-06,
+      "loss": 0.074,
+      "step": 344
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.442065574751732,
+      "learning_rate": 2.315399131755081e-06,
+      "loss": 0.0836,
+      "step": 345
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3959460207104249,
+      "learning_rate": 2.2881324327572336e-06,
+      "loss": 0.0796,
+      "step": 346
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5811148185563871,
+      "learning_rate": 2.260979518291186e-06,
+      "loss": 0.1085,
+      "step": 347
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5786190629593744,
+      "learning_rate": 2.233941527638848e-06,
+      "loss": 0.1267,
+      "step": 348
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0171032000489182,
+      "learning_rate": 2.207019595260154e-06,
+      "loss": 0.1309,
+      "step": 349
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.48160306249010754,
+      "learning_rate": 2.1802148507454675e-06,
+      "loss": 0.0878,
+      "step": 350
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3280483511586015,
+      "learning_rate": 2.1535284187681866e-06,
+      "loss": 0.0693,
+      "step": 351
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.34096842916481573,
+      "learning_rate": 2.1269614190375477e-06,
+      "loss": 0.0499,
+      "step": 352
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.47978813814877763,
+      "learning_rate": 2.1005149662516517e-06,
+      "loss": 0.0625,
+      "step": 353
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7425879124863387,
+      "learning_rate": 2.07419017005069e-06,
+      "loss": 0.1186,
+      "step": 354
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6228364584187716,
+      "learning_rate": 2.0479881349703885e-06,
+      "loss": 0.1052,
+      "step": 355
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4806977117902373,
+      "learning_rate": 2.021909960395661e-06,
+      "loss": 0.1054,
+      "step": 356
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.609915927744076,
+      "learning_rate": 1.9959567405144825e-06,
+      "loss": 0.1285,
+      "step": 357
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3594194411476996,
+      "learning_rate": 1.9701295642719836e-06,
+      "loss": 0.0644,
+      "step": 358
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6102620773665237,
+      "learning_rate": 1.944429515324749e-06,
+      "loss": 0.0967,
+      "step": 359
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.49370723009062395,
+      "learning_rate": 1.9188576719953635e-06,
+      "loss": 0.1382,
+      "step": 360
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9307415830568554,
+      "learning_rate": 1.8934151072271573e-06,
+      "loss": 0.1114,
+      "step": 361
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5873584645988698,
+      "learning_rate": 1.8681028885391905e-06,
+      "loss": 0.1111,
+      "step": 362
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5086410983198352,
+      "learning_rate": 1.8429220779814654e-06,
+      "loss": 0.0813,
+      "step": 363
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7494157766725494,
+      "learning_rate": 1.81787373209036e-06,
+      "loss": 0.143,
+      "step": 364
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.59457257382502,
+      "learning_rate": 1.7929589018443016e-06,
+      "loss": 0.0995,
+      "step": 365
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7171481785607706,
+      "learning_rate": 1.7681786326196665e-06,
+      "loss": 0.135,
+      "step": 366
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6544054085500117,
+      "learning_rate": 1.743533964146924e-06,
+      "loss": 0.1084,
+      "step": 367
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9507053821842658,
+      "learning_rate": 1.7190259304670038e-06,
+      "loss": 0.1547,
+      "step": 368
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5798816104551191,
+      "learning_rate": 1.6946555598879138e-06,
+      "loss": 0.124,
+      "step": 369
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7567771885229603,
+      "learning_rate": 1.6704238749415958e-06,
+      "loss": 0.1481,
+      "step": 370
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3037738050212244,
+      "learning_rate": 1.6463318923410183e-06,
+      "loss": 0.1183,
+      "step": 371
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6640220162663745,
+      "learning_rate": 1.6223806229375182e-06,
+      "loss": 0.122,
+      "step": 372
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.477033527760463,
+      "learning_rate": 1.5985710716783936e-06,
+      "loss": 0.0744,
+      "step": 373
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5646325992453681,
+      "learning_rate": 1.5749042375647261e-06,
+      "loss": 0.1229,
+      "step": 374
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.497343124798125,
+      "learning_rate": 1.5513811136094786e-06,
+      "loss": 0.0914,
+      "step": 375
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4787217350165173,
+      "learning_rate": 1.5280026867958186e-06,
+      "loss": 0.1012,
+      "step": 376
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.7864702413137152,
+      "learning_rate": 1.5047699380357134e-06,
+      "loss": 0.1432,
+      "step": 377
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3630059899374488,
+      "learning_rate": 1.4816838421287693e-06,
+      "loss": 0.0499,
+      "step": 378
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9135185471337351,
+      "learning_rate": 1.4587453677213348e-06,
+      "loss": 0.1565,
+      "step": 379
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.4025025771883063,
+      "learning_rate": 1.4359554772658551e-06,
+      "loss": 0.0813,
+      "step": 380
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8944633586248318,
+      "learning_rate": 1.4133151269804873e-06,
+      "loss": 0.1194,
+      "step": 381
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.4963838870293355,
+      "learning_rate": 1.39082526680899e-06,
+      "loss": 0.1533,
+      "step": 382
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6233327379521121,
+      "learning_rate": 1.368486840380851e-06,
+      "loss": 0.1167,
+      "step": 383
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3567685987225913,
+      "learning_rate": 1.3463007849717035e-06,
+      "loss": 0.0699,
+      "step": 384
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.4907523234293947,
+      "learning_rate": 1.3242680314639995e-06,
+      "loss": 0.0638,
+      "step": 385
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8918356254984783,
+      "learning_rate": 1.3023895043079476e-06,
+      "loss": 0.1236,
+      "step": 386
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.4932187447287198,
+      "learning_rate": 1.2806661214827286e-06,
+      "loss": 0.0744,
+      "step": 387
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7425082873706774,
+      "learning_rate": 1.2590987944579808e-06,
+      "loss": 0.1324,
+      "step": 388
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.773838140527864,
+      "learning_rate": 1.2376884281555485e-06,
+      "loss": 0.133,
+      "step": 389
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5173817236431654,
+      "learning_rate": 1.2164359209115235e-06,
+      "loss": 0.0962,
+      "step": 390
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5641305515634265,
+      "learning_rate": 1.1953421644385444e-06,
+      "loss": 0.1363,
+      "step": 391
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9415313279746156,
+      "learning_rate": 1.1744080437883859e-06,
+      "loss": 0.1398,
+      "step": 392
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.506483121069416,
+      "learning_rate": 1.1536344373148245e-06,
+      "loss": 0.0848,
+      "step": 393
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9255572933453069,
+      "learning_rate": 1.133022216636781e-06,
+      "loss": 0.1123,
+      "step": 394
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6029862576440691,
+      "learning_rate": 1.1125722466017547e-06,
+      "loss": 0.1182,
+      "step": 395
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6376670932752512,
+      "learning_rate": 1.092285385249528e-06,
+      "loss": 0.1217,
+      "step": 396
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6945037548136007,
+      "learning_rate": 1.0721624837761768e-06,
+      "loss": 0.1146,
+      "step": 397
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3933583908849212,
+      "learning_rate": 1.0522043864983428e-06,
+      "loss": 0.0851,
+      "step": 398
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8528313429730696,
+      "learning_rate": 1.0324119308178166e-06,
+      "loss": 0.1346,
+      "step": 399
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7032753581261906,
+      "learning_rate": 1.012785947186397e-06,
+      "loss": 0.1213,
+      "step": 400
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.700851792321212,
+      "learning_rate": 9.933272590710508e-07,
+      "loss": 0.0983,
+      "step": 401
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6810211763792263,
+      "learning_rate": 9.740366829193587e-07,
+      "loss": 0.1227,
+      "step": 402
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7204524580947844,
+      "learning_rate": 9.549150281252633e-07,
+      "loss": 0.0993,
+      "step": 403
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0497265670010762,
+      "learning_rate": 9.359630969951012e-07,
+      "loss": 0.0656,
+      "step": 404
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.4493415113338161,
+      "learning_rate": 9.171816847139447e-07,
+      "loss": 0.1023,
+      "step": 405
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6197686610517851,
+      "learning_rate": 8.985715793122407e-07,
+      "loss": 0.0944,
+      "step": 406
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6454561325542718,
+      "learning_rate": 8.801335616327378e-07,
+      "loss": 0.1135,
+      "step": 407
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8188263044708098,
+      "learning_rate": 8.618684052977305e-07,
+      "loss": 0.1069,
+      "step": 408
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3682638853893349,
+      "learning_rate": 8.437768766765975e-07,
+      "loss": 0.091,
+      "step": 409
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6929729652828701,
+      "learning_rate": 8.258597348536452e-07,
+      "loss": 0.1138,
+      "step": 410
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7147551524391744,
+      "learning_rate": 8.081177315962601e-07,
+      "loss": 0.1293,
+      "step": 411
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7024450782828899,
+      "learning_rate": 7.905516113233652e-07,
+      "loss": 0.1297,
+      "step": 412
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.30317714532691387,
+      "learning_rate": 7.731621110741871e-07,
+      "loss": 0.075,
+      "step": 413
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3697260307910134,
+      "learning_rate": 7.55949960477328e-07,
+      "loss": 0.096,
+      "step": 414
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.4428959709714789,
+      "learning_rate": 7.389158817201541e-07,
+      "loss": 0.0748,
+      "step": 415
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8004523089375409,
+      "learning_rate": 7.220605895184946e-07,
+      "loss": 0.0782,
+      "step": 416
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1444774950837089,
+      "learning_rate": 7.053847910866513e-07,
+      "loss": 0.1297,
+      "step": 417
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7898964558677061,
+      "learning_rate": 6.888891861077301e-07,
+      "loss": 0.095,
+      "step": 418
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.033195499917111,
+      "learning_rate": 6.725744667042778e-07,
+      "loss": 0.1317,
+      "step": 419
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.47792986539711907,
+      "learning_rate": 6.564413174092443e-07,
+      "loss": 0.1169,
+      "step": 420
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6550032710177911,
+      "learning_rate": 6.404904151372649e-07,
+      "loss": 0.0896,
+      "step": 421
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8084734832162704,
+      "learning_rate": 6.24722429156251e-07,
+      "loss": 0.0998,
+      "step": 422
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5357925216101086,
+      "learning_rate": 6.091380210593145e-07,
+      "loss": 0.0814,
+      "step": 423
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.4881261803935285,
+      "learning_rate": 5.937378447370068e-07,
+      "loss": 0.0929,
+      "step": 424
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6924211795087191,
+      "learning_rate": 5.785225463498828e-07,
+      "loss": 0.0986,
+      "step": 425
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7707349079438567,
+      "learning_rate": 5.634927643013899e-07,
+      "loss": 0.1393,
+      "step": 426
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.5765244494687627,
+      "learning_rate": 5.486491292110796e-07,
+      "loss": 0.097,
+      "step": 427
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.4969093516786432,
+      "learning_rate": 5.339922638881545e-07,
+      "loss": 0.0848,
+      "step": 428
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9387402710009252,
+      "learning_rate": 5.195227833053273e-07,
+      "loss": 0.126,
+      "step": 429
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8802367999231513,
+      "learning_rate": 5.05241294573024e-07,
+      "loss": 0.1259,
+      "step": 430
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.46639751531285945,
+      "learning_rate": 4.911483969139086e-07,
+      "loss": 0.1235,
+      "step": 431
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.872214566697276,
+      "learning_rate": 4.772446816377408e-07,
+      "loss": 0.1188,
+      "step": 432
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5081856819309839,
+      "learning_rate": 4.6353073211656886e-07,
+      "loss": 0.0933,
+      "step": 433
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6697265799715091,
+      "learning_rate": 4.5000712376024826e-07,
+      "loss": 0.1028,
+      "step": 434
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5417303256208303,
+      "learning_rate": 4.3667442399229985e-07,
+      "loss": 0.0853,
+      "step": 435
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4919464736847065,
+      "learning_rate": 4.2353319222610265e-07,
+      "loss": 0.1185,
+      "step": 436
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7312868401103222,
+      "learning_rate": 4.1058397984142405e-07,
+      "loss": 0.1091,
+      "step": 437
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8929518208862638,
+      "learning_rate": 3.9782733016128006e-07,
+      "loss": 0.1014,
+      "step": 438
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5064559227512172,
+      "learning_rate": 3.852637784291424e-07,
+      "loss": 0.0812,
+      "step": 439
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7003756083181368,
+      "learning_rate": 3.728938517864794e-07,
+      "loss": 0.0914,
+      "step": 440
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9754357254280254,
+      "learning_rate": 3.60718069250639e-07,
+      "loss": 0.0963,
+      "step": 441
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4166659852770702,
+      "learning_rate": 3.4873694169306915e-07,
+      "loss": 0.0752,
+      "step": 442
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5994165255977032,
+      "learning_rate": 3.369509718178887e-07,
+      "loss": 0.0914,
+      "step": 443
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5532696609253416,
+      "learning_rate": 3.2536065414078724e-07,
+      "loss": 0.1189,
+      "step": 444
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.4898561361634835,
+      "learning_rate": 3.1396647496828245e-07,
+      "loss": 0.1086,
+      "step": 445
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5204425966121,
+      "learning_rate": 3.0276891237731085e-07,
+      "loss": 0.091,
+      "step": 446
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6907546342225368,
+      "learning_rate": 2.917684361951728e-07,
+      "loss": 0.1217,
+      "step": 447
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4176504020041913,
+      "learning_rate": 2.809655079798179e-07,
+      "loss": 0.0508,
+      "step": 448
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.43581953108596455,
+      "learning_rate": 2.7036058100047723e-07,
+      "loss": 0.0787,
+      "step": 449
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8249172088818233,
+      "learning_rate": 2.599541002186479e-07,
+      "loss": 0.1025,
+      "step": 450
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7051733307894276,
+      "learning_rate": 2.497465022694207e-07,
+      "loss": 0.1129,
+      "step": 451
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3685818395664113,
+      "learning_rate": 2.397382154431621e-07,
+      "loss": 0.0623,
+      "step": 452
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7701828038636435,
+      "learning_rate": 2.2992965966754378e-07,
+      "loss": 0.1578,
+      "step": 453
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.47951540454898006,
+      "learning_rate": 2.2032124648992015e-07,
+      "loss": 0.1065,
+      "step": 454
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.967219792124932,
+      "learning_rate": 2.109133790600648e-07,
+      "loss": 0.1079,
+      "step": 455
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7740093949626732,
+      "learning_rate": 2.0170645211325335e-07,
+      "loss": 0.1417,
+      "step": 456
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5131423955626542,
+      "learning_rate": 1.9270085195370048e-07,
+      "loss": 0.1095,
+      "step": 457
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4475424388277173,
+      "learning_rate": 1.838969564383525e-07,
+      "loss": 0.1004,
+      "step": 458
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7253496435886067,
+      "learning_rate": 1.7529513496103322e-07,
+      "loss": 0.1099,
+      "step": 459
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6110013706554089,
+      "learning_rate": 1.6689574843694433e-07,
+      "loss": 0.085,
+      "step": 460
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5522128545402384,
+      "learning_rate": 1.5869914928752117e-07,
+      "loss": 0.1208,
+      "step": 461
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7719280457464354,
+      "learning_rate": 1.5070568142564912e-07,
+      "loss": 0.1264,
+      "step": 462
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1812476336742652,
+      "learning_rate": 1.4291568024122848e-07,
+      "loss": 0.186,
+      "step": 463
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.43320873919423475,
+      "learning_rate": 1.3532947258710905e-07,
+      "loss": 0.0833,
+      "step": 464
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8054906254618007,
+      "learning_rate": 1.2794737676536993e-07,
+      "loss": 0.146,
+      "step": 465
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.50148856743542,
+      "learning_rate": 1.2076970251396593e-07,
+      "loss": 0.0987,
+      "step": 466
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.47531633403256274,
+      "learning_rate": 1.1379675099373489e-07,
+      "loss": 0.0853,
+      "step": 467
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9020478089825394,
+      "learning_rate": 1.0702881477575589e-07,
+      "loss": 0.0893,
+      "step": 468
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3989134945661557,
+      "learning_rate": 1.004661778290783e-07,
+      "loss": 0.0847,
+      "step": 469
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5858494168173927,
+      "learning_rate": 9.410911550880474e-08,
+      "loss": 0.0947,
+      "step": 470
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0751517292115917,
+      "learning_rate": 8.795789454453862e-08,
+      "loss": 0.1158,
+      "step": 471
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7065663950146411,
+      "learning_rate": 8.201277302919086e-08,
+      "loss": 0.0955,
+      "step": 472
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5260547199750213,
+      "learning_rate": 7.627400040815414e-08,
+      "loss": 0.0816,
+      "step": 473
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.1865371271179677,
+      "learning_rate": 7.074181746883402e-08,
+      "loss": 0.0977,
+      "step": 474
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.728326472998703,
+      "learning_rate": 6.54164563305465e-08,
+      "loss": 0.1569,
+      "step": 475
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5759374898440636,
+      "learning_rate": 6.029814043478022e-08,
+      "loss": 0.1095,
+      "step": 476
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5159800707502684,
+      "learning_rate": 5.538708453581787e-08,
+      "loss": 0.1186,
+      "step": 477
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5912916700717128,
+      "learning_rate": 5.068349469173006e-08,
+      "loss": 0.1295,
+      "step": 478
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6628394938861257,
+      "learning_rate": 4.618756825572612e-08,
+      "loss": 0.138,
+      "step": 479
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6669753700150592,
+      "learning_rate": 4.189949386787462e-08,
+      "loss": 0.1209,
+      "step": 480
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7521903097769528,
+      "learning_rate": 3.781945144718912e-08,
+      "loss": 0.1393,
+      "step": 481
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8721516844944895,
+      "learning_rate": 3.394761218407705e-08,
+      "loss": 0.1097,
+      "step": 482
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5187003748899651,
+      "learning_rate": 3.0284138533160924e-08,
+      "loss": 0.0901,
+      "step": 483
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6504180266917741,
+      "learning_rate": 2.6829184206457194e-08,
+      "loss": 0.1314,
+      "step": 484
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5560843324253453,
+      "learning_rate": 2.358289416693027e-08,
+      "loss": 0.0801,
+      "step": 485
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5857924563014624,
+      "learning_rate": 2.0545404622407396e-08,
+      "loss": 0.129,
+      "step": 486
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7418705435308932,
+      "learning_rate": 1.7716843019867646e-08,
+      "loss": 0.0845,
+      "step": 487
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6368611012787012,
+      "learning_rate": 1.509732804009012e-08,
+      "loss": 0.1266,
+      "step": 488
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7396925851613066,
+      "learning_rate": 1.268696959267679e-08,
+      "loss": 0.1313,
+      "step": 489
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5753378740087396,
+      "learning_rate": 1.0485868811441757e-08,
+      "loss": 0.0977,
+      "step": 490
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.511791927295535,
+      "learning_rate": 8.494118050164646e-09,
+      "loss": 0.118,
+      "step": 491
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.4569556080132059,
+      "learning_rate": 6.711800878718144e-09,
+      "loss": 0.0674,
+      "step": 492
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.57891326152388,
+      "learning_rate": 5.138992079561367e-09,
+      "loss": 0.1116,
+      "step": 493
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8006661841027647,
+      "learning_rate": 3.775757644601808e-09,
+      "loss": 0.1464,
+      "step": 494
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5278677382991025,
+      "learning_rate": 2.6221547724253337e-09,
+      "loss": 0.0883,
+      "step": 495
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.43730996329037447,
+      "learning_rate": 1.6782318658992159e-09,
+      "loss": 0.0847,
+      "step": 496
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.42980308851818955,
+      "learning_rate": 9.440285301370865e-10,
+      "loss": 0.0876,
+      "step": 497
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5362614035022275,
+      "learning_rate": 4.1957557084082447e-10,
+      "loss": 0.0951,
+      "step": 498
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.420427750356471,
+      "learning_rate": 1.0489499300603279e-10,
+      "loss": 0.0816,
+      "step": 499
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.724803929021113,
+      "learning_rate": 0.0,
+      "loss": 0.1418,
+      "step": 500
+    },
+    {
+      "epoch": 1.0,
+      "step": 500,
+      "total_flos": 19125432246272.0,
+      "train_loss": 0.13232657791674138,
+      "train_runtime": 1203.9784,
+      "train_samples_per_second": 0.831,
+      "train_steps_per_second": 0.415
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 500,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 50000,
+  "total_flos": 19125432246272.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}
diff --git a/llavaqwen2-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split01_all_mm_tune_olora256_512_llm/README.md b/llavaqwen2-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split01_all_mm_tune_olora256_512_llm/README.md
new file mode 100644
index 0000000000000000000000000000000000000000..d4576fe074287232d3836bf69c21d3f2593290d9
--- /dev/null
+++ b/llavaqwen2-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split01_all_mm_tune_olora256_512_llm/README.md
@@ -0,0 +1,9 @@
+---
+library_name: peft
+---
+## Training procedure
+
+### Framework versions
+
+
+- PEFT 0.4.0
diff --git a/llavaqwen2-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split01_all_mm_tune_olora256_512_llm/adapter_config.json b/llavaqwen2-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split01_all_mm_tune_olora256_512_llm/adapter_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..0ff40233217d5959c9549b350d75ed1d93daffb1
--- /dev/null
+++ b/llavaqwen2-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split01_all_mm_tune_olora256_512_llm/adapter_config.json
@@ -0,0 +1,26 @@
+{
+  "auto_mapping": null,
+  "base_model_name_or_path": "lmms-lab/LLaVA-Video-7B-Qwen2",
+  "bias": "none",
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": "olora",
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "lora_alpha": 512,
+  "lora_dropout": 0.05,
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "r": 256,
+  "revision": null,
+  "target_modules": [
+    "up_proj",
+    "o_proj",
+    "down_proj",
+    "gate_proj",
+    "v_proj",
+    "q_proj",
+    "k_proj"
+  ],
+  "task_type": "CAUSAL_LM"
+}
\ No newline at end of file
diff --git a/llavaqwen2-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split01_all_mm_tune_olora256_512_llm/adapter_model.bin b/llavaqwen2-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split01_all_mm_tune_olora256_512_llm/adapter_model.bin
new file mode 100644
index 0000000000000000000000000000000000000000..3c06f029efd6c3c56a60377177ad0ac54de45e91
--- /dev/null
+++ b/llavaqwen2-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split01_all_mm_tune_olora256_512_llm/adapter_model.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c710d9a1b1f16680adbbb33421b57d61d4baefbf94c194ec80720c97d9bf489b
+size 1384057050
diff --git a/llavaqwen2-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split01_all_mm_tune_olora256_512_llm/config.json b/llavaqwen2-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split01_all_mm_tune_olora256_512_llm/config.json
new file mode 100644
index 0000000000000000000000000000000000000000..1a67f02d0063c3de7740207b9ab2a3eb7be1cbe3
--- /dev/null
+++ b/llavaqwen2-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split01_all_mm_tune_olora256_512_llm/config.json
@@ -0,0 +1,221 @@
+{
+  "_name_or_path": "lmms-lab/LLaVA-Video-7B-Qwen2",
+  "add_faster_video": false,
+  "add_time_instruction": true,
+  "architectures": [
+    "LlavaQwenForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 151643,
+  "eos_token_id": 151645,
+  "faster_token_stride": 10,
+  "force_sample": true,
+  "hidden_act": "silu",
+  "hidden_size": 3584,
+  "ignore_index": -100,
+  "image_aspect_ratio": "anyres_max_9",
+  "image_crop_resolution": null,
+  "image_grid_pinpoints": [
+    [
+      384,
+      384
+    ],
+    [
+      384,
+      768
+    ],
+    [
+      384,
+      1152
+    ],
+    [
+      384,
+      1536
+    ],
+    [
+      384,
+      1920
+    ],
+    [
+      384,
+      2304
+    ],
+    [
+      768,
+      384
+    ],
+    [
+      768,
+      768
+    ],
+    [
+      768,
+      1152
+    ],
+    [
+      768,
+      1536
+    ],
+    [
+      768,
+      1920
+    ],
+    [
+      768,
+      2304
+    ],
+    [
+      1152,
+      384
+    ],
+    [
+      1152,
+      768
+    ],
+    [
+      1152,
+      1152
+    ],
+    [
+      1152,
+      1536
+    ],
+    [
+      1152,
+      1920
+    ],
+    [
+      1152,
+      2304
+    ],
+    [
+      1536,
+      384
+    ],
+    [
+      1536,
+      768
+    ],
+    [
+      1536,
+      1152
+    ],
+    [
+      1536,
+      1536
+    ],
+    [
+      1536,
+      1920
+    ],
+    [
+      1536,
+      2304
+    ],
+    [
+      1920,
+      384
+    ],
+    [
+      1920,
+      768
+    ],
+    [
+      1920,
+      1152
+    ],
+    [
+      1920,
+      1536
+    ],
+    [
+      1920,
+      1920
+    ],
+    [
+      1920,
+      2304
+    ],
+    [
+      2304,
+      384
+    ],
+    [
+      2304,
+      768
+    ],
+    [
+      2304,
+      1152
+    ],
+    [
+      2304,
+      1536
+    ],
+    [
+      2304,
+      1920
+    ],
+    [
+      2304,
+      2304
+    ]
+  ],
+  "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": "grid",
+  "mm_patch_merge_type": "spatial_unpad",
+  "mm_projector_lr": 2e-05,
+  "mm_projector_type": "mlp2x_gelu",
+  "mm_resampler_type": null,
+  "mm_spatial_pool_mode": "bilinear",
+  "mm_spatial_pool_stride": 2,
+  "mm_tunable_parts": "mm_vision_tower,mm_mlp_adapter,mm_language_model",
+  "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",
+  "num_attention_heads": 28,
+  "num_hidden_layers": 28,
+  "num_key_value_heads": 4,
+  "pos_skipping_range": 4096,
+  "projector_hidden_act": "gelu",
+  "rms_norm_eps": 1e-06,
+  "rope_scaling": null,
+  "rope_theta": 1000000.0,
+  "sliding_window": 131072,
+  "text_config": {
+    "model_type": "llama"
+  },
+  "tie_word_embeddings": false,
+  "tokenizer_model_max_length": 32768,
+  "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_config": {
+    "hidden_size": 1024,
+    "image_size": 336,
+    "intermediate_size": 4096,
+    "model_type": "clip_vision_model",
+    "num_attention_heads": 16,
+    "num_hidden_layers": 24,
+    "patch_size": 14,
+    "projection_dim": 768,
+    "vocab_size": 32000
+  },
+  "vision_feature_layer": -2,
+  "vision_feature_select_strategy": "default",
+  "vision_tower_pretrained": null
+}
diff --git a/llavaqwen2-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split01_all_mm_tune_olora256_512_llm/generation_config.json b/llavaqwen2-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split01_all_mm_tune_olora256_512_llm/generation_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..19a297221acb87418d4388a3decef2282c6d7316
--- /dev/null
+++ b/llavaqwen2-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split01_all_mm_tune_olora256_512_llm/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/llavaqwen2-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split01_all_mm_tune_olora256_512_llm/non_lora_trainables.bin b/llavaqwen2-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split01_all_mm_tune_olora256_512_llm/non_lora_trainables.bin
new file mode 100644
index 0000000000000000000000000000000000000000..6f6f801daba98ef297d1415e4afa0b069b5ac103
--- /dev/null
+++ b/llavaqwen2-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split01_all_mm_tune_olora256_512_llm/non_lora_trainables.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6e6b67548d1e0c12a0ba9bc1f7f3fe44a54a6e2601ea37d771260fd162be7419
+size 33964208
diff --git a/llavaqwen2-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split01_all_mm_tune_olora256_512_llm/trainer_state.json b/llavaqwen2-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split01_all_mm_tune_olora256_512_llm/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..1820211b23cfcd9734df8819c9c7a72e7ccc5c45
--- /dev/null
+++ b/llavaqwen2-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split01_all_mm_tune_olora256_512_llm/trainer_state.json
@@ -0,0 +1,7030 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 1000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.3607246746171,
+      "learning_rate": 3.3333333333333335e-07,
+      "loss": 0.5894,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.6015062312526702,
+      "learning_rate": 6.666666666666667e-07,
+      "loss": 0.4503,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.7883022890023197,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 0.564,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.446709129426947,
+      "learning_rate": 1.3333333333333334e-06,
+      "loss": 0.4052,
+      "step": 4
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.950005793454983,
+      "learning_rate": 1.6666666666666667e-06,
+      "loss": 0.5694,
+      "step": 5
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0315383895224093,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 0.6325,
+      "step": 6
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4380940429454823,
+      "learning_rate": 2.3333333333333336e-06,
+      "loss": 0.4073,
+      "step": 7
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7902410169543705,
+      "learning_rate": 2.666666666666667e-06,
+      "loss": 0.496,
+      "step": 8
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.019886784152221,
+      "learning_rate": 3e-06,
+      "loss": 0.2978,
+      "step": 9
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.305448906172369,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 0.6514,
+      "step": 10
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7461464618031717,
+      "learning_rate": 3.6666666666666666e-06,
+      "loss": 0.4841,
+      "step": 11
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.133559780473121,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 0.5899,
+      "step": 12
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9577956061913435,
+      "learning_rate": 4.333333333333334e-06,
+      "loss": 0.5176,
+      "step": 13
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3215929577477967,
+      "learning_rate": 4.666666666666667e-06,
+      "loss": 0.3868,
+      "step": 14
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2001009666150546,
+      "learning_rate": 5e-06,
+      "loss": 0.3958,
+      "step": 15
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9304714590374654,
+      "learning_rate": 5.333333333333334e-06,
+      "loss": 0.306,
+      "step": 16
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7767444580668679,
+      "learning_rate": 5.666666666666667e-06,
+      "loss": 0.408,
+      "step": 17
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3287907694675276,
+      "learning_rate": 6e-06,
+      "loss": 0.3444,
+      "step": 18
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.597203933659128,
+      "learning_rate": 6.333333333333333e-06,
+      "loss": 0.632,
+      "step": 19
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2620348074780825,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 0.294,
+      "step": 20
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5328912209345755,
+      "learning_rate": 7e-06,
+      "loss": 0.3178,
+      "step": 21
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2419857209516763,
+      "learning_rate": 7.333333333333333e-06,
+      "loss": 0.2809,
+      "step": 22
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0122248180630153,
+      "learning_rate": 7.666666666666667e-06,
+      "loss": 0.1807,
+      "step": 23
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3320938267594107,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 0.2579,
+      "step": 24
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2756210972480526,
+      "learning_rate": 8.333333333333334e-06,
+      "loss": 0.2214,
+      "step": 25
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9440953818070815,
+      "learning_rate": 8.666666666666668e-06,
+      "loss": 0.1571,
+      "step": 26
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0654633495505794,
+      "learning_rate": 9e-06,
+      "loss": 0.1946,
+      "step": 27
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0307271012692505,
+      "learning_rate": 9.333333333333334e-06,
+      "loss": 0.2028,
+      "step": 28
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1525380505511964,
+      "learning_rate": 9.666666666666667e-06,
+      "loss": 0.2617,
+      "step": 29
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7953701897420127,
+      "learning_rate": 1e-05,
+      "loss": 0.1606,
+      "step": 30
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7658330397139902,
+      "learning_rate": 9.99997377618298e-06,
+      "loss": 0.1623,
+      "step": 31
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3688769608235227,
+      "learning_rate": 9.999895105006995e-06,
+      "loss": 0.232,
+      "step": 32
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0710880606725555,
+      "learning_rate": 9.999763987297266e-06,
+      "loss": 0.199,
+      "step": 33
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8922477359097567,
+      "learning_rate": 9.99958042442916e-06,
+      "loss": 0.0938,
+      "step": 34
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0519325757175217,
+      "learning_rate": 9.999344418328161e-06,
+      "loss": 0.1081,
+      "step": 35
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.42008932273011,
+      "learning_rate": 9.999055971469864e-06,
+      "loss": 0.2147,
+      "step": 36
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5090612769262195,
+      "learning_rate": 9.998715086879938e-06,
+      "loss": 0.2645,
+      "step": 37
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6386571363100515,
+      "learning_rate": 9.998321768134101e-06,
+      "loss": 0.1775,
+      "step": 38
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.177034629976266,
+      "learning_rate": 9.997876019358083e-06,
+      "loss": 0.2036,
+      "step": 39
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0024062591496823,
+      "learning_rate": 9.997377845227577e-06,
+      "loss": 0.1728,
+      "step": 40
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0929263562861968,
+      "learning_rate": 9.99682725096819e-06,
+      "loss": 0.1565,
+      "step": 41
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3187393755711911,
+      "learning_rate": 9.9962242423554e-06,
+      "loss": 0.1816,
+      "step": 42
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8213087137487407,
+      "learning_rate": 9.995568825714479e-06,
+      "loss": 0.1353,
+      "step": 43
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.649253335581187,
+      "learning_rate": 9.99486100792044e-06,
+      "loss": 0.1324,
+      "step": 44
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5717233913418078,
+      "learning_rate": 9.994100796397954e-06,
+      "loss": 0.1286,
+      "step": 45
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.737297938600789,
+      "learning_rate": 9.993288199121283e-06,
+      "loss": 0.1902,
+      "step": 46
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0028142866905823,
+      "learning_rate": 9.992423224614185e-06,
+      "loss": 0.1567,
+      "step": 47
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0374643268503156,
+      "learning_rate": 9.991505881949837e-06,
+      "loss": 0.1834,
+      "step": 48
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2828809631329192,
+      "learning_rate": 9.990536180750724e-06,
+      "loss": 0.1352,
+      "step": 49
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.807626588306248,
+      "learning_rate": 9.98951413118856e-06,
+      "loss": 0.1295,
+      "step": 50
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6402633907444848,
+      "learning_rate": 9.988439743984155e-06,
+      "loss": 0.2704,
+      "step": 51
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9909565607487354,
+      "learning_rate": 9.987313030407325e-06,
+      "loss": 0.1682,
+      "step": 52
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1824740893896895,
+      "learning_rate": 9.98613400227676e-06,
+      "loss": 0.1557,
+      "step": 53
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5358200613295229,
+      "learning_rate": 9.984902671959911e-06,
+      "loss": 0.1026,
+      "step": 54
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9805793717317114,
+      "learning_rate": 9.983619052372847e-06,
+      "loss": 0.1293,
+      "step": 55
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6228000314436585,
+      "learning_rate": 9.982283156980133e-06,
+      "loss": 0.1249,
+      "step": 56
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6112410766798437,
+      "learning_rate": 9.980894999794678e-06,
+      "loss": 0.1226,
+      "step": 57
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5350250183893485,
+      "learning_rate": 9.979454595377594e-06,
+      "loss": 0.0583,
+      "step": 58
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5595000396004797,
+      "learning_rate": 9.97796195883804e-06,
+      "loss": 0.1351,
+      "step": 59
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7974050724404418,
+      "learning_rate": 9.97641710583307e-06,
+      "loss": 0.163,
+      "step": 60
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7866162745931496,
+      "learning_rate": 9.97482005256746e-06,
+      "loss": 0.1425,
+      "step": 61
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6241895011307711,
+      "learning_rate": 9.973170815793543e-06,
+      "loss": 0.064,
+      "step": 62
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.913102515262216,
+      "learning_rate": 9.971469412811032e-06,
+      "loss": 0.1483,
+      "step": 63
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3697825079458334,
+      "learning_rate": 9.969715861466839e-06,
+      "loss": 0.1984,
+      "step": 64
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.187974862551584,
+      "learning_rate": 9.96791018015489e-06,
+      "loss": 0.1756,
+      "step": 65
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6520756310162947,
+      "learning_rate": 9.966052387815923e-06,
+      "loss": 0.1027,
+      "step": 66
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9336791633038164,
+      "learning_rate": 9.964142503937305e-06,
+      "loss": 0.1663,
+      "step": 67
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7459608541470574,
+      "learning_rate": 9.962180548552812e-06,
+      "loss": 0.1269,
+      "step": 68
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5369458077637503,
+      "learning_rate": 9.96016654224243e-06,
+      "loss": 0.1168,
+      "step": 69
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.3099380290630775,
+      "learning_rate": 9.958100506132127e-06,
+      "loss": 0.1417,
+      "step": 70
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7088466229007135,
+      "learning_rate": 9.955982461893648e-06,
+      "loss": 0.1128,
+      "step": 71
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6273327117991097,
+      "learning_rate": 9.953812431744274e-06,
+      "loss": 0.1235,
+      "step": 72
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5693900280941977,
+      "learning_rate": 9.951590438446597e-06,
+      "loss": 0.0968,
+      "step": 73
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5719038668639742,
+      "learning_rate": 9.94931650530827e-06,
+      "loss": 0.0969,
+      "step": 74
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8622476614437252,
+      "learning_rate": 9.946990656181782e-06,
+      "loss": 0.1384,
+      "step": 75
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6993792887886608,
+      "learning_rate": 9.944612915464183e-06,
+      "loss": 0.1265,
+      "step": 76
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7640565569055416,
+      "learning_rate": 9.942183308096853e-06,
+      "loss": 0.143,
+      "step": 77
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1258360637887,
+      "learning_rate": 9.93970185956522e-06,
+      "loss": 0.1361,
+      "step": 78
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5445199443767731,
+      "learning_rate": 9.93716859589851e-06,
+      "loss": 0.093,
+      "step": 79
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.040171743521041,
+      "learning_rate": 9.934583543669454e-06,
+      "loss": 0.124,
+      "step": 80
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9839597034019067,
+      "learning_rate": 9.93194672999403e-06,
+      "loss": 0.202,
+      "step": 81
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9301812327121488,
+      "learning_rate": 9.929258182531167e-06,
+      "loss": 0.1327,
+      "step": 82
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1345207127377692,
+      "learning_rate": 9.926517929482454e-06,
+      "loss": 0.1526,
+      "step": 83
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6622530400858483,
+      "learning_rate": 9.923725999591846e-06,
+      "loss": 0.1047,
+      "step": 84
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6711658591110805,
+      "learning_rate": 9.920882422145372e-06,
+      "loss": 0.1242,
+      "step": 85
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5879483840752662,
+      "learning_rate": 9.917987226970811e-06,
+      "loss": 0.1041,
+      "step": 86
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8511584440885696,
+      "learning_rate": 9.91504044443739e-06,
+      "loss": 0.1283,
+      "step": 87
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6782669907972337,
+      "learning_rate": 9.912042105455462e-06,
+      "loss": 0.0576,
+      "step": 88
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6457585382953173,
+      "learning_rate": 9.908992241476189e-06,
+      "loss": 0.1026,
+      "step": 89
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8833122772627231,
+      "learning_rate": 9.905890884491196e-06,
+      "loss": 0.1507,
+      "step": 90
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1980970686962742,
+      "learning_rate": 9.902738067032254e-06,
+      "loss": 0.1062,
+      "step": 91
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1334701103868552,
+      "learning_rate": 9.899533822170922e-06,
+      "loss": 0.164,
+      "step": 92
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9011291147381055,
+      "learning_rate": 9.896278183518216e-06,
+      "loss": 0.1399,
+      "step": 93
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7569316599830422,
+      "learning_rate": 9.892971185224244e-06,
+      "loss": 0.1326,
+      "step": 94
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2328121035048378,
+      "learning_rate": 9.889612861977855e-06,
+      "loss": 0.1503,
+      "step": 95
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.596300915381594,
+      "learning_rate": 9.886203249006265e-06,
+      "loss": 0.1142,
+      "step": 96
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6168275952114758,
+      "learning_rate": 9.882742382074707e-06,
+      "loss": 0.1048,
+      "step": 97
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6646867123233097,
+      "learning_rate": 9.879230297486034e-06,
+      "loss": 0.1268,
+      "step": 98
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4624576438541248,
+      "learning_rate": 9.875667032080354e-06,
+      "loss": 0.1636,
+      "step": 99
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.696569485211007,
+      "learning_rate": 9.872052623234632e-06,
+      "loss": 0.0904,
+      "step": 100
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6876476738040996,
+      "learning_rate": 9.868387108862307e-06,
+      "loss": 0.1326,
+      "step": 101
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3128848330139118,
+      "learning_rate": 9.864670527412891e-06,
+      "loss": 0.1819,
+      "step": 102
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9664864009368302,
+      "learning_rate": 9.860902917871566e-06,
+      "loss": 0.0792,
+      "step": 103
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6920287290973034,
+      "learning_rate": 9.857084319758772e-06,
+      "loss": 0.1043,
+      "step": 104
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6092944325732855,
+      "learning_rate": 9.853214773129796e-06,
+      "loss": 0.0876,
+      "step": 105
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0643719636119542,
+      "learning_rate": 9.849294318574353e-06,
+      "loss": 0.1577,
+      "step": 106
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7729367078232964,
+      "learning_rate": 9.845322997216153e-06,
+      "loss": 0.1009,
+      "step": 107
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0307500009173252,
+      "learning_rate": 9.841300850712479e-06,
+      "loss": 0.1418,
+      "step": 108
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6172660361836317,
+      "learning_rate": 9.837227921253747e-06,
+      "loss": 0.1078,
+      "step": 109
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6053934199519216,
+      "learning_rate": 9.833104251563058e-06,
+      "loss": 0.1046,
+      "step": 110
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8492583798544472,
+      "learning_rate": 9.828929884895753e-06,
+      "loss": 0.1615,
+      "step": 111
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7898012278460997,
+      "learning_rate": 9.824704865038967e-06,
+      "loss": 0.1504,
+      "step": 112
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4423348640674584,
+      "learning_rate": 9.820429236311158e-06,
+      "loss": 0.0746,
+      "step": 113
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.43000135393988737,
+      "learning_rate": 9.816103043561648e-06,
+      "loss": 0.0938,
+      "step": 114
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.370505236760701,
+      "learning_rate": 9.811726332170153e-06,
+      "loss": 0.1902,
+      "step": 115
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5419284871881185,
+      "learning_rate": 9.807299148046301e-06,
+      "loss": 0.0816,
+      "step": 116
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.582272008730522,
+      "learning_rate": 9.802821537629162e-06,
+      "loss": 0.093,
+      "step": 117
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5038565292887868,
+      "learning_rate": 9.798293547886748e-06,
+      "loss": 0.1,
+      "step": 118
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.509865304312992,
+      "learning_rate": 9.79371522631553e-06,
+      "loss": 0.0955,
+      "step": 119
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4957796733902061,
+      "learning_rate": 9.789086620939936e-06,
+      "loss": 0.0876,
+      "step": 120
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0857429677099277,
+      "learning_rate": 9.784407780311845e-06,
+      "loss": 0.1366,
+      "step": 121
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3891390182343122,
+      "learning_rate": 9.779678753510082e-06,
+      "loss": 0.1722,
+      "step": 122
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6338060483690361,
+      "learning_rate": 9.774899590139897e-06,
+      "loss": 0.1151,
+      "step": 123
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.38917086192216094,
+      "learning_rate": 9.770070340332457e-06,
+      "loss": 0.0861,
+      "step": 124
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7336168996432986,
+      "learning_rate": 9.765191054744305e-06,
+      "loss": 0.134,
+      "step": 125
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0534518367367267,
+      "learning_rate": 9.76026178455684e-06,
+      "loss": 0.1289,
+      "step": 126
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8079918912060179,
+      "learning_rate": 9.755282581475769e-06,
+      "loss": 0.1913,
+      "step": 127
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4763900468548068,
+      "learning_rate": 9.75025349773058e-06,
+      "loss": 0.0803,
+      "step": 128
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7681534439869117,
+      "learning_rate": 9.745174586073982e-06,
+      "loss": 0.1378,
+      "step": 129
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.750963291621297,
+      "learning_rate": 9.740045899781353e-06,
+      "loss": 0.1259,
+      "step": 130
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4344226462757411,
+      "learning_rate": 9.734867492650187e-06,
+      "loss": 0.0995,
+      "step": 131
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.539997704930958,
+      "learning_rate": 9.729639418999524e-06,
+      "loss": 0.1098,
+      "step": 132
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6762346368086923,
+      "learning_rate": 9.724361733669383e-06,
+      "loss": 0.125,
+      "step": 133
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5485194569881519,
+      "learning_rate": 9.719034492020183e-06,
+      "loss": 0.0991,
+      "step": 134
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3110924332122729,
+      "learning_rate": 9.713657749932172e-06,
+      "loss": 0.1883,
+      "step": 135
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8191799283206096,
+      "learning_rate": 9.708231563804828e-06,
+      "loss": 0.1505,
+      "step": 136
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6577131240580671,
+      "learning_rate": 9.702755990556277e-06,
+      "loss": 0.1412,
+      "step": 137
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.432630262046169,
+      "learning_rate": 9.697231087622691e-06,
+      "loss": 0.1068,
+      "step": 138
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6304900358752927,
+      "learning_rate": 9.691656912957686e-06,
+      "loss": 0.104,
+      "step": 139
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5559743037433238,
+      "learning_rate": 9.68603352503172e-06,
+      "loss": 0.1034,
+      "step": 140
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4505737091122359,
+      "learning_rate": 9.680360982831467e-06,
+      "loss": 0.1156,
+      "step": 141
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0503414847004549,
+      "learning_rate": 9.674639345859213e-06,
+      "loss": 0.1081,
+      "step": 142
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.953092212431232,
+      "learning_rate": 9.668868674132224e-06,
+      "loss": 0.142,
+      "step": 143
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.5773173116053294,
+      "learning_rate": 9.663049028182112e-06,
+      "loss": 0.1665,
+      "step": 144
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5211093890918267,
+      "learning_rate": 9.657180469054213e-06,
+      "loss": 0.1083,
+      "step": 145
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6772333518808913,
+      "learning_rate": 9.651263058306932e-06,
+      "loss": 0.1402,
+      "step": 146
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4728466109803736,
+      "learning_rate": 9.645296858011109e-06,
+      "loss": 0.1,
+      "step": 147
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5686134641449667,
+      "learning_rate": 9.639281930749363e-06,
+      "loss": 0.0981,
+      "step": 148
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5150381464311025,
+      "learning_rate": 9.633218339615433e-06,
+      "loss": 0.0768,
+      "step": 149
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7829238992726227,
+      "learning_rate": 9.627106148213521e-06,
+      "loss": 0.1199,
+      "step": 150
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5634299468247613,
+      "learning_rate": 9.620945420657625e-06,
+      "loss": 0.1284,
+      "step": 151
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7739328263970474,
+      "learning_rate": 9.61473622157086e-06,
+      "loss": 0.0756,
+      "step": 152
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4929993829829465,
+      "learning_rate": 9.608478616084784e-06,
+      "loss": 0.0835,
+      "step": 153
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6627675519129016,
+      "learning_rate": 9.602172669838721e-06,
+      "loss": 0.0857,
+      "step": 154
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.43333238348562697,
+      "learning_rate": 9.595818448979061e-06,
+      "loss": 0.0629,
+      "step": 155
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4318823799173638,
+      "learning_rate": 9.589416020158577e-06,
+      "loss": 0.0726,
+      "step": 156
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3223767412425636,
+      "learning_rate": 9.582965450535716e-06,
+      "loss": 0.0711,
+      "step": 157
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8479275013069608,
+      "learning_rate": 9.5764668077739e-06,
+      "loss": 0.177,
+      "step": 158
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6033706807381652,
+      "learning_rate": 9.569920160040815e-06,
+      "loss": 0.1022,
+      "step": 159
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5960746908872542,
+      "learning_rate": 9.563325576007702e-06,
+      "loss": 0.0917,
+      "step": 160
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5989400193286061,
+      "learning_rate": 9.556683124848624e-06,
+      "loss": 0.1499,
+      "step": 161
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6321761364280136,
+      "learning_rate": 9.549992876239753e-06,
+      "loss": 0.1021,
+      "step": 162
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9456659052294284,
+      "learning_rate": 9.54325490035863e-06,
+      "loss": 0.1732,
+      "step": 163
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2369762916556193,
+      "learning_rate": 9.536469267883432e-06,
+      "loss": 0.1457,
+      "step": 164
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5776155361505679,
+      "learning_rate": 9.529636049992235e-06,
+      "loss": 0.108,
+      "step": 165
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5861181662355707,
+      "learning_rate": 9.52275531836226e-06,
+      "loss": 0.0885,
+      "step": 166
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.466704353066058,
+      "learning_rate": 9.515827145169128e-06,
+      "loss": 0.1213,
+      "step": 167
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9732726445645538,
+      "learning_rate": 9.508851603086094e-06,
+      "loss": 0.1545,
+      "step": 168
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8489210921256994,
+      "learning_rate": 9.501828765283295e-06,
+      "loss": 0.1307,
+      "step": 169
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7185761664234415,
+      "learning_rate": 9.494758705426978e-06,
+      "loss": 0.1384,
+      "step": 170
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7416125172072018,
+      "learning_rate": 9.487641497678724e-06,
+      "loss": 0.1252,
+      "step": 171
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1794527596283202,
+      "learning_rate": 9.480477216694674e-06,
+      "loss": 0.1259,
+      "step": 172
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7402358672676869,
+      "learning_rate": 9.473265937624748e-06,
+      "loss": 0.1198,
+      "step": 173
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6312235522978472,
+      "learning_rate": 9.466007736111846e-06,
+      "loss": 0.1148,
+      "step": 174
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7196515773415135,
+      "learning_rate": 9.458702688291072e-06,
+      "loss": 0.1319,
+      "step": 175
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6850486515308531,
+      "learning_rate": 9.451350870788922e-06,
+      "loss": 0.1245,
+      "step": 176
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5696445336187979,
+      "learning_rate": 9.443952360722477e-06,
+      "loss": 0.1393,
+      "step": 177
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5404192629395442,
+      "learning_rate": 9.436507235698613e-06,
+      "loss": 0.1339,
+      "step": 178
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8471694217332177,
+      "learning_rate": 9.429015573813163e-06,
+      "loss": 0.1648,
+      "step": 179
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.2583760623452127,
+      "learning_rate": 9.421477453650118e-06,
+      "loss": 0.1804,
+      "step": 180
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9438170221888703,
+      "learning_rate": 9.413892954280793e-06,
+      "loss": 0.1331,
+      "step": 181
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5315911820285912,
+      "learning_rate": 9.406262155262995e-06,
+      "loss": 0.1131,
+      "step": 182
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.519208078747679,
+      "learning_rate": 9.398585136640195e-06,
+      "loss": 0.1156,
+      "step": 183
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5286389605503108,
+      "learning_rate": 9.390861978940687e-06,
+      "loss": 0.0938,
+      "step": 184
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.56762669333044,
+      "learning_rate": 9.38309276317674e-06,
+      "loss": 0.1242,
+      "step": 185
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.573116084711049,
+      "learning_rate": 9.37527757084375e-06,
+      "loss": 0.1163,
+      "step": 186
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1280550535208913,
+      "learning_rate": 9.367416483919387e-06,
+      "loss": 0.1394,
+      "step": 187
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7361759351819016,
+      "learning_rate": 9.359509584862735e-06,
+      "loss": 0.0944,
+      "step": 188
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6086432318944858,
+      "learning_rate": 9.351556956613423e-06,
+      "loss": 0.0966,
+      "step": 189
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5310758967896163,
+      "learning_rate": 9.343558682590757e-06,
+      "loss": 0.1755,
+      "step": 190
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5161731418528083,
+      "learning_rate": 9.335514846692846e-06,
+      "loss": 0.1072,
+      "step": 191
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6043224377669838,
+      "learning_rate": 9.327425533295725e-06,
+      "loss": 0.1523,
+      "step": 192
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5596996008010169,
+      "learning_rate": 9.31929082725246e-06,
+      "loss": 0.0894,
+      "step": 193
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4741364703546338,
+      "learning_rate": 9.31111081389227e-06,
+      "loss": 0.1253,
+      "step": 194
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4319905138436581,
+      "learning_rate": 9.302885579019626e-06,
+      "loss": 0.107,
+      "step": 195
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.627710521584554,
+      "learning_rate": 9.29461520891335e-06,
+      "loss": 0.1123,
+      "step": 196
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.36311863871574673,
+      "learning_rate": 9.286299790325708e-06,
+      "loss": 0.0525,
+      "step": 197
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7891128732253475,
+      "learning_rate": 9.277939410481507e-06,
+      "loss": 0.1446,
+      "step": 198
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2654640878518608,
+      "learning_rate": 9.269534157077177e-06,
+      "loss": 0.1481,
+      "step": 199
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5008647483406797,
+      "learning_rate": 9.261084118279846e-06,
+      "loss": 0.0961,
+      "step": 200
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5492177287320504,
+      "learning_rate": 9.252589382726426e-06,
+      "loss": 0.0914,
+      "step": 201
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8402239998184694,
+      "learning_rate": 9.244050039522673e-06,
+      "loss": 0.1617,
+      "step": 202
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5559678389424397,
+      "learning_rate": 9.235466178242255e-06,
+      "loss": 0.0805,
+      "step": 203
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.34769819923378315,
+      "learning_rate": 9.226837888925813e-06,
+      "loss": 0.0627,
+      "step": 204
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5464670400325233,
+      "learning_rate": 9.218165262080024e-06,
+      "loss": 0.0956,
+      "step": 205
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8564112615923611,
+      "learning_rate": 9.209448388676636e-06,
+      "loss": 0.1793,
+      "step": 206
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4637904050413668,
+      "learning_rate": 9.200687360151527e-06,
+      "loss": 0.0837,
+      "step": 207
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5601509840385756,
+      "learning_rate": 9.191882268403743e-06,
+      "loss": 0.1054,
+      "step": 208
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5539453679241664,
+      "learning_rate": 9.183033205794525e-06,
+      "loss": 0.1133,
+      "step": 209
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5577119185386734,
+      "learning_rate": 9.174140265146356e-06,
+      "loss": 0.1092,
+      "step": 210
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7982965587061583,
+      "learning_rate": 9.165203539741976e-06,
+      "loss": 0.0945,
+      "step": 211
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7005488888369035,
+      "learning_rate": 9.156223123323405e-06,
+      "loss": 0.1175,
+      "step": 212
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.49775597774931946,
+      "learning_rate": 9.14719911009096e-06,
+      "loss": 0.1005,
+      "step": 213
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2939083285022925,
+      "learning_rate": 9.13813159470227e-06,
+      "loss": 0.1264,
+      "step": 214
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4817398733660537,
+      "learning_rate": 9.129020672271283e-06,
+      "loss": 0.104,
+      "step": 215
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6234597598880522,
+      "learning_rate": 9.119866438367263e-06,
+      "loss": 0.1971,
+      "step": 216
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.44266381304483315,
+      "learning_rate": 9.11066898901379e-06,
+      "loss": 0.0781,
+      "step": 217
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1016366481857078,
+      "learning_rate": 9.101428420687759e-06,
+      "loss": 0.1109,
+      "step": 218
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7277807395614646,
+      "learning_rate": 9.092144830318357e-06,
+      "loss": 0.111,
+      "step": 219
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.39280208312630055,
+      "learning_rate": 9.082818315286054e-06,
+      "loss": 0.0812,
+      "step": 220
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6050461775054878,
+      "learning_rate": 9.073448973421581e-06,
+      "loss": 0.1104,
+      "step": 221
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7444788675747808,
+      "learning_rate": 9.0640369030049e-06,
+      "loss": 0.1285,
+      "step": 222
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8148118998555378,
+      "learning_rate": 9.054582202764175e-06,
+      "loss": 0.165,
+      "step": 223
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5611328780760997,
+      "learning_rate": 9.045084971874738e-06,
+      "loss": 0.0996,
+      "step": 224
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5613399147559089,
+      "learning_rate": 9.035545309958048e-06,
+      "loss": 0.1338,
+      "step": 225
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.35377812380473267,
+      "learning_rate": 9.025963317080641e-06,
+      "loss": 0.0574,
+      "step": 226
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.124413626677449,
+      "learning_rate": 9.016339093753093e-06,
+      "loss": 0.111,
+      "step": 227
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.681070569595969,
+      "learning_rate": 9.006672740928952e-06,
+      "loss": 0.0982,
+      "step": 228
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5759668179153918,
+      "learning_rate": 8.99696436000368e-06,
+      "loss": 0.0726,
+      "step": 229
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1282044565957114,
+      "learning_rate": 8.987214052813605e-06,
+      "loss": 0.1048,
+      "step": 230
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.45035923851895315,
+      "learning_rate": 8.977421921634833e-06,
+      "loss": 0.0972,
+      "step": 231
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.679954755110806,
+      "learning_rate": 8.967588069182184e-06,
+      "loss": 0.0965,
+      "step": 232
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4958838697150068,
+      "learning_rate": 8.957712598608123e-06,
+      "loss": 0.1084,
+      "step": 233
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.946219199103302,
+      "learning_rate": 8.947795613501658e-06,
+      "loss": 0.1722,
+      "step": 234
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5869393928381269,
+      "learning_rate": 8.937837217887273e-06,
+      "loss": 0.1092,
+      "step": 235
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4229221773571093,
+      "learning_rate": 8.927837516223824e-06,
+      "loss": 0.1151,
+      "step": 236
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4559163753035949,
+      "learning_rate": 8.917796613403451e-06,
+      "loss": 0.0691,
+      "step": 237
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.670297330053043,
+      "learning_rate": 8.907714614750473e-06,
+      "loss": 0.13,
+      "step": 238
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.43386300447049314,
+      "learning_rate": 8.897591626020284e-06,
+      "loss": 0.0955,
+      "step": 239
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6785813412177598,
+      "learning_rate": 8.887427753398249e-06,
+      "loss": 0.1308,
+      "step": 240
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6197078038796717,
+      "learning_rate": 8.877223103498576e-06,
+      "loss": 0.0958,
+      "step": 241
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7076683957063388,
+      "learning_rate": 8.866977783363219e-06,
+      "loss": 0.1035,
+      "step": 242
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6210521343309058,
+      "learning_rate": 8.85669190046074e-06,
+      "loss": 0.0873,
+      "step": 243
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3961276294508894,
+      "learning_rate": 8.846365562685178e-06,
+      "loss": 0.0856,
+      "step": 244
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7986412285531543,
+      "learning_rate": 8.83599887835493e-06,
+      "loss": 0.1862,
+      "step": 245
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4381098035523295,
+      "learning_rate": 8.825591956211614e-06,
+      "loss": 0.0909,
+      "step": 246
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7819035176829826,
+      "learning_rate": 8.815144905418918e-06,
+      "loss": 0.0882,
+      "step": 247
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8340751878964306,
+      "learning_rate": 8.804657835561456e-06,
+      "loss": 0.0708,
+      "step": 248
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4759107242077579,
+      "learning_rate": 8.794130856643635e-06,
+      "loss": 0.1174,
+      "step": 249
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4419037098541667,
+      "learning_rate": 8.783564079088478e-06,
+      "loss": 0.0611,
+      "step": 250
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6838098339826392,
+      "learning_rate": 8.772957613736483e-06,
+      "loss": 0.1318,
+      "step": 251
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6805585810191564,
+      "learning_rate": 8.762311571844453e-06,
+      "loss": 0.1144,
+      "step": 252
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.880891649880843,
+      "learning_rate": 8.751626065084328e-06,
+      "loss": 0.1613,
+      "step": 253
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6484061037081844,
+      "learning_rate": 8.74090120554202e-06,
+      "loss": 0.0721,
+      "step": 254
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5481113998080442,
+      "learning_rate": 8.730137105716231e-06,
+      "loss": 0.09,
+      "step": 255
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8729016284845439,
+      "learning_rate": 8.719333878517274e-06,
+      "loss": 0.1313,
+      "step": 256
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8361974654852907,
+      "learning_rate": 8.708491637265888e-06,
+      "loss": 0.1075,
+      "step": 257
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5774979943719853,
+      "learning_rate": 8.697610495692055e-06,
+      "loss": 0.107,
+      "step": 258
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6649869661098885,
+      "learning_rate": 8.686690567933803e-06,
+      "loss": 0.1118,
+      "step": 259
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7000964181249009,
+      "learning_rate": 8.675731968536004e-06,
+      "loss": 0.1259,
+      "step": 260
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.662137270152306,
+      "learning_rate": 8.66473481244918e-06,
+      "loss": 0.0957,
+      "step": 261
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8378809743130473,
+      "learning_rate": 8.653699215028298e-06,
+      "loss": 0.1436,
+      "step": 262
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6933424859787171,
+      "learning_rate": 8.64262529203155e-06,
+      "loss": 0.1198,
+      "step": 263
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5675466712230994,
+      "learning_rate": 8.63151315961915e-06,
+      "loss": 0.1583,
+      "step": 264
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6733269469920226,
+      "learning_rate": 8.620362934352109e-06,
+      "loss": 0.1049,
+      "step": 265
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.086757466314047,
+      "learning_rate": 8.609174733191012e-06,
+      "loss": 0.1396,
+      "step": 266
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7256776861439452,
+      "learning_rate": 8.597948673494794e-06,
+      "loss": 0.162,
+      "step": 267
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.47995856922745955,
+      "learning_rate": 8.586684873019513e-06,
+      "loss": 0.0983,
+      "step": 268
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8221525546080067,
+      "learning_rate": 8.575383449917103e-06,
+      "loss": 0.1,
+      "step": 269
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6149586237967274,
+      "learning_rate": 8.564044522734147e-06,
+      "loss": 0.1187,
+      "step": 270
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.290261899072581,
+      "learning_rate": 8.552668210410624e-06,
+      "loss": 0.1215,
+      "step": 271
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7103356271751441,
+      "learning_rate": 8.541254632278667e-06,
+      "loss": 0.1128,
+      "step": 272
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8183102626323511,
+      "learning_rate": 8.52980390806131e-06,
+      "loss": 0.0652,
+      "step": 273
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4924364063295098,
+      "learning_rate": 8.518316157871232e-06,
+      "loss": 0.0811,
+      "step": 274
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6379863736427677,
+      "learning_rate": 8.506791502209497e-06,
+      "loss": 0.1421,
+      "step": 275
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5365549756136844,
+      "learning_rate": 8.495230061964289e-06,
+      "loss": 0.1037,
+      "step": 276
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5918213588332102,
+      "learning_rate": 8.483631958409644e-06,
+      "loss": 0.1091,
+      "step": 277
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5910185718944498,
+      "learning_rate": 8.471997313204183e-06,
+      "loss": 0.1008,
+      "step": 278
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8434865299141647,
+      "learning_rate": 8.460326248389825e-06,
+      "loss": 0.1261,
+      "step": 279
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8084790273022816,
+      "learning_rate": 8.448618886390523e-06,
+      "loss": 0.1472,
+      "step": 280
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6064244109977189,
+      "learning_rate": 8.436875350010958e-06,
+      "loss": 0.1378,
+      "step": 281
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8099455428698011,
+      "learning_rate": 8.425095762435274e-06,
+      "loss": 0.1488,
+      "step": 282
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.854892262931246,
+      "learning_rate": 8.41328024722577e-06,
+      "loss": 0.1498,
+      "step": 283
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5185575941538992,
+      "learning_rate": 8.401428928321607e-06,
+      "loss": 0.0902,
+      "step": 284
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5965587314972228,
+      "learning_rate": 8.389541930037516e-06,
+      "loss": 0.106,
+      "step": 285
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.459582914012201,
+      "learning_rate": 8.377619377062483e-06,
+      "loss": 0.0982,
+      "step": 286
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6673096852048638,
+      "learning_rate": 8.365661394458446e-06,
+      "loss": 0.1108,
+      "step": 287
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6698950014998947,
+      "learning_rate": 8.353668107658984e-06,
+      "loss": 0.1296,
+      "step": 288
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9163329174066326,
+      "learning_rate": 8.341639642468002e-06,
+      "loss": 0.1499,
+      "step": 289
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6835073933135426,
+      "learning_rate": 8.329576125058406e-06,
+      "loss": 0.0651,
+      "step": 290
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.47586344778493134,
+      "learning_rate": 8.317477681970786e-06,
+      "loss": 0.0982,
+      "step": 291
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.027887942457216,
+      "learning_rate": 8.305344440112089e-06,
+      "loss": 0.1193,
+      "step": 292
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6883146738724482,
+      "learning_rate": 8.293176526754274e-06,
+      "loss": 0.1636,
+      "step": 293
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.308519284995934,
+      "learning_rate": 8.280974069532999e-06,
+      "loss": 0.1081,
+      "step": 294
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5575442309211142,
+      "learning_rate": 8.268737196446264e-06,
+      "loss": 0.1049,
+      "step": 295
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4295260773986241,
+      "learning_rate": 8.256466035853077e-06,
+      "loss": 0.0751,
+      "step": 296
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6141976073277213,
+      "learning_rate": 8.244160716472109e-06,
+      "loss": 0.1336,
+      "step": 297
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1909104778355368,
+      "learning_rate": 8.231821367380335e-06,
+      "loss": 0.1719,
+      "step": 298
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5340253751607172,
+      "learning_rate": 8.219448118011687e-06,
+      "loss": 0.0899,
+      "step": 299
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.029527908281168,
+      "learning_rate": 8.207041098155701e-06,
+      "loss": 0.1151,
+      "step": 300
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8985930248766892,
+      "learning_rate": 8.19460043795614e-06,
+      "loss": 0.1225,
+      "step": 301
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5206416280199094,
+      "learning_rate": 8.182126267909642e-06,
+      "loss": 0.0987,
+      "step": 302
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9381246435874875,
+      "learning_rate": 8.16961871886435e-06,
+      "loss": 0.2008,
+      "step": 303
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0766922252901279,
+      "learning_rate": 8.157077922018537e-06,
+      "loss": 0.1593,
+      "step": 304
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5950079693273216,
+      "learning_rate": 8.144504008919224e-06,
+      "loss": 0.099,
+      "step": 305
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5071063988827308,
+      "learning_rate": 8.13189711146081e-06,
+      "loss": 0.0884,
+      "step": 306
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6170093430029727,
+      "learning_rate": 8.119257361883686e-06,
+      "loss": 0.0949,
+      "step": 307
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6216963302278578,
+      "learning_rate": 8.106584892772844e-06,
+      "loss": 0.135,
+      "step": 308
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7515361895060123,
+      "learning_rate": 8.093879837056486e-06,
+      "loss": 0.1424,
+      "step": 309
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5301845393435228,
+      "learning_rate": 8.081142328004638e-06,
+      "loss": 0.0842,
+      "step": 310
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.36401321561122435,
+      "learning_rate": 8.068372499227738e-06,
+      "loss": 0.0941,
+      "step": 311
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.43388092832373126,
+      "learning_rate": 8.055570484675252e-06,
+      "loss": 0.0812,
+      "step": 312
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3111677501066997,
+      "learning_rate": 8.042736418634252e-06,
+      "loss": 0.071,
+      "step": 313
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2711927555540456,
+      "learning_rate": 8.029870435728018e-06,
+      "loss": 0.0665,
+      "step": 314
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6158192375387915,
+      "learning_rate": 8.016972670914624e-06,
+      "loss": 0.1366,
+      "step": 315
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5393832711474328,
+      "learning_rate": 8.004043259485519e-06,
+      "loss": 0.1256,
+      "step": 316
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.819073710300546,
+      "learning_rate": 7.99108233706411e-06,
+      "loss": 0.1342,
+      "step": 317
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4881682283012466,
+      "learning_rate": 7.978090039604342e-06,
+      "loss": 0.1098,
+      "step": 318
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.425625759706754,
+      "learning_rate": 7.965066503389264e-06,
+      "loss": 0.1157,
+      "step": 319
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8417583243303938,
+      "learning_rate": 7.952011865029614e-06,
+      "loss": 0.1504,
+      "step": 320
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5241553526731788,
+      "learning_rate": 7.938926261462366e-06,
+      "loss": 0.1047,
+      "step": 321
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.092665792919607,
+      "learning_rate": 7.925809829949312e-06,
+      "loss": 0.1766,
+      "step": 322
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.48982963314229716,
+      "learning_rate": 7.91266270807561e-06,
+      "loss": 0.1012,
+      "step": 323
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5045779559864191,
+      "learning_rate": 7.89948503374835e-06,
+      "loss": 0.0838,
+      "step": 324
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0147095478683965,
+      "learning_rate": 7.886276945195098e-06,
+      "loss": 0.1543,
+      "step": 325
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6721766277581208,
+      "learning_rate": 7.873038580962453e-06,
+      "loss": 0.0797,
+      "step": 326
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2897950063397245,
+      "learning_rate": 7.859770079914592e-06,
+      "loss": 0.1268,
+      "step": 327
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6630020038299413,
+      "learning_rate": 7.846471581231814e-06,
+      "loss": 0.0768,
+      "step": 328
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7728033471228617,
+      "learning_rate": 7.833143224409076e-06,
+      "loss": 0.1191,
+      "step": 329
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4554370312043164,
+      "learning_rate": 7.819785149254534e-06,
+      "loss": 0.0948,
+      "step": 330
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4023913241784609,
+      "learning_rate": 7.806397495888074e-06,
+      "loss": 0.1015,
+      "step": 331
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5534162223569474,
+      "learning_rate": 7.792980404739849e-06,
+      "loss": 0.0938,
+      "step": 332
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.46691388495842806,
+      "learning_rate": 7.779534016548791e-06,
+      "loss": 0.0601,
+      "step": 333
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7957525089538556,
+      "learning_rate": 7.766058472361154e-06,
+      "loss": 0.0786,
+      "step": 334
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6171254836650297,
+      "learning_rate": 7.752553913529019e-06,
+      "loss": 0.1148,
+      "step": 335
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.39913106915745733,
+      "learning_rate": 7.739020481708816e-06,
+      "loss": 0.1068,
+      "step": 336
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8911333363777895,
+      "learning_rate": 7.725458318859842e-06,
+      "loss": 0.102,
+      "step": 337
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4629156522652224,
+      "learning_rate": 7.711867567242769e-06,
+      "loss": 0.0943,
+      "step": 338
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6427640885980827,
+      "learning_rate": 7.698248369418146e-06,
+      "loss": 0.1326,
+      "step": 339
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5659495632220061,
+      "learning_rate": 7.68460086824492e-06,
+      "loss": 0.0995,
+      "step": 340
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6325174096409139,
+      "learning_rate": 7.670925206878917e-06,
+      "loss": 0.1138,
+      "step": 341
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7936200068990925,
+      "learning_rate": 7.657221528771352e-06,
+      "loss": 0.1654,
+      "step": 342
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8356850463342996,
+      "learning_rate": 7.643489977667327e-06,
+      "loss": 0.1486,
+      "step": 343
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7685126012757484,
+      "learning_rate": 7.629730697604314e-06,
+      "loss": 0.1103,
+      "step": 344
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4838660469726114,
+      "learning_rate": 7.61594383291065e-06,
+      "loss": 0.1066,
+      "step": 345
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3483440899453769,
+      "learning_rate": 7.602129528204023e-06,
+      "loss": 0.0903,
+      "step": 346
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8548146759546644,
+      "learning_rate": 7.588287928389952e-06,
+      "loss": 0.1636,
+      "step": 347
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4294305083473748,
+      "learning_rate": 7.574419178660269e-06,
+      "loss": 0.1158,
+      "step": 348
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4115732988087181,
+      "learning_rate": 7.560523424491595e-06,
+      "loss": 0.1021,
+      "step": 349
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0097008417071993,
+      "learning_rate": 7.546600811643816e-06,
+      "loss": 0.1783,
+      "step": 350
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6259202083211147,
+      "learning_rate": 7.532651486158554e-06,
+      "loss": 0.1021,
+      "step": 351
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7320795596317787,
+      "learning_rate": 7.5186755943576324e-06,
+      "loss": 0.1083,
+      "step": 352
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.39660777221329047,
+      "learning_rate": 7.504673282841544e-06,
+      "loss": 0.0782,
+      "step": 353
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5037925010178963,
+      "learning_rate": 7.490644698487909e-06,
+      "loss": 0.1066,
+      "step": 354
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3920215273167089,
+      "learning_rate": 7.476589988449939e-06,
+      "loss": 0.073,
+      "step": 355
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6877910161758389,
+      "learning_rate": 7.462509300154892e-06,
+      "loss": 0.1165,
+      "step": 356
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6193936684890975,
+      "learning_rate": 7.448402781302526e-06,
+      "loss": 0.0897,
+      "step": 357
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6544150672463809,
+      "learning_rate": 7.434270579863549e-06,
+      "loss": 0.0911,
+      "step": 358
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.314171970891291,
+      "learning_rate": 7.420112844078066e-06,
+      "loss": 0.1305,
+      "step": 359
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8041982462278258,
+      "learning_rate": 7.405929722454026e-06,
+      "loss": 0.1192,
+      "step": 360
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5637989804443223,
+      "learning_rate": 7.391721363765664e-06,
+      "loss": 0.1563,
+      "step": 361
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7521330412268286,
+      "learning_rate": 7.3774879170519386e-06,
+      "loss": 0.1577,
+      "step": 362
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.39846411449313535,
+      "learning_rate": 7.363229531614973e-06,
+      "loss": 0.1052,
+      "step": 363
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.46842312605388364,
+      "learning_rate": 7.348946357018479e-06,
+      "loss": 0.0967,
+      "step": 364
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5908090831853243,
+      "learning_rate": 7.334638543086203e-06,
+      "loss": 0.0759,
+      "step": 365
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3214400708242895,
+      "learning_rate": 7.320306239900343e-06,
+      "loss": 0.0385,
+      "step": 366
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.722132793266274,
+      "learning_rate": 7.305949597799976e-06,
+      "loss": 0.1114,
+      "step": 367
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.3269546028744137,
+      "learning_rate": 7.291568767379484e-06,
+      "loss": 0.1547,
+      "step": 368
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.43869990888711935,
+      "learning_rate": 7.277163899486975e-06,
+      "loss": 0.0412,
+      "step": 369
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0948960403311223,
+      "learning_rate": 7.262735145222696e-06,
+      "loss": 0.117,
+      "step": 370
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6680283933724516,
+      "learning_rate": 7.248282655937451e-06,
+      "loss": 0.0918,
+      "step": 371
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5454331336315559,
+      "learning_rate": 7.233806583231012e-06,
+      "loss": 0.0615,
+      "step": 372
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5088438263092467,
+      "learning_rate": 7.219307078950536e-06,
+      "loss": 0.0816,
+      "step": 373
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6588995635030355,
+      "learning_rate": 7.204784295188959e-06,
+      "loss": 0.1221,
+      "step": 374
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9800509058632336,
+      "learning_rate": 7.190238384283413e-06,
+      "loss": 0.1574,
+      "step": 375
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2052741828186413,
+      "learning_rate": 7.1756694988136165e-06,
+      "loss": 0.1793,
+      "step": 376
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7092844962709588,
+      "learning_rate": 7.161077791600288e-06,
+      "loss": 0.1434,
+      "step": 377
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8103864711082105,
+      "learning_rate": 7.14646341570353e-06,
+      "loss": 0.1735,
+      "step": 378
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5186224805906783,
+      "learning_rate": 7.1318265244212305e-06,
+      "loss": 0.1008,
+      "step": 379
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5046095769937358,
+      "learning_rate": 7.117167271287453e-06,
+      "loss": 0.1266,
+      "step": 380
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6291806702745448,
+      "learning_rate": 7.102485810070824e-06,
+      "loss": 0.129,
+      "step": 381
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8706680878449085,
+      "learning_rate": 7.0877822947729265e-06,
+      "loss": 0.1425,
+      "step": 382
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7517343416434489,
+      "learning_rate": 7.073056879626681e-06,
+      "loss": 0.1381,
+      "step": 383
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4850521528617002,
+      "learning_rate": 7.05830971909472e-06,
+      "loss": 0.0837,
+      "step": 384
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.467190916228377,
+      "learning_rate": 7.043540967867782e-06,
+      "loss": 0.1229,
+      "step": 385
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.35442767259956226,
+      "learning_rate": 7.028750780863078e-06,
+      "loss": 0.0604,
+      "step": 386
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.632531620020242,
+      "learning_rate": 7.013939313222669e-06,
+      "loss": 0.0992,
+      "step": 387
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7721017686992716,
+      "learning_rate": 6.999106720311846e-06,
+      "loss": 0.1562,
+      "step": 388
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6265392153084421,
+      "learning_rate": 6.9842531577174865e-06,
+      "loss": 0.138,
+      "step": 389
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8619646556127373,
+      "learning_rate": 6.969378781246436e-06,
+      "loss": 0.1323,
+      "step": 390
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6973949993687323,
+      "learning_rate": 6.954483746923865e-06,
+      "loss": 0.1082,
+      "step": 391
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.48241252511918464,
+      "learning_rate": 6.939568210991633e-06,
+      "loss": 0.1345,
+      "step": 392
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.39274471270934264,
+      "learning_rate": 6.924632329906657e-06,
+      "loss": 0.1055,
+      "step": 393
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5781946342673107,
+      "learning_rate": 6.9096762603392595e-06,
+      "loss": 0.1409,
+      "step": 394
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3525825420240066,
+      "learning_rate": 6.894700159171535e-06,
+      "loss": 0.0883,
+      "step": 395
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5246273586743376,
+      "learning_rate": 6.8797041834956955e-06,
+      "loss": 0.081,
+      "step": 396
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6518585669331103,
+      "learning_rate": 6.8646884906124345e-06,
+      "loss": 0.1007,
+      "step": 397
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3508427750003166,
+      "learning_rate": 6.849653238029261e-06,
+      "loss": 0.0641,
+      "step": 398
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6551363944925317,
+      "learning_rate": 6.834598583458862e-06,
+      "loss": 0.1292,
+      "step": 399
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7451482471986087,
+      "learning_rate": 6.819524684817439e-06,
+      "loss": 0.1621,
+      "step": 400
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5304490064747236,
+      "learning_rate": 6.804431700223057e-06,
+      "loss": 0.0912,
+      "step": 401
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.767445319430965,
+      "learning_rate": 6.78931978799398e-06,
+      "loss": 0.1184,
+      "step": 402
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5097226662887627,
+      "learning_rate": 6.774189106647021e-06,
+      "loss": 0.0819,
+      "step": 403
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.40426010526771866,
+      "learning_rate": 6.7590398148958625e-06,
+      "loss": 0.0621,
+      "step": 404
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6698796219963394,
+      "learning_rate": 6.743872071649411e-06,
+      "loss": 0.0607,
+      "step": 405
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7051411305657742,
+      "learning_rate": 6.728686036010115e-06,
+      "loss": 0.1194,
+      "step": 406
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5192947281652533,
+      "learning_rate": 6.7134818672723005e-06,
+      "loss": 0.1401,
+      "step": 407
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5375041644147377,
+      "learning_rate": 6.698259724920503e-06,
+      "loss": 0.114,
+      "step": 408
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5586400961711949,
+      "learning_rate": 6.6830197686277945e-06,
+      "loss": 0.1078,
+      "step": 409
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1282069492368068,
+      "learning_rate": 6.667762158254104e-06,
+      "loss": 0.1872,
+      "step": 410
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4389746508236176,
+      "learning_rate": 6.652487053844544e-06,
+      "loss": 0.0831,
+      "step": 411
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5030133333814518,
+      "learning_rate": 6.637194615627733e-06,
+      "loss": 0.1168,
+      "step": 412
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5244578006283286,
+      "learning_rate": 6.621885004014113e-06,
+      "loss": 0.1154,
+      "step": 413
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3319122797122553,
+      "learning_rate": 6.6065583795942625e-06,
+      "loss": 0.0897,
+      "step": 414
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8670940267704824,
+      "learning_rate": 6.591214903137221e-06,
+      "loss": 0.1614,
+      "step": 415
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3790472396407466,
+      "learning_rate": 6.5758547355887944e-06,
+      "loss": 0.0799,
+      "step": 416
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6708691225198117,
+      "learning_rate": 6.560478038069873e-06,
+      "loss": 0.1179,
+      "step": 417
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5613038123983146,
+      "learning_rate": 6.545084971874738e-06,
+      "loss": 0.0733,
+      "step": 418
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3911040134070448,
+      "learning_rate": 6.52967569846937e-06,
+      "loss": 0.0664,
+      "step": 419
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7275454662991591,
+      "learning_rate": 6.514250379489754e-06,
+      "loss": 0.1519,
+      "step": 420
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5393692947961118,
+      "learning_rate": 6.49880917674019e-06,
+      "loss": 0.0969,
+      "step": 421
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.44961915653679846,
+      "learning_rate": 6.483352252191585e-06,
+      "loss": 0.0591,
+      "step": 422
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4186506337439106,
+      "learning_rate": 6.467879767979764e-06,
+      "loss": 0.062,
+      "step": 423
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3253164042363115,
+      "learning_rate": 6.452391886403767e-06,
+      "loss": 0.0544,
+      "step": 424
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8390011982461838,
+      "learning_rate": 6.436888769924142e-06,
+      "loss": 0.1382,
+      "step": 425
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.795830650877867,
+      "learning_rate": 6.421370581161244e-06,
+      "loss": 0.119,
+      "step": 426
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3216213141007211,
+      "learning_rate": 6.405837482893529e-06,
+      "loss": 0.2146,
+      "step": 427
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7353417059435452,
+      "learning_rate": 6.390289638055851e-06,
+      "loss": 0.16,
+      "step": 428
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4257100364828846,
+      "learning_rate": 6.374727209737743e-06,
+      "loss": 0.0681,
+      "step": 429
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9308011184791537,
+      "learning_rate": 6.3591503611817155e-06,
+      "loss": 0.1233,
+      "step": 430
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.46600768468553494,
+      "learning_rate": 6.343559255781538e-06,
+      "loss": 0.0968,
+      "step": 431
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.45217837773837133,
+      "learning_rate": 6.3279540570805265e-06,
+      "loss": 0.1168,
+      "step": 432
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8122425030745498,
+      "learning_rate": 6.3123349287698345e-06,
+      "loss": 0.1554,
+      "step": 433
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5831822627226863,
+      "learning_rate": 6.296702034686726e-06,
+      "loss": 0.1024,
+      "step": 434
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6510368385175214,
+      "learning_rate": 6.281055538812861e-06,
+      "loss": 0.1557,
+      "step": 435
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5908036903679669,
+      "learning_rate": 6.265395605272581e-06,
+      "loss": 0.1302,
+      "step": 436
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.083073206243553,
+      "learning_rate": 6.249722398331177e-06,
+      "loss": 0.1211,
+      "step": 437
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5113900815850158,
+      "learning_rate": 6.234036082393171e-06,
+      "loss": 0.1038,
+      "step": 438
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4695176010179583,
+      "learning_rate": 6.218336822000598e-06,
+      "loss": 0.0823,
+      "step": 439
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5152907538075204,
+      "learning_rate": 6.202624781831269e-06,
+      "loss": 0.0954,
+      "step": 440
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.31267507550158435,
+      "learning_rate": 6.18690012669705e-06,
+      "loss": 0.0657,
+      "step": 441
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5493800886906555,
+      "learning_rate": 6.171163021542134e-06,
+      "loss": 0.1007,
+      "step": 442
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.056816520305224,
+      "learning_rate": 6.155413631441307e-06,
+      "loss": 0.148,
+      "step": 443
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6045608645180944,
+      "learning_rate": 6.139652121598219e-06,
+      "loss": 0.1264,
+      "step": 444
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5505192451579597,
+      "learning_rate": 6.123878657343648e-06,
+      "loss": 0.1074,
+      "step": 445
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7797223186850629,
+      "learning_rate": 6.108093404133772e-06,
+      "loss": 0.1064,
+      "step": 446
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7391319365614268,
+      "learning_rate": 6.092296527548427e-06,
+      "loss": 0.1364,
+      "step": 447
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3580599720523544,
+      "learning_rate": 6.076488193289375e-06,
+      "loss": 0.0802,
+      "step": 448
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7448422709657808,
+      "learning_rate": 6.060668567178561e-06,
+      "loss": 0.1341,
+      "step": 449
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6370753832372978,
+      "learning_rate": 6.044837815156377e-06,
+      "loss": 0.0976,
+      "step": 450
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.34856966727165745,
+      "learning_rate": 6.028996103279918e-06,
+      "loss": 0.0905,
+      "step": 451
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5308458686033565,
+      "learning_rate": 6.013143597721252e-06,
+      "loss": 0.1205,
+      "step": 452
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.69040451724238,
+      "learning_rate": 5.997280464765655e-06,
+      "loss": 0.1336,
+      "step": 453
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.48804773229436027,
+      "learning_rate": 5.981406870809889e-06,
+      "loss": 0.1038,
+      "step": 454
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5309662342793907,
+      "learning_rate": 5.965522982360441e-06,
+      "loss": 0.1165,
+      "step": 455
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.45807565342272794,
+      "learning_rate": 5.949628966031785e-06,
+      "loss": 0.1204,
+      "step": 456
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.053483997823371,
+      "learning_rate": 5.933724988544632e-06,
+      "loss": 0.0992,
+      "step": 457
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.628667312080432,
+      "learning_rate": 5.9178112167241805e-06,
+      "loss": 0.0693,
+      "step": 458
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3146694384985098,
+      "learning_rate": 5.9018878174983674e-06,
+      "loss": 0.0792,
+      "step": 459
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.774760113025305,
+      "learning_rate": 5.885954957896115e-06,
+      "loss": 0.1119,
+      "step": 460
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7246436390810435,
+      "learning_rate": 5.87001280504558e-06,
+      "loss": 0.1165,
+      "step": 461
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8854769979402363,
+      "learning_rate": 5.854061526172402e-06,
+      "loss": 0.0715,
+      "step": 462
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6484126267384271,
+      "learning_rate": 5.838101288597951e-06,
+      "loss": 0.112,
+      "step": 463
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8365157565396312,
+      "learning_rate": 5.822132259737565e-06,
+      "loss": 0.139,
+      "step": 464
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3293273384065748,
+      "learning_rate": 5.806154607098799e-06,
+      "loss": 0.0763,
+      "step": 465
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6777090984134903,
+      "learning_rate": 5.7901684982796716e-06,
+      "loss": 0.0652,
+      "step": 466
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3589078794021504,
+      "learning_rate": 5.774174100966899e-06,
+      "loss": 0.0743,
+      "step": 467
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3797961215560621,
+      "learning_rate": 5.75817158293414e-06,
+      "loss": 0.0999,
+      "step": 468
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6213340240753514,
+      "learning_rate": 5.742161112040237e-06,
+      "loss": 0.1444,
+      "step": 469
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5452388616304445,
+      "learning_rate": 5.726142856227453e-06,
+      "loss": 0.1048,
+      "step": 470
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2808905141821018,
+      "learning_rate": 5.7101169835197115e-06,
+      "loss": 0.0678,
+      "step": 471
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3891114334553897,
+      "learning_rate": 5.694083662020835e-06,
+      "loss": 0.0782,
+      "step": 472
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8848874048446539,
+      "learning_rate": 5.678043059912776e-06,
+      "loss": 0.179,
+      "step": 473
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6590686572962577,
+      "learning_rate": 5.661995345453867e-06,
+      "loss": 0.0686,
+      "step": 474
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3629048994912652,
+      "learning_rate": 5.645940686977033e-06,
+      "loss": 0.0751,
+      "step": 475
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.35568250345038716,
+      "learning_rate": 5.629879252888046e-06,
+      "loss": 0.0757,
+      "step": 476
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5521471050077407,
+      "learning_rate": 5.613811211663751e-06,
+      "loss": 0.086,
+      "step": 477
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5984229410801276,
+      "learning_rate": 5.597736731850295e-06,
+      "loss": 0.1275,
+      "step": 478
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6060209143197848,
+      "learning_rate": 5.581655982061367e-06,
+      "loss": 0.1195,
+      "step": 479
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3324520039486348,
+      "learning_rate": 5.5655691309764225e-06,
+      "loss": 0.0839,
+      "step": 480
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5127937592468951,
+      "learning_rate": 5.549476347338915e-06,
+      "loss": 0.0996,
+      "step": 481
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.31282947055083743,
+      "learning_rate": 5.533377799954532e-06,
+      "loss": 0.0699,
+      "step": 482
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5965805392065356,
+      "learning_rate": 5.517273657689419e-06,
+      "loss": 0.1073,
+      "step": 483
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.679873280620491,
+      "learning_rate": 5.501164089468406e-06,
+      "loss": 0.1126,
+      "step": 484
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.23598676421594147,
+      "learning_rate": 5.485049264273241e-06,
+      "loss": 0.0546,
+      "step": 485
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.34477926408670995,
+      "learning_rate": 5.4689293511408155e-06,
+      "loss": 0.0724,
+      "step": 486
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.489492463045434,
+      "learning_rate": 5.45280451916139e-06,
+      "loss": 0.0986,
+      "step": 487
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4499802484193189,
+      "learning_rate": 5.43667493747682e-06,
+      "loss": 0.1027,
+      "step": 488
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5272285944848393,
+      "learning_rate": 5.4205407752787884e-06,
+      "loss": 0.0978,
+      "step": 489
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8062238664270278,
+      "learning_rate": 5.404402201807022e-06,
+      "loss": 0.1298,
+      "step": 490
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5706207633505472,
+      "learning_rate": 5.388259386347518e-06,
+      "loss": 0.1054,
+      "step": 491
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6137644564369118,
+      "learning_rate": 5.372112498230771e-06,
+      "loss": 0.0861,
+      "step": 492
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5348038604556673,
+      "learning_rate": 5.355961706829997e-06,
+      "loss": 0.111,
+      "step": 493
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.726680269615648,
+      "learning_rate": 5.339807181559359e-06,
+      "loss": 0.1462,
+      "step": 494
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8348663322329462,
+      "learning_rate": 5.323649091872179e-06,
+      "loss": 0.1371,
+      "step": 495
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7891876906540254,
+      "learning_rate": 5.307487607259175e-06,
+      "loss": 0.1428,
+      "step": 496
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9819820558100029,
+      "learning_rate": 5.291322897246669e-06,
+      "loss": 0.1564,
+      "step": 497
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6935458072615769,
+      "learning_rate": 5.275155131394825e-06,
+      "loss": 0.12,
+      "step": 498
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6661975848806015,
+      "learning_rate": 5.258984479295853e-06,
+      "loss": 0.1184,
+      "step": 499
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5423519400422018,
+      "learning_rate": 5.242811110572243e-06,
+      "loss": 0.0722,
+      "step": 500
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7463419132209343,
+      "learning_rate": 5.226635194874978e-06,
+      "loss": 0.0835,
+      "step": 501
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3914366947605155,
+      "learning_rate": 5.210456901881761e-06,
+      "loss": 0.1508,
+      "step": 502
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4734523496979647,
+      "learning_rate": 5.194276401295231e-06,
+      "loss": 0.0897,
+      "step": 503
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6083711854607663,
+      "learning_rate": 5.1780938628411795e-06,
+      "loss": 0.0839,
+      "step": 504
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4548664811607439,
+      "learning_rate": 5.161909456266781e-06,
+      "loss": 0.0977,
+      "step": 505
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5423324275876441,
+      "learning_rate": 5.145723351338799e-06,
+      "loss": 0.0907,
+      "step": 506
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8035750104285685,
+      "learning_rate": 5.129535717841818e-06,
+      "loss": 0.1195,
+      "step": 507
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.40132851504371736,
+      "learning_rate": 5.11334672557645e-06,
+      "loss": 0.1208,
+      "step": 508
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6114501978559828,
+      "learning_rate": 5.097156544357567e-06,
+      "loss": 0.1301,
+      "step": 509
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.49583161784800067,
+      "learning_rate": 5.080965344012509e-06,
+      "loss": 0.1156,
+      "step": 510
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.36787842036529156,
+      "learning_rate": 5.064773294379302e-06,
+      "loss": 0.0778,
+      "step": 511
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3817167517079446,
+      "learning_rate": 5.048580565304887e-06,
+      "loss": 0.1222,
+      "step": 512
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.22305406444966328,
+      "learning_rate": 5.032387326643331e-06,
+      "loss": 0.0432,
+      "step": 513
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3409084631457501,
+      "learning_rate": 5.016193748254045e-06,
+      "loss": 0.1031,
+      "step": 514
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.29297434360276947,
+      "learning_rate": 5e-06,
+      "loss": 0.0648,
+      "step": 515
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.326977097196938,
+      "learning_rate": 4.983806251745958e-06,
+      "loss": 0.0799,
+      "step": 516
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5782706967667893,
+      "learning_rate": 4.9676126733566705e-06,
+      "loss": 0.1153,
+      "step": 517
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6163249310474644,
+      "learning_rate": 4.951419434695115e-06,
+      "loss": 0.1516,
+      "step": 518
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.35774650461873464,
+      "learning_rate": 4.935226705620699e-06,
+      "loss": 0.1215,
+      "step": 519
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.38764313060720784,
+      "learning_rate": 4.919034655987493e-06,
+      "loss": 0.0817,
+      "step": 520
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3948929977017619,
+      "learning_rate": 4.9028434556424335e-06,
+      "loss": 0.1035,
+      "step": 521
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6165264593739969,
+      "learning_rate": 4.886653274423551e-06,
+      "loss": 0.1222,
+      "step": 522
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6155385207318133,
+      "learning_rate": 4.870464282158184e-06,
+      "loss": 0.1023,
+      "step": 523
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.30092895692862326,
+      "learning_rate": 4.8542766486612035e-06,
+      "loss": 0.08,
+      "step": 524
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7494847442096868,
+      "learning_rate": 4.838090543733222e-06,
+      "loss": 0.1681,
+      "step": 525
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3984866922106728,
+      "learning_rate": 4.821906137158822e-06,
+      "loss": 0.131,
+      "step": 526
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5422271752011945,
+      "learning_rate": 4.805723598704772e-06,
+      "loss": 0.0787,
+      "step": 527
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5069148743800129,
+      "learning_rate": 4.7895430981182415e-06,
+      "loss": 0.0845,
+      "step": 528
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.325494130964241,
+      "learning_rate": 4.773364805125025e-06,
+      "loss": 0.0803,
+      "step": 529
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.271987833994552,
+      "learning_rate": 4.757188889427761e-06,
+      "loss": 0.0659,
+      "step": 530
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5398440656890574,
+      "learning_rate": 4.741015520704148e-06,
+      "loss": 0.1236,
+      "step": 531
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.41418520065692305,
+      "learning_rate": 4.724844868605176e-06,
+      "loss": 0.0673,
+      "step": 532
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.43921876895037953,
+      "learning_rate": 4.708677102753331e-06,
+      "loss": 0.12,
+      "step": 533
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3679446962756676,
+      "learning_rate": 4.6925123927408265e-06,
+      "loss": 0.0894,
+      "step": 534
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.40484965677896917,
+      "learning_rate": 4.6763509081278215e-06,
+      "loss": 0.0787,
+      "step": 535
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3540453646142033,
+      "learning_rate": 4.660192818440642e-06,
+      "loss": 0.0601,
+      "step": 536
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4955006571394118,
+      "learning_rate": 4.644038293170003e-06,
+      "loss": 0.1223,
+      "step": 537
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3990714576406284,
+      "learning_rate": 4.627887501769231e-06,
+      "loss": 0.0932,
+      "step": 538
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3389236780913822,
+      "learning_rate": 4.611740613652485e-06,
+      "loss": 0.0603,
+      "step": 539
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4683123239462747,
+      "learning_rate": 4.59559779819298e-06,
+      "loss": 0.1154,
+      "step": 540
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5513301800384146,
+      "learning_rate": 4.579459224721212e-06,
+      "loss": 0.0788,
+      "step": 541
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6503069347315288,
+      "learning_rate": 4.5633250625231806e-06,
+      "loss": 0.0957,
+      "step": 542
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.866742745789913,
+      "learning_rate": 4.547195480838612e-06,
+      "loss": 0.0933,
+      "step": 543
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.693403208304385,
+      "learning_rate": 4.531070648859186e-06,
+      "loss": 0.1292,
+      "step": 544
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3536703836603506,
+      "learning_rate": 4.51495073572676e-06,
+      "loss": 0.0801,
+      "step": 545
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4008054007992528,
+      "learning_rate": 4.498835910531595e-06,
+      "loss": 0.0848,
+      "step": 546
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.47284322872739853,
+      "learning_rate": 4.482726342310582e-06,
+      "loss": 0.06,
+      "step": 547
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4175256051238674,
+      "learning_rate": 4.4666222000454685e-06,
+      "loss": 0.1091,
+      "step": 548
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4586154908814644,
+      "learning_rate": 4.450523652661086e-06,
+      "loss": 0.0782,
+      "step": 549
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9556460754355709,
+      "learning_rate": 4.434430869023579e-06,
+      "loss": 0.1668,
+      "step": 550
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3707445337609915,
+      "learning_rate": 4.418344017938634e-06,
+      "loss": 0.0892,
+      "step": 551
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4089291338259481,
+      "learning_rate": 4.402263268149707e-06,
+      "loss": 0.083,
+      "step": 552
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4465819266048553,
+      "learning_rate": 4.386188788336251e-06,
+      "loss": 0.0772,
+      "step": 553
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.534441190046031,
+      "learning_rate": 4.370120747111956e-06,
+      "loss": 0.1126,
+      "step": 554
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3109373606701782,
+      "learning_rate": 4.3540593130229695e-06,
+      "loss": 0.081,
+      "step": 555
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3506086444658416,
+      "learning_rate": 4.338004654546136e-06,
+      "loss": 0.0702,
+      "step": 556
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4009813003806628,
+      "learning_rate": 4.3219569400872244e-06,
+      "loss": 0.1124,
+      "step": 557
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3757141177945834,
+      "learning_rate": 4.3059163379791676e-06,
+      "loss": 0.0705,
+      "step": 558
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.40190096513745965,
+      "learning_rate": 4.289883016480291e-06,
+      "loss": 0.0875,
+      "step": 559
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5700004266893263,
+      "learning_rate": 4.27385714377255e-06,
+      "loss": 0.1128,
+      "step": 560
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8271321860711406,
+      "learning_rate": 4.257838887959764e-06,
+      "loss": 0.2003,
+      "step": 561
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5976969809739466,
+      "learning_rate": 4.24182841706586e-06,
+      "loss": 0.0856,
+      "step": 562
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.44157772328502304,
+      "learning_rate": 4.2258258990331015e-06,
+      "loss": 0.0876,
+      "step": 563
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4450141985659714,
+      "learning_rate": 4.209831501720328e-06,
+      "loss": 0.0731,
+      "step": 564
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6582841589746702,
+      "learning_rate": 4.1938453929012014e-06,
+      "loss": 0.1316,
+      "step": 565
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3533324598426827,
+      "learning_rate": 4.177867740262437e-06,
+      "loss": 0.0644,
+      "step": 566
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3307438638772335,
+      "learning_rate": 4.16189871140205e-06,
+      "loss": 0.0681,
+      "step": 567
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.620825674856218,
+      "learning_rate": 4.145938473827598e-06,
+      "loss": 0.1064,
+      "step": 568
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.43238197956559593,
+      "learning_rate": 4.129987194954421e-06,
+      "loss": 0.078,
+      "step": 569
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7632901889924528,
+      "learning_rate": 4.1140450421038865e-06,
+      "loss": 0.1456,
+      "step": 570
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5004370967842584,
+      "learning_rate": 4.098112182501633e-06,
+      "loss": 0.0879,
+      "step": 571
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7096444469082916,
+      "learning_rate": 4.08218878327582e-06,
+      "loss": 0.1555,
+      "step": 572
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5698177130717139,
+      "learning_rate": 4.066275011455369e-06,
+      "loss": 0.126,
+      "step": 573
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5760084319378798,
+      "learning_rate": 4.050371033968216e-06,
+      "loss": 0.0682,
+      "step": 574
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8730902060488447,
+      "learning_rate": 4.034477017639561e-06,
+      "loss": 0.1257,
+      "step": 575
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8184188635015923,
+      "learning_rate": 4.018593129190113e-06,
+      "loss": 0.1016,
+      "step": 576
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6838885473379398,
+      "learning_rate": 4.002719535234346e-06,
+      "loss": 0.0851,
+      "step": 577
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6176165420319377,
+      "learning_rate": 3.98685640227875e-06,
+      "loss": 0.068,
+      "step": 578
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.41160404509320364,
+      "learning_rate": 3.9710038967200825e-06,
+      "loss": 0.1078,
+      "step": 579
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7288198845829278,
+      "learning_rate": 3.955162184843625e-06,
+      "loss": 0.1327,
+      "step": 580
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.41488840926339043,
+      "learning_rate": 3.93933143282144e-06,
+      "loss": 0.1052,
+      "step": 581
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.42115462528877484,
+      "learning_rate": 3.9235118067106255e-06,
+      "loss": 0.0838,
+      "step": 582
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5249101217958358,
+      "learning_rate": 3.907703472451574e-06,
+      "loss": 0.0598,
+      "step": 583
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5453484651752785,
+      "learning_rate": 3.89190659586623e-06,
+      "loss": 0.1203,
+      "step": 584
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.41905508315746437,
+      "learning_rate": 3.8761213426563546e-06,
+      "loss": 0.0749,
+      "step": 585
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7890185726165332,
+      "learning_rate": 3.8603478784017845e-06,
+      "loss": 0.1391,
+      "step": 586
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7715916565540125,
+      "learning_rate": 3.8445863685586946e-06,
+      "loss": 0.1475,
+      "step": 587
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5767704289056564,
+      "learning_rate": 3.828836978457868e-06,
+      "loss": 0.093,
+      "step": 588
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5642051557086234,
+      "learning_rate": 3.8130998733029517e-06,
+      "loss": 0.0927,
+      "step": 589
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8153865148499294,
+      "learning_rate": 3.7973752181687336e-06,
+      "loss": 0.132,
+      "step": 590
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5126391487727465,
+      "learning_rate": 3.7816631779994018e-06,
+      "loss": 0.0971,
+      "step": 591
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.33503983096330675,
+      "learning_rate": 3.7659639176068287e-06,
+      "loss": 0.043,
+      "step": 592
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7429372469335324,
+      "learning_rate": 3.7502776016688234e-06,
+      "loss": 0.1171,
+      "step": 593
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8137940244667058,
+      "learning_rate": 3.734604394727419e-06,
+      "loss": 0.0901,
+      "step": 594
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6816129739354029,
+      "learning_rate": 3.7189444611871383e-06,
+      "loss": 0.1312,
+      "step": 595
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3915209329433682,
+      "learning_rate": 3.703297965313275e-06,
+      "loss": 0.0829,
+      "step": 596
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6138621407262453,
+      "learning_rate": 3.6876650712301654e-06,
+      "loss": 0.1041,
+      "step": 597
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8133074874470426,
+      "learning_rate": 3.6720459429194743e-06,
+      "loss": 0.1473,
+      "step": 598
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5357114819689096,
+      "learning_rate": 3.656440744218464e-06,
+      "loss": 0.1138,
+      "step": 599
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.48803614493782865,
+      "learning_rate": 3.6408496388182857e-06,
+      "loss": 0.0743,
+      "step": 600
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7808333607409286,
+      "learning_rate": 3.6252727902622575e-06,
+      "loss": 0.1627,
+      "step": 601
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8907522302041608,
+      "learning_rate": 3.6097103619441505e-06,
+      "loss": 0.1136,
+      "step": 602
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6361251500086073,
+      "learning_rate": 3.594162517106472e-06,
+      "loss": 0.1272,
+      "step": 603
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8430236560862056,
+      "learning_rate": 3.578629418838757e-06,
+      "loss": 0.1464,
+      "step": 604
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9563954698231266,
+      "learning_rate": 3.5631112300758595e-06,
+      "loss": 0.126,
+      "step": 605
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5546794613455583,
+      "learning_rate": 3.5476081135962335e-06,
+      "loss": 0.0922,
+      "step": 606
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6877638381128309,
+      "learning_rate": 3.532120232020236e-06,
+      "loss": 0.0916,
+      "step": 607
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5036152867163292,
+      "learning_rate": 3.516647747808417e-06,
+      "loss": 0.1354,
+      "step": 608
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.33842916353091324,
+      "learning_rate": 3.5011908232598124e-06,
+      "loss": 0.0646,
+      "step": 609
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9320829595951963,
+      "learning_rate": 3.4857496205102475e-06,
+      "loss": 0.1311,
+      "step": 610
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8024408164610765,
+      "learning_rate": 3.4703243015306314e-06,
+      "loss": 0.1109,
+      "step": 611
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5114684978236648,
+      "learning_rate": 3.4549150281252635e-06,
+      "loss": 0.0941,
+      "step": 612
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4830183987334865,
+      "learning_rate": 3.4395219619301288e-06,
+      "loss": 0.0973,
+      "step": 613
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7966028580680918,
+      "learning_rate": 3.4241452644112085e-06,
+      "loss": 0.116,
+      "step": 614
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7095415092901032,
+      "learning_rate": 3.4087850968627823e-06,
+      "loss": 0.1453,
+      "step": 615
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7133196430221032,
+      "learning_rate": 3.3934416204057396e-06,
+      "loss": 0.1217,
+      "step": 616
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.42557164050608987,
+      "learning_rate": 3.3781149959858894e-06,
+      "loss": 0.0838,
+      "step": 617
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5936729599711326,
+      "learning_rate": 3.3628053843722674e-06,
+      "loss": 0.1305,
+      "step": 618
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6447716869921588,
+      "learning_rate": 3.3475129461554567e-06,
+      "loss": 0.1263,
+      "step": 619
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5007167642691519,
+      "learning_rate": 3.3322378417458985e-06,
+      "loss": 0.0926,
+      "step": 620
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.587047058888338,
+      "learning_rate": 3.3169802313722076e-06,
+      "loss": 0.111,
+      "step": 621
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8108969289508441,
+      "learning_rate": 3.3017402750794976e-06,
+      "loss": 0.1494,
+      "step": 622
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3957587554261609,
+      "learning_rate": 3.2865181327277007e-06,
+      "loss": 0.0785,
+      "step": 623
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5710083551683404,
+      "learning_rate": 3.271313963989886e-06,
+      "loss": 0.1493,
+      "step": 624
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.388057071644006,
+      "learning_rate": 3.2561279283505888e-06,
+      "loss": 0.0846,
+      "step": 625
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9746132297403017,
+      "learning_rate": 3.240960185104137e-06,
+      "loss": 0.1204,
+      "step": 626
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4718093731794653,
+      "learning_rate": 3.2258108933529808e-06,
+      "loss": 0.1121,
+      "step": 627
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5259558631591478,
+      "learning_rate": 3.2106802120060197e-06,
+      "loss": 0.0992,
+      "step": 628
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4221936820795589,
+      "learning_rate": 3.195568299776945e-06,
+      "loss": 0.0966,
+      "step": 629
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7976862815632301,
+      "learning_rate": 3.180475315182563e-06,
+      "loss": 0.122,
+      "step": 630
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7384611149356484,
+      "learning_rate": 3.16540141654114e-06,
+      "loss": 0.0992,
+      "step": 631
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6931742137365803,
+      "learning_rate": 3.1503467619707407e-06,
+      "loss": 0.128,
+      "step": 632
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5620189111954943,
+      "learning_rate": 3.1353115093875676e-06,
+      "loss": 0.0722,
+      "step": 633
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.37304979034092023,
+      "learning_rate": 3.1202958165043053e-06,
+      "loss": 0.0743,
+      "step": 634
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7052205794973762,
+      "learning_rate": 3.1052998408284664e-06,
+      "loss": 0.1205,
+      "step": 635
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5027489633022588,
+      "learning_rate": 3.090323739660742e-06,
+      "loss": 0.0826,
+      "step": 636
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.46204912764624495,
+      "learning_rate": 3.0753676700933448e-06,
+      "loss": 0.1099,
+      "step": 637
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.45300681594685277,
+      "learning_rate": 3.060431789008368e-06,
+      "loss": 0.0837,
+      "step": 638
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.47759829311418994,
+      "learning_rate": 3.045516253076137e-06,
+      "loss": 0.1125,
+      "step": 639
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.33787287232114194,
+      "learning_rate": 3.0306212187535653e-06,
+      "loss": 0.0608,
+      "step": 640
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.46425912363765914,
+      "learning_rate": 3.0157468422825148e-06,
+      "loss": 0.0816,
+      "step": 641
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4900976922127227,
+      "learning_rate": 3.000893279688155e-06,
+      "loss": 0.1401,
+      "step": 642
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8104387688541158,
+      "learning_rate": 2.9860606867773323e-06,
+      "loss": 0.1081,
+      "step": 643
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3535704802647087,
+      "learning_rate": 2.9712492191369245e-06,
+      "loss": 0.065,
+      "step": 644
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9991322797628877,
+      "learning_rate": 2.9564590321322206e-06,
+      "loss": 0.0665,
+      "step": 645
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0422825768722481,
+      "learning_rate": 2.9416902809052817e-06,
+      "loss": 0.1375,
+      "step": 646
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5299253822155083,
+      "learning_rate": 2.9269431203733213e-06,
+      "loss": 0.0961,
+      "step": 647
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4834405310049794,
+      "learning_rate": 2.912217705227075e-06,
+      "loss": 0.0769,
+      "step": 648
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8731749363014457,
+      "learning_rate": 2.8975141899291777e-06,
+      "loss": 0.1258,
+      "step": 649
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.537391607246855,
+      "learning_rate": 2.882832728712551e-06,
+      "loss": 0.0982,
+      "step": 650
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5282659039608084,
+      "learning_rate": 2.868173475578772e-06,
+      "loss": 0.1295,
+      "step": 651
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0433353406984989,
+      "learning_rate": 2.8535365842964713e-06,
+      "loss": 0.1705,
+      "step": 652
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7453860522113966,
+      "learning_rate": 2.838922208399712e-06,
+      "loss": 0.1035,
+      "step": 653
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6212394757546471,
+      "learning_rate": 2.8243305011863843e-06,
+      "loss": 0.1055,
+      "step": 654
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5694658461876471,
+      "learning_rate": 2.8097616157165886e-06,
+      "loss": 0.1018,
+      "step": 655
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.32345558779100947,
+      "learning_rate": 2.7952157048110406e-06,
+      "loss": 0.0489,
+      "step": 656
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5575135094907753,
+      "learning_rate": 2.780692921049465e-06,
+      "loss": 0.1246,
+      "step": 657
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.38517134000721004,
+      "learning_rate": 2.7661934167689887e-06,
+      "loss": 0.0742,
+      "step": 658
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6099408082762448,
+      "learning_rate": 2.751717344062552e-06,
+      "loss": 0.092,
+      "step": 659
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6631133502618448,
+      "learning_rate": 2.7372648547773063e-06,
+      "loss": 0.0935,
+      "step": 660
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7921567001033962,
+      "learning_rate": 2.722836100513027e-06,
+      "loss": 0.1173,
+      "step": 661
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3708726071175215,
+      "learning_rate": 2.7084312326205164e-06,
+      "loss": 0.0586,
+      "step": 662
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8014839693489464,
+      "learning_rate": 2.6940504022000248e-06,
+      "loss": 0.1147,
+      "step": 663
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3666430283803144,
+      "learning_rate": 2.6796937600996587e-06,
+      "loss": 0.0961,
+      "step": 664
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2623498823649957,
+      "learning_rate": 2.665361456913797e-06,
+      "loss": 0.0542,
+      "step": 665
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.44296477230391784,
+      "learning_rate": 2.6510536429815224e-06,
+      "loss": 0.0861,
+      "step": 666
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8430283865858859,
+      "learning_rate": 2.6367704683850293e-06,
+      "loss": 0.1427,
+      "step": 667
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5750145406212099,
+      "learning_rate": 2.622512082948063e-06,
+      "loss": 0.1356,
+      "step": 668
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5605681913755481,
+      "learning_rate": 2.6082786362343377e-06,
+      "loss": 0.0903,
+      "step": 669
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5671797648361183,
+      "learning_rate": 2.594070277545975e-06,
+      "loss": 0.0829,
+      "step": 670
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.453226327889582,
+      "learning_rate": 2.5798871559219362e-06,
+      "loss": 0.0805,
+      "step": 671
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6601329345109225,
+      "learning_rate": 2.5657294201364526e-06,
+      "loss": 0.0955,
+      "step": 672
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.4707117892944517,
+      "learning_rate": 2.551597218697476e-06,
+      "loss": 0.1027,
+      "step": 673
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5998457328358333,
+      "learning_rate": 2.5374906998451094e-06,
+      "loss": 0.1241,
+      "step": 674
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6802417736597872,
+      "learning_rate": 2.5234100115500643e-06,
+      "loss": 0.1153,
+      "step": 675
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4283427944295628,
+      "learning_rate": 2.5093553015120937e-06,
+      "loss": 0.0893,
+      "step": 676
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8127336103022881,
+      "learning_rate": 2.4953267171584573e-06,
+      "loss": 0.1255,
+      "step": 677
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.45310730028913915,
+      "learning_rate": 2.4813244056423692e-06,
+      "loss": 0.1255,
+      "step": 678
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.42178118278585036,
+      "learning_rate": 2.467348513841447e-06,
+      "loss": 0.0759,
+      "step": 679
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.38756636994650595,
+      "learning_rate": 2.4533991883561868e-06,
+      "loss": 0.0895,
+      "step": 680
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.47015869594707566,
+      "learning_rate": 2.439476575508408e-06,
+      "loss": 0.0948,
+      "step": 681
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7040022486529203,
+      "learning_rate": 2.425580821339733e-06,
+      "loss": 0.1352,
+      "step": 682
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6922280792965334,
+      "learning_rate": 2.4117120716100484e-06,
+      "loss": 0.1328,
+      "step": 683
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5427720510108438,
+      "learning_rate": 2.3978704717959777e-06,
+      "loss": 0.1086,
+      "step": 684
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2945662139998699,
+      "learning_rate": 2.38405616708935e-06,
+      "loss": 0.0737,
+      "step": 685
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3933990917790085,
+      "learning_rate": 2.3702693023956853e-06,
+      "loss": 0.103,
+      "step": 686
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.49956299931638654,
+      "learning_rate": 2.356510022332674e-06,
+      "loss": 0.09,
+      "step": 687
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7508054308369841,
+      "learning_rate": 2.342778471228648e-06,
+      "loss": 0.1146,
+      "step": 688
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5271553950947587,
+      "learning_rate": 2.329074793121085e-06,
+      "loss": 0.0895,
+      "step": 689
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7965928687713791,
+      "learning_rate": 2.315399131755081e-06,
+      "loss": 0.1128,
+      "step": 690
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.49009664769117406,
+      "learning_rate": 2.301751630581855e-06,
+      "loss": 0.1134,
+      "step": 691
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.49315966608767386,
+      "learning_rate": 2.2881324327572336e-06,
+      "loss": 0.0987,
+      "step": 692
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5191502870797712,
+      "learning_rate": 2.274541681140159e-06,
+      "loss": 0.0859,
+      "step": 693
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5078356213753658,
+      "learning_rate": 2.260979518291186e-06,
+      "loss": 0.1592,
+      "step": 694
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5207191641837007,
+      "learning_rate": 2.2474460864709825e-06,
+      "loss": 0.1523,
+      "step": 695
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6670319462170284,
+      "learning_rate": 2.233941527638848e-06,
+      "loss": 0.123,
+      "step": 696
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.40816819748101985,
+      "learning_rate": 2.2204659834512095e-06,
+      "loss": 0.0849,
+      "step": 697
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5867536893857979,
+      "learning_rate": 2.207019595260154e-06,
+      "loss": 0.0888,
+      "step": 698
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4389856821033172,
+      "learning_rate": 2.1936025041119268e-06,
+      "loss": 0.0412,
+      "step": 699
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7263951215397981,
+      "learning_rate": 2.1802148507454675e-06,
+      "loss": 0.0867,
+      "step": 700
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5984255990773606,
+      "learning_rate": 2.1668567755909257e-06,
+      "loss": 0.0962,
+      "step": 701
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.36921996516220956,
+      "learning_rate": 2.1535284187681866e-06,
+      "loss": 0.045,
+      "step": 702
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5018032525251166,
+      "learning_rate": 2.140229920085409e-06,
+      "loss": 0.0922,
+      "step": 703
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.48277919928941554,
+      "learning_rate": 2.1269614190375477e-06,
+      "loss": 0.1087,
+      "step": 704
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.405947855722002,
+      "learning_rate": 2.1137230548049042e-06,
+      "loss": 0.0759,
+      "step": 705
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4198312145346275,
+      "learning_rate": 2.1005149662516517e-06,
+      "loss": 0.0861,
+      "step": 706
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7563228799716432,
+      "learning_rate": 2.08733729192439e-06,
+      "loss": 0.0957,
+      "step": 707
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.38788962473556865,
+      "learning_rate": 2.07419017005069e-06,
+      "loss": 0.0732,
+      "step": 708
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5149575322411407,
+      "learning_rate": 2.061073738537635e-06,
+      "loss": 0.0762,
+      "step": 709
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3955849378216403,
+      "learning_rate": 2.0479881349703885e-06,
+      "loss": 0.0821,
+      "step": 710
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7497119030249353,
+      "learning_rate": 2.0349334966107363e-06,
+      "loss": 0.1142,
+      "step": 711
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.46099566896848265,
+      "learning_rate": 2.021909960395661e-06,
+      "loss": 0.0767,
+      "step": 712
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.782998484314079,
+      "learning_rate": 2.0089176629358904e-06,
+      "loss": 0.1022,
+      "step": 713
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5358045007152038,
+      "learning_rate": 1.9959567405144825e-06,
+      "loss": 0.111,
+      "step": 714
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.48557075109442327,
+      "learning_rate": 1.983027329085377e-06,
+      "loss": 0.1026,
+      "step": 715
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5586759145776173,
+      "learning_rate": 1.9701295642719836e-06,
+      "loss": 0.1102,
+      "step": 716
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.44842535485767154,
+      "learning_rate": 1.957263581365749e-06,
+      "loss": 0.0644,
+      "step": 717
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.94541261269077,
+      "learning_rate": 1.944429515324749e-06,
+      "loss": 0.1162,
+      "step": 718
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6891229860681114,
+      "learning_rate": 1.931627500772263e-06,
+      "loss": 0.1033,
+      "step": 719
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.047583625186985,
+      "learning_rate": 1.9188576719953635e-06,
+      "loss": 0.1685,
+      "step": 720
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.936455102385331,
+      "learning_rate": 1.906120162943515e-06,
+      "loss": 0.1058,
+      "step": 721
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1606737870410293,
+      "learning_rate": 1.8934151072271573e-06,
+      "loss": 0.1194,
+      "step": 722
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.607503861810507,
+      "learning_rate": 1.8807426381163151e-06,
+      "loss": 0.0691,
+      "step": 723
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.4364356890469223,
+      "learning_rate": 1.8681028885391905e-06,
+      "loss": 0.0798,
+      "step": 724
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6835322140988652,
+      "learning_rate": 1.8554959910807773e-06,
+      "loss": 0.1322,
+      "step": 725
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9815104193344515,
+      "learning_rate": 1.8429220779814654e-06,
+      "loss": 0.0908,
+      "step": 726
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3210822865580083,
+      "learning_rate": 1.8303812811356503e-06,
+      "loss": 0.1309,
+      "step": 727
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5785209760215473,
+      "learning_rate": 1.81787373209036e-06,
+      "loss": 0.0876,
+      "step": 728
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.4064375335049112,
+      "learning_rate": 1.8053995620438625e-06,
+      "loss": 0.0931,
+      "step": 729
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.37455731813663073,
+      "learning_rate": 1.7929589018443016e-06,
+      "loss": 0.1118,
+      "step": 730
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5159708041577111,
+      "learning_rate": 1.7805518819883134e-06,
+      "loss": 0.088,
+      "step": 731
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7719412037208926,
+      "learning_rate": 1.7681786326196665e-06,
+      "loss": 0.1429,
+      "step": 732
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6732284452708844,
+      "learning_rate": 1.755839283527893e-06,
+      "loss": 0.1136,
+      "step": 733
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5646313484116418,
+      "learning_rate": 1.743533964146924e-06,
+      "loss": 0.1117,
+      "step": 734
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3371832406390867,
+      "learning_rate": 1.7312628035537388e-06,
+      "loss": 0.049,
+      "step": 735
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.40546197646401577,
+      "learning_rate": 1.7190259304670038e-06,
+      "loss": 0.0614,
+      "step": 736
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7659360231859863,
+      "learning_rate": 1.706823473245729e-06,
+      "loss": 0.1069,
+      "step": 737
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5880219130646935,
+      "learning_rate": 1.6946555598879138e-06,
+      "loss": 0.1132,
+      "step": 738
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.27949475630415416,
+      "learning_rate": 1.6825223180292138e-06,
+      "loss": 0.0547,
+      "step": 739
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5695985268937083,
+      "learning_rate": 1.6704238749415958e-06,
+      "loss": 0.1115,
+      "step": 740
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.4261816246363467,
+      "learning_rate": 1.6583603575320002e-06,
+      "loss": 0.0629,
+      "step": 741
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.42622038967695236,
+      "learning_rate": 1.6463318923410183e-06,
+      "loss": 0.0923,
+      "step": 742
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3976860508185511,
+      "learning_rate": 1.6343386055415545e-06,
+      "loss": 0.0902,
+      "step": 743
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9295478081934369,
+      "learning_rate": 1.6223806229375182e-06,
+      "loss": 0.1128,
+      "step": 744
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3716687173126451,
+      "learning_rate": 1.6104580699624839e-06,
+      "loss": 0.0745,
+      "step": 745
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.43052003327195315,
+      "learning_rate": 1.5985710716783936e-06,
+      "loss": 0.0875,
+      "step": 746
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6567640294936006,
+      "learning_rate": 1.5867197527742312e-06,
+      "loss": 0.1606,
+      "step": 747
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5264679447674188,
+      "learning_rate": 1.5749042375647261e-06,
+      "loss": 0.0651,
+      "step": 748
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.877656014114188,
+      "learning_rate": 1.563124649989043e-06,
+      "loss": 0.1476,
+      "step": 749
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5281814939456732,
+      "learning_rate": 1.5513811136094786e-06,
+      "loss": 0.113,
+      "step": 750
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3476887016609292,
+      "learning_rate": 1.5396737516101757e-06,
+      "loss": 0.0665,
+      "step": 751
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5665787485305963,
+      "learning_rate": 1.5280026867958186e-06,
+      "loss": 0.0674,
+      "step": 752
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5506665105163838,
+      "learning_rate": 1.516368041590358e-06,
+      "loss": 0.1022,
+      "step": 753
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7185287494672667,
+      "learning_rate": 1.5047699380357134e-06,
+      "loss": 0.0783,
+      "step": 754
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.4239042152308306,
+      "learning_rate": 1.4932084977905043e-06,
+      "loss": 0.0628,
+      "step": 755
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3959681085581489,
+      "learning_rate": 1.4816838421287693e-06,
+      "loss": 0.0877,
+      "step": 756
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.41661033848056206,
+      "learning_rate": 1.470196091938691e-06,
+      "loss": 0.0951,
+      "step": 757
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8160533610936686,
+      "learning_rate": 1.4587453677213348e-06,
+      "loss": 0.1162,
+      "step": 758
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6466654348738692,
+      "learning_rate": 1.4473317895893773e-06,
+      "loss": 0.0761,
+      "step": 759
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.4556826559116612,
+      "learning_rate": 1.4359554772658551e-06,
+      "loss": 0.058,
+      "step": 760
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3665097084372266,
+      "learning_rate": 1.4246165500828974e-06,
+      "loss": 0.0659,
+      "step": 761
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3547535384550275,
+      "learning_rate": 1.4133151269804873e-06,
+      "loss": 0.0617,
+      "step": 762
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7574630322327562,
+      "learning_rate": 1.4020513265052072e-06,
+      "loss": 0.1073,
+      "step": 763
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.501896848643187,
+      "learning_rate": 1.39082526680899e-06,
+      "loss": 0.1029,
+      "step": 764
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5675894103495152,
+      "learning_rate": 1.3796370656478936e-06,
+      "loss": 0.0824,
+      "step": 765
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0309323915280348,
+      "learning_rate": 1.368486840380851e-06,
+      "loss": 0.1023,
+      "step": 766
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.32109560258885117,
+      "learning_rate": 1.357374707968452e-06,
+      "loss": 0.0752,
+      "step": 767
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6702479147784276,
+      "learning_rate": 1.3463007849717035e-06,
+      "loss": 0.1248,
+      "step": 768
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5467165079831062,
+      "learning_rate": 1.3352651875508204e-06,
+      "loss": 0.0864,
+      "step": 769
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.47073459355099145,
+      "learning_rate": 1.3242680314639995e-06,
+      "loss": 0.1016,
+      "step": 770
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.36995184837167383,
+      "learning_rate": 1.3133094320662e-06,
+      "loss": 0.093,
+      "step": 771
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.39535081463036353,
+      "learning_rate": 1.3023895043079476e-06,
+      "loss": 0.0753,
+      "step": 772
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5734293156151007,
+      "learning_rate": 1.291508362734113e-06,
+      "loss": 0.124,
+      "step": 773
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3462354071238205,
+      "learning_rate": 1.2806661214827286e-06,
+      "loss": 0.0814,
+      "step": 774
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3499217729523727,
+      "learning_rate": 1.2698628942837698e-06,
+      "loss": 0.0787,
+      "step": 775
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3616177005181913,
+      "learning_rate": 1.2590987944579808e-06,
+      "loss": 0.076,
+      "step": 776
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6637701056151413,
+      "learning_rate": 1.2483739349156726e-06,
+      "loss": 0.1121,
+      "step": 777
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3133002950838982,
+      "learning_rate": 1.2376884281555485e-06,
+      "loss": 0.0566,
+      "step": 778
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5249896746398682,
+      "learning_rate": 1.2270423862635188e-06,
+      "loss": 0.082,
+      "step": 779
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7169971351316476,
+      "learning_rate": 1.2164359209115235e-06,
+      "loss": 0.1448,
+      "step": 780
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6354694159681136,
+      "learning_rate": 1.2058691433563675e-06,
+      "loss": 0.1056,
+      "step": 781
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.36229322611138237,
+      "learning_rate": 1.1953421644385444e-06,
+      "loss": 0.0532,
+      "step": 782
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7440990889116061,
+      "learning_rate": 1.184855094581085e-06,
+      "loss": 0.0838,
+      "step": 783
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.44345146358008225,
+      "learning_rate": 1.1744080437883859e-06,
+      "loss": 0.0506,
+      "step": 784
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6745690110949678,
+      "learning_rate": 1.164001121645069e-06,
+      "loss": 0.1162,
+      "step": 785
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3260271984781166,
+      "learning_rate": 1.1536344373148245e-06,
+      "loss": 0.0703,
+      "step": 786
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6149266796401971,
+      "learning_rate": 1.1433080995392614e-06,
+      "loss": 0.116,
+      "step": 787
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.46876880137370996,
+      "learning_rate": 1.133022216636781e-06,
+      "loss": 0.1031,
+      "step": 788
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.553811515271877,
+      "learning_rate": 1.1227768965014246e-06,
+      "loss": 0.1212,
+      "step": 789
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5099010882096217,
+      "learning_rate": 1.1125722466017547e-06,
+      "loss": 0.0966,
+      "step": 790
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.4437081223496812,
+      "learning_rate": 1.102408373979717e-06,
+      "loss": 0.086,
+      "step": 791
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5357333206322139,
+      "learning_rate": 1.092285385249528e-06,
+      "loss": 0.0581,
+      "step": 792
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6152606654522678,
+      "learning_rate": 1.0822033865965503e-06,
+      "loss": 0.1013,
+      "step": 793
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5097304509814814,
+      "learning_rate": 1.0721624837761768e-06,
+      "loss": 0.0735,
+      "step": 794
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4375681758880017,
+      "learning_rate": 1.062162782112729e-06,
+      "loss": 0.086,
+      "step": 795
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3309292309626842,
+      "learning_rate": 1.0522043864983428e-06,
+      "loss": 0.0877,
+      "step": 796
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.47916005764026764,
+      "learning_rate": 1.0422874013918793e-06,
+      "loss": 0.0942,
+      "step": 797
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4944521220442185,
+      "learning_rate": 1.0324119308178166e-06,
+      "loss": 0.0791,
+      "step": 798
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6332692184131137,
+      "learning_rate": 1.0225780783651689e-06,
+      "loss": 0.1008,
+      "step": 799
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5720309815606217,
+      "learning_rate": 1.012785947186397e-06,
+      "loss": 0.0862,
+      "step": 800
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6044429021896905,
+      "learning_rate": 1.0030356399963204e-06,
+      "loss": 0.1115,
+      "step": 801
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.36518784495833523,
+      "learning_rate": 9.933272590710508e-07,
+      "loss": 0.0721,
+      "step": 802
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6836496553279878,
+      "learning_rate": 9.836609062469066e-07,
+      "loss": 0.1074,
+      "step": 803
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5199729931620068,
+      "learning_rate": 9.740366829193587e-07,
+      "loss": 0.1431,
+      "step": 804
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.42167201694835255,
+      "learning_rate": 9.644546900419533e-07,
+      "loss": 0.0852,
+      "step": 805
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.572680728037601,
+      "learning_rate": 9.549150281252633e-07,
+      "loss": 0.1054,
+      "step": 806
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.47298662378855444,
+      "learning_rate": 9.454177972358258e-07,
+      "loss": 0.0973,
+      "step": 807
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8017043505628718,
+      "learning_rate": 9.359630969951012e-07,
+      "loss": 0.143,
+      "step": 808
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9306228479938863,
+      "learning_rate": 9.265510265784189e-07,
+      "loss": 0.1303,
+      "step": 809
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5950675192192034,
+      "learning_rate": 9.171816847139447e-07,
+      "loss": 0.0757,
+      "step": 810
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.4359479937637749,
+      "learning_rate": 9.078551696816434e-07,
+      "loss": 0.0946,
+      "step": 811
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.4436620372001849,
+      "learning_rate": 8.985715793122407e-07,
+      "loss": 0.0724,
+      "step": 812
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3667687933779485,
+      "learning_rate": 8.893310109862102e-07,
+      "loss": 0.0764,
+      "step": 813
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.550835015205489,
+      "learning_rate": 8.801335616327378e-07,
+      "loss": 0.0817,
+      "step": 814
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3821253923796869,
+      "learning_rate": 8.709793277287182e-07,
+      "loss": 0.054,
+      "step": 815
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3377257006076869,
+      "learning_rate": 8.618684052977305e-07,
+      "loss": 0.0455,
+      "step": 816
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.392984261268136,
+      "learning_rate": 8.528008899090412e-07,
+      "loss": 0.0906,
+      "step": 817
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6319458856582114,
+      "learning_rate": 8.437768766765975e-07,
+      "loss": 0.0999,
+      "step": 818
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5997712785383388,
+      "learning_rate": 8.347964602580245e-07,
+      "loss": 0.1051,
+      "step": 819
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.34809517658133093,
+      "learning_rate": 8.258597348536452e-07,
+      "loss": 0.0708,
+      "step": 820
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5256431514802758,
+      "learning_rate": 8.16966794205476e-07,
+      "loss": 0.0838,
+      "step": 821
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.4075759755235258,
+      "learning_rate": 8.081177315962601e-07,
+      "loss": 0.0825,
+      "step": 822
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.76627654803204,
+      "learning_rate": 7.993126398484741e-07,
+      "loss": 0.1359,
+      "step": 823
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3222103505297358,
+      "learning_rate": 7.905516113233652e-07,
+      "loss": 0.0509,
+      "step": 824
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.4714922164002309,
+      "learning_rate": 7.818347379199781e-07,
+      "loss": 0.1038,
+      "step": 825
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5186381434617623,
+      "learning_rate": 7.731621110741871e-07,
+      "loss": 0.0692,
+      "step": 826
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5364703858338294,
+      "learning_rate": 7.645338217577474e-07,
+      "loss": 0.1117,
+      "step": 827
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.4335026860798553,
+      "learning_rate": 7.55949960477328e-07,
+      "loss": 0.0974,
+      "step": 828
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.755360241582875,
+      "learning_rate": 7.474106172735746e-07,
+      "loss": 0.1068,
+      "step": 829
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.48524350549450007,
+      "learning_rate": 7.389158817201541e-07,
+      "loss": 0.083,
+      "step": 830
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.42491558152066805,
+      "learning_rate": 7.304658429228245e-07,
+      "loss": 0.0857,
+      "step": 831
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.4861633779871085,
+      "learning_rate": 7.220605895184946e-07,
+      "loss": 0.0789,
+      "step": 832
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6358496471179419,
+      "learning_rate": 7.13700209674294e-07,
+      "loss": 0.0934,
+      "step": 833
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.44338981475516126,
+      "learning_rate": 7.053847910866513e-07,
+      "loss": 0.0771,
+      "step": 834
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6714763536022414,
+      "learning_rate": 6.971144209803738e-07,
+      "loss": 0.1262,
+      "step": 835
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.41897166828047444,
+      "learning_rate": 6.888891861077301e-07,
+      "loss": 0.0481,
+      "step": 836
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4255996435276327,
+      "learning_rate": 6.807091727475412e-07,
+      "loss": 0.0525,
+      "step": 837
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.474536598786256,
+      "learning_rate": 6.725744667042778e-07,
+      "loss": 0.0923,
+      "step": 838
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4896460204492975,
+      "learning_rate": 6.644851533071556e-07,
+      "loss": 0.1151,
+      "step": 839
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.38425843847511787,
+      "learning_rate": 6.564413174092443e-07,
+      "loss": 0.0755,
+      "step": 840
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3938923759762976,
+      "learning_rate": 6.484430433865785e-07,
+      "loss": 0.1674,
+      "step": 841
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5023810794843392,
+      "learning_rate": 6.404904151372649e-07,
+      "loss": 0.1079,
+      "step": 842
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6340913982384018,
+      "learning_rate": 6.325835160806132e-07,
+      "loss": 0.112,
+      "step": 843
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5225108859572722,
+      "learning_rate": 6.24722429156251e-07,
+      "loss": 0.0989,
+      "step": 844
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2990337541490649,
+      "learning_rate": 6.16907236823262e-07,
+      "loss": 0.0878,
+      "step": 845
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.32466614878115196,
+      "learning_rate": 6.091380210593145e-07,
+      "loss": 0.0719,
+      "step": 846
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5267810752911024,
+      "learning_rate": 6.014148633598055e-07,
+      "loss": 0.125,
+      "step": 847
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.4634858411250408,
+      "learning_rate": 5.937378447370068e-07,
+      "loss": 0.0675,
+      "step": 848
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.41201627699477966,
+      "learning_rate": 5.861070457192081e-07,
+      "loss": 0.0974,
+      "step": 849
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0069609403070547,
+      "learning_rate": 5.785225463498828e-07,
+      "loss": 0.1299,
+      "step": 850
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5473304468749308,
+      "learning_rate": 5.709844261868381e-07,
+      "loss": 0.1042,
+      "step": 851
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3672255838242061,
+      "learning_rate": 5.634927643013899e-07,
+      "loss": 0.0995,
+      "step": 852
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7117682001734832,
+      "learning_rate": 5.560476392775239e-07,
+      "loss": 0.1007,
+      "step": 853
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.310169949837104,
+      "learning_rate": 5.486491292110796e-07,
+      "loss": 0.0942,
+      "step": 854
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.44696554803595046,
+      "learning_rate": 5.412973117089288e-07,
+      "loss": 0.0632,
+      "step": 855
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5791409344657856,
+      "learning_rate": 5.339922638881545e-07,
+      "loss": 0.0993,
+      "step": 856
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5469183216245861,
+      "learning_rate": 5.267340623752554e-07,
+      "loss": 0.1217,
+      "step": 857
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3715118970165225,
+      "learning_rate": 5.195227833053273e-07,
+      "loss": 0.0748,
+      "step": 858
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6057710253253564,
+      "learning_rate": 5.123585023212785e-07,
+      "loss": 0.1089,
+      "step": 859
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.4238894048911,
+      "learning_rate": 5.05241294573024e-07,
+      "loss": 0.0788,
+      "step": 860
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5997134947258506,
+      "learning_rate": 4.981712347167061e-07,
+      "loss": 0.0713,
+      "step": 861
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5534958553812986,
+      "learning_rate": 4.911483969139086e-07,
+      "loss": 0.098,
+      "step": 862
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8066007062748421,
+      "learning_rate": 4.841728548308744e-07,
+      "loss": 0.1272,
+      "step": 863
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3825555555448329,
+      "learning_rate": 4.772446816377408e-07,
+      "loss": 0.0629,
+      "step": 864
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5436519823519242,
+      "learning_rate": 4.7036395000776556e-07,
+      "loss": 0.0661,
+      "step": 865
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6999572333406373,
+      "learning_rate": 4.6353073211656886e-07,
+      "loss": 0.0782,
+      "step": 866
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6665398985207782,
+      "learning_rate": 4.5674509964137136e-07,
+      "loss": 0.1244,
+      "step": 867
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4102524219578372,
+      "learning_rate": 4.5000712376024826e-07,
+      "loss": 0.0725,
+      "step": 868
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.31539447100958523,
+      "learning_rate": 4.4331687515137614e-07,
+      "loss": 0.0389,
+      "step": 869
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.32079003707334564,
+      "learning_rate": 4.3667442399229985e-07,
+      "loss": 0.0723,
+      "step": 870
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7334622696512892,
+      "learning_rate": 4.30079839959186e-07,
+      "loss": 0.1115,
+      "step": 871
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5104145536374513,
+      "learning_rate": 4.2353319222610265e-07,
+      "loss": 0.1117,
+      "step": 872
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.407631950449968,
+      "learning_rate": 4.1703454946428635e-07,
+      "loss": 0.098,
+      "step": 873
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5052468200921778,
+      "learning_rate": 4.1058397984142405e-07,
+      "loss": 0.1079,
+      "step": 874
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5647002734695262,
+      "learning_rate": 4.041815510209396e-07,
+      "loss": 0.0969,
+      "step": 875
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9680459120711749,
+      "learning_rate": 3.9782733016128006e-07,
+      "loss": 0.1692,
+      "step": 876
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5588469027832113,
+      "learning_rate": 3.9152138391521766e-07,
+      "loss": 0.095,
+      "step": 877
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3323981905313235,
+      "learning_rate": 3.852637784291424e-07,
+      "loss": 0.0641,
+      "step": 878
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8081738368564643,
+      "learning_rate": 3.790545793423761e-07,
+      "loss": 0.1244,
+      "step": 879
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6412500408187927,
+      "learning_rate": 3.728938517864794e-07,
+      "loss": 0.1266,
+      "step": 880
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.28687232827929193,
+      "learning_rate": 3.667816603845681e-07,
+      "loss": 0.0747,
+      "step": 881
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.4982165667621581,
+      "learning_rate": 3.60718069250639e-07,
+      "loss": 0.1053,
+      "step": 882
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5464715501708446,
+      "learning_rate": 3.547031419888919e-07,
+      "loss": 0.1301,
+      "step": 883
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5942245125339878,
+      "learning_rate": 3.4873694169306915e-07,
+      "loss": 0.0986,
+      "step": 884
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5399557216992824,
+      "learning_rate": 3.4281953094578877e-07,
+      "loss": 0.0904,
+      "step": 885
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7068041293977915,
+      "learning_rate": 3.369509718178887e-07,
+      "loss": 0.1054,
+      "step": 886
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5820835803293782,
+      "learning_rate": 3.3113132586777786e-07,
+      "loss": 0.0814,
+      "step": 887
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.540465794297693,
+      "learning_rate": 3.2536065414078724e-07,
+      "loss": 0.1207,
+      "step": 888
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7907977007084812,
+      "learning_rate": 3.196390171685343e-07,
+      "loss": 0.1397,
+      "step": 889
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7159037109972106,
+      "learning_rate": 3.1396647496828245e-07,
+      "loss": 0.1585,
+      "step": 890
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.38466650660591445,
+      "learning_rate": 3.0834308704231485e-07,
+      "loss": 0.0882,
+      "step": 891
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8151401337946305,
+      "learning_rate": 3.0276891237731085e-07,
+      "loss": 0.1342,
+      "step": 892
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.47364592790571436,
+      "learning_rate": 2.97244009443724e-07,
+      "loss": 0.0863,
+      "step": 893
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.30711784182012725,
+      "learning_rate": 2.917684361951728e-07,
+      "loss": 0.0549,
+      "step": 894
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.47101524283620005,
+      "learning_rate": 2.8634225006782867e-07,
+      "loss": 0.0903,
+      "step": 895
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.30967047878137277,
+      "learning_rate": 2.809655079798179e-07,
+      "loss": 0.0533,
+      "step": 896
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5143493600993974,
+      "learning_rate": 2.75638266330619e-07,
+      "loss": 0.088,
+      "step": 897
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6743754796337982,
+      "learning_rate": 2.7036058100047723e-07,
+      "loss": 0.1103,
+      "step": 898
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6095538912751249,
+      "learning_rate": 2.65132507349814e-07,
+      "loss": 0.0781,
+      "step": 899
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3958113201950967,
+      "learning_rate": 2.599541002186479e-07,
+      "loss": 0.0779,
+      "step": 900
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.047123808357352,
+      "learning_rate": 2.5482541392601924e-07,
+      "loss": 0.0555,
+      "step": 901
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4420409227261539,
+      "learning_rate": 2.497465022694207e-07,
+      "loss": 0.0777,
+      "step": 902
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5935850991221072,
+      "learning_rate": 2.447174185242324e-07,
+      "loss": 0.0846,
+      "step": 903
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5221496297926413,
+      "learning_rate": 2.397382154431621e-07,
+      "loss": 0.0739,
+      "step": 904
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5415360874192054,
+      "learning_rate": 2.3480894525569564e-07,
+      "loss": 0.0866,
+      "step": 905
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.4109552316884023,
+      "learning_rate": 2.2992965966754378e-07,
+      "loss": 0.083,
+      "step": 906
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.909885268355602,
+      "learning_rate": 2.251004098601034e-07,
+      "loss": 0.0825,
+      "step": 907
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.4658964372668037,
+      "learning_rate": 2.2032124648992015e-07,
+      "loss": 0.1046,
+      "step": 908
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.33553011782146736,
+      "learning_rate": 2.1559221968815547e-07,
+      "loss": 0.0899,
+      "step": 909
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.4834566512135711,
+      "learning_rate": 2.109133790600648e-07,
+      "loss": 0.0851,
+      "step": 910
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6260157288294408,
+      "learning_rate": 2.062847736844703e-07,
+      "loss": 0.1829,
+      "step": 911
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.4284786562742925,
+      "learning_rate": 2.0170645211325335e-07,
+      "loss": 0.0632,
+      "step": 912
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.42840539869243016,
+      "learning_rate": 1.9717846237084005e-07,
+      "loss": 0.0902,
+      "step": 913
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.39495047949288664,
+      "learning_rate": 1.9270085195370048e-07,
+      "loss": 0.0784,
+      "step": 914
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6895093352181992,
+      "learning_rate": 1.8827366782984913e-07,
+      "loss": 0.1401,
+      "step": 915
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4341918274282859,
+      "learning_rate": 1.838969564383525e-07,
+      "loss": 0.0991,
+      "step": 916
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7336846337260968,
+      "learning_rate": 1.7957076368884274e-07,
+      "loss": 0.1068,
+      "step": 917
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.33843907932125694,
+      "learning_rate": 1.7529513496103322e-07,
+      "loss": 0.098,
+      "step": 918
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.641426534079293,
+      "learning_rate": 1.7107011510424766e-07,
+      "loss": 0.1121,
+      "step": 919
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6088894766654794,
+      "learning_rate": 1.6689574843694433e-07,
+      "loss": 0.049,
+      "step": 920
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7638243262874345,
+      "learning_rate": 1.6277207874625444e-07,
+      "loss": 0.1407,
+      "step": 921
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.38273022218175334,
+      "learning_rate": 1.5869914928752117e-07,
+      "loss": 0.0869,
+      "step": 922
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.47809969343613545,
+      "learning_rate": 1.546770027838479e-07,
+      "loss": 0.1059,
+      "step": 923
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3328505710537316,
+      "learning_rate": 1.5070568142564912e-07,
+      "loss": 0.0824,
+      "step": 924
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3412516515646284,
+      "learning_rate": 1.4678522687020414e-07,
+      "loss": 0.0673,
+      "step": 925
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3582880986734356,
+      "learning_rate": 1.4291568024122848e-07,
+      "loss": 0.0813,
+      "step": 926
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.4627890159848611,
+      "learning_rate": 1.390970821284343e-07,
+      "loss": 0.0972,
+      "step": 927
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.4058521199725231,
+      "learning_rate": 1.3532947258710905e-07,
+      "loss": 0.0477,
+      "step": 928
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5285869257912966,
+      "learning_rate": 1.3161289113769405e-07,
+      "loss": 0.069,
+      "step": 929
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.886532533745455,
+      "learning_rate": 1.2794737676536993e-07,
+      "loss": 0.1483,
+      "step": 930
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.48656899155350514,
+      "learning_rate": 1.2433296791964754e-07,
+      "loss": 0.1062,
+      "step": 931
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7968133314235069,
+      "learning_rate": 1.2076970251396593e-07,
+      "loss": 0.1477,
+      "step": 932
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5725663468018927,
+      "learning_rate": 1.1725761792529378e-07,
+      "loss": 0.1135,
+      "step": 933
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5170245470451504,
+      "learning_rate": 1.1379675099373489e-07,
+      "loss": 0.1299,
+      "step": 934
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6572490103667149,
+      "learning_rate": 1.1038713802214718e-07,
+      "loss": 0.1291,
+      "step": 935
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6625741117878644,
+      "learning_rate": 1.0702881477575589e-07,
+      "loss": 0.0883,
+      "step": 936
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.4741783508324799,
+      "learning_rate": 1.0372181648178436e-07,
+      "loss": 0.1157,
+      "step": 937
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.35230005784167795,
+      "learning_rate": 1.004661778290783e-07,
+      "loss": 0.059,
+      "step": 938
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.38374367424824235,
+      "learning_rate": 9.726193296774767e-08,
+      "loss": 0.075,
+      "step": 939
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3103326494714687,
+      "learning_rate": 9.410911550880474e-08,
+      "loss": 0.0384,
+      "step": 940
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.32360309647427055,
+      "learning_rate": 9.100775852381227e-08,
+      "loss": 0.0957,
+      "step": 941
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.4687389813884961,
+      "learning_rate": 8.795789454453862e-08,
+      "loss": 0.0634,
+      "step": 942
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6463975180537007,
+      "learning_rate": 8.495955556261204e-08,
+      "loss": 0.1049,
+      "step": 943
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3725035351846233,
+      "learning_rate": 8.201277302919086e-08,
+      "loss": 0.0687,
+      "step": 944
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5327315777783594,
+      "learning_rate": 7.911757785462882e-08,
+      "loss": 0.0863,
+      "step": 945
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5649959636734481,
+      "learning_rate": 7.627400040815414e-08,
+      "loss": 0.1035,
+      "step": 946
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1124113720729487,
+      "learning_rate": 7.34820705175482e-08,
+      "loss": 0.1088,
+      "step": 947
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2575662592302751,
+      "learning_rate": 7.074181746883402e-08,
+      "loss": 0.0597,
+      "step": 948
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4217917378062126,
+      "learning_rate": 6.805327000596995e-08,
+      "loss": 0.1233,
+      "step": 949
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5900202418657501,
+      "learning_rate": 6.54164563305465e-08,
+      "loss": 0.1005,
+      "step": 950
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.4903148765855846,
+      "learning_rate": 6.283140410149213e-08,
+      "loss": 0.102,
+      "step": 951
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6283680733361539,
+      "learning_rate": 6.029814043478022e-08,
+      "loss": 0.0986,
+      "step": 952
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6437189610067745,
+      "learning_rate": 5.781669190314809e-08,
+      "loss": 0.1144,
+      "step": 953
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.40301108037963757,
+      "learning_rate": 5.538708453581787e-08,
+      "loss": 0.0823,
+      "step": 954
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.46077593051803917,
+      "learning_rate": 5.3009343818219985e-08,
+      "loss": 0.1069,
+      "step": 955
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1633111722577614,
+      "learning_rate": 5.068349469173006e-08,
+      "loss": 0.1768,
+      "step": 956
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5676334727677624,
+      "learning_rate": 4.840956155340415e-08,
+      "loss": 0.1069,
+      "step": 957
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1418603533214473,
+      "learning_rate": 4.618756825572612e-08,
+      "loss": 0.1313,
+      "step": 958
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4998266333127885,
+      "learning_rate": 4.40175381063529e-08,
+      "loss": 0.1141,
+      "step": 959
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.49451116960113756,
+      "learning_rate": 4.189949386787462e-08,
+      "loss": 0.1055,
+      "step": 960
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6605021406745498,
+      "learning_rate": 3.9833457757572636e-08,
+      "loss": 0.1312,
+      "step": 961
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7611722108034722,
+      "learning_rate": 3.781945144718912e-08,
+      "loss": 0.1127,
+      "step": 962
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4052079041989755,
+      "learning_rate": 3.585749606269562e-08,
+      "loss": 0.0762,
+      "step": 963
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2812946717625918,
+      "learning_rate": 3.394761218407705e-08,
+      "loss": 0.0453,
+      "step": 964
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6666698061897918,
+      "learning_rate": 3.2089819845111946e-08,
+      "loss": 0.0599,
+      "step": 965
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6616369308089431,
+      "learning_rate": 3.0284138533160924e-08,
+      "loss": 0.099,
+      "step": 966
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5595168583725473,
+      "learning_rate": 2.8530587188968508e-08,
+      "loss": 0.1096,
+      "step": 967
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3862726620479047,
+      "learning_rate": 2.6829184206457194e-08,
+      "loss": 0.064,
+      "step": 968
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7776881783133217,
+      "learning_rate": 2.5179947432540376e-08,
+      "loss": 0.1087,
+      "step": 969
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8070875336917069,
+      "learning_rate": 2.358289416693027e-08,
+      "loss": 0.1547,
+      "step": 970
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4034578997229353,
+      "learning_rate": 2.2038041161960288e-08,
+      "loss": 0.0961,
+      "step": 971
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5057471723242044,
+      "learning_rate": 2.0545404622407396e-08,
+      "loss": 0.1123,
+      "step": 972
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.35190909630950856,
+      "learning_rate": 1.91050002053228e-08,
+      "loss": 0.0635,
+      "step": 973
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5187828907818753,
+      "learning_rate": 1.7716843019867646e-08,
+      "loss": 0.0856,
+      "step": 974
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.394444506384258,
+      "learning_rate": 1.6380947627153143e-08,
+      "loss": 0.0623,
+      "step": 975
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5285270173606262,
+      "learning_rate": 1.509732804009012e-08,
+      "loss": 0.1073,
+      "step": 976
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.37338286137353993,
+      "learning_rate": 1.386599772324082e-08,
+      "loss": 0.082,
+      "step": 977
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.43264207890271816,
+      "learning_rate": 1.268696959267679e-08,
+      "loss": 0.1035,
+      "step": 978
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7496118580227054,
+      "learning_rate": 1.156025601584676e-08,
+      "loss": 0.1461,
+      "step": 979
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.36412351519153907,
+      "learning_rate": 1.0485868811441757e-08,
+      "loss": 0.1018,
+      "step": 980
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5314038519295458,
+      "learning_rate": 9.463819249275751e-09,
+      "loss": 0.1252,
+      "step": 981
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.48112137285779566,
+      "learning_rate": 8.494118050164646e-09,
+      "loss": 0.1104,
+      "step": 982
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7189714211724794,
+      "learning_rate": 7.576775385815249e-09,
+      "loss": 0.1468,
+      "step": 983
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.40292170595375787,
+      "learning_rate": 6.711800878718144e-09,
+      "loss": 0.0745,
+      "step": 984
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.44643651712628796,
+      "learning_rate": 5.899203602046655e-09,
+      "loss": 0.0931,
+      "step": 985
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7086989217011592,
+      "learning_rate": 5.138992079561367e-09,
+      "loss": 0.1206,
+      "step": 986
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6598387711819119,
+      "learning_rate": 4.431174285521866e-09,
+      "loss": 0.1148,
+      "step": 987
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6104258040113265,
+      "learning_rate": 3.775757644601808e-09,
+      "loss": 0.0875,
+      "step": 988
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.48595597670970386,
+      "learning_rate": 3.1727490318111953e-09,
+      "loss": 0.1145,
+      "step": 989
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6209649139140617,
+      "learning_rate": 2.6221547724253337e-09,
+      "loss": 0.1113,
+      "step": 990
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.410496344724791,
+      "learning_rate": 2.1239806419176556e-09,
+      "loss": 0.059,
+      "step": 991
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7022484208897792,
+      "learning_rate": 1.6782318658992159e-09,
+      "loss": 0.0865,
+      "step": 992
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.4556693550827079,
+      "learning_rate": 1.2849131200631804e-09,
+      "loss": 0.0795,
+      "step": 993
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.31377306177553776,
+      "learning_rate": 9.440285301370865e-10,
+      "loss": 0.0617,
+      "step": 994
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5835806283311138,
+      "learning_rate": 6.555816718389896e-10,
+      "loss": 0.1197,
+      "step": 995
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6526393593582267,
+      "learning_rate": 4.1957557084082447e-10,
+      "loss": 0.1179,
+      "step": 996
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6533521512496714,
+      "learning_rate": 2.360127027339898e-10,
+      "loss": 0.0879,
+      "step": 997
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6049139152852736,
+      "learning_rate": 1.0489499300603279e-10,
+      "loss": 0.0594,
+      "step": 998
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7328945103620341,
+      "learning_rate": 2.622381702066523e-11,
+      "loss": 0.1141,
+      "step": 999
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.759905161698802,
+      "learning_rate": 0.0,
+      "loss": 0.1072,
+      "step": 1000
+    },
+    {
+      "epoch": 1.0,
+      "step": 1000,
+      "total_flos": 38113260066816.0,
+      "train_loss": 0.11589447382092476,
+      "train_runtime": 2391.3181,
+      "train_samples_per_second": 0.836,
+      "train_steps_per_second": 0.418
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 1000,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 50000,
+  "total_flos": 38113260066816.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}
diff --git a/llavaqwen2-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split02_all_mm_tune_olora256_512_llm/README.md b/llavaqwen2-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split02_all_mm_tune_olora256_512_llm/README.md
new file mode 100644
index 0000000000000000000000000000000000000000..84ec7fb74b98cfd0816d69a85e53a402a3fe9252
--- /dev/null
+++ b/llavaqwen2-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split02_all_mm_tune_olora256_512_llm/README.md
@@ -0,0 +1,10 @@
+---
+library_name: peft
+---
+## Training procedure
+
+### Framework versions
+
+- PEFT 0.4.0
+
+- PEFT 0.4.0
diff --git a/llavaqwen2-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split02_all_mm_tune_olora256_512_llm/adapter_config.json b/llavaqwen2-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split02_all_mm_tune_olora256_512_llm/adapter_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..c4919754736b9e4bdfbc56e33ca2c7472a8a8c96
--- /dev/null
+++ b/llavaqwen2-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split02_all_mm_tune_olora256_512_llm/adapter_config.json
@@ -0,0 +1,26 @@
+{
+  "auto_mapping": null,
+  "base_model_name_or_path": "lmms-lab/LLaVA-Video-7B-Qwen2",
+  "bias": "none",
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": "olora",
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "lora_alpha": 512,
+  "lora_dropout": 0.05,
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "r": 256,
+  "revision": null,
+  "target_modules": [
+    "v_proj",
+    "gate_proj",
+    "q_proj",
+    "k_proj",
+    "up_proj",
+    "o_proj",
+    "down_proj"
+  ],
+  "task_type": "CAUSAL_LM"
+}
\ No newline at end of file
diff --git a/llavaqwen2-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split02_all_mm_tune_olora256_512_llm/adapter_model.bin b/llavaqwen2-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split02_all_mm_tune_olora256_512_llm/adapter_model.bin
new file mode 100644
index 0000000000000000000000000000000000000000..72fbc4a2ed0e5ba7ba7bbfbf939cab44e6632f6e
--- /dev/null
+++ b/llavaqwen2-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split02_all_mm_tune_olora256_512_llm/adapter_model.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c000cb164df104a9753016cf678b8b51addb141609a1fbb90e3a9c42e17ffde7
+size 1384057050
diff --git a/llavaqwen2-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split02_all_mm_tune_olora256_512_llm/config.json b/llavaqwen2-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split02_all_mm_tune_olora256_512_llm/config.json
new file mode 100644
index 0000000000000000000000000000000000000000..1a67f02d0063c3de7740207b9ab2a3eb7be1cbe3
--- /dev/null
+++ b/llavaqwen2-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split02_all_mm_tune_olora256_512_llm/config.json
@@ -0,0 +1,221 @@
+{
+  "_name_or_path": "lmms-lab/LLaVA-Video-7B-Qwen2",
+  "add_faster_video": false,
+  "add_time_instruction": true,
+  "architectures": [
+    "LlavaQwenForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 151643,
+  "eos_token_id": 151645,
+  "faster_token_stride": 10,
+  "force_sample": true,
+  "hidden_act": "silu",
+  "hidden_size": 3584,
+  "ignore_index": -100,
+  "image_aspect_ratio": "anyres_max_9",
+  "image_crop_resolution": null,
+  "image_grid_pinpoints": [
+    [
+      384,
+      384
+    ],
+    [
+      384,
+      768
+    ],
+    [
+      384,
+      1152
+    ],
+    [
+      384,
+      1536
+    ],
+    [
+      384,
+      1920
+    ],
+    [
+      384,
+      2304
+    ],
+    [
+      768,
+      384
+    ],
+    [
+      768,
+      768
+    ],
+    [
+      768,
+      1152
+    ],
+    [
+      768,
+      1536
+    ],
+    [
+      768,
+      1920
+    ],
+    [
+      768,
+      2304
+    ],
+    [
+      1152,
+      384
+    ],
+    [
+      1152,
+      768
+    ],
+    [
+      1152,
+      1152
+    ],
+    [
+      1152,
+      1536
+    ],
+    [
+      1152,
+      1920
+    ],
+    [
+      1152,
+      2304
+    ],
+    [
+      1536,
+      384
+    ],
+    [
+      1536,
+      768
+    ],
+    [
+      1536,
+      1152
+    ],
+    [
+      1536,
+      1536
+    ],
+    [
+      1536,
+      1920
+    ],
+    [
+      1536,
+      2304
+    ],
+    [
+      1920,
+      384
+    ],
+    [
+      1920,
+      768
+    ],
+    [
+      1920,
+      1152
+    ],
+    [
+      1920,
+      1536
+    ],
+    [
+      1920,
+      1920
+    ],
+    [
+      1920,
+      2304
+    ],
+    [
+      2304,
+      384
+    ],
+    [
+      2304,
+      768
+    ],
+    [
+      2304,
+      1152
+    ],
+    [
+      2304,
+      1536
+    ],
+    [
+      2304,
+      1920
+    ],
+    [
+      2304,
+      2304
+    ]
+  ],
+  "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": "grid",
+  "mm_patch_merge_type": "spatial_unpad",
+  "mm_projector_lr": 2e-05,
+  "mm_projector_type": "mlp2x_gelu",
+  "mm_resampler_type": null,
+  "mm_spatial_pool_mode": "bilinear",
+  "mm_spatial_pool_stride": 2,
+  "mm_tunable_parts": "mm_vision_tower,mm_mlp_adapter,mm_language_model",
+  "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",
+  "num_attention_heads": 28,
+  "num_hidden_layers": 28,
+  "num_key_value_heads": 4,
+  "pos_skipping_range": 4096,
+  "projector_hidden_act": "gelu",
+  "rms_norm_eps": 1e-06,
+  "rope_scaling": null,
+  "rope_theta": 1000000.0,
+  "sliding_window": 131072,
+  "text_config": {
+    "model_type": "llama"
+  },
+  "tie_word_embeddings": false,
+  "tokenizer_model_max_length": 32768,
+  "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_config": {
+    "hidden_size": 1024,
+    "image_size": 336,
+    "intermediate_size": 4096,
+    "model_type": "clip_vision_model",
+    "num_attention_heads": 16,
+    "num_hidden_layers": 24,
+    "patch_size": 14,
+    "projection_dim": 768,
+    "vocab_size": 32000
+  },
+  "vision_feature_layer": -2,
+  "vision_feature_select_strategy": "default",
+  "vision_tower_pretrained": null
+}
diff --git a/llavaqwen2-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split02_all_mm_tune_olora256_512_llm/generation_config.json b/llavaqwen2-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split02_all_mm_tune_olora256_512_llm/generation_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..19a297221acb87418d4388a3decef2282c6d7316
--- /dev/null
+++ b/llavaqwen2-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split02_all_mm_tune_olora256_512_llm/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/llavaqwen2-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split02_all_mm_tune_olora256_512_llm/non_lora_trainables.bin b/llavaqwen2-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split02_all_mm_tune_olora256_512_llm/non_lora_trainables.bin
new file mode 100644
index 0000000000000000000000000000000000000000..d0650a24e63145778fe77dc7c6625b87a067eab7
--- /dev/null
+++ b/llavaqwen2-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split02_all_mm_tune_olora256_512_llm/non_lora_trainables.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7dad0f447903e090b103776926817a34a460998c74753362a5680b0f8b59028a
+size 33964208
diff --git a/llavaqwen2-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split02_all_mm_tune_olora256_512_llm/trainer_state.json b/llavaqwen2-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split02_all_mm_tune_olora256_512_llm/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..5848f7926361e1f5f77a5f31915c9b98745fc882
--- /dev/null
+++ b/llavaqwen2-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split02_all_mm_tune_olora256_512_llm/trainer_state.json
@@ -0,0 +1,83106 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 11868,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.378236731435547,
+      "learning_rate": 2.8011204481792718e-08,
+      "loss": 0.6801,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.722609869464893,
+      "learning_rate": 5.6022408963585437e-08,
+      "loss": 0.7218,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.847724876204175,
+      "learning_rate": 8.403361344537815e-08,
+      "loss": 0.4376,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.5297435835569315,
+      "learning_rate": 1.1204481792717087e-07,
+      "loss": 0.4518,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.4594532441008154,
+      "learning_rate": 1.400560224089636e-07,
+      "loss": 0.5368,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.342484129298085,
+      "learning_rate": 1.680672268907563e-07,
+      "loss": 0.7457,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.7258716398365765,
+      "learning_rate": 1.9607843137254904e-07,
+      "loss": 0.4883,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.374800919761394,
+      "learning_rate": 2.2408963585434175e-07,
+      "loss": 0.6331,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.931840058665886,
+      "learning_rate": 2.5210084033613445e-07,
+      "loss": 0.5522,
+      "step": 9
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.5580990973108273,
+      "learning_rate": 2.801120448179272e-07,
+      "loss": 0.4187,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.7699991769572647,
+      "learning_rate": 3.081232492997199e-07,
+      "loss": 0.7044,
+      "step": 11
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.8607488429310421,
+      "learning_rate": 3.361344537815126e-07,
+      "loss": 0.4766,
+      "step": 12
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.314201389132091,
+      "learning_rate": 3.641456582633054e-07,
+      "loss": 0.5408,
+      "step": 13
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.7867487593176063,
+      "learning_rate": 3.921568627450981e-07,
+      "loss": 0.4655,
+      "step": 14
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.083342201941341,
+      "learning_rate": 4.201680672268908e-07,
+      "loss": 0.5379,
+      "step": 15
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.98994471655251,
+      "learning_rate": 4.481792717086835e-07,
+      "loss": 0.762,
+      "step": 16
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.3005144423857664,
+      "learning_rate": 4.7619047619047623e-07,
+      "loss": 0.6111,
+      "step": 17
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.28068264709896,
+      "learning_rate": 5.042016806722689e-07,
+      "loss": 0.879,
+      "step": 18
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.669347349205939,
+      "learning_rate": 5.322128851540616e-07,
+      "loss": 0.4334,
+      "step": 19
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.274896716929672,
+      "learning_rate": 5.602240896358544e-07,
+      "loss": 0.6438,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.8954774848253007,
+      "learning_rate": 5.882352941176471e-07,
+      "loss": 0.2637,
+      "step": 21
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.5306507550466755,
+      "learning_rate": 6.162464985994398e-07,
+      "loss": 0.9185,
+      "step": 22
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.5144969672407966,
+      "learning_rate": 6.442577030812325e-07,
+      "loss": 0.7177,
+      "step": 23
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.5397609499677072,
+      "learning_rate": 6.722689075630252e-07,
+      "loss": 0.4216,
+      "step": 24
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.147361661302807,
+      "learning_rate": 7.002801120448179e-07,
+      "loss": 0.6331,
+      "step": 25
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.0814448143364994,
+      "learning_rate": 7.282913165266108e-07,
+      "loss": 0.7564,
+      "step": 26
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.02008979730368,
+      "learning_rate": 7.563025210084034e-07,
+      "loss": 0.4948,
+      "step": 27
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.8232678476497801,
+      "learning_rate": 7.843137254901962e-07,
+      "loss": 0.5,
+      "step": 28
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.3284249466759057,
+      "learning_rate": 8.123249299719889e-07,
+      "loss": 0.5935,
+      "step": 29
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.223237815984025,
+      "learning_rate": 8.403361344537816e-07,
+      "loss": 0.4647,
+      "step": 30
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.1156301417954375,
+      "learning_rate": 8.683473389355742e-07,
+      "loss": 0.9354,
+      "step": 31
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.3338272351732452,
+      "learning_rate": 8.96358543417367e-07,
+      "loss": 0.386,
+      "step": 32
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.214014387713159,
+      "learning_rate": 9.243697478991598e-07,
+      "loss": 0.8075,
+      "step": 33
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.77956706524366,
+      "learning_rate": 9.523809523809525e-07,
+      "loss": 0.7149,
+      "step": 34
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.9983819491132484,
+      "learning_rate": 9.80392156862745e-07,
+      "loss": 0.5854,
+      "step": 35
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.6280142558712318,
+      "learning_rate": 1.0084033613445378e-06,
+      "loss": 0.4886,
+      "step": 36
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.1777859145289016,
+      "learning_rate": 1.0364145658263308e-06,
+      "loss": 0.8269,
+      "step": 37
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.9289542432724536,
+      "learning_rate": 1.0644257703081233e-06,
+      "loss": 0.641,
+      "step": 38
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.555043912185022,
+      "learning_rate": 1.092436974789916e-06,
+      "loss": 0.4924,
+      "step": 39
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.1517709724637335,
+      "learning_rate": 1.1204481792717088e-06,
+      "loss": 0.3612,
+      "step": 40
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.3090280563058787,
+      "learning_rate": 1.1484593837535015e-06,
+      "loss": 0.3707,
+      "step": 41
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.5348881241692107,
+      "learning_rate": 1.1764705882352942e-06,
+      "loss": 0.4788,
+      "step": 42
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.1642073571650549,
+      "learning_rate": 1.204481792717087e-06,
+      "loss": 0.3499,
+      "step": 43
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.561620555582239,
+      "learning_rate": 1.2324929971988797e-06,
+      "loss": 0.6534,
+      "step": 44
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.5200680863783844,
+      "learning_rate": 1.2605042016806724e-06,
+      "loss": 0.4172,
+      "step": 45
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.0672131028851704,
+      "learning_rate": 1.288515406162465e-06,
+      "loss": 0.5117,
+      "step": 46
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.3267050525804174,
+      "learning_rate": 1.316526610644258e-06,
+      "loss": 0.392,
+      "step": 47
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.6082344603552319,
+      "learning_rate": 1.3445378151260504e-06,
+      "loss": 0.386,
+      "step": 48
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.5539462317860868,
+      "learning_rate": 1.3725490196078434e-06,
+      "loss": 0.3556,
+      "step": 49
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.309493873051941,
+      "learning_rate": 1.4005602240896359e-06,
+      "loss": 0.2866,
+      "step": 50
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.5679101753087994,
+      "learning_rate": 1.4285714285714286e-06,
+      "loss": 0.3349,
+      "step": 51
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.4037561690173512,
+      "learning_rate": 1.4565826330532216e-06,
+      "loss": 0.3631,
+      "step": 52
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.149439501665514,
+      "learning_rate": 1.484593837535014e-06,
+      "loss": 0.3144,
+      "step": 53
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.9156445398777989,
+      "learning_rate": 1.5126050420168068e-06,
+      "loss": 0.2594,
+      "step": 54
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.8554957368940027,
+      "learning_rate": 1.5406162464985996e-06,
+      "loss": 0.2235,
+      "step": 55
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.304353447332407,
+      "learning_rate": 1.5686274509803923e-06,
+      "loss": 0.3232,
+      "step": 56
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.9474417020162798,
+      "learning_rate": 1.5966386554621848e-06,
+      "loss": 0.2953,
+      "step": 57
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.9925423302626352,
+      "learning_rate": 1.6246498599439778e-06,
+      "loss": 0.2353,
+      "step": 58
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.806722828112331,
+      "learning_rate": 1.6526610644257705e-06,
+      "loss": 0.4393,
+      "step": 59
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3421913952877822,
+      "learning_rate": 1.6806722689075632e-06,
+      "loss": 0.2885,
+      "step": 60
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.629878207455927,
+      "learning_rate": 1.708683473389356e-06,
+      "loss": 0.3663,
+      "step": 61
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.1623322808354322,
+      "learning_rate": 1.7366946778711485e-06,
+      "loss": 0.2524,
+      "step": 62
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2713641551250345,
+      "learning_rate": 1.7647058823529414e-06,
+      "loss": 0.3329,
+      "step": 63
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6220897652081994,
+      "learning_rate": 1.792717086834734e-06,
+      "loss": 0.2722,
+      "step": 64
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.480713887381015,
+      "learning_rate": 1.8207282913165267e-06,
+      "loss": 0.2783,
+      "step": 65
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3330905139520488,
+      "learning_rate": 1.8487394957983196e-06,
+      "loss": 0.2439,
+      "step": 66
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8873443176784379,
+      "learning_rate": 1.8767507002801122e-06,
+      "loss": 0.157,
+      "step": 67
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.165198341695851,
+      "learning_rate": 1.904761904761905e-06,
+      "loss": 0.3187,
+      "step": 68
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6964579458207625,
+      "learning_rate": 1.932773109243698e-06,
+      "loss": 0.3121,
+      "step": 69
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.1187192308380378,
+      "learning_rate": 1.96078431372549e-06,
+      "loss": 0.2171,
+      "step": 70
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0276764807060643,
+      "learning_rate": 1.988795518207283e-06,
+      "loss": 0.2625,
+      "step": 71
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9027873230036005,
+      "learning_rate": 2.0168067226890756e-06,
+      "loss": 0.3217,
+      "step": 72
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5165163685998264,
+      "learning_rate": 2.0448179271708684e-06,
+      "loss": 0.2961,
+      "step": 73
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.1473792353335586,
+      "learning_rate": 2.0728291316526615e-06,
+      "loss": 0.2281,
+      "step": 74
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9349009675846548,
+      "learning_rate": 2.100840336134454e-06,
+      "loss": 0.2056,
+      "step": 75
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0032690191243625,
+      "learning_rate": 2.1288515406162466e-06,
+      "loss": 0.1932,
+      "step": 76
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4498611743782313,
+      "learning_rate": 2.1568627450980393e-06,
+      "loss": 0.2399,
+      "step": 77
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5705516137473403,
+      "learning_rate": 2.184873949579832e-06,
+      "loss": 0.1891,
+      "step": 78
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3027285742443064,
+      "learning_rate": 2.2128851540616248e-06,
+      "loss": 0.2051,
+      "step": 79
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.008999876619946,
+      "learning_rate": 2.2408963585434175e-06,
+      "loss": 0.4214,
+      "step": 80
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.283035163456896,
+      "learning_rate": 2.2689075630252102e-06,
+      "loss": 0.1946,
+      "step": 81
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.577756531888979,
+      "learning_rate": 2.296918767507003e-06,
+      "loss": 0.1838,
+      "step": 82
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7382106351764461,
+      "learning_rate": 2.3249299719887957e-06,
+      "loss": 0.217,
+      "step": 83
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2549335402945008,
+      "learning_rate": 2.3529411764705885e-06,
+      "loss": 0.1845,
+      "step": 84
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0818358619557094,
+      "learning_rate": 2.380952380952381e-06,
+      "loss": 0.2088,
+      "step": 85
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.1092305795090334,
+      "learning_rate": 2.408963585434174e-06,
+      "loss": 0.223,
+      "step": 86
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0046575088900542,
+      "learning_rate": 2.4369747899159667e-06,
+      "loss": 0.1588,
+      "step": 87
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.543139589180529,
+      "learning_rate": 2.4649859943977594e-06,
+      "loss": 0.25,
+      "step": 88
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0862284836177165,
+      "learning_rate": 2.492997198879552e-06,
+      "loss": 0.1834,
+      "step": 89
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0951286709007708,
+      "learning_rate": 2.521008403361345e-06,
+      "loss": 0.1843,
+      "step": 90
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0854365027544433,
+      "learning_rate": 2.549019607843137e-06,
+      "loss": 0.1839,
+      "step": 91
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8741260935022611,
+      "learning_rate": 2.57703081232493e-06,
+      "loss": 0.1462,
+      "step": 92
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.900664319990146,
+      "learning_rate": 2.605042016806723e-06,
+      "loss": 0.2132,
+      "step": 93
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.002053960580304,
+      "learning_rate": 2.633053221288516e-06,
+      "loss": 0.0995,
+      "step": 94
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.070235803762512,
+      "learning_rate": 2.6610644257703085e-06,
+      "loss": 0.1864,
+      "step": 95
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.994410676021949,
+      "learning_rate": 2.689075630252101e-06,
+      "loss": 0.2112,
+      "step": 96
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8435942764081367,
+      "learning_rate": 2.7170868347338936e-06,
+      "loss": 0.1225,
+      "step": 97
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5227951998275673,
+      "learning_rate": 2.7450980392156867e-06,
+      "loss": 0.1688,
+      "step": 98
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.569301304363513,
+      "learning_rate": 2.7731092436974795e-06,
+      "loss": 0.2324,
+      "step": 99
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.122581084598405,
+      "learning_rate": 2.8011204481792718e-06,
+      "loss": 0.1808,
+      "step": 100
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7171633676092366,
+      "learning_rate": 2.8291316526610645e-06,
+      "loss": 0.1331,
+      "step": 101
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9068398782740705,
+      "learning_rate": 2.8571428571428573e-06,
+      "loss": 0.2535,
+      "step": 102
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0038063649347526,
+      "learning_rate": 2.88515406162465e-06,
+      "loss": 0.1292,
+      "step": 103
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.49553597269004,
+      "learning_rate": 2.913165266106443e-06,
+      "loss": 0.1794,
+      "step": 104
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9147845385812657,
+      "learning_rate": 2.9411764705882355e-06,
+      "loss": 0.1347,
+      "step": 105
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0340199089798479,
+      "learning_rate": 2.969187675070028e-06,
+      "loss": 0.2023,
+      "step": 106
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9518522137890731,
+      "learning_rate": 2.997198879551821e-06,
+      "loss": 0.1584,
+      "step": 107
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3008290897927282,
+      "learning_rate": 3.0252100840336137e-06,
+      "loss": 0.1693,
+      "step": 108
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7663463604846564,
+      "learning_rate": 3.053221288515407e-06,
+      "loss": 0.1533,
+      "step": 109
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.1969616595733,
+      "learning_rate": 3.081232492997199e-06,
+      "loss": 0.2056,
+      "step": 110
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.3165158068159264,
+      "learning_rate": 3.109243697478992e-06,
+      "loss": 0.165,
+      "step": 111
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6738693608018019,
+      "learning_rate": 3.1372549019607846e-06,
+      "loss": 0.1323,
+      "step": 112
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3285311317453075,
+      "learning_rate": 3.1652661064425773e-06,
+      "loss": 0.2615,
+      "step": 113
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7602131341712919,
+      "learning_rate": 3.1932773109243696e-06,
+      "loss": 0.2569,
+      "step": 114
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5686472084206783,
+      "learning_rate": 3.221288515406163e-06,
+      "loss": 0.0984,
+      "step": 115
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0151729646603311,
+      "learning_rate": 3.2492997198879555e-06,
+      "loss": 0.1424,
+      "step": 116
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3321196500506631,
+      "learning_rate": 3.2773109243697483e-06,
+      "loss": 0.2019,
+      "step": 117
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7005549079921075,
+      "learning_rate": 3.305322128851541e-06,
+      "loss": 0.0449,
+      "step": 118
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9280575675822933,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 0.094,
+      "step": 119
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8187209837179708,
+      "learning_rate": 3.3613445378151265e-06,
+      "loss": 0.1091,
+      "step": 120
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7228045751843665,
+      "learning_rate": 3.3893557422969192e-06,
+      "loss": 0.1129,
+      "step": 121
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.1373406248756526,
+      "learning_rate": 3.417366946778712e-06,
+      "loss": 0.1797,
+      "step": 122
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9688545946817783,
+      "learning_rate": 3.4453781512605043e-06,
+      "loss": 0.1431,
+      "step": 123
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9018344392656461,
+      "learning_rate": 3.473389355742297e-06,
+      "loss": 0.1278,
+      "step": 124
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.692740577546455,
+      "learning_rate": 3.5014005602240897e-06,
+      "loss": 0.1488,
+      "step": 125
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.1873256092562878,
+      "learning_rate": 3.529411764705883e-06,
+      "loss": 0.1357,
+      "step": 126
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.627583002919607,
+      "learning_rate": 3.5574229691876756e-06,
+      "loss": 0.1115,
+      "step": 127
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6408287701248372,
+      "learning_rate": 3.585434173669468e-06,
+      "loss": 0.1252,
+      "step": 128
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7829349104719872,
+      "learning_rate": 3.6134453781512607e-06,
+      "loss": 0.1353,
+      "step": 129
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8150973685572598,
+      "learning_rate": 3.6414565826330534e-06,
+      "loss": 0.1319,
+      "step": 130
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.909265927877615,
+      "learning_rate": 3.669467787114846e-06,
+      "loss": 0.0976,
+      "step": 131
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6399903356863155,
+      "learning_rate": 3.6974789915966393e-06,
+      "loss": 0.0924,
+      "step": 132
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9810165314757194,
+      "learning_rate": 3.7254901960784316e-06,
+      "loss": 0.1374,
+      "step": 133
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.1169033864692144,
+      "learning_rate": 3.7535014005602243e-06,
+      "loss": 0.1201,
+      "step": 134
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0230195969558176,
+      "learning_rate": 3.781512605042017e-06,
+      "loss": 0.1623,
+      "step": 135
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8201776917789521,
+      "learning_rate": 3.80952380952381e-06,
+      "loss": 0.1038,
+      "step": 136
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7854349579918295,
+      "learning_rate": 3.8375350140056026e-06,
+      "loss": 0.1404,
+      "step": 137
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9473898549020787,
+      "learning_rate": 3.865546218487396e-06,
+      "loss": 0.1913,
+      "step": 138
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0816682569185128,
+      "learning_rate": 3.893557422969188e-06,
+      "loss": 0.145,
+      "step": 139
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8972569274975063,
+      "learning_rate": 3.92156862745098e-06,
+      "loss": 0.1494,
+      "step": 140
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7977795459044014,
+      "learning_rate": 3.9495798319327735e-06,
+      "loss": 0.1063,
+      "step": 141
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4936965473968211,
+      "learning_rate": 3.977591036414566e-06,
+      "loss": 0.1487,
+      "step": 142
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3648362102115557,
+      "learning_rate": 4.005602240896359e-06,
+      "loss": 0.1043,
+      "step": 143
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.660370415184799,
+      "learning_rate": 4.033613445378151e-06,
+      "loss": 0.1203,
+      "step": 144
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9550080631057337,
+      "learning_rate": 4.0616246498599444e-06,
+      "loss": 0.1464,
+      "step": 145
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2467657727213688,
+      "learning_rate": 4.089635854341737e-06,
+      "loss": 0.1712,
+      "step": 146
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9738315271510968,
+      "learning_rate": 4.11764705882353e-06,
+      "loss": 0.1435,
+      "step": 147
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2557624756002042,
+      "learning_rate": 4.145658263305323e-06,
+      "loss": 0.1748,
+      "step": 148
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.1322387036992898,
+      "learning_rate": 4.173669467787115e-06,
+      "loss": 0.1507,
+      "step": 149
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7925609196162391,
+      "learning_rate": 4.201680672268908e-06,
+      "loss": 0.1517,
+      "step": 150
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9363183396165131,
+      "learning_rate": 4.229691876750701e-06,
+      "loss": 0.1182,
+      "step": 151
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.036075343362631,
+      "learning_rate": 4.257703081232493e-06,
+      "loss": 0.1257,
+      "step": 152
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.292902569474106,
+      "learning_rate": 4.2857142857142855e-06,
+      "loss": 0.1682,
+      "step": 153
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.1189764303753709,
+      "learning_rate": 4.313725490196079e-06,
+      "loss": 0.1104,
+      "step": 154
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8997710102528754,
+      "learning_rate": 4.341736694677872e-06,
+      "loss": 0.1483,
+      "step": 155
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3888439514053874,
+      "learning_rate": 4.369747899159664e-06,
+      "loss": 0.1388,
+      "step": 156
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6241812478721267,
+      "learning_rate": 4.397759103641457e-06,
+      "loss": 0.1529,
+      "step": 157
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0750061034225609,
+      "learning_rate": 4.4257703081232496e-06,
+      "loss": 0.1183,
+      "step": 158
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.1806226563310234,
+      "learning_rate": 4.453781512605043e-06,
+      "loss": 0.1158,
+      "step": 159
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.012194765371845,
+      "learning_rate": 4.481792717086835e-06,
+      "loss": 0.1374,
+      "step": 160
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.491833971454559,
+      "learning_rate": 4.509803921568628e-06,
+      "loss": 0.1667,
+      "step": 161
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6814837935597085,
+      "learning_rate": 4.5378151260504205e-06,
+      "loss": 0.1395,
+      "step": 162
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0534935397219367,
+      "learning_rate": 4.565826330532213e-06,
+      "loss": 0.1329,
+      "step": 163
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.181140819390502,
+      "learning_rate": 4.593837535014006e-06,
+      "loss": 0.137,
+      "step": 164
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6721444437160121,
+      "learning_rate": 4.621848739495799e-06,
+      "loss": 0.1604,
+      "step": 165
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.1232194654735719,
+      "learning_rate": 4.6498599439775914e-06,
+      "loss": 0.1325,
+      "step": 166
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9063000227821515,
+      "learning_rate": 4.677871148459384e-06,
+      "loss": 0.1329,
+      "step": 167
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7880981332151767,
+      "learning_rate": 4.705882352941177e-06,
+      "loss": 0.0799,
+      "step": 168
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.75882075731128,
+      "learning_rate": 4.733893557422969e-06,
+      "loss": 0.1193,
+      "step": 169
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.707775289332377,
+      "learning_rate": 4.761904761904762e-06,
+      "loss": 0.1268,
+      "step": 170
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8654630841941064,
+      "learning_rate": 4.7899159663865555e-06,
+      "loss": 0.1466,
+      "step": 171
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4331089362823357,
+      "learning_rate": 4.817927170868348e-06,
+      "loss": 0.1496,
+      "step": 172
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7788799339916578,
+      "learning_rate": 4.84593837535014e-06,
+      "loss": 0.1042,
+      "step": 173
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.851418831279099,
+      "learning_rate": 4.873949579831933e-06,
+      "loss": 0.2319,
+      "step": 174
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7548200282053826,
+      "learning_rate": 4.901960784313726e-06,
+      "loss": 0.0723,
+      "step": 175
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8352632652003819,
+      "learning_rate": 4.929971988795519e-06,
+      "loss": 0.1174,
+      "step": 176
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0768429613158856,
+      "learning_rate": 4.957983193277311e-06,
+      "loss": 0.0984,
+      "step": 177
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6860551834344599,
+      "learning_rate": 4.985994397759104e-06,
+      "loss": 0.1275,
+      "step": 178
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5072187058489912,
+      "learning_rate": 5.0140056022408966e-06,
+      "loss": 0.1237,
+      "step": 179
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5693532801629468,
+      "learning_rate": 5.04201680672269e-06,
+      "loss": 0.1045,
+      "step": 180
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5909399961449435,
+      "learning_rate": 5.070028011204482e-06,
+      "loss": 0.1024,
+      "step": 181
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3184753528297817,
+      "learning_rate": 5.098039215686274e-06,
+      "loss": 0.1499,
+      "step": 182
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7387783165179622,
+      "learning_rate": 5.1260504201680675e-06,
+      "loss": 0.1134,
+      "step": 183
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1874169089723954,
+      "learning_rate": 5.15406162464986e-06,
+      "loss": 0.1405,
+      "step": 184
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.108865943119381,
+      "learning_rate": 5.182072829131654e-06,
+      "loss": 0.1333,
+      "step": 185
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.793094650111033,
+      "learning_rate": 5.210084033613446e-06,
+      "loss": 0.1374,
+      "step": 186
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9491133219463196,
+      "learning_rate": 5.2380952380952384e-06,
+      "loss": 0.1241,
+      "step": 187
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0412064473071794,
+      "learning_rate": 5.266106442577032e-06,
+      "loss": 0.0947,
+      "step": 188
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5987650433717931,
+      "learning_rate": 5.294117647058824e-06,
+      "loss": 0.087,
+      "step": 189
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7088693304879308,
+      "learning_rate": 5.322128851540617e-06,
+      "loss": 0.0629,
+      "step": 190
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0136297711338809,
+      "learning_rate": 5.350140056022409e-06,
+      "loss": 0.0825,
+      "step": 191
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6908771420245972,
+      "learning_rate": 5.378151260504202e-06,
+      "loss": 0.1026,
+      "step": 192
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6181606512498035,
+      "learning_rate": 5.406162464985995e-06,
+      "loss": 0.0725,
+      "step": 193
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0198534062141151,
+      "learning_rate": 5.434173669467787e-06,
+      "loss": 0.1137,
+      "step": 194
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7407742939780293,
+      "learning_rate": 5.4621848739495795e-06,
+      "loss": 0.1234,
+      "step": 195
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9864376614302909,
+      "learning_rate": 5.4901960784313735e-06,
+      "loss": 0.1083,
+      "step": 196
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3105056476116779,
+      "learning_rate": 5.518207282913166e-06,
+      "loss": 0.1462,
+      "step": 197
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8500959694566836,
+      "learning_rate": 5.546218487394959e-06,
+      "loss": 0.1293,
+      "step": 198
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8049639630293183,
+      "learning_rate": 5.574229691876751e-06,
+      "loss": 0.1427,
+      "step": 199
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8175086191610695,
+      "learning_rate": 5.6022408963585436e-06,
+      "loss": 0.1326,
+      "step": 200
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0738426113487498,
+      "learning_rate": 5.630252100840337e-06,
+      "loss": 0.1197,
+      "step": 201
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9445056867957289,
+      "learning_rate": 5.658263305322129e-06,
+      "loss": 0.1176,
+      "step": 202
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0965738365735191,
+      "learning_rate": 5.686274509803922e-06,
+      "loss": 0.1501,
+      "step": 203
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5956119796110726,
+      "learning_rate": 5.7142857142857145e-06,
+      "loss": 0.0626,
+      "step": 204
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9839322325853606,
+      "learning_rate": 5.742296918767507e-06,
+      "loss": 0.1541,
+      "step": 205
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0196458103635884,
+      "learning_rate": 5.7703081232493e-06,
+      "loss": 0.1291,
+      "step": 206
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9156257111669147,
+      "learning_rate": 5.798319327731093e-06,
+      "loss": 0.0676,
+      "step": 207
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1691888058139204,
+      "learning_rate": 5.826330532212886e-06,
+      "loss": 0.1102,
+      "step": 208
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2281207163385925,
+      "learning_rate": 5.854341736694679e-06,
+      "loss": 0.1248,
+      "step": 209
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7956677084983999,
+      "learning_rate": 5.882352941176471e-06,
+      "loss": 0.1344,
+      "step": 210
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3420020663082877,
+      "learning_rate": 5.910364145658264e-06,
+      "loss": 0.173,
+      "step": 211
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9516047545964721,
+      "learning_rate": 5.938375350140056e-06,
+      "loss": 0.1542,
+      "step": 212
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5357263932212112,
+      "learning_rate": 5.9663865546218495e-06,
+      "loss": 0.1013,
+      "step": 213
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2253766818994591,
+      "learning_rate": 5.994397759103642e-06,
+      "loss": 0.1472,
+      "step": 214
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.368394434203125,
+      "learning_rate": 6.022408963585434e-06,
+      "loss": 0.1352,
+      "step": 215
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.564022696136383,
+      "learning_rate": 6.050420168067227e-06,
+      "loss": 0.2411,
+      "step": 216
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9832068902828558,
+      "learning_rate": 6.07843137254902e-06,
+      "loss": 0.1309,
+      "step": 217
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6125673936716415,
+      "learning_rate": 6.106442577030814e-06,
+      "loss": 0.1245,
+      "step": 218
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.375934875362054,
+      "learning_rate": 6.134453781512606e-06,
+      "loss": 0.1915,
+      "step": 219
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4056305921320742,
+      "learning_rate": 6.162464985994398e-06,
+      "loss": 0.1264,
+      "step": 220
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8682730619311089,
+      "learning_rate": 6.1904761904761914e-06,
+      "loss": 0.101,
+      "step": 221
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7539398805306745,
+      "learning_rate": 6.218487394957984e-06,
+      "loss": 0.0843,
+      "step": 222
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.004436289611301,
+      "learning_rate": 6.246498599439776e-06,
+      "loss": 0.1376,
+      "step": 223
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1607485522612968,
+      "learning_rate": 6.274509803921569e-06,
+      "loss": 0.1281,
+      "step": 224
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0786201456835138,
+      "learning_rate": 6.3025210084033615e-06,
+      "loss": 0.1534,
+      "step": 225
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8102842019977521,
+      "learning_rate": 6.330532212885155e-06,
+      "loss": 0.1062,
+      "step": 226
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.494194292716641,
+      "learning_rate": 6.358543417366947e-06,
+      "loss": 0.1243,
+      "step": 227
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7577938538854214,
+      "learning_rate": 6.386554621848739e-06,
+      "loss": 0.0875,
+      "step": 228
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8883762694307261,
+      "learning_rate": 6.414565826330533e-06,
+      "loss": 0.124,
+      "step": 229
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2912673146126228,
+      "learning_rate": 6.442577030812326e-06,
+      "loss": 0.1347,
+      "step": 230
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2536637158520805,
+      "learning_rate": 6.470588235294119e-06,
+      "loss": 0.1159,
+      "step": 231
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.887432612038577,
+      "learning_rate": 6.498599439775911e-06,
+      "loss": 0.0902,
+      "step": 232
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9958418038296715,
+      "learning_rate": 6.526610644257703e-06,
+      "loss": 0.1671,
+      "step": 233
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.4256230440377394,
+      "learning_rate": 6.5546218487394966e-06,
+      "loss": 0.0776,
+      "step": 234
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.792639600779407,
+      "learning_rate": 6.582633053221289e-06,
+      "loss": 0.1737,
+      "step": 235
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3531430409489649,
+      "learning_rate": 6.610644257703082e-06,
+      "loss": 0.1908,
+      "step": 236
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6160059180993964,
+      "learning_rate": 6.638655462184874e-06,
+      "loss": 0.1,
+      "step": 237
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.44218621076890513,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 0.0905,
+      "step": 238
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7934067501977317,
+      "learning_rate": 6.69467787114846e-06,
+      "loss": 0.1075,
+      "step": 239
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.015273688117325,
+      "learning_rate": 6.722689075630253e-06,
+      "loss": 0.1017,
+      "step": 240
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8589450725959116,
+      "learning_rate": 6.750700280112046e-06,
+      "loss": 0.1343,
+      "step": 241
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1868668159709224,
+      "learning_rate": 6.7787114845938384e-06,
+      "loss": 0.0735,
+      "step": 242
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.624749218438332,
+      "learning_rate": 6.806722689075631e-06,
+      "loss": 0.1153,
+      "step": 243
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9117904723909608,
+      "learning_rate": 6.834733893557424e-06,
+      "loss": 0.1397,
+      "step": 244
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.348056610920841,
+      "learning_rate": 6.862745098039216e-06,
+      "loss": 0.1342,
+      "step": 245
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9788193021114568,
+      "learning_rate": 6.8907563025210085e-06,
+      "loss": 0.2195,
+      "step": 246
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8805606669229958,
+      "learning_rate": 6.918767507002802e-06,
+      "loss": 0.1245,
+      "step": 247
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1252650337746726,
+      "learning_rate": 6.946778711484594e-06,
+      "loss": 0.1508,
+      "step": 248
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1502642953663642,
+      "learning_rate": 6.974789915966387e-06,
+      "loss": 0.143,
+      "step": 249
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7265820033419227,
+      "learning_rate": 7.0028011204481795e-06,
+      "loss": 0.0952,
+      "step": 250
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0408077277033767,
+      "learning_rate": 7.030812324929972e-06,
+      "loss": 0.118,
+      "step": 251
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7152894868756683,
+      "learning_rate": 7.058823529411766e-06,
+      "loss": 0.1179,
+      "step": 252
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1771133756751315,
+      "learning_rate": 7.086834733893558e-06,
+      "loss": 0.1871,
+      "step": 253
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7868140613650079,
+      "learning_rate": 7.114845938375351e-06,
+      "loss": 0.102,
+      "step": 254
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0963641998512772,
+      "learning_rate": 7.1428571428571436e-06,
+      "loss": 0.1601,
+      "step": 255
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.8494884016013327,
+      "learning_rate": 7.170868347338936e-06,
+      "loss": 0.1118,
+      "step": 256
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7955504742627728,
+      "learning_rate": 7.198879551820729e-06,
+      "loss": 0.1205,
+      "step": 257
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5823455860412632,
+      "learning_rate": 7.226890756302521e-06,
+      "loss": 0.106,
+      "step": 258
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9676164660232698,
+      "learning_rate": 7.2549019607843145e-06,
+      "loss": 0.1386,
+      "step": 259
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.654979177276987,
+      "learning_rate": 7.282913165266107e-06,
+      "loss": 0.1002,
+      "step": 260
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8320532160587607,
+      "learning_rate": 7.310924369747899e-06,
+      "loss": 0.0843,
+      "step": 261
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6435451195097509,
+      "learning_rate": 7.338935574229692e-06,
+      "loss": 0.0826,
+      "step": 262
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.660236816111505,
+      "learning_rate": 7.3669467787114854e-06,
+      "loss": 0.1055,
+      "step": 263
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0586532395876818,
+      "learning_rate": 7.394957983193279e-06,
+      "loss": 0.1433,
+      "step": 264
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7180157872264912,
+      "learning_rate": 7.422969187675071e-06,
+      "loss": 0.133,
+      "step": 265
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6586616577809415,
+      "learning_rate": 7.450980392156863e-06,
+      "loss": 0.1014,
+      "step": 266
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.909167436704611,
+      "learning_rate": 7.478991596638656e-06,
+      "loss": 0.1247,
+      "step": 267
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5155235546294923,
+      "learning_rate": 7.507002801120449e-06,
+      "loss": 0.1787,
+      "step": 268
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8862318907564707,
+      "learning_rate": 7.535014005602241e-06,
+      "loss": 0.1103,
+      "step": 269
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7961777666694813,
+      "learning_rate": 7.563025210084034e-06,
+      "loss": 0.1391,
+      "step": 270
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2268121697537753,
+      "learning_rate": 7.5910364145658265e-06,
+      "loss": 0.104,
+      "step": 271
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8424893528228262,
+      "learning_rate": 7.61904761904762e-06,
+      "loss": 0.1293,
+      "step": 272
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5945027275520673,
+      "learning_rate": 7.647058823529411e-06,
+      "loss": 0.106,
+      "step": 273
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9020478674324943,
+      "learning_rate": 7.675070028011205e-06,
+      "loss": 0.1501,
+      "step": 274
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6602743166751268,
+      "learning_rate": 7.703081232492997e-06,
+      "loss": 0.1164,
+      "step": 275
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7128662762896094,
+      "learning_rate": 7.731092436974791e-06,
+      "loss": 0.0765,
+      "step": 276
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6716073336746646,
+      "learning_rate": 7.759103641456584e-06,
+      "loss": 0.0912,
+      "step": 277
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.3602254932010835,
+      "learning_rate": 7.787114845938376e-06,
+      "loss": 0.0646,
+      "step": 278
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7922401097154961,
+      "learning_rate": 7.815126050420168e-06,
+      "loss": 0.0816,
+      "step": 279
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8965502776188748,
+      "learning_rate": 7.84313725490196e-06,
+      "loss": 0.1042,
+      "step": 280
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.006229806736457,
+      "learning_rate": 7.871148459383755e-06,
+      "loss": 0.1536,
+      "step": 281
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8036842584260245,
+      "learning_rate": 7.899159663865547e-06,
+      "loss": 0.1008,
+      "step": 282
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7641755373566564,
+      "learning_rate": 7.92717086834734e-06,
+      "loss": 0.1019,
+      "step": 283
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5435840865677047,
+      "learning_rate": 7.955182072829132e-06,
+      "loss": 0.0879,
+      "step": 284
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.117229913816576,
+      "learning_rate": 7.983193277310926e-06,
+      "loss": 0.111,
+      "step": 285
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4255462180416747,
+      "learning_rate": 8.011204481792718e-06,
+      "loss": 0.1196,
+      "step": 286
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8560337611042066,
+      "learning_rate": 8.03921568627451e-06,
+      "loss": 0.1192,
+      "step": 287
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7453938132777229,
+      "learning_rate": 8.067226890756303e-06,
+      "loss": 0.0938,
+      "step": 288
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5344160036818146,
+      "learning_rate": 8.095238095238097e-06,
+      "loss": 0.1389,
+      "step": 289
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2445530419872426,
+      "learning_rate": 8.123249299719889e-06,
+      "loss": 0.1857,
+      "step": 290
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8201105729160768,
+      "learning_rate": 8.151260504201681e-06,
+      "loss": 0.1204,
+      "step": 291
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.68602184828112,
+      "learning_rate": 8.179271708683473e-06,
+      "loss": 0.13,
+      "step": 292
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.159022500857207,
+      "learning_rate": 8.207282913165266e-06,
+      "loss": 0.1394,
+      "step": 293
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.482898982638148,
+      "learning_rate": 8.23529411764706e-06,
+      "loss": 0.0969,
+      "step": 294
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0976033158870477,
+      "learning_rate": 8.263305322128852e-06,
+      "loss": 0.1186,
+      "step": 295
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.219906914646565,
+      "learning_rate": 8.291316526610646e-06,
+      "loss": 0.1286,
+      "step": 296
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8883998652416756,
+      "learning_rate": 8.319327731092438e-06,
+      "loss": 0.1233,
+      "step": 297
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.683376475279875,
+      "learning_rate": 8.34733893557423e-06,
+      "loss": 0.0934,
+      "step": 298
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6815811997423906,
+      "learning_rate": 8.375350140056023e-06,
+      "loss": 0.0999,
+      "step": 299
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6507939989312181,
+      "learning_rate": 8.403361344537815e-06,
+      "loss": 0.0894,
+      "step": 300
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5029467799578613,
+      "learning_rate": 8.43137254901961e-06,
+      "loss": 0.1208,
+      "step": 301
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.59289985631983,
+      "learning_rate": 8.459383753501402e-06,
+      "loss": 0.1556,
+      "step": 302
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7435023481452023,
+      "learning_rate": 8.487394957983194e-06,
+      "loss": 0.117,
+      "step": 303
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8797687851595912,
+      "learning_rate": 8.515406162464986e-06,
+      "loss": 0.1056,
+      "step": 304
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5876011082652602,
+      "learning_rate": 8.543417366946779e-06,
+      "loss": 0.0953,
+      "step": 305
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4171970763131865,
+      "learning_rate": 8.571428571428571e-06,
+      "loss": 0.198,
+      "step": 306
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9413837759523089,
+      "learning_rate": 8.599439775910365e-06,
+      "loss": 0.1018,
+      "step": 307
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2430758571541691,
+      "learning_rate": 8.627450980392157e-06,
+      "loss": 0.1277,
+      "step": 308
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.3602081613156978,
+      "learning_rate": 8.655462184873951e-06,
+      "loss": 0.1276,
+      "step": 309
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8684694085341427,
+      "learning_rate": 8.683473389355744e-06,
+      "loss": 0.1278,
+      "step": 310
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.341288434386308,
+      "learning_rate": 8.711484593837536e-06,
+      "loss": 0.1625,
+      "step": 311
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5657386807369579,
+      "learning_rate": 8.739495798319328e-06,
+      "loss": 0.0655,
+      "step": 312
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7675959063816352,
+      "learning_rate": 8.76750700280112e-06,
+      "loss": 0.0876,
+      "step": 313
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5677678167017767,
+      "learning_rate": 8.795518207282914e-06,
+      "loss": 0.0872,
+      "step": 314
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7331556013163614,
+      "learning_rate": 8.823529411764707e-06,
+      "loss": 0.1371,
+      "step": 315
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.538261767586086,
+      "learning_rate": 8.851540616246499e-06,
+      "loss": 0.1135,
+      "step": 316
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6061951403038772,
+      "learning_rate": 8.879551820728291e-06,
+      "loss": 0.0837,
+      "step": 317
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.853638420522851,
+      "learning_rate": 8.907563025210085e-06,
+      "loss": 0.1701,
+      "step": 318
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7491503645158919,
+      "learning_rate": 8.935574229691878e-06,
+      "loss": 0.142,
+      "step": 319
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5753420716760586,
+      "learning_rate": 8.96358543417367e-06,
+      "loss": 0.1067,
+      "step": 320
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0994723241815523,
+      "learning_rate": 8.991596638655462e-06,
+      "loss": 0.0896,
+      "step": 321
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7933813065611962,
+      "learning_rate": 9.019607843137256e-06,
+      "loss": 0.09,
+      "step": 322
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8507879001937253,
+      "learning_rate": 9.047619047619049e-06,
+      "loss": 0.078,
+      "step": 323
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4023159099396283,
+      "learning_rate": 9.075630252100841e-06,
+      "loss": 0.1311,
+      "step": 324
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6680112670310601,
+      "learning_rate": 9.103641456582633e-06,
+      "loss": 0.1076,
+      "step": 325
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9929825899425084,
+      "learning_rate": 9.131652661064426e-06,
+      "loss": 0.1197,
+      "step": 326
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5771122052184955,
+      "learning_rate": 9.15966386554622e-06,
+      "loss": 0.0558,
+      "step": 327
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.74524960917958,
+      "learning_rate": 9.187675070028012e-06,
+      "loss": 0.118,
+      "step": 328
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2818665090537797,
+      "learning_rate": 9.215686274509804e-06,
+      "loss": 0.1356,
+      "step": 329
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5463510671824718,
+      "learning_rate": 9.243697478991598e-06,
+      "loss": 0.0669,
+      "step": 330
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.71357449243926,
+      "learning_rate": 9.27170868347339e-06,
+      "loss": 0.1034,
+      "step": 331
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7278246158003135,
+      "learning_rate": 9.299719887955183e-06,
+      "loss": 0.0962,
+      "step": 332
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.011311704762003,
+      "learning_rate": 9.327731092436975e-06,
+      "loss": 0.1192,
+      "step": 333
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8045394187677906,
+      "learning_rate": 9.355742296918767e-06,
+      "loss": 0.0728,
+      "step": 334
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.45846489036131516,
+      "learning_rate": 9.383753501400561e-06,
+      "loss": 0.0756,
+      "step": 335
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7539729227805693,
+      "learning_rate": 9.411764705882354e-06,
+      "loss": 0.1236,
+      "step": 336
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0510971035035583,
+      "learning_rate": 9.439775910364146e-06,
+      "loss": 0.1329,
+      "step": 337
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8656009252285208,
+      "learning_rate": 9.467787114845938e-06,
+      "loss": 0.1086,
+      "step": 338
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5420911667172628,
+      "learning_rate": 9.49579831932773e-06,
+      "loss": 0.0789,
+      "step": 339
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4770627197230506,
+      "learning_rate": 9.523809523809525e-06,
+      "loss": 0.0716,
+      "step": 340
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8958014457497894,
+      "learning_rate": 9.551820728291317e-06,
+      "loss": 0.1056,
+      "step": 341
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.643953139246371,
+      "learning_rate": 9.579831932773111e-06,
+      "loss": 0.1224,
+      "step": 342
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.9714742292981904,
+      "learning_rate": 9.607843137254903e-06,
+      "loss": 0.1436,
+      "step": 343
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6220072129182658,
+      "learning_rate": 9.635854341736696e-06,
+      "loss": 0.0871,
+      "step": 344
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5303088272034027,
+      "learning_rate": 9.663865546218488e-06,
+      "loss": 0.1703,
+      "step": 345
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7226529697999378,
+      "learning_rate": 9.69187675070028e-06,
+      "loss": 0.1156,
+      "step": 346
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2260748489126005,
+      "learning_rate": 9.719887955182074e-06,
+      "loss": 0.1271,
+      "step": 347
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6748616643203599,
+      "learning_rate": 9.747899159663867e-06,
+      "loss": 0.1232,
+      "step": 348
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1013718233309193,
+      "learning_rate": 9.775910364145659e-06,
+      "loss": 0.111,
+      "step": 349
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.052337693001522,
+      "learning_rate": 9.803921568627451e-06,
+      "loss": 0.1555,
+      "step": 350
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4536110593980989,
+      "learning_rate": 9.831932773109244e-06,
+      "loss": 0.0726,
+      "step": 351
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8457898329587078,
+      "learning_rate": 9.859943977591038e-06,
+      "loss": 0.1324,
+      "step": 352
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5120053623169777,
+      "learning_rate": 9.88795518207283e-06,
+      "loss": 0.0838,
+      "step": 353
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5504843035063868,
+      "learning_rate": 9.915966386554622e-06,
+      "loss": 0.0818,
+      "step": 354
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.423564606144916,
+      "learning_rate": 9.943977591036416e-06,
+      "loss": 0.15,
+      "step": 355
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0587957908183863,
+      "learning_rate": 9.971988795518209e-06,
+      "loss": 0.123,
+      "step": 356
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8578592733638776,
+      "learning_rate": 1e-05,
+      "loss": 0.1127,
+      "step": 357
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8815493803001586,
+      "learning_rate": 9.999999813785437e-06,
+      "loss": 0.1771,
+      "step": 358
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.113227727342441,
+      "learning_rate": 9.99999925514176e-06,
+      "loss": 0.1462,
+      "step": 359
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7915549138912488,
+      "learning_rate": 9.99999832406901e-06,
+      "loss": 0.0968,
+      "step": 360
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0723579158386007,
+      "learning_rate": 9.99999702056726e-06,
+      "loss": 0.102,
+      "step": 361
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1235711365058287,
+      "learning_rate": 9.999995344636605e-06,
+      "loss": 0.1583,
+      "step": 362
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2759531404786768,
+      "learning_rate": 9.999993296277168e-06,
+      "loss": 0.1653,
+      "step": 363
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.294747411897277,
+      "learning_rate": 9.999990875489104e-06,
+      "loss": 0.1602,
+      "step": 364
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6076115319143451,
+      "learning_rate": 9.999988082272593e-06,
+      "loss": 0.1052,
+      "step": 365
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1569298403268538,
+      "learning_rate": 9.999984916627841e-06,
+      "loss": 0.0671,
+      "step": 366
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5323238762100674,
+      "learning_rate": 9.999981378555086e-06,
+      "loss": 0.0727,
+      "step": 367
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.3576882479079333,
+      "learning_rate": 9.999977468054592e-06,
+      "loss": 0.0951,
+      "step": 368
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2830878862708373,
+      "learning_rate": 9.999973185126647e-06,
+      "loss": 0.1444,
+      "step": 369
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9986121372720049,
+      "learning_rate": 9.999968529771573e-06,
+      "loss": 0.1175,
+      "step": 370
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6712629723993542,
+      "learning_rate": 9.999963501989717e-06,
+      "loss": 0.0976,
+      "step": 371
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8805471005107806,
+      "learning_rate": 9.99995810178145e-06,
+      "loss": 0.1135,
+      "step": 372
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7892499685388796,
+      "learning_rate": 9.99995232914718e-06,
+      "loss": 0.1991,
+      "step": 373
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8026901664725075,
+      "learning_rate": 9.999946184087332e-06,
+      "loss": 0.1455,
+      "step": 374
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7530310671355417,
+      "learning_rate": 9.999939666602365e-06,
+      "loss": 0.1207,
+      "step": 375
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9191083699721888,
+      "learning_rate": 9.999932776692765e-06,
+      "loss": 0.1476,
+      "step": 376
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7009997924776623,
+      "learning_rate": 9.999925514359045e-06,
+      "loss": 0.0995,
+      "step": 377
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7074357470778695,
+      "learning_rate": 9.999917879601746e-06,
+      "loss": 0.1111,
+      "step": 378
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9624155972736683,
+      "learning_rate": 9.999909872421439e-06,
+      "loss": 0.1333,
+      "step": 379
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.684838351137313,
+      "learning_rate": 9.999901492818714e-06,
+      "loss": 0.1212,
+      "step": 380
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6902930353429366,
+      "learning_rate": 9.999892740794201e-06,
+      "loss": 0.1143,
+      "step": 381
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3764351813451206,
+      "learning_rate": 9.999883616348551e-06,
+      "loss": 0.1448,
+      "step": 382
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0654760412384587,
+      "learning_rate": 9.999874119482442e-06,
+      "loss": 0.067,
+      "step": 383
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3977401109537015,
+      "learning_rate": 9.999864250196582e-06,
+      "loss": 0.1374,
+      "step": 384
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4598036942142179,
+      "learning_rate": 9.999854008491707e-06,
+      "loss": 0.1202,
+      "step": 385
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8363603571549727,
+      "learning_rate": 9.999843394368579e-06,
+      "loss": 0.103,
+      "step": 386
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.642018463149415,
+      "learning_rate": 9.999832407827988e-06,
+      "loss": 0.1287,
+      "step": 387
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0812781152617352,
+      "learning_rate": 9.999821048870753e-06,
+      "loss": 0.1305,
+      "step": 388
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7260835054899538,
+      "learning_rate": 9.99980931749772e-06,
+      "loss": 0.043,
+      "step": 389
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0679017072726107,
+      "learning_rate": 9.999797213709764e-06,
+      "loss": 0.1643,
+      "step": 390
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8172198916508853,
+      "learning_rate": 9.999784737507786e-06,
+      "loss": 0.1368,
+      "step": 391
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4832709124642445,
+      "learning_rate": 9.999771888892713e-06,
+      "loss": 0.154,
+      "step": 392
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.60807905681673,
+      "learning_rate": 9.999758667865506e-06,
+      "loss": 0.0706,
+      "step": 393
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.60879834660372,
+      "learning_rate": 9.999745074427147e-06,
+      "loss": 0.1284,
+      "step": 394
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8675368859356152,
+      "learning_rate": 9.999731108578648e-06,
+      "loss": 0.1633,
+      "step": 395
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7573311793662983,
+      "learning_rate": 9.999716770321052e-06,
+      "loss": 0.0923,
+      "step": 396
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9524590206768031,
+      "learning_rate": 9.999702059655423e-06,
+      "loss": 0.1,
+      "step": 397
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4696946536083038,
+      "learning_rate": 9.999686976582862e-06,
+      "loss": 0.1012,
+      "step": 398
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8518966433696187,
+      "learning_rate": 9.999671521104489e-06,
+      "loss": 0.1396,
+      "step": 399
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6195946184783215,
+      "learning_rate": 9.999655693221454e-06,
+      "loss": 0.0824,
+      "step": 400
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.591151422291874,
+      "learning_rate": 9.99963949293494e-06,
+      "loss": 0.1209,
+      "step": 401
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5062533740828177,
+      "learning_rate": 9.99962292024615e-06,
+      "loss": 0.0727,
+      "step": 402
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1864507035069163,
+      "learning_rate": 9.999605975156322e-06,
+      "loss": 0.169,
+      "step": 403
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5787858074675994,
+      "learning_rate": 9.999588657666715e-06,
+      "loss": 0.0771,
+      "step": 404
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7089851677108531,
+      "learning_rate": 9.999570967778618e-06,
+      "loss": 0.08,
+      "step": 405
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6368272158985858,
+      "learning_rate": 9.999552905493352e-06,
+      "loss": 0.1221,
+      "step": 406
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5856673895295825,
+      "learning_rate": 9.999534470812262e-06,
+      "loss": 0.0935,
+      "step": 407
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0956286449896735,
+      "learning_rate": 9.99951566373672e-06,
+      "loss": 0.1548,
+      "step": 408
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7936770844101599,
+      "learning_rate": 9.999496484268127e-06,
+      "loss": 0.1466,
+      "step": 409
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8895310467872868,
+      "learning_rate": 9.999476932407913e-06,
+      "loss": 0.136,
+      "step": 410
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.7262630524416394,
+      "learning_rate": 9.999457008157529e-06,
+      "loss": 0.2041,
+      "step": 411
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7063052207104966,
+      "learning_rate": 9.999436711518467e-06,
+      "loss": 0.0903,
+      "step": 412
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7829596724890455,
+      "learning_rate": 9.999416042492233e-06,
+      "loss": 0.1272,
+      "step": 413
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5782971535803838,
+      "learning_rate": 9.999395001080368e-06,
+      "loss": 0.1055,
+      "step": 414
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4408465690017831,
+      "learning_rate": 9.99937358728444e-06,
+      "loss": 0.0793,
+      "step": 415
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5914949158707651,
+      "learning_rate": 9.999351801106044e-06,
+      "loss": 0.1087,
+      "step": 416
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8869791593426875,
+      "learning_rate": 9.999329642546802e-06,
+      "loss": 0.1237,
+      "step": 417
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6626500279143978,
+      "learning_rate": 9.999307111608366e-06,
+      "loss": 0.1417,
+      "step": 418
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4083491663954509,
+      "learning_rate": 9.999284208292412e-06,
+      "loss": 0.1189,
+      "step": 419
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.520968320487048,
+      "learning_rate": 9.999260932600649e-06,
+      "loss": 0.083,
+      "step": 420
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8475956354539327,
+      "learning_rate": 9.999237284534808e-06,
+      "loss": 0.1649,
+      "step": 421
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0004759207779281,
+      "learning_rate": 9.999213264096651e-06,
+      "loss": 0.1081,
+      "step": 422
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4284016872736085,
+      "learning_rate": 9.999188871287967e-06,
+      "loss": 0.0607,
+      "step": 423
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.942209140728215,
+      "learning_rate": 9.999164106110574e-06,
+      "loss": 0.1863,
+      "step": 424
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8303710619051764,
+      "learning_rate": 9.999138968566317e-06,
+      "loss": 0.1237,
+      "step": 425
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8797360781869421,
+      "learning_rate": 9.999113458657066e-06,
+      "loss": 0.1533,
+      "step": 426
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7520200202175943,
+      "learning_rate": 9.999087576384723e-06,
+      "loss": 0.1147,
+      "step": 427
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7763533929093924,
+      "learning_rate": 9.999061321751215e-06,
+      "loss": 0.1033,
+      "step": 428
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6182054681852013,
+      "learning_rate": 9.9990346947585e-06,
+      "loss": 0.0984,
+      "step": 429
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6269097783989711,
+      "learning_rate": 9.999007695408559e-06,
+      "loss": 0.1082,
+      "step": 430
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9902443934513061,
+      "learning_rate": 9.998980323703403e-06,
+      "loss": 0.1062,
+      "step": 431
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.987425391169721,
+      "learning_rate": 9.998952579645068e-06,
+      "loss": 0.145,
+      "step": 432
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8298876706657803,
+      "learning_rate": 9.998924463235628e-06,
+      "loss": 0.123,
+      "step": 433
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5956116378358924,
+      "learning_rate": 9.998895974477172e-06,
+      "loss": 0.1033,
+      "step": 434
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.44934521687809664,
+      "learning_rate": 9.998867113371824e-06,
+      "loss": 0.1016,
+      "step": 435
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5759296768779403,
+      "learning_rate": 9.998837879921732e-06,
+      "loss": 0.1086,
+      "step": 436
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3565615328440879,
+      "learning_rate": 9.998808274129073e-06,
+      "loss": 0.0673,
+      "step": 437
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5510718520623004,
+      "learning_rate": 9.998778295996054e-06,
+      "loss": 0.0964,
+      "step": 438
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.611371042558083,
+      "learning_rate": 9.99874794552491e-06,
+      "loss": 0.1054,
+      "step": 439
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8257973609669831,
+      "learning_rate": 9.998717222717896e-06,
+      "loss": 0.1108,
+      "step": 440
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5596408984719556,
+      "learning_rate": 9.998686127577305e-06,
+      "loss": 0.1169,
+      "step": 441
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.48547098410358214,
+      "learning_rate": 9.998654660105451e-06,
+      "loss": 0.1063,
+      "step": 442
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.519789135168508,
+      "learning_rate": 9.99862282030468e-06,
+      "loss": 0.1251,
+      "step": 443
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3180794764397854,
+      "learning_rate": 9.998590608177361e-06,
+      "loss": 0.1794,
+      "step": 444
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.739902398693344,
+      "learning_rate": 9.998558023725896e-06,
+      "loss": 0.1398,
+      "step": 445
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6045677753121406,
+      "learning_rate": 9.998525066952709e-06,
+      "loss": 0.0828,
+      "step": 446
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7988531564079917,
+      "learning_rate": 9.998491737860256e-06,
+      "loss": 0.1092,
+      "step": 447
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.518060320605162,
+      "learning_rate": 9.998458036451022e-06,
+      "loss": 0.0864,
+      "step": 448
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2612545325155107,
+      "learning_rate": 9.998423962727514e-06,
+      "loss": 0.1481,
+      "step": 449
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6699689889857784,
+      "learning_rate": 9.998389516692272e-06,
+      "loss": 0.1069,
+      "step": 450
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.579477288461418,
+      "learning_rate": 9.99835469834786e-06,
+      "loss": 0.1243,
+      "step": 451
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.47259436832250945,
+      "learning_rate": 9.998319507696874e-06,
+      "loss": 0.0842,
+      "step": 452
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6461597455139227,
+      "learning_rate": 9.998283944741932e-06,
+      "loss": 0.1012,
+      "step": 453
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.296265985877647,
+      "learning_rate": 9.998248009485685e-06,
+      "loss": 0.146,
+      "step": 454
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5403351160481193,
+      "learning_rate": 9.99821170193081e-06,
+      "loss": 0.0893,
+      "step": 455
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.48404375981303693,
+      "learning_rate": 9.99817502208001e-06,
+      "loss": 0.1249,
+      "step": 456
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5452138186241835,
+      "learning_rate": 9.99813796993602e-06,
+      "loss": 0.0968,
+      "step": 457
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5408153715381283,
+      "learning_rate": 9.998100545501595e-06,
+      "loss": 0.0878,
+      "step": 458
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4637347253448656,
+      "learning_rate": 9.998062748779527e-06,
+      "loss": 0.0818,
+      "step": 459
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6002717791081874,
+      "learning_rate": 9.998024579772628e-06,
+      "loss": 0.0913,
+      "step": 460
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7778546802377903,
+      "learning_rate": 9.997986038483744e-06,
+      "loss": 0.1096,
+      "step": 461
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6823411599203874,
+      "learning_rate": 9.997947124915743e-06,
+      "loss": 0.1447,
+      "step": 462
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.41059589679698616,
+      "learning_rate": 9.997907839071526e-06,
+      "loss": 0.0736,
+      "step": 463
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3746874131971565,
+      "learning_rate": 9.997868180954017e-06,
+      "loss": 0.0583,
+      "step": 464
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4084417863796041,
+      "learning_rate": 9.997828150566174e-06,
+      "loss": 0.0836,
+      "step": 465
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4031936106293697,
+      "learning_rate": 9.997787747910974e-06,
+      "loss": 0.0881,
+      "step": 466
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.598120977661575,
+      "learning_rate": 9.997746972991427e-06,
+      "loss": 0.1031,
+      "step": 467
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4838991095112319,
+      "learning_rate": 9.997705825810574e-06,
+      "loss": 0.0466,
+      "step": 468
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.49334636352735917,
+      "learning_rate": 9.997664306371476e-06,
+      "loss": 0.1106,
+      "step": 469
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3554698052882315,
+      "learning_rate": 9.997622414677227e-06,
+      "loss": 0.0826,
+      "step": 470
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6461637048522569,
+      "learning_rate": 9.997580150730947e-06,
+      "loss": 0.1079,
+      "step": 471
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3680252416183355,
+      "learning_rate": 9.997537514535785e-06,
+      "loss": 0.0813,
+      "step": 472
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.801902627365545,
+      "learning_rate": 9.997494506094916e-06,
+      "loss": 0.1191,
+      "step": 473
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3337940715156686,
+      "learning_rate": 9.997451125411542e-06,
+      "loss": 0.159,
+      "step": 474
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.49171510201932156,
+      "learning_rate": 9.997407372488898e-06,
+      "loss": 0.1082,
+      "step": 475
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7138644142831517,
+      "learning_rate": 9.99736324733024e-06,
+      "loss": 0.1343,
+      "step": 476
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7483748916796715,
+      "learning_rate": 9.997318749938856e-06,
+      "loss": 0.1601,
+      "step": 477
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5438634056067543,
+      "learning_rate": 9.997273880318058e-06,
+      "loss": 0.07,
+      "step": 478
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5391960919956674,
+      "learning_rate": 9.997228638471192e-06,
+      "loss": 0.086,
+      "step": 479
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5503328137892898,
+      "learning_rate": 9.997183024401627e-06,
+      "loss": 0.1121,
+      "step": 480
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4563977164051309,
+      "learning_rate": 9.997137038112758e-06,
+      "loss": 0.1021,
+      "step": 481
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5319076475615525,
+      "learning_rate": 9.99709067960801e-06,
+      "loss": 0.0828,
+      "step": 482
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5297629239747339,
+      "learning_rate": 9.997043948890841e-06,
+      "loss": 0.1022,
+      "step": 483
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6175352559175136,
+      "learning_rate": 9.996996845964729e-06,
+      "loss": 0.1081,
+      "step": 484
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6259564128021599,
+      "learning_rate": 9.99694937083318e-06,
+      "loss": 0.1111,
+      "step": 485
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.34471655188850864,
+      "learning_rate": 9.996901523499733e-06,
+      "loss": 0.0704,
+      "step": 486
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5776546753090909,
+      "learning_rate": 9.996853303967953e-06,
+      "loss": 0.1054,
+      "step": 487
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4511868296874108,
+      "learning_rate": 9.99680471224143e-06,
+      "loss": 0.0983,
+      "step": 488
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.515678683484427,
+      "learning_rate": 9.996755748323782e-06,
+      "loss": 0.1256,
+      "step": 489
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6800153674811197,
+      "learning_rate": 9.996706412218658e-06,
+      "loss": 0.118,
+      "step": 490
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3968468937166093,
+      "learning_rate": 9.996656703929733e-06,
+      "loss": 0.0773,
+      "step": 491
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.464086865109732,
+      "learning_rate": 9.99660662346071e-06,
+      "loss": 0.0842,
+      "step": 492
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7641447353183326,
+      "learning_rate": 9.996556170815315e-06,
+      "loss": 0.1386,
+      "step": 493
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.47166397522099524,
+      "learning_rate": 9.996505345997311e-06,
+      "loss": 0.0838,
+      "step": 494
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6278954677632359,
+      "learning_rate": 9.996454149010484e-06,
+      "loss": 0.0897,
+      "step": 495
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3604775455384319,
+      "learning_rate": 9.996402579858644e-06,
+      "loss": 0.068,
+      "step": 496
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5408802140052029,
+      "learning_rate": 9.996350638545633e-06,
+      "loss": 0.1074,
+      "step": 497
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.2781717586890425,
+      "learning_rate": 9.996298325075322e-06,
+      "loss": 0.0835,
+      "step": 498
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4597361347141018,
+      "learning_rate": 9.996245639451603e-06,
+      "loss": 0.0422,
+      "step": 499
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.13132198533703,
+      "learning_rate": 9.996192581678407e-06,
+      "loss": 0.1446,
+      "step": 500
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.046582299432322,
+      "learning_rate": 9.99613915175968e-06,
+      "loss": 0.158,
+      "step": 501
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6711420785578908,
+      "learning_rate": 9.996085349699407e-06,
+      "loss": 0.1441,
+      "step": 502
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9910923739073177,
+      "learning_rate": 9.99603117550159e-06,
+      "loss": 0.1802,
+      "step": 503
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5913690128537314,
+      "learning_rate": 9.99597662917027e-06,
+      "loss": 0.0949,
+      "step": 504
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.724501776402321,
+      "learning_rate": 9.995921710709504e-06,
+      "loss": 0.1272,
+      "step": 505
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5097582103514099,
+      "learning_rate": 9.995866420123387e-06,
+      "loss": 0.0997,
+      "step": 506
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.621934622231527,
+      "learning_rate": 9.995810757416036e-06,
+      "loss": 0.1105,
+      "step": 507
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5202017740838007,
+      "learning_rate": 9.995754722591597e-06,
+      "loss": 0.0665,
+      "step": 508
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3834096933119617,
+      "learning_rate": 9.995698315654244e-06,
+      "loss": 0.1656,
+      "step": 509
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3892741297485457,
+      "learning_rate": 9.995641536608178e-06,
+      "loss": 0.0449,
+      "step": 510
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6551174259633928,
+      "learning_rate": 9.995584385457629e-06,
+      "loss": 0.1135,
+      "step": 511
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.220152564097502,
+      "learning_rate": 9.995526862206852e-06,
+      "loss": 0.1165,
+      "step": 512
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0568416329525472,
+      "learning_rate": 9.995468966860135e-06,
+      "loss": 0.1487,
+      "step": 513
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.39048557738709233,
+      "learning_rate": 9.995410699421788e-06,
+      "loss": 0.0748,
+      "step": 514
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.832719304203821,
+      "learning_rate": 9.995352059896152e-06,
+      "loss": 0.1791,
+      "step": 515
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6313956435549932,
+      "learning_rate": 9.995293048287595e-06,
+      "loss": 0.1626,
+      "step": 516
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5337349838505776,
+      "learning_rate": 9.99523366460051e-06,
+      "loss": 0.1036,
+      "step": 517
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6544270540260293,
+      "learning_rate": 9.995173908839324e-06,
+      "loss": 0.1071,
+      "step": 518
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.559257811059123,
+      "learning_rate": 9.995113781008485e-06,
+      "loss": 0.0871,
+      "step": 519
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5850860312546887,
+      "learning_rate": 9.995053281112475e-06,
+      "loss": 0.0689,
+      "step": 520
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.256963241911466,
+      "learning_rate": 9.994992409155797e-06,
+      "loss": 0.1538,
+      "step": 521
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5090878933377821,
+      "learning_rate": 9.994931165142986e-06,
+      "loss": 0.0882,
+      "step": 522
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4294758309767924,
+      "learning_rate": 9.994869549078605e-06,
+      "loss": 0.0718,
+      "step": 523
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.456826803356107,
+      "learning_rate": 9.994807560967241e-06,
+      "loss": 0.1033,
+      "step": 524
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.830167603889245,
+      "learning_rate": 9.994745200813515e-06,
+      "loss": 0.105,
+      "step": 525
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5719666718827081,
+      "learning_rate": 9.99468246862207e-06,
+      "loss": 0.0735,
+      "step": 526
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.677575709393069,
+      "learning_rate": 9.994619364397576e-06,
+      "loss": 0.1179,
+      "step": 527
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7125015681410454,
+      "learning_rate": 9.994555888144738e-06,
+      "loss": 0.1122,
+      "step": 528
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4108344536293684,
+      "learning_rate": 9.994492039868283e-06,
+      "loss": 0.0707,
+      "step": 529
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3965546390441736,
+      "learning_rate": 9.994427819572963e-06,
+      "loss": 0.0896,
+      "step": 530
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6251902964897077,
+      "learning_rate": 9.994363227263565e-06,
+      "loss": 0.0715,
+      "step": 531
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8056165087410867,
+      "learning_rate": 9.994298262944902e-06,
+      "loss": 0.1144,
+      "step": 532
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.838992011088451,
+      "learning_rate": 9.994232926621807e-06,
+      "loss": 0.1068,
+      "step": 533
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7116641563129352,
+      "learning_rate": 9.994167218299152e-06,
+      "loss": 0.1399,
+      "step": 534
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7378263274066403,
+      "learning_rate": 9.994101137981829e-06,
+      "loss": 0.1186,
+      "step": 535
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7041449410903895,
+      "learning_rate": 9.99403468567476e-06,
+      "loss": 0.0993,
+      "step": 536
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6254325213413032,
+      "learning_rate": 9.993967861382896e-06,
+      "loss": 0.1032,
+      "step": 537
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0533206631189598,
+      "learning_rate": 9.993900665111213e-06,
+      "loss": 0.1755,
+      "step": 538
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7112542349578724,
+      "learning_rate": 9.993833096864719e-06,
+      "loss": 0.1283,
+      "step": 539
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.274831365207608,
+      "learning_rate": 9.993765156648442e-06,
+      "loss": 0.1978,
+      "step": 540
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6463996387109753,
+      "learning_rate": 9.993696844467447e-06,
+      "loss": 0.1263,
+      "step": 541
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5078792777730835,
+      "learning_rate": 9.99362816032682e-06,
+      "loss": 0.0816,
+      "step": 542
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.36945632584032145,
+      "learning_rate": 9.993559104231676e-06,
+      "loss": 0.0818,
+      "step": 543
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7584613739323615,
+      "learning_rate": 9.993489676187162e-06,
+      "loss": 0.1156,
+      "step": 544
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5550081327051679,
+      "learning_rate": 9.993419876198448e-06,
+      "loss": 0.0782,
+      "step": 545
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4500914806166052,
+      "learning_rate": 9.993349704270733e-06,
+      "loss": 0.1601,
+      "step": 546
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6270229749906052,
+      "learning_rate": 9.993279160409243e-06,
+      "loss": 0.1007,
+      "step": 547
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7585445841588694,
+      "learning_rate": 9.993208244619232e-06,
+      "loss": 0.0851,
+      "step": 548
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3796296026306707,
+      "learning_rate": 9.993136956905985e-06,
+      "loss": 0.0519,
+      "step": 549
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9476177420044567,
+      "learning_rate": 9.99306529727481e-06,
+      "loss": 0.1707,
+      "step": 550
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4961212889927388,
+      "learning_rate": 9.992993265731044e-06,
+      "loss": 0.1061,
+      "step": 551
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6792133972369617,
+      "learning_rate": 9.992920862280054e-06,
+      "loss": 0.103,
+      "step": 552
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5190310895252396,
+      "learning_rate": 9.992848086927232e-06,
+      "loss": 0.0986,
+      "step": 553
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6887901071033682,
+      "learning_rate": 9.992774939677998e-06,
+      "loss": 0.0914,
+      "step": 554
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.602808068195729,
+      "learning_rate": 9.992701420537804e-06,
+      "loss": 0.1282,
+      "step": 555
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.42715537528399167,
+      "learning_rate": 9.992627529512122e-06,
+      "loss": 0.0874,
+      "step": 556
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3866382323268875,
+      "learning_rate": 9.992553266606457e-06,
+      "loss": 0.0833,
+      "step": 557
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.45723410507274154,
+      "learning_rate": 9.99247863182634e-06,
+      "loss": 0.0973,
+      "step": 558
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8460176996161588,
+      "learning_rate": 9.992403625177333e-06,
+      "loss": 0.0958,
+      "step": 559
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.049788559783224,
+      "learning_rate": 9.992328246665021e-06,
+      "loss": 0.1721,
+      "step": 560
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.6202660791611607,
+      "learning_rate": 9.992252496295018e-06,
+      "loss": 0.1782,
+      "step": 561
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.43680908563113335,
+      "learning_rate": 9.992176374072968e-06,
+      "loss": 0.0685,
+      "step": 562
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8545196911296337,
+      "learning_rate": 9.99209988000454e-06,
+      "loss": 0.1737,
+      "step": 563
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8039961883982933,
+      "learning_rate": 9.992023014095431e-06,
+      "loss": 0.1486,
+      "step": 564
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5587173575844564,
+      "learning_rate": 9.991945776351368e-06,
+      "loss": 0.0994,
+      "step": 565
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5825429344800608,
+      "learning_rate": 9.991868166778103e-06,
+      "loss": 0.1219,
+      "step": 566
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5269676403472533,
+      "learning_rate": 9.991790185381416e-06,
+      "loss": 0.0909,
+      "step": 567
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7607855977635978,
+      "learning_rate": 9.991711832167119e-06,
+      "loss": 0.1251,
+      "step": 568
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.36124544967025113,
+      "learning_rate": 9.991633107141043e-06,
+      "loss": 0.0754,
+      "step": 569
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6700259840727764,
+      "learning_rate": 9.991554010309057e-06,
+      "loss": 0.116,
+      "step": 570
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.34117193375791594,
+      "learning_rate": 9.99147454167705e-06,
+      "loss": 0.0892,
+      "step": 571
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4645072200501844,
+      "learning_rate": 9.991394701250941e-06,
+      "loss": 0.1495,
+      "step": 572
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5039230574002541,
+      "learning_rate": 9.991314489036678e-06,
+      "loss": 0.1411,
+      "step": 573
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.36907899959379753,
+      "learning_rate": 9.991233905040234e-06,
+      "loss": 0.0802,
+      "step": 574
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4813627763686513,
+      "learning_rate": 9.991152949267615e-06,
+      "loss": 0.0659,
+      "step": 575
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6080267806649663,
+      "learning_rate": 9.991071621724847e-06,
+      "loss": 0.1213,
+      "step": 576
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3296477756766429,
+      "learning_rate": 9.990989922417991e-06,
+      "loss": 0.0534,
+      "step": 577
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7949673619884388,
+      "learning_rate": 9.990907851353129e-06,
+      "loss": 0.1226,
+      "step": 578
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.41403372438903074,
+      "learning_rate": 9.990825408536377e-06,
+      "loss": 0.0711,
+      "step": 579
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5409182312537043,
+      "learning_rate": 9.990742593973876e-06,
+      "loss": 0.0884,
+      "step": 580
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5879315809835352,
+      "learning_rate": 9.990659407671793e-06,
+      "loss": 0.1144,
+      "step": 581
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6959366535210851,
+      "learning_rate": 9.990575849636322e-06,
+      "loss": 0.1165,
+      "step": 582
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4382921231669729,
+      "learning_rate": 9.990491919873692e-06,
+      "loss": 0.0852,
+      "step": 583
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3746800474911327,
+      "learning_rate": 9.990407618390151e-06,
+      "loss": 0.0798,
+      "step": 584
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4233637593178525,
+      "learning_rate": 9.990322945191979e-06,
+      "loss": 0.0757,
+      "step": 585
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7095894367893713,
+      "learning_rate": 9.990237900285484e-06,
+      "loss": 0.1227,
+      "step": 586
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4073519461855776,
+      "learning_rate": 9.990152483677e-06,
+      "loss": 0.0933,
+      "step": 587
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6994301134078731,
+      "learning_rate": 9.99006669537289e-06,
+      "loss": 0.1362,
+      "step": 588
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3497621479101028,
+      "learning_rate": 9.989980535379542e-06,
+      "loss": 0.0966,
+      "step": 589
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.39985364697513043,
+      "learning_rate": 9.989894003703375e-06,
+      "loss": 0.0855,
+      "step": 590
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4836760904604064,
+      "learning_rate": 9.989807100350834e-06,
+      "loss": 0.0865,
+      "step": 591
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.29820842002707637,
+      "learning_rate": 9.989719825328394e-06,
+      "loss": 0.0601,
+      "step": 592
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8453628926786679,
+      "learning_rate": 9.989632178642553e-06,
+      "loss": 0.1502,
+      "step": 593
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5234738780012702,
+      "learning_rate": 9.98954416029984e-06,
+      "loss": 0.1107,
+      "step": 594
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4967032282665953,
+      "learning_rate": 9.989455770306812e-06,
+      "loss": 0.1074,
+      "step": 595
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.34743376077546084,
+      "learning_rate": 9.989367008670054e-06,
+      "loss": 0.0871,
+      "step": 596
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4492626716605562,
+      "learning_rate": 9.989277875396176e-06,
+      "loss": 0.119,
+      "step": 597
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6897675475179305,
+      "learning_rate": 9.989188370491814e-06,
+      "loss": 0.0949,
+      "step": 598
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.38203174895636355,
+      "learning_rate": 9.98909849396364e-06,
+      "loss": 0.082,
+      "step": 599
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3659272969348084,
+      "learning_rate": 9.989008245818349e-06,
+      "loss": 0.0852,
+      "step": 600
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5540785818848548,
+      "learning_rate": 9.988917626062658e-06,
+      "loss": 0.1162,
+      "step": 601
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7174871771534248,
+      "learning_rate": 9.988826634703318e-06,
+      "loss": 0.0908,
+      "step": 602
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.46028336734572634,
+      "learning_rate": 9.988735271747111e-06,
+      "loss": 0.0806,
+      "step": 603
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5341200542113298,
+      "learning_rate": 9.988643537200839e-06,
+      "loss": 0.0707,
+      "step": 604
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5459506599509913,
+      "learning_rate": 9.988551431071334e-06,
+      "loss": 0.1133,
+      "step": 605
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8650308693532739,
+      "learning_rate": 9.988458953365458e-06,
+      "loss": 0.1335,
+      "step": 606
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5167271814754328,
+      "learning_rate": 9.988366104090101e-06,
+      "loss": 0.0895,
+      "step": 607
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4757146134343915,
+      "learning_rate": 9.988272883252175e-06,
+      "loss": 0.0956,
+      "step": 608
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5979190251152173,
+      "learning_rate": 9.988179290858628e-06,
+      "loss": 0.1015,
+      "step": 609
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5421151955831349,
+      "learning_rate": 9.988085326916427e-06,
+      "loss": 0.1125,
+      "step": 610
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4138660629175644,
+      "learning_rate": 9.987990991432575e-06,
+      "loss": 0.0716,
+      "step": 611
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4352447007847622,
+      "learning_rate": 9.987896284414096e-06,
+      "loss": 0.0423,
+      "step": 612
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8483519854557902,
+      "learning_rate": 9.987801205868046e-06,
+      "loss": 0.1476,
+      "step": 613
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8653367733360023,
+      "learning_rate": 9.987705755801505e-06,
+      "loss": 0.1446,
+      "step": 614
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7884598250327272,
+      "learning_rate": 9.987609934221584e-06,
+      "loss": 0.1331,
+      "step": 615
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5768077773682908,
+      "learning_rate": 9.987513741135419e-06,
+      "loss": 0.0597,
+      "step": 616
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4950915967755527,
+      "learning_rate": 9.987417176550177e-06,
+      "loss": 0.1771,
+      "step": 617
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7672729682728879,
+      "learning_rate": 9.98732024047305e-06,
+      "loss": 0.1332,
+      "step": 618
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5124256765111024,
+      "learning_rate": 9.98722293291126e-06,
+      "loss": 0.0828,
+      "step": 619
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.46115873788036593,
+      "learning_rate": 9.98712525387205e-06,
+      "loss": 0.0986,
+      "step": 620
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.43851838026958595,
+      "learning_rate": 9.987027203362702e-06,
+      "loss": 0.1341,
+      "step": 621
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4933692636942046,
+      "learning_rate": 9.986928781390515e-06,
+      "loss": 0.1154,
+      "step": 622
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7596838157927859,
+      "learning_rate": 9.986829987962821e-06,
+      "loss": 0.1853,
+      "step": 623
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4482223467953727,
+      "learning_rate": 9.98673082308698e-06,
+      "loss": 0.0391,
+      "step": 624
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.546819435988106,
+      "learning_rate": 9.986631286770375e-06,
+      "loss": 0.1259,
+      "step": 625
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.31016185358042764,
+      "learning_rate": 9.986531379020425e-06,
+      "loss": 0.0689,
+      "step": 626
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3366705821812166,
+      "learning_rate": 9.986431099844567e-06,
+      "loss": 0.1201,
+      "step": 627
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5972035151395412,
+      "learning_rate": 9.986330449250275e-06,
+      "loss": 0.124,
+      "step": 628
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3784447366098721,
+      "learning_rate": 9.986229427245042e-06,
+      "loss": 0.0613,
+      "step": 629
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4599583131512941,
+      "learning_rate": 9.986128033836395e-06,
+      "loss": 0.1014,
+      "step": 630
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5184754615425619,
+      "learning_rate": 9.986026269031886e-06,
+      "loss": 0.1079,
+      "step": 631
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3998366126981325,
+      "learning_rate": 9.985924132839094e-06,
+      "loss": 0.073,
+      "step": 632
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.34278854993948016,
+      "learning_rate": 9.985821625265628e-06,
+      "loss": 0.0873,
+      "step": 633
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.37036264861234747,
+      "learning_rate": 9.985718746319121e-06,
+      "loss": 0.0909,
+      "step": 634
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4368701474913475,
+      "learning_rate": 9.98561549600724e-06,
+      "loss": 0.0732,
+      "step": 635
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3949499507523411,
+      "learning_rate": 9.985511874337674e-06,
+      "loss": 0.0771,
+      "step": 636
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3786770072654766,
+      "learning_rate": 9.985407881318139e-06,
+      "loss": 0.1032,
+      "step": 637
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8247300873352073,
+      "learning_rate": 9.985303516956384e-06,
+      "loss": 0.1549,
+      "step": 638
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.53503488372417,
+      "learning_rate": 9.98519878126018e-06,
+      "loss": 0.0852,
+      "step": 639
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1380528987805165,
+      "learning_rate": 9.985093674237331e-06,
+      "loss": 0.1739,
+      "step": 640
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5647741339837101,
+      "learning_rate": 9.984988195895665e-06,
+      "loss": 0.1006,
+      "step": 641
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7907468355414724,
+      "learning_rate": 9.98488234624304e-06,
+      "loss": 0.0976,
+      "step": 642
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.44103242732956977,
+      "learning_rate": 9.984776125287334e-06,
+      "loss": 0.101,
+      "step": 643
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4763927534225128,
+      "learning_rate": 9.984669533036468e-06,
+      "loss": 0.0854,
+      "step": 644
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4364591360546529,
+      "learning_rate": 9.984562569498376e-06,
+      "loss": 0.0791,
+      "step": 645
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8354452825737708,
+      "learning_rate": 9.984455234681026e-06,
+      "loss": 0.1304,
+      "step": 646
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6315167822769249,
+      "learning_rate": 9.984347528592414e-06,
+      "loss": 0.1318,
+      "step": 647
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9837174420911087,
+      "learning_rate": 9.984239451240563e-06,
+      "loss": 0.126,
+      "step": 648
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7110140817816601,
+      "learning_rate": 9.984131002633522e-06,
+      "loss": 0.1276,
+      "step": 649
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7804572355713342,
+      "learning_rate": 9.984022182779368e-06,
+      "loss": 0.1353,
+      "step": 650
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.37308411809973047,
+      "learning_rate": 9.983912991686209e-06,
+      "loss": 0.0953,
+      "step": 651
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7812706705060706,
+      "learning_rate": 9.983803429362176e-06,
+      "loss": 0.0928,
+      "step": 652
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7794512009178994,
+      "learning_rate": 9.983693495815431e-06,
+      "loss": 0.0883,
+      "step": 653
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.39391040063477734,
+      "learning_rate": 9.983583191054162e-06,
+      "loss": 0.0759,
+      "step": 654
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7524525592446534,
+      "learning_rate": 9.983472515086587e-06,
+      "loss": 0.163,
+      "step": 655
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5645054399105317,
+      "learning_rate": 9.983361467920946e-06,
+      "loss": 0.0933,
+      "step": 656
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5776931264675441,
+      "learning_rate": 9.983250049565513e-06,
+      "loss": 0.096,
+      "step": 657
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6488340876420681,
+      "learning_rate": 9.983138260028588e-06,
+      "loss": 0.126,
+      "step": 658
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.29527027694565605,
+      "learning_rate": 9.983026099318496e-06,
+      "loss": 0.0724,
+      "step": 659
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4209397093066584,
+      "learning_rate": 9.982913567443593e-06,
+      "loss": 0.0899,
+      "step": 660
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8849909341407453,
+      "learning_rate": 9.982800664412258e-06,
+      "loss": 0.1422,
+      "step": 661
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3016102935108095,
+      "learning_rate": 9.982687390232903e-06,
+      "loss": 0.0713,
+      "step": 662
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8462776193051589,
+      "learning_rate": 9.982573744913967e-06,
+      "loss": 0.1406,
+      "step": 663
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5974186029855596,
+      "learning_rate": 9.982459728463909e-06,
+      "loss": 0.1229,
+      "step": 664
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5287798021722847,
+      "learning_rate": 9.982345340891228e-06,
+      "loss": 0.1011,
+      "step": 665
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5024606088830081,
+      "learning_rate": 9.98223058220444e-06,
+      "loss": 0.09,
+      "step": 666
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4176777456679742,
+      "learning_rate": 9.982115452412096e-06,
+      "loss": 0.0889,
+      "step": 667
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4783718986253215,
+      "learning_rate": 9.98199995152277e-06,
+      "loss": 0.1249,
+      "step": 668
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5746502172418447,
+      "learning_rate": 9.981884079545065e-06,
+      "loss": 0.1019,
+      "step": 669
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7838284704959871,
+      "learning_rate": 9.981767836487612e-06,
+      "loss": 0.1312,
+      "step": 670
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.46211289563329305,
+      "learning_rate": 9.98165122235907e-06,
+      "loss": 0.1046,
+      "step": 671
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5180169877701982,
+      "learning_rate": 9.981534237168125e-06,
+      "loss": 0.0977,
+      "step": 672
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.648313147331964,
+      "learning_rate": 9.98141688092349e-06,
+      "loss": 0.102,
+      "step": 673
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4555003707249415,
+      "learning_rate": 9.981299153633907e-06,
+      "loss": 0.1086,
+      "step": 674
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6594244793104176,
+      "learning_rate": 9.981181055308144e-06,
+      "loss": 0.1495,
+      "step": 675
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3910892291954089,
+      "learning_rate": 9.981062585954998e-06,
+      "loss": 0.0981,
+      "step": 676
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.500056338395118,
+      "learning_rate": 9.980943745583295e-06,
+      "loss": 0.0858,
+      "step": 677
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4409433771060219,
+      "learning_rate": 9.980824534201887e-06,
+      "loss": 0.0682,
+      "step": 678
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.025807500082504,
+      "learning_rate": 9.98070495181965e-06,
+      "loss": 0.1602,
+      "step": 679
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.060495146342171,
+      "learning_rate": 9.980584998445494e-06,
+      "loss": 0.1455,
+      "step": 680
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4472523554821826,
+      "learning_rate": 9.980464674088356e-06,
+      "loss": 0.0979,
+      "step": 681
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7558408616564564,
+      "learning_rate": 9.980343978757192e-06,
+      "loss": 0.1388,
+      "step": 682
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6134617548087573,
+      "learning_rate": 9.980222912460998e-06,
+      "loss": 0.1134,
+      "step": 683
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5487507413098076,
+      "learning_rate": 9.980101475208788e-06,
+      "loss": 0.0838,
+      "step": 684
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9217127138530026,
+      "learning_rate": 9.97997966700961e-06,
+      "loss": 0.1463,
+      "step": 685
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5189865907994974,
+      "learning_rate": 9.979857487872534e-06,
+      "loss": 0.0796,
+      "step": 686
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5158497852699863,
+      "learning_rate": 9.979734937806665e-06,
+      "loss": 0.0602,
+      "step": 687
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4948210244474041,
+      "learning_rate": 9.979612016821127e-06,
+      "loss": 0.0933,
+      "step": 688
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.589607784339615,
+      "learning_rate": 9.979488724925078e-06,
+      "loss": 0.1498,
+      "step": 689
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.47301440797198707,
+      "learning_rate": 9.979365062127702e-06,
+      "loss": 0.0957,
+      "step": 690
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5454882542083043,
+      "learning_rate": 9.979241028438208e-06,
+      "loss": 0.0699,
+      "step": 691
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8554320147287323,
+      "learning_rate": 9.979116623865837e-06,
+      "loss": 0.1319,
+      "step": 692
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3730520495259766,
+      "learning_rate": 9.978991848419853e-06,
+      "loss": 0.0534,
+      "step": 693
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.40773601959232847,
+      "learning_rate": 9.978866702109553e-06,
+      "loss": 0.0944,
+      "step": 694
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4089814641648594,
+      "learning_rate": 9.978741184944256e-06,
+      "loss": 0.0902,
+      "step": 695
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.38094078688763744,
+      "learning_rate": 9.978615296933311e-06,
+      "loss": 0.0655,
+      "step": 696
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6948070486021376,
+      "learning_rate": 9.978489038086099e-06,
+      "loss": 0.1081,
+      "step": 697
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5206503957453157,
+      "learning_rate": 9.97836240841202e-06,
+      "loss": 0.116,
+      "step": 698
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4573488794539956,
+      "learning_rate": 9.978235407920506e-06,
+      "loss": 0.0652,
+      "step": 699
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5829010355688334,
+      "learning_rate": 9.978108036621021e-06,
+      "loss": 0.104,
+      "step": 700
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.444850397072115,
+      "learning_rate": 9.977980294523048e-06,
+      "loss": 0.0983,
+      "step": 701
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5762184524949143,
+      "learning_rate": 9.977852181636106e-06,
+      "loss": 0.1087,
+      "step": 702
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.40349017557780686,
+      "learning_rate": 9.977723697969735e-06,
+      "loss": 0.0665,
+      "step": 703
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.39232734971983557,
+      "learning_rate": 9.977594843533505e-06,
+      "loss": 0.1036,
+      "step": 704
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4854036632535074,
+      "learning_rate": 9.977465618337013e-06,
+      "loss": 0.1259,
+      "step": 705
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7149773205926221,
+      "learning_rate": 9.977336022389888e-06,
+      "loss": 0.1694,
+      "step": 706
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2564009534439236,
+      "learning_rate": 9.97720605570178e-06,
+      "loss": 0.0911,
+      "step": 707
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5250941327411177,
+      "learning_rate": 9.97707571828237e-06,
+      "loss": 0.1032,
+      "step": 708
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4310308484906928,
+      "learning_rate": 9.976945010141369e-06,
+      "loss": 0.0994,
+      "step": 709
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4801171159821674,
+      "learning_rate": 9.976813931288508e-06,
+      "loss": 0.0929,
+      "step": 710
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9408550585340312,
+      "learning_rate": 9.976682481733554e-06,
+      "loss": 0.1812,
+      "step": 711
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6670035667101979,
+      "learning_rate": 9.9765506614863e-06,
+      "loss": 0.1066,
+      "step": 712
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.35806788360972625,
+      "learning_rate": 9.97641847055656e-06,
+      "loss": 0.0884,
+      "step": 713
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5234161323169547,
+      "learning_rate": 9.976285908954182e-06,
+      "loss": 0.1177,
+      "step": 714
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8088895483824468,
+      "learning_rate": 9.976152976689043e-06,
+      "loss": 0.1351,
+      "step": 715
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5605482943163758,
+      "learning_rate": 9.976019673771038e-06,
+      "loss": 0.0891,
+      "step": 716
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7380956337822607,
+      "learning_rate": 9.975886000210104e-06,
+      "loss": 0.1371,
+      "step": 717
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.36673155123878615,
+      "learning_rate": 9.975751956016192e-06,
+      "loss": 0.0666,
+      "step": 718
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6197419593903347,
+      "learning_rate": 9.975617541199289e-06,
+      "loss": 0.1003,
+      "step": 719
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.35300926269928473,
+      "learning_rate": 9.975482755769406e-06,
+      "loss": 0.0543,
+      "step": 720
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5669314921832739,
+      "learning_rate": 9.975347599736585e-06,
+      "loss": 0.1069,
+      "step": 721
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6374110848146446,
+      "learning_rate": 9.97521207311089e-06,
+      "loss": 0.1072,
+      "step": 722
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.394069298750776,
+      "learning_rate": 9.975076175902415e-06,
+      "loss": 0.0487,
+      "step": 723
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5194233371124842,
+      "learning_rate": 9.974939908121285e-06,
+      "loss": 0.1205,
+      "step": 724
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.122963800657039,
+      "learning_rate": 9.974803269777654e-06,
+      "loss": 0.1005,
+      "step": 725
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7066368936197616,
+      "learning_rate": 9.97466626088169e-06,
+      "loss": 0.1129,
+      "step": 726
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1486603337554333,
+      "learning_rate": 9.974528881443606e-06,
+      "loss": 0.1346,
+      "step": 727
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5422940997853235,
+      "learning_rate": 9.974391131473631e-06,
+      "loss": 0.1238,
+      "step": 728
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9430505268457784,
+      "learning_rate": 9.974253010982026e-06,
+      "loss": 0.1637,
+      "step": 729
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9332423259531186,
+      "learning_rate": 9.974114519979081e-06,
+      "loss": 0.1428,
+      "step": 730
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.622667978830865,
+      "learning_rate": 9.973975658475109e-06,
+      "loss": 0.0909,
+      "step": 731
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3169498690315978,
+      "learning_rate": 9.973836426480455e-06,
+      "loss": 0.0616,
+      "step": 732
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.45812757108662544,
+      "learning_rate": 9.97369682400549e-06,
+      "loss": 0.0831,
+      "step": 733
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6578225858005559,
+      "learning_rate": 9.973556851060611e-06,
+      "loss": 0.1465,
+      "step": 734
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.46958628611936376,
+      "learning_rate": 9.973416507656245e-06,
+      "loss": 0.1151,
+      "step": 735
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6162633795441038,
+      "learning_rate": 9.973275793802844e-06,
+      "loss": 0.1466,
+      "step": 736
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5197208235937061,
+      "learning_rate": 9.97313470951089e-06,
+      "loss": 0.1214,
+      "step": 737
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.47561863006047034,
+      "learning_rate": 9.972993254790894e-06,
+      "loss": 0.0924,
+      "step": 738
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6123899806423115,
+      "learning_rate": 9.97285142965339e-06,
+      "loss": 0.1492,
+      "step": 739
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.45631134029699005,
+      "learning_rate": 9.97270923410894e-06,
+      "loss": 0.0544,
+      "step": 740
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.464961830420518,
+      "learning_rate": 9.972566668168143e-06,
+      "loss": 0.0863,
+      "step": 741
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5072664529808102,
+      "learning_rate": 9.972423731841609e-06,
+      "loss": 0.1157,
+      "step": 742
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4255338961588323,
+      "learning_rate": 9.972280425139992e-06,
+      "loss": 0.1098,
+      "step": 743
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8164375088772227,
+      "learning_rate": 9.972136748073962e-06,
+      "loss": 0.201,
+      "step": 744
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4902943607398687,
+      "learning_rate": 9.971992700654221e-06,
+      "loss": 0.0876,
+      "step": 745
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6964868623743768,
+      "learning_rate": 9.971848282891501e-06,
+      "loss": 0.1097,
+      "step": 746
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3916933419423451,
+      "learning_rate": 9.971703494796556e-06,
+      "loss": 0.0764,
+      "step": 747
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.31562422960077924,
+      "learning_rate": 9.971558336380173e-06,
+      "loss": 0.0857,
+      "step": 748
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5892484859725993,
+      "learning_rate": 9.971412807653166e-06,
+      "loss": 0.1265,
+      "step": 749
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5717862042182116,
+      "learning_rate": 9.971266908626372e-06,
+      "loss": 0.1293,
+      "step": 750
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0155816934403537,
+      "learning_rate": 9.971120639310656e-06,
+      "loss": 0.126,
+      "step": 751
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5987383696405469,
+      "learning_rate": 9.970973999716917e-06,
+      "loss": 0.0842,
+      "step": 752
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5039857009747198,
+      "learning_rate": 9.970826989856079e-06,
+      "loss": 0.0974,
+      "step": 753
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.39264581261524684,
+      "learning_rate": 9.970679609739086e-06,
+      "loss": 0.0622,
+      "step": 754
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5287540907886646,
+      "learning_rate": 9.970531859376921e-06,
+      "loss": 0.0974,
+      "step": 755
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7477820966847073,
+      "learning_rate": 9.970383738780589e-06,
+      "loss": 0.0626,
+      "step": 756
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7232686230853854,
+      "learning_rate": 9.97023524796112e-06,
+      "loss": 0.1323,
+      "step": 757
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5850992577571005,
+      "learning_rate": 9.970086386929576e-06,
+      "loss": 0.1041,
+      "step": 758
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.563005733848986,
+      "learning_rate": 9.969937155697045e-06,
+      "loss": 0.0999,
+      "step": 759
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4076292540871964,
+      "learning_rate": 9.969787554274642e-06,
+      "loss": 0.0828,
+      "step": 760
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5199660949872195,
+      "learning_rate": 9.969637582673512e-06,
+      "loss": 0.0997,
+      "step": 761
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9267861775973697,
+      "learning_rate": 9.969487240904823e-06,
+      "loss": 0.1426,
+      "step": 762
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4133904853641572,
+      "learning_rate": 9.969336528979777e-06,
+      "loss": 0.064,
+      "step": 763
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7482741881440251,
+      "learning_rate": 9.969185446909596e-06,
+      "loss": 0.1113,
+      "step": 764
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.508486674947673,
+      "learning_rate": 9.969033994705536e-06,
+      "loss": 0.0924,
+      "step": 765
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.989977905168848,
+      "learning_rate": 9.968882172378878e-06,
+      "loss": 0.1558,
+      "step": 766
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3554746217510501,
+      "learning_rate": 9.968729979940929e-06,
+      "loss": 0.0783,
+      "step": 767
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3976754946972572,
+      "learning_rate": 9.968577417403025e-06,
+      "loss": 0.1003,
+      "step": 768
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.29254594131500583,
+      "learning_rate": 9.968424484776534e-06,
+      "loss": 0.066,
+      "step": 769
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.48740356214464453,
+      "learning_rate": 9.968271182072842e-06,
+      "loss": 0.0991,
+      "step": 770
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4838852190172392,
+      "learning_rate": 9.96811750930337e-06,
+      "loss": 0.1309,
+      "step": 771
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5184392328414776,
+      "learning_rate": 9.967963466479564e-06,
+      "loss": 0.1424,
+      "step": 772
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.35862278131174913,
+      "learning_rate": 9.9678090536129e-06,
+      "loss": 0.078,
+      "step": 773
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6999597335122942,
+      "learning_rate": 9.967654270714878e-06,
+      "loss": 0.1601,
+      "step": 774
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9557062808268679,
+      "learning_rate": 9.967499117797026e-06,
+      "loss": 0.1623,
+      "step": 775
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.48354655525256973,
+      "learning_rate": 9.967343594870901e-06,
+      "loss": 0.116,
+      "step": 776
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4157235700044118,
+      "learning_rate": 9.967187701948092e-06,
+      "loss": 0.08,
+      "step": 777
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3874809401929965,
+      "learning_rate": 9.967031439040203e-06,
+      "loss": 0.0862,
+      "step": 778
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8514931067533913,
+      "learning_rate": 9.96687480615888e-06,
+      "loss": 0.1254,
+      "step": 779
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.833859617600702,
+      "learning_rate": 9.966717803315787e-06,
+      "loss": 0.1099,
+      "step": 780
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3880532590790286,
+      "learning_rate": 9.966560430522616e-06,
+      "loss": 0.0668,
+      "step": 781
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.25742737410472455,
+      "learning_rate": 9.966402687791094e-06,
+      "loss": 0.0409,
+      "step": 782
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6584616390619324,
+      "learning_rate": 9.966244575132969e-06,
+      "loss": 0.1036,
+      "step": 783
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5591847199464093,
+      "learning_rate": 9.966086092560016e-06,
+      "loss": 0.13,
+      "step": 784
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5588944667741114,
+      "learning_rate": 9.965927240084041e-06,
+      "loss": 0.1165,
+      "step": 785
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8185564454858302,
+      "learning_rate": 9.965768017716878e-06,
+      "loss": 0.1371,
+      "step": 786
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5201127036839857,
+      "learning_rate": 9.965608425470384e-06,
+      "loss": 0.1201,
+      "step": 787
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5525468657747358,
+      "learning_rate": 9.965448463356449e-06,
+      "loss": 0.1474,
+      "step": 788
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3800766597777952,
+      "learning_rate": 9.965288131386985e-06,
+      "loss": 0.0927,
+      "step": 789
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.603956284448405,
+      "learning_rate": 9.965127429573938e-06,
+      "loss": 0.0998,
+      "step": 790
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5069560620465234,
+      "learning_rate": 9.964966357929275e-06,
+      "loss": 0.1127,
+      "step": 791
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5105246084781327,
+      "learning_rate": 9.964804916464994e-06,
+      "loss": 0.0795,
+      "step": 792
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.31809322360310843,
+      "learning_rate": 9.964643105193122e-06,
+      "loss": 0.0584,
+      "step": 793
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4639321177618638,
+      "learning_rate": 9.964480924125708e-06,
+      "loss": 0.0613,
+      "step": 794
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.47966608852474757,
+      "learning_rate": 9.964318373274838e-06,
+      "loss": 0.1025,
+      "step": 795
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6848677806647402,
+      "learning_rate": 9.964155452652613e-06,
+      "loss": 0.1018,
+      "step": 796
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7153338135643892,
+      "learning_rate": 9.963992162271172e-06,
+      "loss": 0.1306,
+      "step": 797
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4935492060472502,
+      "learning_rate": 9.963828502142679e-06,
+      "loss": 0.1168,
+      "step": 798
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.46902027441710226,
+      "learning_rate": 9.963664472279323e-06,
+      "loss": 0.1033,
+      "step": 799
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.2995033907721179,
+      "learning_rate": 9.96350007269332e-06,
+      "loss": 0.0528,
+      "step": 800
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5788015858082936,
+      "learning_rate": 9.963335303396918e-06,
+      "loss": 0.1304,
+      "step": 801
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5831250618080749,
+      "learning_rate": 9.963170164402388e-06,
+      "loss": 0.131,
+      "step": 802
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.37525764909733145,
+      "learning_rate": 9.963004655722034e-06,
+      "loss": 0.0756,
+      "step": 803
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7108094217246574,
+      "learning_rate": 9.962838777368178e-06,
+      "loss": 0.1184,
+      "step": 804
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.46155342771347846,
+      "learning_rate": 9.962672529353182e-06,
+      "loss": 0.073,
+      "step": 805
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7613256091599936,
+      "learning_rate": 9.962505911689424e-06,
+      "loss": 0.1227,
+      "step": 806
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0658788782319246,
+      "learning_rate": 9.962338924389319e-06,
+      "loss": 0.1931,
+      "step": 807
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4292166439445262,
+      "learning_rate": 9.962171567465301e-06,
+      "loss": 0.1004,
+      "step": 808
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6077807835596586,
+      "learning_rate": 9.962003840929841e-06,
+      "loss": 0.1096,
+      "step": 809
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.674077113531982,
+      "learning_rate": 9.961835744795428e-06,
+      "loss": 0.1595,
+      "step": 810
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5617260265709009,
+      "learning_rate": 9.961667279074582e-06,
+      "loss": 0.1449,
+      "step": 811
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.599243693948226,
+      "learning_rate": 9.961498443779855e-06,
+      "loss": 0.1588,
+      "step": 812
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7899581434039957,
+      "learning_rate": 9.96132923892382e-06,
+      "loss": 0.0812,
+      "step": 813
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7247643298857532,
+      "learning_rate": 9.961159664519082e-06,
+      "loss": 0.1047,
+      "step": 814
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4901747082118883,
+      "learning_rate": 9.960989720578272e-06,
+      "loss": 0.1228,
+      "step": 815
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5687435358143359,
+      "learning_rate": 9.960819407114046e-06,
+      "loss": 0.0892,
+      "step": 816
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6877339802997794,
+      "learning_rate": 9.960648724139092e-06,
+      "loss": 0.0954,
+      "step": 817
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0281489558541008,
+      "learning_rate": 9.960477671666125e-06,
+      "loss": 0.1,
+      "step": 818
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6030054399436926,
+      "learning_rate": 9.960306249707883e-06,
+      "loss": 0.1285,
+      "step": 819
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.42562933145037557,
+      "learning_rate": 9.960134458277135e-06,
+      "loss": 0.0641,
+      "step": 820
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.505524835148414,
+      "learning_rate": 9.959962297386678e-06,
+      "loss": 0.1171,
+      "step": 821
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.525205893688321,
+      "learning_rate": 9.959789767049336e-06,
+      "loss": 0.1018,
+      "step": 822
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.493912094701411,
+      "learning_rate": 9.959616867277959e-06,
+      "loss": 0.1097,
+      "step": 823
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6974269836147844,
+      "learning_rate": 9.959443598085425e-06,
+      "loss": 0.1525,
+      "step": 824
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.42286614725709104,
+      "learning_rate": 9.959269959484641e-06,
+      "loss": 0.0796,
+      "step": 825
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.30912254854988763,
+      "learning_rate": 9.95909595148854e-06,
+      "loss": 0.0585,
+      "step": 826
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4996509572761813,
+      "learning_rate": 9.958921574110085e-06,
+      "loss": 0.1148,
+      "step": 827
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5104671959358027,
+      "learning_rate": 9.958746827362263e-06,
+      "loss": 0.0834,
+      "step": 828
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4363796896506406,
+      "learning_rate": 9.95857171125809e-06,
+      "loss": 0.1233,
+      "step": 829
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3754810621808287,
+      "learning_rate": 9.95839622581061e-06,
+      "loss": 0.0866,
+      "step": 830
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4561248286320553,
+      "learning_rate": 9.958220371032895e-06,
+      "loss": 0.0763,
+      "step": 831
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.39857532983299565,
+      "learning_rate": 9.958044146938042e-06,
+      "loss": 0.0893,
+      "step": 832
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3354060942880233,
+      "learning_rate": 9.95786755353918e-06,
+      "loss": 0.0972,
+      "step": 833
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.946659022183752,
+      "learning_rate": 9.95769059084946e-06,
+      "loss": 0.0904,
+      "step": 834
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4998077954112448,
+      "learning_rate": 9.957513258882064e-06,
+      "loss": 0.0841,
+      "step": 835
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.29125879380710573,
+      "learning_rate": 9.9573355576502e-06,
+      "loss": 0.0667,
+      "step": 836
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5981195999337393,
+      "learning_rate": 9.957157487167106e-06,
+      "loss": 0.1174,
+      "step": 837
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3126495227050807,
+      "learning_rate": 9.956979047446043e-06,
+      "loss": 0.0601,
+      "step": 838
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5803070385773836,
+      "learning_rate": 9.956800238500306e-06,
+      "loss": 0.1091,
+      "step": 839
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1782036550619492,
+      "learning_rate": 9.956621060343211e-06,
+      "loss": 0.1121,
+      "step": 840
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.46202716532304794,
+      "learning_rate": 9.956441512988107e-06,
+      "loss": 0.11,
+      "step": 841
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7369296510727624,
+      "learning_rate": 9.956261596448362e-06,
+      "loss": 0.1169,
+      "step": 842
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.40711809229195994,
+      "learning_rate": 9.956081310737383e-06,
+      "loss": 0.0856,
+      "step": 843
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.426743422911565,
+      "learning_rate": 9.955900655868598e-06,
+      "loss": 0.1114,
+      "step": 844
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8743227901529752,
+      "learning_rate": 9.95571963185546e-06,
+      "loss": 0.0708,
+      "step": 845
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.45198407629943704,
+      "learning_rate": 9.955538238711456e-06,
+      "loss": 0.1034,
+      "step": 846
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5720107964880582,
+      "learning_rate": 9.955356476450093e-06,
+      "loss": 0.0838,
+      "step": 847
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.176480626776132,
+      "learning_rate": 9.955174345084915e-06,
+      "loss": 0.1089,
+      "step": 848
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.40222387309825214,
+      "learning_rate": 9.954991844629485e-06,
+      "loss": 0.0708,
+      "step": 849
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5034536977136195,
+      "learning_rate": 9.954808975097397e-06,
+      "loss": 0.0872,
+      "step": 850
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8317191033338598,
+      "learning_rate": 9.954625736502274e-06,
+      "loss": 0.1161,
+      "step": 851
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5612592392382804,
+      "learning_rate": 9.954442128857763e-06,
+      "loss": 0.1164,
+      "step": 852
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6781380989260859,
+      "learning_rate": 9.95425815217754e-06,
+      "loss": 0.0819,
+      "step": 853
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4829270128483685,
+      "learning_rate": 9.954073806475308e-06,
+      "loss": 0.1162,
+      "step": 854
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7324520660795716,
+      "learning_rate": 9.953889091764801e-06,
+      "loss": 0.1379,
+      "step": 855
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5232788888718319,
+      "learning_rate": 9.953704008059776e-06,
+      "loss": 0.0898,
+      "step": 856
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.44170488636008465,
+      "learning_rate": 9.953518555374018e-06,
+      "loss": 0.0984,
+      "step": 857
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.37657221779276345,
+      "learning_rate": 9.953332733721341e-06,
+      "loss": 0.0761,
+      "step": 858
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.43349959330673243,
+      "learning_rate": 9.953146543115588e-06,
+      "loss": 0.0698,
+      "step": 859
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3425556857627271,
+      "learning_rate": 9.952959983570627e-06,
+      "loss": 0.049,
+      "step": 860
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4695565298025845,
+      "learning_rate": 9.952773055100352e-06,
+      "loss": 0.0697,
+      "step": 861
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.40545502464561173,
+      "learning_rate": 9.952585757718687e-06,
+      "loss": 0.0749,
+      "step": 862
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7098123897720013,
+      "learning_rate": 9.952398091439586e-06,
+      "loss": 0.0891,
+      "step": 863
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.646234243570636,
+      "learning_rate": 9.952210056277025e-06,
+      "loss": 0.0957,
+      "step": 864
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7637499983151023,
+      "learning_rate": 9.952021652245008e-06,
+      "loss": 0.1016,
+      "step": 865
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6106644158140575,
+      "learning_rate": 9.951832879357572e-06,
+      "loss": 0.1057,
+      "step": 866
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.40197049986377603,
+      "learning_rate": 9.951643737628778e-06,
+      "loss": 0.1083,
+      "step": 867
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7298684604904201,
+      "learning_rate": 9.951454227072711e-06,
+      "loss": 0.1361,
+      "step": 868
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5917815728060243,
+      "learning_rate": 9.95126434770349e-06,
+      "loss": 0.1043,
+      "step": 869
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.066448645734056,
+      "learning_rate": 9.951074099535256e-06,
+      "loss": 0.1714,
+      "step": 870
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8085208911013361,
+      "learning_rate": 9.950883482582182e-06,
+      "loss": 0.1591,
+      "step": 871
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8884557855233239,
+      "learning_rate": 9.950692496858466e-06,
+      "loss": 0.0688,
+      "step": 872
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5421268698685046,
+      "learning_rate": 9.95050114237833e-06,
+      "loss": 0.0914,
+      "step": 873
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4836414387467889,
+      "learning_rate": 9.950309419156033e-06,
+      "loss": 0.1282,
+      "step": 874
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6255306531869191,
+      "learning_rate": 9.95011732720585e-06,
+      "loss": 0.1279,
+      "step": 875
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.572004815392286,
+      "learning_rate": 9.949924866542095e-06,
+      "loss": 0.1153,
+      "step": 876
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9162023296829667,
+      "learning_rate": 9.949732037179098e-06,
+      "loss": 0.1246,
+      "step": 877
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3783815537144979,
+      "learning_rate": 9.949538839131225e-06,
+      "loss": 0.0698,
+      "step": 878
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7957663871932426,
+      "learning_rate": 9.949345272412866e-06,
+      "loss": 0.1136,
+      "step": 879
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9810703595829444,
+      "learning_rate": 9.94915133703844e-06,
+      "loss": 0.107,
+      "step": 880
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3183140183894449,
+      "learning_rate": 9.948957033022392e-06,
+      "loss": 0.0828,
+      "step": 881
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9714247586876741,
+      "learning_rate": 9.948762360379193e-06,
+      "loss": 0.138,
+      "step": 882
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5387107070689463,
+      "learning_rate": 9.948567319123345e-06,
+      "loss": 0.1302,
+      "step": 883
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5732754717833087,
+      "learning_rate": 9.948371909269376e-06,
+      "loss": 0.1251,
+      "step": 884
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4060247681359065,
+      "learning_rate": 9.948176130831842e-06,
+      "loss": 0.0641,
+      "step": 885
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5297539830451976,
+      "learning_rate": 9.947979983825322e-06,
+      "loss": 0.0792,
+      "step": 886
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5361088800134761,
+      "learning_rate": 9.94778346826443e-06,
+      "loss": 0.0808,
+      "step": 887
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5276001713735403,
+      "learning_rate": 9.947586584163802e-06,
+      "loss": 0.1271,
+      "step": 888
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5471188989182744,
+      "learning_rate": 9.947389331538105e-06,
+      "loss": 0.1177,
+      "step": 889
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5407686032517404,
+      "learning_rate": 9.947191710402027e-06,
+      "loss": 0.0862,
+      "step": 890
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4135873825953957,
+      "learning_rate": 9.946993720770293e-06,
+      "loss": 0.0895,
+      "step": 891
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5771589147281307,
+      "learning_rate": 9.946795362657647e-06,
+      "loss": 0.1021,
+      "step": 892
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6657756866399844,
+      "learning_rate": 9.946596636078866e-06,
+      "loss": 0.1135,
+      "step": 893
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4902035318995722,
+      "learning_rate": 9.946397541048751e-06,
+      "loss": 0.08,
+      "step": 894
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6956332496909995,
+      "learning_rate": 9.946198077582132e-06,
+      "loss": 0.1385,
+      "step": 895
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.249718410001851,
+      "learning_rate": 9.945998245693867e-06,
+      "loss": 0.0543,
+      "step": 896
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.45862268046293814,
+      "learning_rate": 9.94579804539884e-06,
+      "loss": 0.1106,
+      "step": 897
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.39291985700628385,
+      "learning_rate": 9.945597476711964e-06,
+      "loss": 0.0495,
+      "step": 898
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.27893399204500074,
+      "learning_rate": 9.945396539648176e-06,
+      "loss": 0.0594,
+      "step": 899
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7587112220764212,
+      "learning_rate": 9.945195234222446e-06,
+      "loss": 0.1263,
+      "step": 900
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5297318027887641,
+      "learning_rate": 9.944993560449768e-06,
+      "loss": 0.1083,
+      "step": 901
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4693822298788793,
+      "learning_rate": 9.94479151834516e-06,
+      "loss": 0.1078,
+      "step": 902
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.508028839304705,
+      "learning_rate": 9.944589107923676e-06,
+      "loss": 0.1133,
+      "step": 903
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8253543588625101,
+      "learning_rate": 9.944386329200391e-06,
+      "loss": 0.1025,
+      "step": 904
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8330281963970702,
+      "learning_rate": 9.944183182190408e-06,
+      "loss": 0.1134,
+      "step": 905
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6005433302126857,
+      "learning_rate": 9.943979666908861e-06,
+      "loss": 0.0847,
+      "step": 906
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7003510994167398,
+      "learning_rate": 9.943775783370906e-06,
+      "loss": 0.1316,
+      "step": 907
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.39904863185992173,
+      "learning_rate": 9.943571531591731e-06,
+      "loss": 0.0963,
+      "step": 908
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.47613970749742557,
+      "learning_rate": 9.943366911586552e-06,
+      "loss": 0.1008,
+      "step": 909
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4679149234016634,
+      "learning_rate": 9.943161923370606e-06,
+      "loss": 0.0715,
+      "step": 910
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.2728335271209151,
+      "learning_rate": 9.942956566959166e-06,
+      "loss": 0.0726,
+      "step": 911
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.35873063294503904,
+      "learning_rate": 9.942750842367523e-06,
+      "loss": 0.0729,
+      "step": 912
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4369990270673701,
+      "learning_rate": 9.942544749611006e-06,
+      "loss": 0.0707,
+      "step": 913
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1984256340236767,
+      "learning_rate": 9.942338288704963e-06,
+      "loss": 0.0732,
+      "step": 914
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3918726128129519,
+      "learning_rate": 9.942131459664773e-06,
+      "loss": 0.0934,
+      "step": 915
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4557919381988022,
+      "learning_rate": 9.941924262505842e-06,
+      "loss": 0.0734,
+      "step": 916
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.32419548515158414,
+      "learning_rate": 9.941716697243602e-06,
+      "loss": 0.0691,
+      "step": 917
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4902540754809835,
+      "learning_rate": 9.941508763893517e-06,
+      "loss": 0.0917,
+      "step": 918
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6426430143885548,
+      "learning_rate": 9.941300462471072e-06,
+      "loss": 0.1201,
+      "step": 919
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.367322313248295,
+      "learning_rate": 9.941091792991784e-06,
+      "loss": 0.0639,
+      "step": 920
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5801069845898432,
+      "learning_rate": 9.940882755471194e-06,
+      "loss": 0.0692,
+      "step": 921
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6903967584542379,
+      "learning_rate": 9.940673349924871e-06,
+      "loss": 0.1211,
+      "step": 922
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4469218576415177,
+      "learning_rate": 9.94046357636842e-06,
+      "loss": 0.0754,
+      "step": 923
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6154049308241845,
+      "learning_rate": 9.94025343481746e-06,
+      "loss": 0.0497,
+      "step": 924
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.37459787815271606,
+      "learning_rate": 9.940042925287644e-06,
+      "loss": 0.0636,
+      "step": 925
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5556325154089106,
+      "learning_rate": 9.939832047794654e-06,
+      "loss": 0.1139,
+      "step": 926
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5146092489930145,
+      "learning_rate": 9.939620802354196e-06,
+      "loss": 0.0919,
+      "step": 927
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4175804510343337,
+      "learning_rate": 9.939409188982005e-06,
+      "loss": 0.1157,
+      "step": 928
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.35352566125519436,
+      "learning_rate": 9.939197207693846e-06,
+      "loss": 0.0844,
+      "step": 929
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.442223706881349,
+      "learning_rate": 9.938984858505502e-06,
+      "loss": 0.1058,
+      "step": 930
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5304650289380314,
+      "learning_rate": 9.938772141432797e-06,
+      "loss": 0.0987,
+      "step": 931
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.49839618608713415,
+      "learning_rate": 9.93855905649157e-06,
+      "loss": 0.1016,
+      "step": 932
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8508442700002354,
+      "learning_rate": 9.938345603697697e-06,
+      "loss": 0.17,
+      "step": 933
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5583779962048626,
+      "learning_rate": 9.938131783067076e-06,
+      "loss": 0.105,
+      "step": 934
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7386437313777376,
+      "learning_rate": 9.937917594615631e-06,
+      "loss": 0.1293,
+      "step": 935
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4899393831383726,
+      "learning_rate": 9.937703038359318e-06,
+      "loss": 0.1151,
+      "step": 936
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5959114212642134,
+      "learning_rate": 9.937488114314121e-06,
+      "loss": 0.0915,
+      "step": 937
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4070676867806631,
+      "learning_rate": 9.937272822496045e-06,
+      "loss": 0.0786,
+      "step": 938
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4454390145846921,
+      "learning_rate": 9.937057162921127e-06,
+      "loss": 0.089,
+      "step": 939
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5180370602475179,
+      "learning_rate": 9.936841135605431e-06,
+      "loss": 0.0997,
+      "step": 940
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5010252045878886,
+      "learning_rate": 9.936624740565049e-06,
+      "loss": 0.063,
+      "step": 941
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.687365184479091,
+      "learning_rate": 9.936407977816097e-06,
+      "loss": 0.115,
+      "step": 942
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.517935305443366,
+      "learning_rate": 9.936190847374722e-06,
+      "loss": 0.0744,
+      "step": 943
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5101032241181154,
+      "learning_rate": 9.935973349257099e-06,
+      "loss": 0.0982,
+      "step": 944
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4784360275682416,
+      "learning_rate": 9.935755483479426e-06,
+      "loss": 0.1002,
+      "step": 945
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.39323509229697784,
+      "learning_rate": 9.935537250057932e-06,
+      "loss": 0.0797,
+      "step": 946
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3495108135403981,
+      "learning_rate": 9.93531864900887e-06,
+      "loss": 0.0848,
+      "step": 947
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.46824795453218726,
+      "learning_rate": 9.935099680348527e-06,
+      "loss": 0.1003,
+      "step": 948
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.32784264690700776,
+      "learning_rate": 9.93488034409321e-06,
+      "loss": 0.0489,
+      "step": 949
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5803597963753291,
+      "learning_rate": 9.934660640259258e-06,
+      "loss": 0.0734,
+      "step": 950
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.25319968410962723,
+      "learning_rate": 9.934440568863033e-06,
+      "loss": 0.0633,
+      "step": 951
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.42073432948788536,
+      "learning_rate": 9.93422012992093e-06,
+      "loss": 0.1026,
+      "step": 952
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.34130394400624925,
+      "learning_rate": 9.933999323449367e-06,
+      "loss": 0.0817,
+      "step": 953
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3629272970857018,
+      "learning_rate": 9.933778149464794e-06,
+      "loss": 0.0758,
+      "step": 954
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.27860397693080197,
+      "learning_rate": 9.93355660798368e-06,
+      "loss": 0.0902,
+      "step": 955
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3867955268411471,
+      "learning_rate": 9.933334699022532e-06,
+      "loss": 0.0765,
+      "step": 956
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6563781763193827,
+      "learning_rate": 9.933112422597878e-06,
+      "loss": 0.0878,
+      "step": 957
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.46182489337509625,
+      "learning_rate": 9.932889778726272e-06,
+      "loss": 0.0869,
+      "step": 958
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6221781367469542,
+      "learning_rate": 9.932666767424297e-06,
+      "loss": 0.1242,
+      "step": 959
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.2744635190303476,
+      "learning_rate": 9.932443388708569e-06,
+      "loss": 0.0784,
+      "step": 960
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4467377845641006,
+      "learning_rate": 9.932219642595721e-06,
+      "loss": 0.0783,
+      "step": 961
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5214324734498765,
+      "learning_rate": 9.931995529102424e-06,
+      "loss": 0.1351,
+      "step": 962
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.35273439576889903,
+      "learning_rate": 9.931771048245368e-06,
+      "loss": 0.0466,
+      "step": 963
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.37684358051357436,
+      "learning_rate": 9.931546200041275e-06,
+      "loss": 0.094,
+      "step": 964
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5229111731524864,
+      "learning_rate": 9.931320984506893e-06,
+      "loss": 0.1248,
+      "step": 965
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.48753037028244073,
+      "learning_rate": 9.931095401658997e-06,
+      "loss": 0.0986,
+      "step": 966
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5456155379414487,
+      "learning_rate": 9.930869451514389e-06,
+      "loss": 0.1154,
+      "step": 967
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.546772992013245,
+      "learning_rate": 9.930643134089902e-06,
+      "loss": 0.1335,
+      "step": 968
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.43968362181394827,
+      "learning_rate": 9.930416449402388e-06,
+      "loss": 0.0891,
+      "step": 969
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3959336312149664,
+      "learning_rate": 9.930189397468737e-06,
+      "loss": 0.0943,
+      "step": 970
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5100112507332921,
+      "learning_rate": 9.929961978305857e-06,
+      "loss": 0.088,
+      "step": 971
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6399681127358918,
+      "learning_rate": 9.929734191930694e-06,
+      "loss": 0.1012,
+      "step": 972
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4281279157910534,
+      "learning_rate": 9.929506038360206e-06,
+      "loss": 0.1391,
+      "step": 973
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4418617836166927,
+      "learning_rate": 9.929277517611396e-06,
+      "loss": 0.0929,
+      "step": 974
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.42468150051652975,
+      "learning_rate": 9.929048629701278e-06,
+      "loss": 0.0795,
+      "step": 975
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4038052014589253,
+      "learning_rate": 9.928819374646905e-06,
+      "loss": 0.0679,
+      "step": 976
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6334702580966,
+      "learning_rate": 9.928589752465352e-06,
+      "loss": 0.0968,
+      "step": 977
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3467438326724044,
+      "learning_rate": 9.928359763173725e-06,
+      "loss": 0.0984,
+      "step": 978
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7294607316698436,
+      "learning_rate": 9.92812940678915e-06,
+      "loss": 0.132,
+      "step": 979
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5790548331738603,
+      "learning_rate": 9.927898683328792e-06,
+      "loss": 0.0962,
+      "step": 980
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.43412295110140336,
+      "learning_rate": 9.927667592809831e-06,
+      "loss": 0.1282,
+      "step": 981
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6415324164016171,
+      "learning_rate": 9.927436135249482e-06,
+      "loss": 0.0979,
+      "step": 982
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.45416709276651535,
+      "learning_rate": 9.927204310664985e-06,
+      "loss": 0.0685,
+      "step": 983
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5067231909032829,
+      "learning_rate": 9.926972119073608e-06,
+      "loss": 0.0901,
+      "step": 984
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5980457141880093,
+      "learning_rate": 9.926739560492646e-06,
+      "loss": 0.088,
+      "step": 985
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5618607065814243,
+      "learning_rate": 9.926506634939421e-06,
+      "loss": 0.117,
+      "step": 986
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6693119391162285,
+      "learning_rate": 9.926273342431283e-06,
+      "loss": 0.1533,
+      "step": 987
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.40272871118215053,
+      "learning_rate": 9.92603968298561e-06,
+      "loss": 0.1014,
+      "step": 988
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5045596281735716,
+      "learning_rate": 9.925805656619803e-06,
+      "loss": 0.0921,
+      "step": 989
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6801264648777448,
+      "learning_rate": 9.925571263351298e-06,
+      "loss": 0.1318,
+      "step": 990
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.43553563582117355,
+      "learning_rate": 9.925336503197549e-06,
+      "loss": 0.0836,
+      "step": 991
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5421283320729043,
+      "learning_rate": 9.925101376176045e-06,
+      "loss": 0.1341,
+      "step": 992
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3604632761717536,
+      "learning_rate": 9.924865882304302e-06,
+      "loss": 0.095,
+      "step": 993
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3705815192143719,
+      "learning_rate": 9.924630021599857e-06,
+      "loss": 0.0545,
+      "step": 994
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.542093934049776,
+      "learning_rate": 9.92439379408028e-06,
+      "loss": 0.1168,
+      "step": 995
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5724655742366311,
+      "learning_rate": 9.924157199763167e-06,
+      "loss": 0.1044,
+      "step": 996
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.65819677817133,
+      "learning_rate": 9.92392023866614e-06,
+      "loss": 0.0873,
+      "step": 997
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6581220702580841,
+      "learning_rate": 9.923682910806851e-06,
+      "loss": 0.1279,
+      "step": 998
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.476709092484444,
+      "learning_rate": 9.923445216202976e-06,
+      "loss": 0.082,
+      "step": 999
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5575590150644242,
+      "learning_rate": 9.923207154872218e-06,
+      "loss": 0.1138,
+      "step": 1000
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6128587692683094,
+      "learning_rate": 9.922968726832313e-06,
+      "loss": 0.1005,
+      "step": 1001
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5071336045917442,
+      "learning_rate": 9.92272993210102e-06,
+      "loss": 0.0656,
+      "step": 1002
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.29330233235088676,
+      "learning_rate": 9.922490770696123e-06,
+      "loss": 0.0878,
+      "step": 1003
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.2943264595247416,
+      "learning_rate": 9.922251242635436e-06,
+      "loss": 0.0674,
+      "step": 1004
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.49734799853936823,
+      "learning_rate": 9.922011347936806e-06,
+      "loss": 0.1322,
+      "step": 1005
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9336663184860172,
+      "learning_rate": 9.921771086618095e-06,
+      "loss": 0.1563,
+      "step": 1006
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5539684019936046,
+      "learning_rate": 9.921530458697205e-06,
+      "loss": 0.1084,
+      "step": 1007
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3655790323392977,
+      "learning_rate": 9.921289464192054e-06,
+      "loss": 0.0686,
+      "step": 1008
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7679399812751885,
+      "learning_rate": 9.921048103120595e-06,
+      "loss": 0.1128,
+      "step": 1009
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.22764904500185687,
+      "learning_rate": 9.920806375500806e-06,
+      "loss": 0.0476,
+      "step": 1010
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.35019030363796383,
+      "learning_rate": 9.920564281350694e-06,
+      "loss": 0.0725,
+      "step": 1011
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5336785567156759,
+      "learning_rate": 9.92032182068829e-06,
+      "loss": 0.1403,
+      "step": 1012
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8570179840004625,
+      "learning_rate": 9.920078993531653e-06,
+      "loss": 0.1512,
+      "step": 1013
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6480474164209786,
+      "learning_rate": 9.919835799898869e-06,
+      "loss": 0.1009,
+      "step": 1014
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3648490110596108,
+      "learning_rate": 9.919592239808055e-06,
+      "loss": 0.1028,
+      "step": 1015
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.49863438479900063,
+      "learning_rate": 9.919348313277354e-06,
+      "loss": 0.095,
+      "step": 1016
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3228715815147967,
+      "learning_rate": 9.919104020324934e-06,
+      "loss": 0.0666,
+      "step": 1017
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.20053838825238074,
+      "learning_rate": 9.918859360968988e-06,
+      "loss": 0.0355,
+      "step": 1018
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0283913521891452,
+      "learning_rate": 9.918614335227745e-06,
+      "loss": 0.1414,
+      "step": 1019
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6893735105584714,
+      "learning_rate": 9.918368943119452e-06,
+      "loss": 0.1006,
+      "step": 1020
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4733481721002276,
+      "learning_rate": 9.918123184662388e-06,
+      "loss": 0.0794,
+      "step": 1021
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.34936358745974627,
+      "learning_rate": 9.917877059874859e-06,
+      "loss": 0.0382,
+      "step": 1022
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6798964286310596,
+      "learning_rate": 9.917630568775199e-06,
+      "loss": 0.1216,
+      "step": 1023
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5058318622927875,
+      "learning_rate": 9.917383711381765e-06,
+      "loss": 0.0975,
+      "step": 1024
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3422759490296814,
+      "learning_rate": 9.917136487712949e-06,
+      "loss": 0.084,
+      "step": 1025
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5241334085758241,
+      "learning_rate": 9.916888897787162e-06,
+      "loss": 0.0918,
+      "step": 1026
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6320630403952675,
+      "learning_rate": 9.916640941622846e-06,
+      "loss": 0.1606,
+      "step": 1027
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3102672092593467,
+      "learning_rate": 9.916392619238471e-06,
+      "loss": 0.0809,
+      "step": 1028
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4547331000766038,
+      "learning_rate": 9.916143930652535e-06,
+      "loss": 0.0823,
+      "step": 1029
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3847028658721494,
+      "learning_rate": 9.915894875883558e-06,
+      "loss": 0.0777,
+      "step": 1030
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6285060513806519,
+      "learning_rate": 9.915645454950095e-06,
+      "loss": 0.1094,
+      "step": 1031
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3964929309546552,
+      "learning_rate": 9.915395667870725e-06,
+      "loss": 0.1103,
+      "step": 1032
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3337567372507969,
+      "learning_rate": 9.915145514664048e-06,
+      "loss": 0.0621,
+      "step": 1033
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2195503200503558,
+      "learning_rate": 9.9148949953487e-06,
+      "loss": 0.0566,
+      "step": 1034
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.24031435739653853,
+      "learning_rate": 9.914644109943344e-06,
+      "loss": 0.0664,
+      "step": 1035
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.48036937905615945,
+      "learning_rate": 9.914392858466663e-06,
+      "loss": 0.0886,
+      "step": 1036
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3410370303034023,
+      "learning_rate": 9.914141240937374e-06,
+      "loss": 0.1427,
+      "step": 1037
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3828778225793709,
+      "learning_rate": 9.913889257374219e-06,
+      "loss": 0.1187,
+      "step": 1038
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4221240304971868,
+      "learning_rate": 9.913636907795965e-06,
+      "loss": 0.0758,
+      "step": 1039
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4195364712905812,
+      "learning_rate": 9.913384192221412e-06,
+      "loss": 0.0602,
+      "step": 1040
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4151616454292022,
+      "learning_rate": 9.913131110669381e-06,
+      "loss": 0.0936,
+      "step": 1041
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.34605635321497047,
+      "learning_rate": 9.912877663158722e-06,
+      "loss": 0.0887,
+      "step": 1042
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6411776233300411,
+      "learning_rate": 9.912623849708318e-06,
+      "loss": 0.0875,
+      "step": 1043
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4810940811302461,
+      "learning_rate": 9.91236967033707e-06,
+      "loss": 0.1379,
+      "step": 1044
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9707370391468721,
+      "learning_rate": 9.912115125063912e-06,
+      "loss": 0.1712,
+      "step": 1045
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7202689951462978,
+      "learning_rate": 9.911860213907804e-06,
+      "loss": 0.1253,
+      "step": 1046
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6792711543976911,
+      "learning_rate": 9.911604936887735e-06,
+      "loss": 0.0773,
+      "step": 1047
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5135716960783598,
+      "learning_rate": 9.911349294022715e-06,
+      "loss": 0.0846,
+      "step": 1048
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4515634359489102,
+      "learning_rate": 9.911093285331791e-06,
+      "loss": 0.1342,
+      "step": 1049
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.43380882297670487,
+      "learning_rate": 9.910836910834032e-06,
+      "loss": 0.0915,
+      "step": 1050
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9496643062501857,
+      "learning_rate": 9.910580170548529e-06,
+      "loss": 0.1456,
+      "step": 1051
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6626901464785682,
+      "learning_rate": 9.910323064494412e-06,
+      "loss": 0.1378,
+      "step": 1052
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4920832816123536,
+      "learning_rate": 9.910065592690825e-06,
+      "loss": 0.0774,
+      "step": 1053
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3921658583848475,
+      "learning_rate": 9.90980775515695e-06,
+      "loss": 0.0847,
+      "step": 1054
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4366545173875459,
+      "learning_rate": 9.909549551911992e-06,
+      "loss": 0.1247,
+      "step": 1055
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5231216282483795,
+      "learning_rate": 9.909290982975184e-06,
+      "loss": 0.1228,
+      "step": 1056
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4802973975040235,
+      "learning_rate": 9.909032048365785e-06,
+      "loss": 0.1007,
+      "step": 1057
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5210139266584852,
+      "learning_rate": 9.908772748103081e-06,
+      "loss": 0.0907,
+      "step": 1058
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3845995009249425,
+      "learning_rate": 9.908513082206386e-06,
+      "loss": 0.106,
+      "step": 1059
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.38921245445975855,
+      "learning_rate": 9.908253050695045e-06,
+      "loss": 0.0789,
+      "step": 1060
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3381470836574768,
+      "learning_rate": 9.907992653588422e-06,
+      "loss": 0.073,
+      "step": 1061
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3196970133080993,
+      "learning_rate": 9.907731890905916e-06,
+      "loss": 0.0717,
+      "step": 1062
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5663861820185041,
+      "learning_rate": 9.907470762666949e-06,
+      "loss": 0.1119,
+      "step": 1063
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4061645650360372,
+      "learning_rate": 9.90720926889097e-06,
+      "loss": 0.0668,
+      "step": 1064
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2823553293931554,
+      "learning_rate": 9.90694740959746e-06,
+      "loss": 0.0525,
+      "step": 1065
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.31584063344088514,
+      "learning_rate": 9.906685184805921e-06,
+      "loss": 0.1093,
+      "step": 1066
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3740090404312078,
+      "learning_rate": 9.906422594535886e-06,
+      "loss": 0.0846,
+      "step": 1067
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4108001189983455,
+      "learning_rate": 9.906159638806914e-06,
+      "loss": 0.0846,
+      "step": 1068
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4246002189512503,
+      "learning_rate": 9.905896317638592e-06,
+      "loss": 0.0735,
+      "step": 1069
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.37410289406831143,
+      "learning_rate": 9.905632631050533e-06,
+      "loss": 0.0645,
+      "step": 1070
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6430449137310531,
+      "learning_rate": 9.905368579062378e-06,
+      "loss": 0.141,
+      "step": 1071
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.574593934666374,
+      "learning_rate": 9.905104161693794e-06,
+      "loss": 0.1142,
+      "step": 1072
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5436024484444233,
+      "learning_rate": 9.90483937896448e-06,
+      "loss": 0.1001,
+      "step": 1073
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7986589416527862,
+      "learning_rate": 9.904574230894157e-06,
+      "loss": 0.1362,
+      "step": 1074
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6530614296960888,
+      "learning_rate": 9.904308717502571e-06,
+      "loss": 0.1364,
+      "step": 1075
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4907342606682929,
+      "learning_rate": 9.904042838809504e-06,
+      "loss": 0.0905,
+      "step": 1076
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.47899382328311724,
+      "learning_rate": 9.903776594834758e-06,
+      "loss": 0.077,
+      "step": 1077
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5307033651809876,
+      "learning_rate": 9.903509985598165e-06,
+      "loss": 0.0768,
+      "step": 1078
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.554800074581126,
+      "learning_rate": 9.903243011119582e-06,
+      "loss": 0.1056,
+      "step": 1079
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6025689043752012,
+      "learning_rate": 9.902975671418897e-06,
+      "loss": 0.105,
+      "step": 1080
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5074997704067946,
+      "learning_rate": 9.902707966516021e-06,
+      "loss": 0.0778,
+      "step": 1081
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.36432689618268954,
+      "learning_rate": 9.902439896430896e-06,
+      "loss": 0.0943,
+      "step": 1082
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.40341345823234764,
+      "learning_rate": 9.902171461183489e-06,
+      "loss": 0.099,
+      "step": 1083
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.19663950841937716,
+      "learning_rate": 9.901902660793793e-06,
+      "loss": 0.0432,
+      "step": 1084
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6729331397948428,
+      "learning_rate": 9.901633495281834e-06,
+      "loss": 0.1004,
+      "step": 1085
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.39267698882367663,
+      "learning_rate": 9.901363964667653e-06,
+      "loss": 0.0966,
+      "step": 1086
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7728266937268466,
+      "learning_rate": 9.901094068971336e-06,
+      "loss": 0.1152,
+      "step": 1087
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6320378172217919,
+      "learning_rate": 9.90082380821298e-06,
+      "loss": 0.1183,
+      "step": 1088
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.45407875871636105,
+      "learning_rate": 9.900553182412717e-06,
+      "loss": 0.0816,
+      "step": 1089
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8374015950204459,
+      "learning_rate": 9.900282191590707e-06,
+      "loss": 0.1751,
+      "step": 1090
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2432323114130031,
+      "learning_rate": 9.900010835767132e-06,
+      "loss": 0.0404,
+      "step": 1091
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5980545050136095,
+      "learning_rate": 9.899739114962206e-06,
+      "loss": 0.1222,
+      "step": 1092
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4742334049973117,
+      "learning_rate": 9.899467029196166e-06,
+      "loss": 0.0817,
+      "step": 1093
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7255131631360962,
+      "learning_rate": 9.899194578489281e-06,
+      "loss": 0.155,
+      "step": 1094
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6483315687049885,
+      "learning_rate": 9.898921762861843e-06,
+      "loss": 0.1297,
+      "step": 1095
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5426605287173824,
+      "learning_rate": 9.898648582334175e-06,
+      "loss": 0.116,
+      "step": 1096
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.38890868008493357,
+      "learning_rate": 9.898375036926625e-06,
+      "loss": 0.0612,
+      "step": 1097
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.47528887741967374,
+      "learning_rate": 9.898101126659564e-06,
+      "loss": 0.1551,
+      "step": 1098
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3676167719489773,
+      "learning_rate": 9.897826851553401e-06,
+      "loss": 0.1065,
+      "step": 1099
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.29939583031476164,
+      "learning_rate": 9.897552211628558e-06,
+      "loss": 0.0667,
+      "step": 1100
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.49107681883207727,
+      "learning_rate": 9.897277206905498e-06,
+      "loss": 0.101,
+      "step": 1101
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7821488056466414,
+      "learning_rate": 9.897001837404705e-06,
+      "loss": 0.0982,
+      "step": 1102
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.33758831116814153,
+      "learning_rate": 9.896726103146686e-06,
+      "loss": 0.1019,
+      "step": 1103
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.27090015565925035,
+      "learning_rate": 9.89645000415198e-06,
+      "loss": 0.0506,
+      "step": 1104
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5900120450662142,
+      "learning_rate": 9.896173540441155e-06,
+      "loss": 0.1049,
+      "step": 1105
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5562054274684947,
+      "learning_rate": 9.895896712034803e-06,
+      "loss": 0.0958,
+      "step": 1106
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4602306865983874,
+      "learning_rate": 9.895619518953544e-06,
+      "loss": 0.0881,
+      "step": 1107
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3625220082706862,
+      "learning_rate": 9.895341961218022e-06,
+      "loss": 0.0812,
+      "step": 1108
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.37358857690179625,
+      "learning_rate": 9.895064038848913e-06,
+      "loss": 0.0912,
+      "step": 1109
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.26740423534999935,
+      "learning_rate": 9.894785751866921e-06,
+      "loss": 0.0636,
+      "step": 1110
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4705011652756792,
+      "learning_rate": 9.89450710029277e-06,
+      "loss": 0.1264,
+      "step": 1111
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.46104239660480945,
+      "learning_rate": 9.89422808414722e-06,
+      "loss": 0.0997,
+      "step": 1112
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6012631479694955,
+      "learning_rate": 9.893948703451049e-06,
+      "loss": 0.1593,
+      "step": 1113
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5669736618486356,
+      "learning_rate": 9.893668958225069e-06,
+      "loss": 0.08,
+      "step": 1114
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.38975958090898793,
+      "learning_rate": 9.893388848490118e-06,
+      "loss": 0.0791,
+      "step": 1115
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5197548169998257,
+      "learning_rate": 9.89310837426706e-06,
+      "loss": 0.1175,
+      "step": 1116
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7452840133651398,
+      "learning_rate": 9.892827535576785e-06,
+      "loss": 0.1192,
+      "step": 1117
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7749975719884958,
+      "learning_rate": 9.892546332440211e-06,
+      "loss": 0.1133,
+      "step": 1118
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6099187626325029,
+      "learning_rate": 9.892264764878288e-06,
+      "loss": 0.1118,
+      "step": 1119
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6987266537787064,
+      "learning_rate": 9.891982832911983e-06,
+      "loss": 0.1286,
+      "step": 1120
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5503263203440083,
+      "learning_rate": 9.891700536562301e-06,
+      "loss": 0.1122,
+      "step": 1121
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3560344337887794,
+      "learning_rate": 9.891417875850265e-06,
+      "loss": 0.0453,
+      "step": 1122
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4069213099516218,
+      "learning_rate": 9.891134850796934e-06,
+      "loss": 0.0642,
+      "step": 1123
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5398333626397479,
+      "learning_rate": 9.890851461423383e-06,
+      "loss": 0.1073,
+      "step": 1124
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5647557233551089,
+      "learning_rate": 9.890567707750725e-06,
+      "loss": 0.1096,
+      "step": 1125
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4907833881619203,
+      "learning_rate": 9.890283589800094e-06,
+      "loss": 0.1133,
+      "step": 1126
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.27304577465887153,
+      "learning_rate": 9.889999107592653e-06,
+      "loss": 0.0832,
+      "step": 1127
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3205087174695364,
+      "learning_rate": 9.889714261149591e-06,
+      "loss": 0.0693,
+      "step": 1128
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5843199898546776,
+      "learning_rate": 9.88942905049213e-06,
+      "loss": 0.0827,
+      "step": 1129
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6719320751535182,
+      "learning_rate": 9.889143475641506e-06,
+      "loss": 0.0862,
+      "step": 1130
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.750216030671962,
+      "learning_rate": 9.888857536619e-06,
+      "loss": 0.1208,
+      "step": 1131
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5238715789547718,
+      "learning_rate": 9.8885712334459e-06,
+      "loss": 0.076,
+      "step": 1132
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5224572329628987,
+      "learning_rate": 9.888284566143538e-06,
+      "loss": 0.0693,
+      "step": 1133
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5031333413533658,
+      "learning_rate": 9.887997534733266e-06,
+      "loss": 0.1339,
+      "step": 1134
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6215835782837629,
+      "learning_rate": 9.887710139236461e-06,
+      "loss": 0.1229,
+      "step": 1135
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6207397443184434,
+      "learning_rate": 9.887422379674534e-06,
+      "loss": 0.1339,
+      "step": 1136
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.33522001258654616,
+      "learning_rate": 9.887134256068916e-06,
+      "loss": 0.0658,
+      "step": 1137
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3609935475436488,
+      "learning_rate": 9.88684576844107e-06,
+      "loss": 0.0996,
+      "step": 1138
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.328336900849856,
+      "learning_rate": 9.886556916812481e-06,
+      "loss": 0.1488,
+      "step": 1139
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.582741176431571,
+      "learning_rate": 9.886267701204669e-06,
+      "loss": 0.0873,
+      "step": 1140
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5414026411886947,
+      "learning_rate": 9.88597812163917e-06,
+      "loss": 0.0868,
+      "step": 1141
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4116039991930575,
+      "learning_rate": 9.885688178137561e-06,
+      "loss": 0.1106,
+      "step": 1142
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.39267213767335374,
+      "learning_rate": 9.885397870721435e-06,
+      "loss": 0.0905,
+      "step": 1143
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5699378615988014,
+      "learning_rate": 9.885107199412415e-06,
+      "loss": 0.1101,
+      "step": 1144
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5228437474031826,
+      "learning_rate": 9.884816164232154e-06,
+      "loss": 0.1025,
+      "step": 1145
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5192292638363653,
+      "learning_rate": 9.884524765202328e-06,
+      "loss": 0.0987,
+      "step": 1146
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6943531096398174,
+      "learning_rate": 9.884233002344643e-06,
+      "loss": 0.0948,
+      "step": 1147
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.51343320059455,
+      "learning_rate": 9.883940875680831e-06,
+      "loss": 0.1288,
+      "step": 1148
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.48947464946427705,
+      "learning_rate": 9.883648385232654e-06,
+      "loss": 0.0841,
+      "step": 1149
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.104629679103235,
+      "learning_rate": 9.883355531021891e-06,
+      "loss": 0.1189,
+      "step": 1150
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.48213675048635046,
+      "learning_rate": 9.883062313070365e-06,
+      "loss": 0.1169,
+      "step": 1151
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.794538595708785,
+      "learning_rate": 9.88276873139991e-06,
+      "loss": 0.1583,
+      "step": 1152
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5923944037002727,
+      "learning_rate": 9.882474786032397e-06,
+      "loss": 0.1206,
+      "step": 1153
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4622035275391892,
+      "learning_rate": 9.882180476989718e-06,
+      "loss": 0.0973,
+      "step": 1154
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5846682430323663,
+      "learning_rate": 9.881885804293798e-06,
+      "loss": 0.0814,
+      "step": 1155
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3759972462195817,
+      "learning_rate": 9.881590767966582e-06,
+      "loss": 0.0505,
+      "step": 1156
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.33017253680889497,
+      "learning_rate": 9.881295368030048e-06,
+      "loss": 0.0843,
+      "step": 1157
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.308779573536669,
+      "learning_rate": 9.880999604506201e-06,
+      "loss": 0.0671,
+      "step": 1158
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3654092837792867,
+      "learning_rate": 9.88070347741707e-06,
+      "loss": 0.0819,
+      "step": 1159
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.647926691460691,
+      "learning_rate": 9.880406986784709e-06,
+      "loss": 0.0793,
+      "step": 1160
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.46083453391305856,
+      "learning_rate": 9.880110132631209e-06,
+      "loss": 0.1042,
+      "step": 1161
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5350901454148727,
+      "learning_rate": 9.879812914978675e-06,
+      "loss": 0.0953,
+      "step": 1162
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.20464063363408663,
+      "learning_rate": 9.879515333849248e-06,
+      "loss": 0.0584,
+      "step": 1163
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7130866518658852,
+      "learning_rate": 9.879217389265094e-06,
+      "loss": 0.1708,
+      "step": 1164
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.36667307433873825,
+      "learning_rate": 9.878919081248406e-06,
+      "loss": 0.0938,
+      "step": 1165
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.45798058844203876,
+      "learning_rate": 9.878620409821403e-06,
+      "loss": 0.1248,
+      "step": 1166
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4414945823284431,
+      "learning_rate": 9.878321375006332e-06,
+      "loss": 0.1045,
+      "step": 1167
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.30677925442786375,
+      "learning_rate": 9.878021976825464e-06,
+      "loss": 0.0783,
+      "step": 1168
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.34492404236771507,
+      "learning_rate": 9.877722215301107e-06,
+      "loss": 0.0846,
+      "step": 1169
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6027907222313164,
+      "learning_rate": 9.877422090455582e-06,
+      "loss": 0.0856,
+      "step": 1170
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8015382227240772,
+      "learning_rate": 9.877121602311246e-06,
+      "loss": 0.1614,
+      "step": 1171
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.39537338573458475,
+      "learning_rate": 9.876820750890484e-06,
+      "loss": 0.0747,
+      "step": 1172
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.42933799499545644,
+      "learning_rate": 9.876519536215701e-06,
+      "loss": 0.0895,
+      "step": 1173
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3343277895465831,
+      "learning_rate": 9.876217958309336e-06,
+      "loss": 0.0866,
+      "step": 1174
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3688364508073806,
+      "learning_rate": 9.875916017193849e-06,
+      "loss": 0.1058,
+      "step": 1175
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5198635824863487,
+      "learning_rate": 9.875613712891736e-06,
+      "loss": 0.1131,
+      "step": 1176
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4361188917910848,
+      "learning_rate": 9.87531104542551e-06,
+      "loss": 0.0862,
+      "step": 1177
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2526501907374999,
+      "learning_rate": 9.875008014817716e-06,
+      "loss": 0.044,
+      "step": 1178
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4864521841889567,
+      "learning_rate": 9.874704621090927e-06,
+      "loss": 0.0714,
+      "step": 1179
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3346313632751537,
+      "learning_rate": 9.87440086426774e-06,
+      "loss": 0.0874,
+      "step": 1180
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.472202182650732,
+      "learning_rate": 9.87409674437078e-06,
+      "loss": 0.1217,
+      "step": 1181
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3449391693934795,
+      "learning_rate": 9.873792261422702e-06,
+      "loss": 0.0856,
+      "step": 1182
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4827059193578475,
+      "learning_rate": 9.873487415446185e-06,
+      "loss": 0.1037,
+      "step": 1183
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.35985001785208826,
+      "learning_rate": 9.873182206463933e-06,
+      "loss": 0.0735,
+      "step": 1184
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.46723099621947195,
+      "learning_rate": 9.872876634498685e-06,
+      "loss": 0.1232,
+      "step": 1185
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4008016354186275,
+      "learning_rate": 9.872570699573196e-06,
+      "loss": 0.0725,
+      "step": 1186
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6539894709708747,
+      "learning_rate": 9.87226440171026e-06,
+      "loss": 0.0822,
+      "step": 1187
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.47733502515948617,
+      "learning_rate": 9.871957740932684e-06,
+      "loss": 0.0893,
+      "step": 1188
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6038160961625617,
+      "learning_rate": 9.871650717263318e-06,
+      "loss": 0.1146,
+      "step": 1189
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4996875005266435,
+      "learning_rate": 9.871343330725023e-06,
+      "loss": 0.1062,
+      "step": 1190
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.39134989921917007,
+      "learning_rate": 9.871035581340702e-06,
+      "loss": 0.0385,
+      "step": 1191
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.26909527128357086,
+      "learning_rate": 9.870727469133275e-06,
+      "loss": 0.0698,
+      "step": 1192
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.34614202734817223,
+      "learning_rate": 9.870418994125691e-06,
+      "loss": 0.0921,
+      "step": 1193
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.34364987840075567,
+      "learning_rate": 9.870110156340928e-06,
+      "loss": 0.0586,
+      "step": 1194
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.40773919101470035,
+      "learning_rate": 9.86980095580199e-06,
+      "loss": 0.0678,
+      "step": 1195
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3318687276539297,
+      "learning_rate": 9.869491392531908e-06,
+      "loss": 0.0723,
+      "step": 1196
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.36136255398538897,
+      "learning_rate": 9.869181466553742e-06,
+      "loss": 0.0827,
+      "step": 1197
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4334417512648093,
+      "learning_rate": 9.868871177890573e-06,
+      "loss": 0.0932,
+      "step": 1198
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7050453338847832,
+      "learning_rate": 9.868560526565516e-06,
+      "loss": 0.1079,
+      "step": 1199
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5398061543496713,
+      "learning_rate": 9.86824951260171e-06,
+      "loss": 0.0787,
+      "step": 1200
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3049898176892678,
+      "learning_rate": 9.86793813602232e-06,
+      "loss": 0.0513,
+      "step": 1201
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.31541742675691486,
+      "learning_rate": 9.867626396850541e-06,
+      "loss": 0.0456,
+      "step": 1202
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5775713344156369,
+      "learning_rate": 9.867314295109592e-06,
+      "loss": 0.1574,
+      "step": 1203
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.40296162967650134,
+      "learning_rate": 9.867001830822717e-06,
+      "loss": 0.0919,
+      "step": 1204
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7315004777882081,
+      "learning_rate": 9.866689004013196e-06,
+      "loss": 0.1468,
+      "step": 1205
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3298272281379322,
+      "learning_rate": 9.866375814704328e-06,
+      "loss": 0.0731,
+      "step": 1206
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4528189488768842,
+      "learning_rate": 9.86606226291944e-06,
+      "loss": 0.1006,
+      "step": 1207
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5477172243983482,
+      "learning_rate": 9.865748348681888e-06,
+      "loss": 0.1113,
+      "step": 1208
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.32433308395056626,
+      "learning_rate": 9.865434072015051e-06,
+      "loss": 0.0751,
+      "step": 1209
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.48099240620608635,
+      "learning_rate": 9.865119432942344e-06,
+      "loss": 0.0941,
+      "step": 1210
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.48147512410297716,
+      "learning_rate": 9.864804431487201e-06,
+      "loss": 0.0863,
+      "step": 1211
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4860049504634328,
+      "learning_rate": 9.864489067673082e-06,
+      "loss": 0.0948,
+      "step": 1212
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3669669481662148,
+      "learning_rate": 9.86417334152348e-06,
+      "loss": 0.0651,
+      "step": 1213
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.511229237392076,
+      "learning_rate": 9.863857253061913e-06,
+      "loss": 0.1042,
+      "step": 1214
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.38455495902357195,
+      "learning_rate": 9.863540802311923e-06,
+      "loss": 0.1037,
+      "step": 1215
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4074745569179462,
+      "learning_rate": 9.863223989297081e-06,
+      "loss": 0.0965,
+      "step": 1216
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7209706079615362,
+      "learning_rate": 9.862906814040987e-06,
+      "loss": 0.1246,
+      "step": 1217
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.47932788755397343,
+      "learning_rate": 9.862589276567263e-06,
+      "loss": 0.0949,
+      "step": 1218
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4540557136197089,
+      "learning_rate": 9.862271376899564e-06,
+      "loss": 0.1039,
+      "step": 1219
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3220968944594814,
+      "learning_rate": 9.861953115061568e-06,
+      "loss": 0.0987,
+      "step": 1220
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.33041159784337465,
+      "learning_rate": 9.861634491076984e-06,
+      "loss": 0.0688,
+      "step": 1221
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2704855424471933,
+      "learning_rate": 9.861315504969538e-06,
+      "loss": 0.0877,
+      "step": 1222
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3046283034984283,
+      "learning_rate": 9.860996156762996e-06,
+      "loss": 0.0671,
+      "step": 1223
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3617303088067,
+      "learning_rate": 9.860676446481142e-06,
+      "loss": 0.1307,
+      "step": 1224
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3985994219892503,
+      "learning_rate": 9.860356374147791e-06,
+      "loss": 0.0969,
+      "step": 1225
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5880517326984759,
+      "learning_rate": 9.860035939786783e-06,
+      "loss": 0.0872,
+      "step": 1226
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.360573780651818,
+      "learning_rate": 9.859715143421988e-06,
+      "loss": 0.09,
+      "step": 1227
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.495907523916959,
+      "learning_rate": 9.859393985077298e-06,
+      "loss": 0.1039,
+      "step": 1228
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.41052371290696726,
+      "learning_rate": 9.859072464776636e-06,
+      "loss": 0.1004,
+      "step": 1229
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.36047426973629865,
+      "learning_rate": 9.85875058254395e-06,
+      "loss": 0.0683,
+      "step": 1230
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.45866046373925823,
+      "learning_rate": 9.858428338403217e-06,
+      "loss": 0.0935,
+      "step": 1231
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4449135382546556,
+      "learning_rate": 9.85810573237844e-06,
+      "loss": 0.0766,
+      "step": 1232
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.569013932285049,
+      "learning_rate": 9.857782764493647e-06,
+      "loss": 0.1356,
+      "step": 1233
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.1886984855398039,
+      "learning_rate": 9.857459434772894e-06,
+      "loss": 0.0288,
+      "step": 1234
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0051714802112395,
+      "learning_rate": 9.857135743240264e-06,
+      "loss": 0.1358,
+      "step": 1235
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6961789002776256,
+      "learning_rate": 9.856811689919872e-06,
+      "loss": 0.0788,
+      "step": 1236
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.33539951876672575,
+      "learning_rate": 9.856487274835851e-06,
+      "loss": 0.105,
+      "step": 1237
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4288445489380213,
+      "learning_rate": 9.856162498012367e-06,
+      "loss": 0.0576,
+      "step": 1238
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3838948328955162,
+      "learning_rate": 9.855837359473611e-06,
+      "loss": 0.0936,
+      "step": 1239
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4027046159241728,
+      "learning_rate": 9.8555118592438e-06,
+      "loss": 0.0843,
+      "step": 1240
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5091845706480014,
+      "learning_rate": 9.855185997347183e-06,
+      "loss": 0.122,
+      "step": 1241
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.39853686330473076,
+      "learning_rate": 9.854859773808027e-06,
+      "loss": 0.0832,
+      "step": 1242
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.46219044288858957,
+      "learning_rate": 9.854533188650635e-06,
+      "loss": 0.1113,
+      "step": 1243
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.37033310759865845,
+      "learning_rate": 9.85420624189933e-06,
+      "loss": 0.0945,
+      "step": 1244
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.23326756366660525,
+      "learning_rate": 9.853878933578466e-06,
+      "loss": 0.0728,
+      "step": 1245
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.47002898762042333,
+      "learning_rate": 9.853551263712423e-06,
+      "loss": 0.0911,
+      "step": 1246
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4351633588514563,
+      "learning_rate": 9.853223232325608e-06,
+      "loss": 0.0791,
+      "step": 1247
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7623362002711589,
+      "learning_rate": 9.852894839442455e-06,
+      "loss": 0.1256,
+      "step": 1248
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.24963041638359088,
+      "learning_rate": 9.852566085087426e-06,
+      "loss": 0.0634,
+      "step": 1249
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.45241593489806203,
+      "learning_rate": 9.852236969285005e-06,
+      "loss": 0.1335,
+      "step": 1250
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.46437209591780204,
+      "learning_rate": 9.851907492059707e-06,
+      "loss": 0.0808,
+      "step": 1251
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.41086595534838877,
+      "learning_rate": 9.851577653436075e-06,
+      "loss": 0.1129,
+      "step": 1252
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5962031113872454,
+      "learning_rate": 9.851247453438678e-06,
+      "loss": 0.1286,
+      "step": 1253
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5939079794126101,
+      "learning_rate": 9.85091689209211e-06,
+      "loss": 0.1219,
+      "step": 1254
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3017344258040954,
+      "learning_rate": 9.850585969420991e-06,
+      "loss": 0.0691,
+      "step": 1255
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5315405208041578,
+      "learning_rate": 9.850254685449976e-06,
+      "loss": 0.1097,
+      "step": 1256
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3511845162819431,
+      "learning_rate": 9.849923040203734e-06,
+      "loss": 0.0665,
+      "step": 1257
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5540155088885919,
+      "learning_rate": 9.849591033706971e-06,
+      "loss": 0.1069,
+      "step": 1258
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.26963567578609043,
+      "learning_rate": 9.849258665984421e-06,
+      "loss": 0.0772,
+      "step": 1259
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.46404570136747586,
+      "learning_rate": 9.848925937060832e-06,
+      "loss": 0.1032,
+      "step": 1260
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4293370949482611,
+      "learning_rate": 9.848592846960994e-06,
+      "loss": 0.0946,
+      "step": 1261
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.45579782108471345,
+      "learning_rate": 9.848259395709715e-06,
+      "loss": 0.1343,
+      "step": 1262
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.43670842211234284,
+      "learning_rate": 9.847925583331834e-06,
+      "loss": 0.1187,
+      "step": 1263
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4089442402537916,
+      "learning_rate": 9.847591409852213e-06,
+      "loss": 0.0836,
+      "step": 1264
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5382432399614083,
+      "learning_rate": 9.847256875295746e-06,
+      "loss": 0.1279,
+      "step": 1265
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3925785502754681,
+      "learning_rate": 9.846921979687348e-06,
+      "loss": 0.0543,
+      "step": 1266
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4414898908721815,
+      "learning_rate": 9.846586723051966e-06,
+      "loss": 0.0639,
+      "step": 1267
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.41429972405256316,
+      "learning_rate": 9.846251105414572e-06,
+      "loss": 0.1118,
+      "step": 1268
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7072948621088979,
+      "learning_rate": 9.845915126800164e-06,
+      "loss": 0.1373,
+      "step": 1269
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.34578062340533344,
+      "learning_rate": 9.845578787233767e-06,
+      "loss": 0.0881,
+      "step": 1270
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4362800768538075,
+      "learning_rate": 9.845242086740436e-06,
+      "loss": 0.1128,
+      "step": 1271
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5682282971247081,
+      "learning_rate": 9.844905025345247e-06,
+      "loss": 0.1243,
+      "step": 1272
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4859135283334354,
+      "learning_rate": 9.84456760307331e-06,
+      "loss": 0.1186,
+      "step": 1273
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.47992981971449605,
+      "learning_rate": 9.844229819949754e-06,
+      "loss": 0.1246,
+      "step": 1274
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.508534007981971,
+      "learning_rate": 9.843891675999742e-06,
+      "loss": 0.1171,
+      "step": 1275
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5665705368711783,
+      "learning_rate": 9.84355317124846e-06,
+      "loss": 0.1172,
+      "step": 1276
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1373088949594616,
+      "learning_rate": 9.843214305721124e-06,
+      "loss": 0.1557,
+      "step": 1277
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.39902945956492036,
+      "learning_rate": 9.842875079442971e-06,
+      "loss": 0.1054,
+      "step": 1278
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5074053889789976,
+      "learning_rate": 9.84253549243927e-06,
+      "loss": 0.1109,
+      "step": 1279
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.428133882775729,
+      "learning_rate": 9.842195544735316e-06,
+      "loss": 0.12,
+      "step": 1280
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3341129443101746,
+      "learning_rate": 9.84185523635643e-06,
+      "loss": 0.0516,
+      "step": 1281
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4310996618024957,
+      "learning_rate": 9.84151456732796e-06,
+      "loss": 0.0951,
+      "step": 1282
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3570990504991548,
+      "learning_rate": 9.841173537675281e-06,
+      "loss": 0.0802,
+      "step": 1283
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3346941823062726,
+      "learning_rate": 9.840832147423797e-06,
+      "loss": 0.0873,
+      "step": 1284
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5668108591973479,
+      "learning_rate": 9.840490396598933e-06,
+      "loss": 0.0868,
+      "step": 1285
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.39383363487851014,
+      "learning_rate": 9.840148285226145e-06,
+      "loss": 0.1166,
+      "step": 1286
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.39024185833878483,
+      "learning_rate": 9.83980581333092e-06,
+      "loss": 0.1306,
+      "step": 1287
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6024058693222207,
+      "learning_rate": 9.83946298093876e-06,
+      "loss": 0.0887,
+      "step": 1288
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4576632950069507,
+      "learning_rate": 9.83911978807521e-06,
+      "loss": 0.075,
+      "step": 1289
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3344338753325829,
+      "learning_rate": 9.838776234765824e-06,
+      "loss": 0.076,
+      "step": 1290
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.46870497420785795,
+      "learning_rate": 9.838432321036198e-06,
+      "loss": 0.0929,
+      "step": 1291
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3780791653476731,
+      "learning_rate": 9.838088046911946e-06,
+      "loss": 0.1106,
+      "step": 1292
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.46173724271856387,
+      "learning_rate": 9.837743412418714e-06,
+      "loss": 0.0595,
+      "step": 1293
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7624166420334992,
+      "learning_rate": 9.837398417582169e-06,
+      "loss": 0.0999,
+      "step": 1294
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.45669878246743767,
+      "learning_rate": 9.83705306242801e-06,
+      "loss": 0.0537,
+      "step": 1295
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7503163272939328,
+      "learning_rate": 9.836707346981962e-06,
+      "loss": 0.1638,
+      "step": 1296
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5718920516006625,
+      "learning_rate": 9.836361271269773e-06,
+      "loss": 0.1184,
+      "step": 1297
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2638980797443833,
+      "learning_rate": 9.836014835317227e-06,
+      "loss": 0.0849,
+      "step": 1298
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5333538831225642,
+      "learning_rate": 9.835668039150119e-06,
+      "loss": 0.0726,
+      "step": 1299
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4678499855716581,
+      "learning_rate": 9.835320882794289e-06,
+      "loss": 0.0694,
+      "step": 1300
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5893940976654404,
+      "learning_rate": 9.83497336627559e-06,
+      "loss": 0.1115,
+      "step": 1301
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.477520473673102,
+      "learning_rate": 9.83462548961991e-06,
+      "loss": 0.1114,
+      "step": 1302
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.38339227477600996,
+      "learning_rate": 9.834277252853159e-06,
+      "loss": 0.0903,
+      "step": 1303
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.40315468654635633,
+      "learning_rate": 9.833928656001277e-06,
+      "loss": 0.0734,
+      "step": 1304
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.51326273500667,
+      "learning_rate": 9.833579699090228e-06,
+      "loss": 0.094,
+      "step": 1305
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4324557356298144,
+      "learning_rate": 9.833230382146007e-06,
+      "loss": 0.1127,
+      "step": 1306
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.43447306946349384,
+      "learning_rate": 9.83288070519463e-06,
+      "loss": 0.0804,
+      "step": 1307
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.37318183925231924,
+      "learning_rate": 9.832530668262146e-06,
+      "loss": 0.0755,
+      "step": 1308
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.393211348658576,
+      "learning_rate": 9.832180271374625e-06,
+      "loss": 0.1092,
+      "step": 1309
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4063557255190413,
+      "learning_rate": 9.831829514558167e-06,
+      "loss": 0.1061,
+      "step": 1310
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3765834886207673,
+      "learning_rate": 9.8314783978389e-06,
+      "loss": 0.1059,
+      "step": 1311
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3213958970422227,
+      "learning_rate": 9.831126921242977e-06,
+      "loss": 0.0835,
+      "step": 1312
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3510965502673136,
+      "learning_rate": 9.830775084796578e-06,
+      "loss": 0.1019,
+      "step": 1313
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3664726620910326,
+      "learning_rate": 9.830422888525908e-06,
+      "loss": 0.0804,
+      "step": 1314
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5505245121910278,
+      "learning_rate": 9.830070332457203e-06,
+      "loss": 0.0466,
+      "step": 1315
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4574474443870217,
+      "learning_rate": 9.829717416616723e-06,
+      "loss": 0.1107,
+      "step": 1316
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.28426321714867947,
+      "learning_rate": 9.829364141030753e-06,
+      "loss": 0.0651,
+      "step": 1317
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4561678875086119,
+      "learning_rate": 9.82901050572561e-06,
+      "loss": 0.1108,
+      "step": 1318
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3974288253269227,
+      "learning_rate": 9.828656510727632e-06,
+      "loss": 0.1201,
+      "step": 1319
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6102372108891713,
+      "learning_rate": 9.82830215606319e-06,
+      "loss": 0.1389,
+      "step": 1320
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4923301976604247,
+      "learning_rate": 9.827947441758675e-06,
+      "loss": 0.1188,
+      "step": 1321
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3768478512027931,
+      "learning_rate": 9.827592367840509e-06,
+      "loss": 0.0811,
+      "step": 1322
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4950674192325829,
+      "learning_rate": 9.827236934335142e-06,
+      "loss": 0.1114,
+      "step": 1323
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3985044521628323,
+      "learning_rate": 9.826881141269046e-06,
+      "loss": 0.0896,
+      "step": 1324
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4718674135071454,
+      "learning_rate": 9.826524988668727e-06,
+      "loss": 0.0702,
+      "step": 1325
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4688557173646293,
+      "learning_rate": 9.826168476560707e-06,
+      "loss": 0.0917,
+      "step": 1326
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4702698560538927,
+      "learning_rate": 9.825811604971546e-06,
+      "loss": 0.0953,
+      "step": 1327
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2966485367414335,
+      "learning_rate": 9.825454373927824e-06,
+      "loss": 0.0561,
+      "step": 1328
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4606986284646067,
+      "learning_rate": 9.82509678345615e-06,
+      "loss": 0.0979,
+      "step": 1329
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3882208342924487,
+      "learning_rate": 9.824738833583158e-06,
+      "loss": 0.0962,
+      "step": 1330
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3769168571737713,
+      "learning_rate": 9.824380524335513e-06,
+      "loss": 0.1003,
+      "step": 1331
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.43139468527300157,
+      "learning_rate": 9.824021855739902e-06,
+      "loss": 0.1085,
+      "step": 1332
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5252124543080456,
+      "learning_rate": 9.823662827823041e-06,
+      "loss": 0.1134,
+      "step": 1333
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2838724211694334,
+      "learning_rate": 9.823303440611673e-06,
+      "loss": 0.071,
+      "step": 1334
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5720282344966142,
+      "learning_rate": 9.822943694132566e-06,
+      "loss": 0.1014,
+      "step": 1335
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5149843547110945,
+      "learning_rate": 9.82258358841252e-06,
+      "loss": 0.1188,
+      "step": 1336
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4758153241786766,
+      "learning_rate": 9.822223123478352e-06,
+      "loss": 0.1067,
+      "step": 1337
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.497953222052105,
+      "learning_rate": 9.821862299356915e-06,
+      "loss": 0.0819,
+      "step": 1338
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.35332746669853893,
+      "learning_rate": 9.821501116075083e-06,
+      "loss": 0.0825,
+      "step": 1339
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2903086115613855,
+      "learning_rate": 9.821139573659762e-06,
+      "loss": 0.0585,
+      "step": 1340
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6606346464925431,
+      "learning_rate": 9.82077767213788e-06,
+      "loss": 0.1531,
+      "step": 1341
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.39219895899805657,
+      "learning_rate": 9.820415411536393e-06,
+      "loss": 0.0583,
+      "step": 1342
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7343511264949762,
+      "learning_rate": 9.820052791882286e-06,
+      "loss": 0.1182,
+      "step": 1343
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.46109748738601714,
+      "learning_rate": 9.819689813202568e-06,
+      "loss": 0.0826,
+      "step": 1344
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3476703006514118,
+      "learning_rate": 9.819326475524278e-06,
+      "loss": 0.071,
+      "step": 1345
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4376253786638165,
+      "learning_rate": 9.818962778874474e-06,
+      "loss": 0.0956,
+      "step": 1346
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.35602944680191645,
+      "learning_rate": 9.818598723280252e-06,
+      "loss": 0.109,
+      "step": 1347
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5374322508878858,
+      "learning_rate": 9.818234308768725e-06,
+      "loss": 0.1135,
+      "step": 1348
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.38168161229482556,
+      "learning_rate": 9.81786953536704e-06,
+      "loss": 0.1008,
+      "step": 1349
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4381885207331545,
+      "learning_rate": 9.817504403102366e-06,
+      "loss": 0.0812,
+      "step": 1350
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5372639360914435,
+      "learning_rate": 9.8171389120019e-06,
+      "loss": 0.0988,
+      "step": 1351
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.40129652875460275,
+      "learning_rate": 9.816773062092863e-06,
+      "loss": 0.1021,
+      "step": 1352
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3508178829775055,
+      "learning_rate": 9.81640685340251e-06,
+      "loss": 0.084,
+      "step": 1353
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.362383745159872,
+      "learning_rate": 9.81604028595812e-06,
+      "loss": 0.067,
+      "step": 1354
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5963392936872175,
+      "learning_rate": 9.815673359786991e-06,
+      "loss": 0.1207,
+      "step": 1355
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3822166658436363,
+      "learning_rate": 9.815306074916458e-06,
+      "loss": 0.0442,
+      "step": 1356
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3639690691632694,
+      "learning_rate": 9.814938431373877e-06,
+      "loss": 0.0764,
+      "step": 1357
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.40571712875780264,
+      "learning_rate": 9.814570429186631e-06,
+      "loss": 0.0743,
+      "step": 1358
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3526182468905231,
+      "learning_rate": 9.814202068382134e-06,
+      "loss": 0.0526,
+      "step": 1359
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6044733411657046,
+      "learning_rate": 9.813833348987823e-06,
+      "loss": 0.1172,
+      "step": 1360
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2741004641602592,
+      "learning_rate": 9.81346427103116e-06,
+      "loss": 0.063,
+      "step": 1361
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2146410708673758,
+      "learning_rate": 9.813094834539638e-06,
+      "loss": 0.138,
+      "step": 1362
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4414791627373951,
+      "learning_rate": 9.812725039540774e-06,
+      "loss": 0.1242,
+      "step": 1363
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5122967093166859,
+      "learning_rate": 9.812354886062115e-06,
+      "loss": 0.0983,
+      "step": 1364
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4213130077276111,
+      "learning_rate": 9.811984374131227e-06,
+      "loss": 0.0917,
+      "step": 1365
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3029624905098351,
+      "learning_rate": 9.811613503775712e-06,
+      "loss": 0.0603,
+      "step": 1366
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9774825380133516,
+      "learning_rate": 9.811242275023194e-06,
+      "loss": 0.1562,
+      "step": 1367
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7891023083916583,
+      "learning_rate": 9.810870687901324e-06,
+      "loss": 0.1114,
+      "step": 1368
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.23904246159783715,
+      "learning_rate": 9.81049874243778e-06,
+      "loss": 0.0631,
+      "step": 1369
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3502714078561316,
+      "learning_rate": 9.810126438660265e-06,
+      "loss": 0.078,
+      "step": 1370
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4784032461045025,
+      "learning_rate": 9.809753776596512e-06,
+      "loss": 0.1252,
+      "step": 1371
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5541415928311386,
+      "learning_rate": 9.80938075627428e-06,
+      "loss": 0.1196,
+      "step": 1372
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3907293810137037,
+      "learning_rate": 9.80900737772135e-06,
+      "loss": 0.0822,
+      "step": 1373
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5785635564179468,
+      "learning_rate": 9.808633640965538e-06,
+      "loss": 0.1094,
+      "step": 1374
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4639965579304603,
+      "learning_rate": 9.808259546034682e-06,
+      "loss": 0.0891,
+      "step": 1375
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.29863717705606946,
+      "learning_rate": 9.807885092956642e-06,
+      "loss": 0.0603,
+      "step": 1376
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.44227780939442596,
+      "learning_rate": 9.807510281759314e-06,
+      "loss": 0.0852,
+      "step": 1377
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8270860320808165,
+      "learning_rate": 9.807135112470612e-06,
+      "loss": 0.1399,
+      "step": 1378
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5926787271392061,
+      "learning_rate": 9.806759585118486e-06,
+      "loss": 0.1182,
+      "step": 1379
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.32695645529656936,
+      "learning_rate": 9.806383699730904e-06,
+      "loss": 0.0828,
+      "step": 1380
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3262736264844147,
+      "learning_rate": 9.806007456335863e-06,
+      "loss": 0.0674,
+      "step": 1381
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4038810033427228,
+      "learning_rate": 9.805630854961391e-06,
+      "loss": 0.0995,
+      "step": 1382
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.26936718077686644,
+      "learning_rate": 9.805253895635537e-06,
+      "loss": 0.0759,
+      "step": 1383
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6008822625180089,
+      "learning_rate": 9.804876578386381e-06,
+      "loss": 0.1454,
+      "step": 1384
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3649423515980822,
+      "learning_rate": 9.804498903242028e-06,
+      "loss": 0.0993,
+      "step": 1385
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7992837295237342,
+      "learning_rate": 9.804120870230606e-06,
+      "loss": 0.1526,
+      "step": 1386
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.48882359371714723,
+      "learning_rate": 9.803742479380277e-06,
+      "loss": 0.0973,
+      "step": 1387
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4174530827679763,
+      "learning_rate": 9.803363730719224e-06,
+      "loss": 0.1082,
+      "step": 1388
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.49091856576080906,
+      "learning_rate": 9.80298462427566e-06,
+      "loss": 0.1306,
+      "step": 1389
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.35527812814084153,
+      "learning_rate": 9.802605160077821e-06,
+      "loss": 0.1086,
+      "step": 1390
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.44082350753822475,
+      "learning_rate": 9.80222533815397e-06,
+      "loss": 0.1009,
+      "step": 1391
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.26992322616007103,
+      "learning_rate": 9.801845158532404e-06,
+      "loss": 0.0636,
+      "step": 1392
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4130224338156468,
+      "learning_rate": 9.801464621241437e-06,
+      "loss": 0.094,
+      "step": 1393
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3125403628838735,
+      "learning_rate": 9.801083726309412e-06,
+      "loss": 0.0779,
+      "step": 1394
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5812029215382618,
+      "learning_rate": 9.800702473764706e-06,
+      "loss": 0.083,
+      "step": 1395
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3707747195477804,
+      "learning_rate": 9.800320863635711e-06,
+      "loss": 0.0979,
+      "step": 1396
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4489089865020385,
+      "learning_rate": 9.799938895950856e-06,
+      "loss": 0.0889,
+      "step": 1397
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.21992947453486855,
+      "learning_rate": 9.799556570738589e-06,
+      "loss": 0.0574,
+      "step": 1398
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6087719667274801,
+      "learning_rate": 9.79917388802739e-06,
+      "loss": 0.1298,
+      "step": 1399
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.41023285218101163,
+      "learning_rate": 9.798790847845763e-06,
+      "loss": 0.1107,
+      "step": 1400
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.38169174577252807,
+      "learning_rate": 9.798407450222238e-06,
+      "loss": 0.1154,
+      "step": 1401
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.39207411104779605,
+      "learning_rate": 9.798023695185372e-06,
+      "loss": 0.0927,
+      "step": 1402
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3215208090495359,
+      "learning_rate": 9.797639582763752e-06,
+      "loss": 0.0862,
+      "step": 1403
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3958789135768561,
+      "learning_rate": 9.797255112985988e-06,
+      "loss": 0.1204,
+      "step": 1404
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.44916089442550017,
+      "learning_rate": 9.796870285880717e-06,
+      "loss": 0.0949,
+      "step": 1405
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5431412185923403,
+      "learning_rate": 9.796485101476603e-06,
+      "loss": 0.09,
+      "step": 1406
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3045440325237794,
+      "learning_rate": 9.796099559802337e-06,
+      "loss": 0.0727,
+      "step": 1407
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8148253383767128,
+      "learning_rate": 9.795713660886636e-06,
+      "loss": 0.1166,
+      "step": 1408
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6464125069707106,
+      "learning_rate": 9.795327404758244e-06,
+      "loss": 0.1447,
+      "step": 1409
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.42511849008305275,
+      "learning_rate": 9.794940791445933e-06,
+      "loss": 0.1147,
+      "step": 1410
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.325528962918417,
+      "learning_rate": 9.7945538209785e-06,
+      "loss": 0.1047,
+      "step": 1411
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.28709517419611796,
+      "learning_rate": 9.794166493384766e-06,
+      "loss": 0.0704,
+      "step": 1412
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6045332845168807,
+      "learning_rate": 9.793778808693586e-06,
+      "loss": 0.1178,
+      "step": 1413
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4247763840892641,
+      "learning_rate": 9.793390766933832e-06,
+      "loss": 0.0937,
+      "step": 1414
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2320937815419936,
+      "learning_rate": 9.793002368134412e-06,
+      "loss": 0.0432,
+      "step": 1415
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2501713040377906,
+      "learning_rate": 9.792613612324253e-06,
+      "loss": 0.0772,
+      "step": 1416
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.282612051125166,
+      "learning_rate": 9.792224499532315e-06,
+      "loss": 0.0843,
+      "step": 1417
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3194330895928499,
+      "learning_rate": 9.791835029787578e-06,
+      "loss": 0.0907,
+      "step": 1418
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.41584849893677234,
+      "learning_rate": 9.791445203119054e-06,
+      "loss": 0.0864,
+      "step": 1419
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6302388247600794,
+      "learning_rate": 9.791055019555777e-06,
+      "loss": 0.1078,
+      "step": 1420
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2695493146066337,
+      "learning_rate": 9.790664479126815e-06,
+      "loss": 0.0714,
+      "step": 1421
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.34442803169797387,
+      "learning_rate": 9.790273581861255e-06,
+      "loss": 0.0962,
+      "step": 1422
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.30369544184414865,
+      "learning_rate": 9.78988232778821e-06,
+      "loss": 0.0432,
+      "step": 1423
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.27589792218212256,
+      "learning_rate": 9.78949071693683e-06,
+      "loss": 0.076,
+      "step": 1424
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5188388731361412,
+      "learning_rate": 9.789098749336276e-06,
+      "loss": 0.1149,
+      "step": 1425
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2800051036238471,
+      "learning_rate": 9.788706425015752e-06,
+      "loss": 0.0783,
+      "step": 1426
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4115534793843664,
+      "learning_rate": 9.788313744004476e-06,
+      "loss": 0.09,
+      "step": 1427
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2957672713403208,
+      "learning_rate": 9.7879207063317e-06,
+      "loss": 0.0852,
+      "step": 1428
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4437042497339269,
+      "learning_rate": 9.787527312026697e-06,
+      "loss": 0.0897,
+      "step": 1429
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.570968232312105,
+      "learning_rate": 9.78713356111877e-06,
+      "loss": 0.1192,
+      "step": 1430
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3822117959482978,
+      "learning_rate": 9.786739453637249e-06,
+      "loss": 0.0655,
+      "step": 1431
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.27197603581802404,
+      "learning_rate": 9.786344989611488e-06,
+      "loss": 0.0577,
+      "step": 1432
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.604075883660989,
+      "learning_rate": 9.78595016907087e-06,
+      "loss": 0.1456,
+      "step": 1433
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4808961241870662,
+      "learning_rate": 9.785554992044804e-06,
+      "loss": 0.1304,
+      "step": 1434
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.38574323916759173,
+      "learning_rate": 9.785159458562721e-06,
+      "loss": 0.0763,
+      "step": 1435
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3827481021051543,
+      "learning_rate": 9.78476356865409e-06,
+      "loss": 0.0606,
+      "step": 1436
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6247257813028966,
+      "learning_rate": 9.784367322348392e-06,
+      "loss": 0.0985,
+      "step": 1437
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4474587601874192,
+      "learning_rate": 9.783970719675146e-06,
+      "loss": 0.1282,
+      "step": 1438
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4393677720124891,
+      "learning_rate": 9.783573760663893e-06,
+      "loss": 0.1201,
+      "step": 1439
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.30710021209559485,
+      "learning_rate": 9.783176445344198e-06,
+      "loss": 0.0731,
+      "step": 1440
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3748985205420724,
+      "learning_rate": 9.78277877374566e-06,
+      "loss": 0.0966,
+      "step": 1441
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3210672923903104,
+      "learning_rate": 9.782380745897894e-06,
+      "loss": 0.1068,
+      "step": 1442
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.26672478574070213,
+      "learning_rate": 9.781982361830553e-06,
+      "loss": 0.089,
+      "step": 1443
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.362470859534081,
+      "learning_rate": 9.781583621573309e-06,
+      "loss": 0.0669,
+      "step": 1444
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3917291628731751,
+      "learning_rate": 9.78118452515586e-06,
+      "loss": 0.0647,
+      "step": 1445
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3293239345465708,
+      "learning_rate": 9.780785072607937e-06,
+      "loss": 0.0691,
+      "step": 1446
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4526544953189884,
+      "learning_rate": 9.780385263959291e-06,
+      "loss": 0.0966,
+      "step": 1447
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5524286790280818,
+      "learning_rate": 9.779985099239702e-06,
+      "loss": 0.1709,
+      "step": 1448
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5913113163272414,
+      "learning_rate": 9.779584578478978e-06,
+      "loss": 0.126,
+      "step": 1449
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2589046747675721,
+      "learning_rate": 9.779183701706952e-06,
+      "loss": 0.0756,
+      "step": 1450
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5108901921863935,
+      "learning_rate": 9.778782468953482e-06,
+      "loss": 0.0973,
+      "step": 1451
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5747416209585868,
+      "learning_rate": 9.778380880248455e-06,
+      "loss": 0.0904,
+      "step": 1452
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.29778006043222877,
+      "learning_rate": 9.777978935621785e-06,
+      "loss": 0.0901,
+      "step": 1453
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.29081784685120615,
+      "learning_rate": 9.77757663510341e-06,
+      "loss": 0.0569,
+      "step": 1454
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3465458995387425,
+      "learning_rate": 9.777173978723294e-06,
+      "loss": 0.125,
+      "step": 1455
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2110996505630479,
+      "learning_rate": 9.776770966511433e-06,
+      "loss": 0.0442,
+      "step": 1456
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3458274942899966,
+      "learning_rate": 9.776367598497842e-06,
+      "loss": 0.0923,
+      "step": 1457
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4062644811860463,
+      "learning_rate": 9.775963874712569e-06,
+      "loss": 0.0922,
+      "step": 1458
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3847518792444242,
+      "learning_rate": 9.775559795185686e-06,
+      "loss": 0.098,
+      "step": 1459
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.39431533621486803,
+      "learning_rate": 9.775155359947287e-06,
+      "loss": 0.0857,
+      "step": 1460
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2839981351990572,
+      "learning_rate": 9.774750569027501e-06,
+      "loss": 0.0687,
+      "step": 1461
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.49376392735683666,
+      "learning_rate": 9.774345422456477e-06,
+      "loss": 0.1218,
+      "step": 1462
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.41792793792763144,
+      "learning_rate": 9.773939920264395e-06,
+      "loss": 0.1042,
+      "step": 1463
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.40335358598153637,
+      "learning_rate": 9.773534062481455e-06,
+      "loss": 0.0859,
+      "step": 1464
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.36173800192963784,
+      "learning_rate": 9.77312784913789e-06,
+      "loss": 0.0829,
+      "step": 1465
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3627514492403434,
+      "learning_rate": 9.772721280263959e-06,
+      "loss": 0.1098,
+      "step": 1466
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4137791524888132,
+      "learning_rate": 9.772314355889945e-06,
+      "loss": 0.0991,
+      "step": 1467
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3965765294153142,
+      "learning_rate": 9.771907076046155e-06,
+      "loss": 0.0629,
+      "step": 1468
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.35825354795333814,
+      "learning_rate": 9.771499440762929e-06,
+      "loss": 0.0927,
+      "step": 1469
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4480681115067083,
+      "learning_rate": 9.771091450070627e-06,
+      "loss": 0.0669,
+      "step": 1470
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.29450196884852836,
+      "learning_rate": 9.77068310399964e-06,
+      "loss": 0.0605,
+      "step": 1471
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5615277442962547,
+      "learning_rate": 9.770274402580385e-06,
+      "loss": 0.1131,
+      "step": 1472
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2560883737250006,
+      "learning_rate": 9.769865345843304e-06,
+      "loss": 0.0698,
+      "step": 1473
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5999024434049399,
+      "learning_rate": 9.769455933818863e-06,
+      "loss": 0.1164,
+      "step": 1474
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5807593233572422,
+      "learning_rate": 9.769046166537563e-06,
+      "loss": 0.1082,
+      "step": 1475
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4211429316245326,
+      "learning_rate": 9.768636044029923e-06,
+      "loss": 0.0827,
+      "step": 1476
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2663639396427144,
+      "learning_rate": 9.768225566326489e-06,
+      "loss": 0.0377,
+      "step": 1477
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3946744858762572,
+      "learning_rate": 9.767814733457838e-06,
+      "loss": 0.1337,
+      "step": 1478
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8243941062998293,
+      "learning_rate": 9.767403545454572e-06,
+      "loss": 0.1504,
+      "step": 1479
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5917378094697343,
+      "learning_rate": 9.766992002347318e-06,
+      "loss": 0.053,
+      "step": 1480
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.591172829632348,
+      "learning_rate": 9.766580104166727e-06,
+      "loss": 0.139,
+      "step": 1481
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.33226585307682716,
+      "learning_rate": 9.766167850943486e-06,
+      "loss": 0.0572,
+      "step": 1482
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.895786496231328,
+      "learning_rate": 9.765755242708296e-06,
+      "loss": 0.1375,
+      "step": 1483
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.35871154484747436,
+      "learning_rate": 9.765342279491895e-06,
+      "loss": 0.0897,
+      "step": 1484
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3370585936636479,
+      "learning_rate": 9.76492896132504e-06,
+      "loss": 0.0818,
+      "step": 1485
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5676033380940825,
+      "learning_rate": 9.76451528823852e-06,
+      "loss": 0.1279,
+      "step": 1486
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3823880623261568,
+      "learning_rate": 9.764101260263143e-06,
+      "loss": 0.0655,
+      "step": 1487
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7229106454474415,
+      "learning_rate": 9.763686877429752e-06,
+      "loss": 0.0985,
+      "step": 1488
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6770218201080149,
+      "learning_rate": 9.763272139769212e-06,
+      "loss": 0.1686,
+      "step": 1489
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5656188493193873,
+      "learning_rate": 9.762857047312414e-06,
+      "loss": 0.1082,
+      "step": 1490
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9419062253241671,
+      "learning_rate": 9.76244160009028e-06,
+      "loss": 0.1541,
+      "step": 1491
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4856216970244555,
+      "learning_rate": 9.762025798133751e-06,
+      "loss": 0.081,
+      "step": 1492
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4626360808782635,
+      "learning_rate": 9.7616096414738e-06,
+      "loss": 0.1202,
+      "step": 1493
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6374693807164482,
+      "learning_rate": 9.761193130141425e-06,
+      "loss": 0.1243,
+      "step": 1494
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.504121248979558,
+      "learning_rate": 9.760776264167647e-06,
+      "loss": 0.108,
+      "step": 1495
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7417105794810317,
+      "learning_rate": 9.760359043583521e-06,
+      "loss": 0.1274,
+      "step": 1496
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4185271924397774,
+      "learning_rate": 9.759941468420123e-06,
+      "loss": 0.1067,
+      "step": 1497
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3808547581628291,
+      "learning_rate": 9.759523538708555e-06,
+      "loss": 0.0775,
+      "step": 1498
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3599704757331957,
+      "learning_rate": 9.759105254479948e-06,
+      "loss": 0.0541,
+      "step": 1499
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.619795090551407,
+      "learning_rate": 9.758686615765457e-06,
+      "loss": 0.1144,
+      "step": 1500
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5866436827248638,
+      "learning_rate": 9.758267622596266e-06,
+      "loss": 0.1448,
+      "step": 1501
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.36986589084907723,
+      "learning_rate": 9.757848275003583e-06,
+      "loss": 0.0563,
+      "step": 1502
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4771788093118218,
+      "learning_rate": 9.757428573018645e-06,
+      "loss": 0.1292,
+      "step": 1503
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4522316304240206,
+      "learning_rate": 9.757008516672712e-06,
+      "loss": 0.0994,
+      "step": 1504
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.32293143705640487,
+      "learning_rate": 9.756588105997074e-06,
+      "loss": 0.0685,
+      "step": 1505
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6007072822506229,
+      "learning_rate": 9.756167341023043e-06,
+      "loss": 0.145,
+      "step": 1506
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6066406062507865,
+      "learning_rate": 9.755746221781963e-06,
+      "loss": 0.1176,
+      "step": 1507
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.24047478063062486,
+      "learning_rate": 9.7553247483052e-06,
+      "loss": 0.0723,
+      "step": 1508
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.24376344534015634,
+      "learning_rate": 9.754902920624148e-06,
+      "loss": 0.0849,
+      "step": 1509
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4762194888847118,
+      "learning_rate": 9.754480738770225e-06,
+      "loss": 0.0785,
+      "step": 1510
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4410817876108083,
+      "learning_rate": 9.754058202774882e-06,
+      "loss": 0.1276,
+      "step": 1511
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3560200765797871,
+      "learning_rate": 9.75363531266959e-06,
+      "loss": 0.0654,
+      "step": 1512
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.21710132432765883,
+      "learning_rate": 9.753212068485847e-06,
+      "loss": 0.0525,
+      "step": 1513
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6206189282015673,
+      "learning_rate": 9.752788470255179e-06,
+      "loss": 0.1095,
+      "step": 1514
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3617918284075503,
+      "learning_rate": 9.752364518009138e-06,
+      "loss": 0.0952,
+      "step": 1515
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9390791114575996,
+      "learning_rate": 9.751940211779304e-06,
+      "loss": 0.1227,
+      "step": 1516
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5058449185075965,
+      "learning_rate": 9.751515551597284e-06,
+      "loss": 0.0791,
+      "step": 1517
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5204834868719281,
+      "learning_rate": 9.751090537494702e-06,
+      "loss": 0.1038,
+      "step": 1518
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4013979355915156,
+      "learning_rate": 9.750665169503221e-06,
+      "loss": 0.109,
+      "step": 1519
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4063206696959503,
+      "learning_rate": 9.750239447654524e-06,
+      "loss": 0.0866,
+      "step": 1520
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5973066775160845,
+      "learning_rate": 9.749813371980321e-06,
+      "loss": 0.0708,
+      "step": 1521
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3726841628857404,
+      "learning_rate": 9.749386942512348e-06,
+      "loss": 0.0911,
+      "step": 1522
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.48477416900487214,
+      "learning_rate": 9.748960159282369e-06,
+      "loss": 0.1189,
+      "step": 1523
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.32388441148074815,
+      "learning_rate": 9.748533022322171e-06,
+      "loss": 0.052,
+      "step": 1524
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.30040425248581754,
+      "learning_rate": 9.748105531663571e-06,
+      "loss": 0.0727,
+      "step": 1525
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3790364973174266,
+      "learning_rate": 9.747677687338412e-06,
+      "loss": 0.0786,
+      "step": 1526
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.32418721650143434,
+      "learning_rate": 9.747249489378563e-06,
+      "loss": 0.0717,
+      "step": 1527
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5253222859510696,
+      "learning_rate": 9.746820937815915e-06,
+      "loss": 0.1177,
+      "step": 1528
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4208861165379929,
+      "learning_rate": 9.746392032682392e-06,
+      "loss": 0.0634,
+      "step": 1529
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.33086106031564055,
+      "learning_rate": 9.745962774009941e-06,
+      "loss": 0.0874,
+      "step": 1530
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4422043077432377,
+      "learning_rate": 9.745533161830536e-06,
+      "loss": 0.0974,
+      "step": 1531
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3339923775304685,
+      "learning_rate": 9.745103196176175e-06,
+      "loss": 0.076,
+      "step": 1532
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.35224397746724745,
+      "learning_rate": 9.744672877078887e-06,
+      "loss": 0.0817,
+      "step": 1533
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.45663814164725286,
+      "learning_rate": 9.744242204570722e-06,
+      "loss": 0.1194,
+      "step": 1534
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3875681469634877,
+      "learning_rate": 9.74381117868376e-06,
+      "loss": 0.109,
+      "step": 1535
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.35698716416209236,
+      "learning_rate": 9.743379799450107e-06,
+      "loss": 0.083,
+      "step": 1536
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6386712034474948,
+      "learning_rate": 9.742948066901895e-06,
+      "loss": 0.1668,
+      "step": 1537
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9720948229774626,
+      "learning_rate": 9.742515981071283e-06,
+      "loss": 0.117,
+      "step": 1538
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.2693157611288496,
+      "learning_rate": 9.74208354199045e-06,
+      "loss": 0.0693,
+      "step": 1539
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.45875415095510885,
+      "learning_rate": 9.741650749691613e-06,
+      "loss": 0.1143,
+      "step": 1540
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5354958449492344,
+      "learning_rate": 9.741217604207005e-06,
+      "loss": 0.1117,
+      "step": 1541
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5151087560697634,
+      "learning_rate": 9.740784105568889e-06,
+      "loss": 0.0964,
+      "step": 1542
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3328494758019325,
+      "learning_rate": 9.740350253809556e-06,
+      "loss": 0.0797,
+      "step": 1543
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.515019355627681,
+      "learning_rate": 9.739916048961323e-06,
+      "loss": 0.1488,
+      "step": 1544
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.38798563056935104,
+      "learning_rate": 9.73948149105653e-06,
+      "loss": 0.0819,
+      "step": 1545
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5629205624905728,
+      "learning_rate": 9.739046580127546e-06,
+      "loss": 0.0774,
+      "step": 1546
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3937594729089547,
+      "learning_rate": 9.738611316206766e-06,
+      "loss": 0.0808,
+      "step": 1547
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.224608788639192,
+      "learning_rate": 9.73817569932661e-06,
+      "loss": 0.0619,
+      "step": 1548
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.39501652421630695,
+      "learning_rate": 9.737739729519526e-06,
+      "loss": 0.1287,
+      "step": 1549
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.44491014834985676,
+      "learning_rate": 9.737303406817989e-06,
+      "loss": 0.1076,
+      "step": 1550
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6025989047629982,
+      "learning_rate": 9.736866731254496e-06,
+      "loss": 0.123,
+      "step": 1551
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.1704103586611847,
+      "learning_rate": 9.736429702861573e-06,
+      "loss": 0.0413,
+      "step": 1552
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4137527597554859,
+      "learning_rate": 9.735992321671776e-06,
+      "loss": 0.105,
+      "step": 1553
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.2878628426595022,
+      "learning_rate": 9.735554587717683e-06,
+      "loss": 0.0995,
+      "step": 1554
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3141072292274884,
+      "learning_rate": 9.735116501031896e-06,
+      "loss": 0.0919,
+      "step": 1555
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3478366982179641,
+      "learning_rate": 9.734678061647048e-06,
+      "loss": 0.1003,
+      "step": 1556
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5283391463332364,
+      "learning_rate": 9.734239269595797e-06,
+      "loss": 0.1633,
+      "step": 1557
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.33077142836322027,
+      "learning_rate": 9.733800124910825e-06,
+      "loss": 0.0629,
+      "step": 1558
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.21220712946516898,
+      "learning_rate": 9.733360627624844e-06,
+      "loss": 0.061,
+      "step": 1559
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.29246371399706167,
+      "learning_rate": 9.73292077777059e-06,
+      "loss": 0.0603,
+      "step": 1560
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.38209044511873574,
+      "learning_rate": 9.732480575380825e-06,
+      "loss": 0.092,
+      "step": 1561
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5845083078941006,
+      "learning_rate": 9.732040020488338e-06,
+      "loss": 0.1317,
+      "step": 1562
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5188095227098074,
+      "learning_rate": 9.731599113125943e-06,
+      "loss": 0.1195,
+      "step": 1563
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.33028395468363947,
+      "learning_rate": 9.731157853326483e-06,
+      "loss": 0.0828,
+      "step": 1564
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3254887091671216,
+      "learning_rate": 9.730716241122826e-06,
+      "loss": 0.0746,
+      "step": 1565
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6073445183841407,
+      "learning_rate": 9.730274276547864e-06,
+      "loss": 0.1195,
+      "step": 1566
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.49404946944528094,
+      "learning_rate": 9.72983195963452e-06,
+      "loss": 0.0801,
+      "step": 1567
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3207447856836685,
+      "learning_rate": 9.729389290415734e-06,
+      "loss": 0.0936,
+      "step": 1568
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.26726629706690463,
+      "learning_rate": 9.728946268924485e-06,
+      "loss": 0.0538,
+      "step": 1569
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8465957917259104,
+      "learning_rate": 9.728502895193772e-06,
+      "loss": 0.1659,
+      "step": 1570
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.32026298132815856,
+      "learning_rate": 9.728059169256616e-06,
+      "loss": 0.0577,
+      "step": 1571
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6773602149451668,
+      "learning_rate": 9.72761509114607e-06,
+      "loss": 0.1323,
+      "step": 1572
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4173663515117522,
+      "learning_rate": 9.727170660895213e-06,
+      "loss": 0.1018,
+      "step": 1573
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.36720059562511537,
+      "learning_rate": 9.726725878537145e-06,
+      "loss": 0.064,
+      "step": 1574
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.506879016939399,
+      "learning_rate": 9.726280744105e-06,
+      "loss": 0.111,
+      "step": 1575
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4718081298190497,
+      "learning_rate": 9.725835257631932e-06,
+      "loss": 0.1242,
+      "step": 1576
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5437057884310103,
+      "learning_rate": 9.725389419151123e-06,
+      "loss": 0.0901,
+      "step": 1577
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3801182594710486,
+      "learning_rate": 9.724943228695784e-06,
+      "loss": 0.0509,
+      "step": 1578
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4162171144579076,
+      "learning_rate": 9.724496686299149e-06,
+      "loss": 0.1017,
+      "step": 1579
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.42879282016265013,
+      "learning_rate": 9.724049791994477e-06,
+      "loss": 0.0792,
+      "step": 1580
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3443438077519364,
+      "learning_rate": 9.723602545815058e-06,
+      "loss": 0.0801,
+      "step": 1581
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.2574610545867587,
+      "learning_rate": 9.723154947794204e-06,
+      "loss": 0.0584,
+      "step": 1582
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4380456248419518,
+      "learning_rate": 9.722706997965256e-06,
+      "loss": 0.1078,
+      "step": 1583
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5373266055449576,
+      "learning_rate": 9.722258696361577e-06,
+      "loss": 0.1192,
+      "step": 1584
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4940203295391308,
+      "learning_rate": 9.721810043016563e-06,
+      "loss": 0.111,
+      "step": 1585
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4604345272918427,
+      "learning_rate": 9.72136103796363e-06,
+      "loss": 0.118,
+      "step": 1586
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.33524699694115173,
+      "learning_rate": 9.720911681236223e-06,
+      "loss": 0.0687,
+      "step": 1587
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.20428752717733256,
+      "learning_rate": 9.720461972867812e-06,
+      "loss": 0.0481,
+      "step": 1588
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.31385879426425145,
+      "learning_rate": 9.720011912891894e-06,
+      "loss": 0.0824,
+      "step": 1589
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4480870699341634,
+      "learning_rate": 9.719561501341994e-06,
+      "loss": 0.0762,
+      "step": 1590
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4369119093390214,
+      "learning_rate": 9.719110738251657e-06,
+      "loss": 0.1067,
+      "step": 1591
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3681110168145045,
+      "learning_rate": 9.718659623654465e-06,
+      "loss": 0.083,
+      "step": 1592
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6103455432805954,
+      "learning_rate": 9.718208157584015e-06,
+      "loss": 0.1201,
+      "step": 1593
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.360903371312486,
+      "learning_rate": 9.717756340073935e-06,
+      "loss": 0.0769,
+      "step": 1594
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3555906285119128,
+      "learning_rate": 9.717304171157881e-06,
+      "loss": 0.0943,
+      "step": 1595
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.34280619971552534,
+      "learning_rate": 9.716851650869532e-06,
+      "loss": 0.045,
+      "step": 1596
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.34696396352755465,
+      "learning_rate": 9.716398779242593e-06,
+      "loss": 0.0951,
+      "step": 1597
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1773188411823499,
+      "learning_rate": 9.715945556310797e-06,
+      "loss": 0.1677,
+      "step": 1598
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.46980452502374365,
+      "learning_rate": 9.715491982107905e-06,
+      "loss": 0.1069,
+      "step": 1599
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.2802689734647612,
+      "learning_rate": 9.715038056667702e-06,
+      "loss": 0.0825,
+      "step": 1600
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4770422624692102,
+      "learning_rate": 9.714583780023995e-06,
+      "loss": 0.1046,
+      "step": 1601
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5785270908229381,
+      "learning_rate": 9.714129152210625e-06,
+      "loss": 0.0973,
+      "step": 1602
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8306559959486894,
+      "learning_rate": 9.713674173261454e-06,
+      "loss": 0.1116,
+      "step": 1603
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2927128651918903,
+      "learning_rate": 9.71321884321037e-06,
+      "loss": 0.0956,
+      "step": 1604
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.34662603339291476,
+      "learning_rate": 9.71276316209129e-06,
+      "loss": 0.0608,
+      "step": 1605
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.41762795518733203,
+      "learning_rate": 9.712307129938157e-06,
+      "loss": 0.0614,
+      "step": 1606
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4105634720377538,
+      "learning_rate": 9.711850746784937e-06,
+      "loss": 0.0776,
+      "step": 1607
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4564513774918546,
+      "learning_rate": 9.711394012665626e-06,
+      "loss": 0.0813,
+      "step": 1608
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5229489516213977,
+      "learning_rate": 9.710936927614241e-06,
+      "loss": 0.0904,
+      "step": 1609
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5909146669430879,
+      "learning_rate": 9.710479491664832e-06,
+      "loss": 0.1313,
+      "step": 1610
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5268795931281853,
+      "learning_rate": 9.71002170485147e-06,
+      "loss": 0.0984,
+      "step": 1611
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.49806451081059294,
+      "learning_rate": 9.709563567208254e-06,
+      "loss": 0.1407,
+      "step": 1612
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3505897504550937,
+      "learning_rate": 9.709105078769306e-06,
+      "loss": 0.1429,
+      "step": 1613
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.40472986301641434,
+      "learning_rate": 9.708646239568781e-06,
+      "loss": 0.0906,
+      "step": 1614
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3330593198849907,
+      "learning_rate": 9.708187049640853e-06,
+      "loss": 0.073,
+      "step": 1615
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.355026167414063,
+      "learning_rate": 9.707727509019726e-06,
+      "loss": 0.0567,
+      "step": 1616
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.31540416563774126,
+      "learning_rate": 9.707267617739632e-06,
+      "loss": 0.0839,
+      "step": 1617
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9833310243637143,
+      "learning_rate": 9.706807375834823e-06,
+      "loss": 0.0937,
+      "step": 1618
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.41699340574494925,
+      "learning_rate": 9.70634678333958e-06,
+      "loss": 0.1076,
+      "step": 1619
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.47901272549747814,
+      "learning_rate": 9.705885840288214e-06,
+      "loss": 0.0935,
+      "step": 1620
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8794044809776304,
+      "learning_rate": 9.705424546715055e-06,
+      "loss": 0.1308,
+      "step": 1621
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2991152516043269,
+      "learning_rate": 9.704962902654466e-06,
+      "loss": 0.0658,
+      "step": 1622
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4470362129046527,
+      "learning_rate": 9.70450090814083e-06,
+      "loss": 0.0628,
+      "step": 1623
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.28368805335921876,
+      "learning_rate": 9.704038563208562e-06,
+      "loss": 0.0677,
+      "step": 1624
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4211828588435363,
+      "learning_rate": 9.703575867892099e-06,
+      "loss": 0.1121,
+      "step": 1625
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6006620620086295,
+      "learning_rate": 9.703112822225903e-06,
+      "loss": 0.1335,
+      "step": 1626
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.34250629363217006,
+      "learning_rate": 9.702649426244469e-06,
+      "loss": 0.0726,
+      "step": 1627
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3446047913915509,
+      "learning_rate": 9.70218567998231e-06,
+      "loss": 0.0934,
+      "step": 1628
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4541490697934343,
+      "learning_rate": 9.701721583473968e-06,
+      "loss": 0.0815,
+      "step": 1629
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2800957868785917,
+      "learning_rate": 9.701257136754014e-06,
+      "loss": 0.0827,
+      "step": 1630
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5938545803444883,
+      "learning_rate": 9.700792339857042e-06,
+      "loss": 0.1428,
+      "step": 1631
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.39735801907682106,
+      "learning_rate": 9.700327192817672e-06,
+      "loss": 0.0907,
+      "step": 1632
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.33372457722583365,
+      "learning_rate": 9.69986169567055e-06,
+      "loss": 0.0947,
+      "step": 1633
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5399399754100918,
+      "learning_rate": 9.699395848450352e-06,
+      "loss": 0.085,
+      "step": 1634
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4788246806715867,
+      "learning_rate": 9.698929651191775e-06,
+      "loss": 0.0588,
+      "step": 1635
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3959951015948476,
+      "learning_rate": 9.698463103929542e-06,
+      "loss": 0.0847,
+      "step": 1636
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.22307701210655395,
+      "learning_rate": 9.69799620669841e-06,
+      "loss": 0.0522,
+      "step": 1637
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5629817844772499,
+      "learning_rate": 9.69752895953315e-06,
+      "loss": 0.1118,
+      "step": 1638
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.35079367337136536,
+      "learning_rate": 9.697061362468569e-06,
+      "loss": 0.0866,
+      "step": 1639
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4512904538653238,
+      "learning_rate": 9.696593415539494e-06,
+      "loss": 0.122,
+      "step": 1640
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.39639752028183983,
+      "learning_rate": 9.696125118780784e-06,
+      "loss": 0.0946,
+      "step": 1641
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.23958162538929298,
+      "learning_rate": 9.695656472227316e-06,
+      "loss": 0.0522,
+      "step": 1642
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.30159306538982017,
+      "learning_rate": 9.695187475914e-06,
+      "loss": 0.0697,
+      "step": 1643
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7782686441904154,
+      "learning_rate": 9.694718129875772e-06,
+      "loss": 0.1254,
+      "step": 1644
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3518363797443654,
+      "learning_rate": 9.694248434147587e-06,
+      "loss": 0.109,
+      "step": 1645
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.33831510922193114,
+      "learning_rate": 9.693778388764431e-06,
+      "loss": 0.0963,
+      "step": 1646
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.24911569462044597,
+      "learning_rate": 9.69330799376132e-06,
+      "loss": 0.0735,
+      "step": 1647
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3983830652292169,
+      "learning_rate": 9.692837249173287e-06,
+      "loss": 0.0925,
+      "step": 1648
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2808850915461096,
+      "learning_rate": 9.692366155035399e-06,
+      "loss": 0.0649,
+      "step": 1649
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2760690129922068,
+      "learning_rate": 9.691894711382745e-06,
+      "loss": 0.0526,
+      "step": 1650
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.34117336217318783,
+      "learning_rate": 9.691422918250441e-06,
+      "loss": 0.0951,
+      "step": 1651
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5470083354239057,
+      "learning_rate": 9.690950775673626e-06,
+      "loss": 0.0672,
+      "step": 1652
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.32479946815877864,
+      "learning_rate": 9.690478283687473e-06,
+      "loss": 0.0578,
+      "step": 1653
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4500612467145754,
+      "learning_rate": 9.690005442327172e-06,
+      "loss": 0.1245,
+      "step": 1654
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7493651018061379,
+      "learning_rate": 9.689532251627946e-06,
+      "loss": 0.0945,
+      "step": 1655
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2083604300294076,
+      "learning_rate": 9.689058711625037e-06,
+      "loss": 0.0408,
+      "step": 1656
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.37823970187962863,
+      "learning_rate": 9.688584822353721e-06,
+      "loss": 0.08,
+      "step": 1657
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3346852404506423,
+      "learning_rate": 9.688110583849296e-06,
+      "loss": 0.094,
+      "step": 1658
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3766079384996788,
+      "learning_rate": 9.687635996147081e-06,
+      "loss": 0.1067,
+      "step": 1659
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3973157988932218,
+      "learning_rate": 9.687161059282431e-06,
+      "loss": 0.0749,
+      "step": 1660
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.31183440566978293,
+      "learning_rate": 9.68668577329072e-06,
+      "loss": 0.0912,
+      "step": 1661
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3539385255413199,
+      "learning_rate": 9.686210138207352e-06,
+      "loss": 0.0899,
+      "step": 1662
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.18797687484048944,
+      "learning_rate": 9.685734154067752e-06,
+      "loss": 0.0404,
+      "step": 1663
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3722093333403882,
+      "learning_rate": 9.685257820907377e-06,
+      "loss": 0.0768,
+      "step": 1664
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3793665651940612,
+      "learning_rate": 9.684781138761705e-06,
+      "loss": 0.0753,
+      "step": 1665
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.48333684582428976,
+      "learning_rate": 9.684304107666245e-06,
+      "loss": 0.0954,
+      "step": 1666
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3694805092446567,
+      "learning_rate": 9.683826727656525e-06,
+      "loss": 0.1027,
+      "step": 1667
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3859817412515084,
+      "learning_rate": 9.683348998768106e-06,
+      "loss": 0.0458,
+      "step": 1668
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4560717759024917,
+      "learning_rate": 9.682870921036569e-06,
+      "loss": 0.0914,
+      "step": 1669
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.48765569775599016,
+      "learning_rate": 9.682392494497528e-06,
+      "loss": 0.1094,
+      "step": 1670
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.29488885185697933,
+      "learning_rate": 9.681913719186616e-06,
+      "loss": 0.0725,
+      "step": 1671
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5631546267800813,
+      "learning_rate": 9.681434595139496e-06,
+      "loss": 0.1699,
+      "step": 1672
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9017945529494749,
+      "learning_rate": 9.680955122391858e-06,
+      "loss": 0.1346,
+      "step": 1673
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3227088044276552,
+      "learning_rate": 9.680475300979412e-06,
+      "loss": 0.0717,
+      "step": 1674
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5813438072495212,
+      "learning_rate": 9.6799951309379e-06,
+      "loss": 0.0853,
+      "step": 1675
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.41135836425111183,
+      "learning_rate": 9.679514612303088e-06,
+      "loss": 0.0878,
+      "step": 1676
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.36621261302916064,
+      "learning_rate": 9.679033745110767e-06,
+      "loss": 0.1101,
+      "step": 1677
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3549293253532294,
+      "learning_rate": 9.678552529396757e-06,
+      "loss": 0.0818,
+      "step": 1678
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4394260590726517,
+      "learning_rate": 9.678070965196897e-06,
+      "loss": 0.0864,
+      "step": 1679
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9490709657429738,
+      "learning_rate": 9.677589052547063e-06,
+      "loss": 0.1304,
+      "step": 1680
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.45403516007725714,
+      "learning_rate": 9.677106791483147e-06,
+      "loss": 0.082,
+      "step": 1681
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3489199906452373,
+      "learning_rate": 9.676624182041067e-06,
+      "loss": 0.0674,
+      "step": 1682
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5438823422306638,
+      "learning_rate": 9.676141224256778e-06,
+      "loss": 0.1462,
+      "step": 1683
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.850743670015194,
+      "learning_rate": 9.675657918166251e-06,
+      "loss": 0.0983,
+      "step": 1684
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.25491265592182877,
+      "learning_rate": 9.675174263805483e-06,
+      "loss": 0.0496,
+      "step": 1685
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.48138606967626524,
+      "learning_rate": 9.6746902612105e-06,
+      "loss": 0.0853,
+      "step": 1686
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4577760492863844,
+      "learning_rate": 9.674205910417356e-06,
+      "loss": 0.0967,
+      "step": 1687
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3258968424402221,
+      "learning_rate": 9.673721211462125e-06,
+      "loss": 0.1017,
+      "step": 1688
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.35809153124559706,
+      "learning_rate": 9.673236164380912e-06,
+      "loss": 0.1029,
+      "step": 1689
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.44933256820846557,
+      "learning_rate": 9.672750769209847e-06,
+      "loss": 0.1082,
+      "step": 1690
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3382113975920095,
+      "learning_rate": 9.672265025985083e-06,
+      "loss": 0.0745,
+      "step": 1691
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.37860175862659473,
+      "learning_rate": 9.671778934742803e-06,
+      "loss": 0.0929,
+      "step": 1692
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.943085121444794,
+      "learning_rate": 9.67129249551921e-06,
+      "loss": 0.1203,
+      "step": 1693
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5332242438493886,
+      "learning_rate": 9.670805708350544e-06,
+      "loss": 0.1086,
+      "step": 1694
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7995730641515192,
+      "learning_rate": 9.670318573273056e-06,
+      "loss": 0.1611,
+      "step": 1695
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.36393982251294177,
+      "learning_rate": 9.669831090323036e-06,
+      "loss": 0.0755,
+      "step": 1696
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5420918455984371,
+      "learning_rate": 9.669343259536792e-06,
+      "loss": 0.1529,
+      "step": 1697
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3258567145449449,
+      "learning_rate": 9.668855080950661e-06,
+      "loss": 0.0909,
+      "step": 1698
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.472418532031593,
+      "learning_rate": 9.668366554601005e-06,
+      "loss": 0.1407,
+      "step": 1699
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3067907705928447,
+      "learning_rate": 9.667877680524213e-06,
+      "loss": 0.0817,
+      "step": 1700
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4783910353119245,
+      "learning_rate": 9.667388458756699e-06,
+      "loss": 0.1296,
+      "step": 1701
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6832021852971457,
+      "learning_rate": 9.666898889334903e-06,
+      "loss": 0.1709,
+      "step": 1702
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2539010809356441,
+      "learning_rate": 9.666408972295291e-06,
+      "loss": 0.0473,
+      "step": 1703
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3915263050142795,
+      "learning_rate": 9.665918707674356e-06,
+      "loss": 0.0965,
+      "step": 1704
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.443688343543515,
+      "learning_rate": 9.665428095508612e-06,
+      "loss": 0.0805,
+      "step": 1705
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2462069906348091,
+      "learning_rate": 9.664937135834607e-06,
+      "loss": 0.0528,
+      "step": 1706
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3059833800395932,
+      "learning_rate": 9.664445828688909e-06,
+      "loss": 0.0779,
+      "step": 1707
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.34997345092857235,
+      "learning_rate": 9.663954174108114e-06,
+      "loss": 0.0466,
+      "step": 1708
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.34688101090722157,
+      "learning_rate": 9.66346217212884e-06,
+      "loss": 0.0691,
+      "step": 1709
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4635148450913152,
+      "learning_rate": 9.66296982278774e-06,
+      "loss": 0.0909,
+      "step": 1710
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.40611009037167684,
+      "learning_rate": 9.662477126121481e-06,
+      "loss": 0.058,
+      "step": 1711
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.545865533814014,
+      "learning_rate": 9.661984082166766e-06,
+      "loss": 0.0715,
+      "step": 1712
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.465371240062029,
+      "learning_rate": 9.661490690960318e-06,
+      "loss": 0.1325,
+      "step": 1713
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3366913122905652,
+      "learning_rate": 9.66099695253889e-06,
+      "loss": 0.0809,
+      "step": 1714
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.39896688225891824,
+      "learning_rate": 9.660502866939254e-06,
+      "loss": 0.0627,
+      "step": 1715
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6788029674185972,
+      "learning_rate": 9.660008434198217e-06,
+      "loss": 0.1011,
+      "step": 1716
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.31861354983673584,
+      "learning_rate": 9.659513654352603e-06,
+      "loss": 0.0941,
+      "step": 1717
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5258465189832534,
+      "learning_rate": 9.65901852743927e-06,
+      "loss": 0.1099,
+      "step": 1718
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2710032818313362,
+      "learning_rate": 9.658523053495096e-06,
+      "loss": 0.0883,
+      "step": 1719
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.40871914245954305,
+      "learning_rate": 9.658027232556989e-06,
+      "loss": 0.1211,
+      "step": 1720
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.2716850623023346,
+      "learning_rate": 9.657531064661875e-06,
+      "loss": 0.0515,
+      "step": 1721
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.2772492049281881,
+      "learning_rate": 9.657034549846719e-06,
+      "loss": 0.0867,
+      "step": 1722
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4174072165238044,
+      "learning_rate": 9.656537688148499e-06,
+      "loss": 0.1076,
+      "step": 1723
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3469617477897984,
+      "learning_rate": 9.656040479604225e-06,
+      "loss": 0.0854,
+      "step": 1724
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7115411063230116,
+      "learning_rate": 9.655542924250933e-06,
+      "loss": 0.11,
+      "step": 1725
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4116103739510056,
+      "learning_rate": 9.655045022125683e-06,
+      "loss": 0.1242,
+      "step": 1726
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3540199019327835,
+      "learning_rate": 9.654546773265562e-06,
+      "loss": 0.1175,
+      "step": 1727
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.2450499848005601,
+      "learning_rate": 9.654048177707684e-06,
+      "loss": 0.0498,
+      "step": 1728
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.23230475104288978,
+      "learning_rate": 9.653549235489184e-06,
+      "loss": 0.0498,
+      "step": 1729
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.33021047725469893,
+      "learning_rate": 9.65304994664723e-06,
+      "loss": 0.0554,
+      "step": 1730
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3298519310981313,
+      "learning_rate": 9.652550311219008e-06,
+      "loss": 0.0717,
+      "step": 1731
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6040181845310911,
+      "learning_rate": 9.652050329241737e-06,
+      "loss": 0.0909,
+      "step": 1732
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.37257839442426416,
+      "learning_rate": 9.651550000752656e-06,
+      "loss": 0.1009,
+      "step": 1733
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4375964409520717,
+      "learning_rate": 9.651049325789035e-06,
+      "loss": 0.1221,
+      "step": 1734
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3517558742518417,
+      "learning_rate": 9.650548304388166e-06,
+      "loss": 0.0812,
+      "step": 1735
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.48173316990309206,
+      "learning_rate": 9.650046936587367e-06,
+      "loss": 0.1067,
+      "step": 1736
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7544284724774956,
+      "learning_rate": 9.649545222423984e-06,
+      "loss": 0.1244,
+      "step": 1737
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.2041707673219208,
+      "learning_rate": 9.649043161935387e-06,
+      "loss": 0.0609,
+      "step": 1738
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4313445397604857,
+      "learning_rate": 9.648540755158973e-06,
+      "loss": 0.0968,
+      "step": 1739
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3546779767254469,
+      "learning_rate": 9.648038002132164e-06,
+      "loss": 0.1146,
+      "step": 1740
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6625767198256146,
+      "learning_rate": 9.647534902892405e-06,
+      "loss": 0.107,
+      "step": 1741
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.49935009692641946,
+      "learning_rate": 9.647031457477176e-06,
+      "loss": 0.1197,
+      "step": 1742
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5357819950506627,
+      "learning_rate": 9.646527665923973e-06,
+      "loss": 0.0958,
+      "step": 1743
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4040191738680696,
+      "learning_rate": 9.646023528270319e-06,
+      "loss": 0.1034,
+      "step": 1744
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4773618109906908,
+      "learning_rate": 9.64551904455377e-06,
+      "loss": 0.1,
+      "step": 1745
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3539996100809835,
+      "learning_rate": 9.645014214811899e-06,
+      "loss": 0.0793,
+      "step": 1746
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3542687730299365,
+      "learning_rate": 9.64450903908231e-06,
+      "loss": 0.1013,
+      "step": 1747
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.35096286233135393,
+      "learning_rate": 9.644003517402631e-06,
+      "loss": 0.1027,
+      "step": 1748
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.2733711946774802,
+      "learning_rate": 9.643497649810518e-06,
+      "loss": 0.0713,
+      "step": 1749
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3692470032729203,
+      "learning_rate": 9.64299143634365e-06,
+      "loss": 0.0749,
+      "step": 1750
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4059296918235959,
+      "learning_rate": 9.642484877039732e-06,
+      "loss": 0.1344,
+      "step": 1751
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4066474635625967,
+      "learning_rate": 9.641977971936496e-06,
+      "loss": 0.0738,
+      "step": 1752
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.43514945540588296,
+      "learning_rate": 9.641470721071698e-06,
+      "loss": 0.0766,
+      "step": 1753
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.47303169059026534,
+      "learning_rate": 9.640963124483123e-06,
+      "loss": 0.114,
+      "step": 1754
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4220763483611639,
+      "learning_rate": 9.640455182208578e-06,
+      "loss": 0.1252,
+      "step": 1755
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.29092101217812594,
+      "learning_rate": 9.639946894285898e-06,
+      "loss": 0.0647,
+      "step": 1756
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4122689323009027,
+      "learning_rate": 9.639438260752945e-06,
+      "loss": 0.1138,
+      "step": 1757
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.501291769446736,
+      "learning_rate": 9.638929281647604e-06,
+      "loss": 0.0911,
+      "step": 1758
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.24670073911623808,
+      "learning_rate": 9.638419957007784e-06,
+      "loss": 0.0422,
+      "step": 1759
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5215980826502019,
+      "learning_rate": 9.637910286871428e-06,
+      "loss": 0.1397,
+      "step": 1760
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.30612686064646755,
+      "learning_rate": 9.637400271276495e-06,
+      "loss": 0.0769,
+      "step": 1761
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3953106733774438,
+      "learning_rate": 9.636889910260974e-06,
+      "loss": 0.1056,
+      "step": 1762
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.46752411724635234,
+      "learning_rate": 9.63637920386288e-06,
+      "loss": 0.0741,
+      "step": 1763
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5339227032316735,
+      "learning_rate": 9.635868152120256e-06,
+      "loss": 0.0925,
+      "step": 1764
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.2254874092089242,
+      "learning_rate": 9.635356755071165e-06,
+      "loss": 0.0582,
+      "step": 1765
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7991655491318462,
+      "learning_rate": 9.6348450127537e-06,
+      "loss": 0.1336,
+      "step": 1766
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.47281557900482096,
+      "learning_rate": 9.634332925205978e-06,
+      "loss": 0.108,
+      "step": 1767
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3818476949818449,
+      "learning_rate": 9.633820492466144e-06,
+      "loss": 0.0891,
+      "step": 1768
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3763393573070956,
+      "learning_rate": 9.633307714572363e-06,
+      "loss": 0.1092,
+      "step": 1769
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4202087038113615,
+      "learning_rate": 9.632794591562835e-06,
+      "loss": 0.0798,
+      "step": 1770
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3985458833864818,
+      "learning_rate": 9.632281123475778e-06,
+      "loss": 0.1118,
+      "step": 1771
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3821740888533187,
+      "learning_rate": 9.631767310349438e-06,
+      "loss": 0.1079,
+      "step": 1772
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3284128242194129,
+      "learning_rate": 9.631253152222085e-06,
+      "loss": 0.0373,
+      "step": 1773
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9914885394284557,
+      "learning_rate": 9.630738649132018e-06,
+      "loss": 0.1046,
+      "step": 1774
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5969830749512147,
+      "learning_rate": 9.630223801117562e-06,
+      "loss": 0.1283,
+      "step": 1775
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6101035610014983,
+      "learning_rate": 9.629708608217063e-06,
+      "loss": 0.1518,
+      "step": 1776
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.38428681715152785,
+      "learning_rate": 9.629193070468898e-06,
+      "loss": 0.0988,
+      "step": 1777
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.2636046602213241,
+      "learning_rate": 9.628677187911466e-06,
+      "loss": 0.0684,
+      "step": 1778
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.463994127607072,
+      "learning_rate": 9.628160960583193e-06,
+      "loss": 0.1169,
+      "step": 1779
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6542644911398723,
+      "learning_rate": 9.627644388522528e-06,
+      "loss": 0.1483,
+      "step": 1780
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.32888198276290986,
+      "learning_rate": 9.627127471767955e-06,
+      "loss": 0.0648,
+      "step": 1781
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.27175643443963027,
+      "learning_rate": 9.626610210357969e-06,
+      "loss": 0.0559,
+      "step": 1782
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4597624521746198,
+      "learning_rate": 9.626092604331105e-06,
+      "loss": 0.138,
+      "step": 1783
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4700774146925981,
+      "learning_rate": 9.625574653725916e-06,
+      "loss": 0.1012,
+      "step": 1784
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5737118216114742,
+      "learning_rate": 9.625056358580978e-06,
+      "loss": 0.1498,
+      "step": 1785
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3181079907006066,
+      "learning_rate": 9.624537718934901e-06,
+      "loss": 0.0861,
+      "step": 1786
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.33637802760195745,
+      "learning_rate": 9.624018734826314e-06,
+      "loss": 0.1093,
+      "step": 1787
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.27508410210893214,
+      "learning_rate": 9.623499406293874e-06,
+      "loss": 0.0555,
+      "step": 1788
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4343609478317466,
+      "learning_rate": 9.622979733376266e-06,
+      "loss": 0.0982,
+      "step": 1789
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3448739760588774,
+      "learning_rate": 9.622459716112195e-06,
+      "loss": 0.0561,
+      "step": 1790
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.23680149695828268,
+      "learning_rate": 9.621939354540399e-06,
+      "loss": 0.073,
+      "step": 1791
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.255239398428975,
+      "learning_rate": 9.621418648699634e-06,
+      "loss": 0.0338,
+      "step": 1792
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.42482429853238646,
+      "learning_rate": 9.620897598628685e-06,
+      "loss": 0.0618,
+      "step": 1793
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4804070310811205,
+      "learning_rate": 9.620376204366365e-06,
+      "loss": 0.0904,
+      "step": 1794
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.2975749195132777,
+      "learning_rate": 9.61985446595151e-06,
+      "loss": 0.0826,
+      "step": 1795
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.66197686965244,
+      "learning_rate": 9.619332383422981e-06,
+      "loss": 0.1226,
+      "step": 1796
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3546357737399377,
+      "learning_rate": 9.618809956819667e-06,
+      "loss": 0.0767,
+      "step": 1797
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.575304126887479,
+      "learning_rate": 9.618287186180481e-06,
+      "loss": 0.1163,
+      "step": 1798
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.45701609110518454,
+      "learning_rate": 9.617764071544361e-06,
+      "loss": 0.1201,
+      "step": 1799
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.33695926938051457,
+      "learning_rate": 9.617240612950274e-06,
+      "loss": 0.0768,
+      "step": 1800
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5728228430650636,
+      "learning_rate": 9.616716810437208e-06,
+      "loss": 0.1186,
+      "step": 1801
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5054069417482399,
+      "learning_rate": 9.616192664044179e-06,
+      "loss": 0.0869,
+      "step": 1802
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.607561965061396,
+      "learning_rate": 9.615668173810228e-06,
+      "loss": 0.1013,
+      "step": 1803
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.489446526027828,
+      "learning_rate": 9.615143339774426e-06,
+      "loss": 0.1139,
+      "step": 1804
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4652410196145404,
+      "learning_rate": 9.614618161975861e-06,
+      "loss": 0.1208,
+      "step": 1805
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3307487510723781,
+      "learning_rate": 9.614092640453654e-06,
+      "loss": 0.0764,
+      "step": 1806
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4834579931552204,
+      "learning_rate": 9.613566775246948e-06,
+      "loss": 0.0963,
+      "step": 1807
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0709741162931112,
+      "learning_rate": 9.613040566394911e-06,
+      "loss": 0.1493,
+      "step": 1808
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7491351902539422,
+      "learning_rate": 9.612514013936741e-06,
+      "loss": 0.1409,
+      "step": 1809
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4938599196296888,
+      "learning_rate": 9.611987117911657e-06,
+      "loss": 0.0731,
+      "step": 1810
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.153649357147798,
+      "learning_rate": 9.611459878358906e-06,
+      "loss": 0.1837,
+      "step": 1811
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.43122340317346164,
+      "learning_rate": 9.610932295317762e-06,
+      "loss": 0.0889,
+      "step": 1812
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.37993865832808865,
+      "learning_rate": 9.610404368827515e-06,
+      "loss": 0.0912,
+      "step": 1813
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.523529343254003,
+      "learning_rate": 9.609876098927496e-06,
+      "loss": 0.1041,
+      "step": 1814
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4902970994945036,
+      "learning_rate": 9.60934748565705e-06,
+      "loss": 0.1151,
+      "step": 1815
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4062905869418258,
+      "learning_rate": 9.608818529055552e-06,
+      "loss": 0.0852,
+      "step": 1816
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.46650697904878186,
+      "learning_rate": 9.608289229162401e-06,
+      "loss": 0.0766,
+      "step": 1817
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7142306543567551,
+      "learning_rate": 9.607759586017025e-06,
+      "loss": 0.0946,
+      "step": 1818
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.37701850761138117,
+      "learning_rate": 9.60722959965887e-06,
+      "loss": 0.1182,
+      "step": 1819
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5784341156163959,
+      "learning_rate": 9.606699270127417e-06,
+      "loss": 0.1008,
+      "step": 1820
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.37996851033851764,
+      "learning_rate": 9.606168597462167e-06,
+      "loss": 0.1154,
+      "step": 1821
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6904213355326778,
+      "learning_rate": 9.605637581702644e-06,
+      "loss": 0.1053,
+      "step": 1822
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4632611818469938,
+      "learning_rate": 9.605106222888408e-06,
+      "loss": 0.0924,
+      "step": 1823
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.29790325861550254,
+      "learning_rate": 9.604574521059031e-06,
+      "loss": 0.0769,
+      "step": 1824
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.2935469359880675,
+      "learning_rate": 9.60404247625412e-06,
+      "loss": 0.0848,
+      "step": 1825
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.35847531601618876,
+      "learning_rate": 9.603510088513306e-06,
+      "loss": 0.1,
+      "step": 1826
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.30672484585996274,
+      "learning_rate": 9.602977357876243e-06,
+      "loss": 0.0917,
+      "step": 1827
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8760810347049858,
+      "learning_rate": 9.602444284382613e-06,
+      "loss": 0.1615,
+      "step": 1828
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.33654705848736377,
+      "learning_rate": 9.60191086807212e-06,
+      "loss": 0.0878,
+      "step": 1829
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4711618548231413,
+      "learning_rate": 9.601377108984498e-06,
+      "loss": 0.1196,
+      "step": 1830
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4243693448127547,
+      "learning_rate": 9.600843007159504e-06,
+      "loss": 0.1094,
+      "step": 1831
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.33136047238713895,
+      "learning_rate": 9.60030856263692e-06,
+      "loss": 0.1052,
+      "step": 1832
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.2866495821779736,
+      "learning_rate": 9.599773775456556e-06,
+      "loss": 0.0861,
+      "step": 1833
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4270524679787621,
+      "learning_rate": 9.599238645658247e-06,
+      "loss": 0.0793,
+      "step": 1834
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4709436703134125,
+      "learning_rate": 9.59870317328185e-06,
+      "loss": 0.1241,
+      "step": 1835
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6749150505946065,
+      "learning_rate": 9.59816735836725e-06,
+      "loss": 0.1411,
+      "step": 1836
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3885439690384418,
+      "learning_rate": 9.597631200954361e-06,
+      "loss": 0.0824,
+      "step": 1837
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.29366077540168367,
+      "learning_rate": 9.597094701083114e-06,
+      "loss": 0.0765,
+      "step": 1838
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3109636267916011,
+      "learning_rate": 9.596557858793476e-06,
+      "loss": 0.0851,
+      "step": 1839
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5061885870624682,
+      "learning_rate": 9.59602067412543e-06,
+      "loss": 0.1226,
+      "step": 1840
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.22946441238905588,
+      "learning_rate": 9.595483147118993e-06,
+      "loss": 0.0746,
+      "step": 1841
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.33264233471642923,
+      "learning_rate": 9.5949452778142e-06,
+      "loss": 0.0781,
+      "step": 1842
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.40897360664440763,
+      "learning_rate": 9.594407066251114e-06,
+      "loss": 0.1093,
+      "step": 1843
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3643547803657084,
+      "learning_rate": 9.593868512469824e-06,
+      "loss": 0.1223,
+      "step": 1844
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3726987785926976,
+      "learning_rate": 9.59332961651045e-06,
+      "loss": 0.1127,
+      "step": 1845
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4545366482154216,
+      "learning_rate": 9.592790378413125e-06,
+      "loss": 0.1041,
+      "step": 1846
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0369463398658216,
+      "learning_rate": 9.592250798218018e-06,
+      "loss": 0.1405,
+      "step": 1847
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.2849198775900309,
+      "learning_rate": 9.59171087596532e-06,
+      "loss": 0.0846,
+      "step": 1848
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.2600027101766995,
+      "learning_rate": 9.591170611695246e-06,
+      "loss": 0.0424,
+      "step": 1849
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5776973722820054,
+      "learning_rate": 9.590630005448041e-06,
+      "loss": 0.1157,
+      "step": 1850
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.37225970150218674,
+      "learning_rate": 9.590089057263972e-06,
+      "loss": 0.0796,
+      "step": 1851
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.49731182522790984,
+      "learning_rate": 9.589547767183328e-06,
+      "loss": 0.093,
+      "step": 1852
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.2215884233876604,
+      "learning_rate": 9.58900613524643e-06,
+      "loss": 0.0551,
+      "step": 1853
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.30276416906795556,
+      "learning_rate": 9.588464161493624e-06,
+      "loss": 0.0831,
+      "step": 1854
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.29190265813288707,
+      "learning_rate": 9.587921845965278e-06,
+      "loss": 0.057,
+      "step": 1855
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.23759949715209405,
+      "learning_rate": 9.587379188701785e-06,
+      "loss": 0.0487,
+      "step": 1856
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.34838472796213904,
+      "learning_rate": 9.586836189743567e-06,
+      "loss": 0.1108,
+      "step": 1857
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.26541310087255937,
+      "learning_rate": 9.58629284913107e-06,
+      "loss": 0.0677,
+      "step": 1858
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3078904514761348,
+      "learning_rate": 9.585749166904762e-06,
+      "loss": 0.0803,
+      "step": 1859
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.36310516759145084,
+      "learning_rate": 9.585205143105144e-06,
+      "loss": 0.1302,
+      "step": 1860
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.2847219970624417,
+      "learning_rate": 9.584660777772736e-06,
+      "loss": 0.0868,
+      "step": 1861
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.16860687460144833,
+      "learning_rate": 9.584116070948083e-06,
+      "loss": 0.0571,
+      "step": 1862
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.331191898363734,
+      "learning_rate": 9.583571022671765e-06,
+      "loss": 0.09,
+      "step": 1863
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.20552531767640034,
+      "learning_rate": 9.583025632984374e-06,
+      "loss": 0.0605,
+      "step": 1864
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4004598143455329,
+      "learning_rate": 9.582479901926536e-06,
+      "loss": 0.0986,
+      "step": 1865
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.19408586707790806,
+      "learning_rate": 9.581933829538899e-06,
+      "loss": 0.0211,
+      "step": 1866
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.33038882294102445,
+      "learning_rate": 9.581387415862139e-06,
+      "loss": 0.0799,
+      "step": 1867
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4257771852146185,
+      "learning_rate": 9.580840660936957e-06,
+      "loss": 0.1007,
+      "step": 1868
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.2914026436021699,
+      "learning_rate": 9.580293564804074e-06,
+      "loss": 0.0798,
+      "step": 1869
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3996582725614731,
+      "learning_rate": 9.579746127504247e-06,
+      "loss": 0.1092,
+      "step": 1870
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4070319055361208,
+      "learning_rate": 9.579198349078248e-06,
+      "loss": 0.1136,
+      "step": 1871
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3973404828997548,
+      "learning_rate": 9.578650229566883e-06,
+      "loss": 0.1048,
+      "step": 1872
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.2563788951519117,
+      "learning_rate": 9.578101769010974e-06,
+      "loss": 0.0806,
+      "step": 1873
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.47528229881893647,
+      "learning_rate": 9.577552967451376e-06,
+      "loss": 0.0897,
+      "step": 1874
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2667795499584422,
+      "learning_rate": 9.577003824928967e-06,
+      "loss": 0.1193,
+      "step": 1875
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5736140290053159,
+      "learning_rate": 9.57645434148465e-06,
+      "loss": 0.1038,
+      "step": 1876
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.2772419181084563,
+      "learning_rate": 9.575904517159354e-06,
+      "loss": 0.1191,
+      "step": 1877
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.2882163210317352,
+      "learning_rate": 9.575354351994032e-06,
+      "loss": 0.0774,
+      "step": 1878
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3667216601081788,
+      "learning_rate": 9.574803846029666e-06,
+      "loss": 0.0974,
+      "step": 1879
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.30618162963866624,
+      "learning_rate": 9.574252999307258e-06,
+      "loss": 0.0635,
+      "step": 1880
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.37307176651415713,
+      "learning_rate": 9.57370181186784e-06,
+      "loss": 0.101,
+      "step": 1881
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.2719634094345342,
+      "learning_rate": 9.573150283752466e-06,
+      "loss": 0.0704,
+      "step": 1882
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.359281631744184,
+      "learning_rate": 9.57259841500222e-06,
+      "loss": 0.1096,
+      "step": 1883
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.2814151280573117,
+      "learning_rate": 9.572046205658207e-06,
+      "loss": 0.0738,
+      "step": 1884
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5749776765851902,
+      "learning_rate": 9.571493655761558e-06,
+      "loss": 0.106,
+      "step": 1885
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.44029755218977173,
+      "learning_rate": 9.570940765353431e-06,
+      "loss": 0.0966,
+      "step": 1886
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3976276617886006,
+      "learning_rate": 9.570387534475008e-06,
+      "loss": 0.1008,
+      "step": 1887
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4744751750152143,
+      "learning_rate": 9.569833963167495e-06,
+      "loss": 0.1221,
+      "step": 1888
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.22816683516551559,
+      "learning_rate": 9.569280051472129e-06,
+      "loss": 0.0689,
+      "step": 1889
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.48673119316387864,
+      "learning_rate": 9.568725799430165e-06,
+      "loss": 0.0819,
+      "step": 1890
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3699284551314705,
+      "learning_rate": 9.568171207082891e-06,
+      "loss": 0.0885,
+      "step": 1891
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.22307421615007708,
+      "learning_rate": 9.567616274471614e-06,
+      "loss": 0.0524,
+      "step": 1892
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5485866441109828,
+      "learning_rate": 9.567061001637666e-06,
+      "loss": 0.1043,
+      "step": 1893
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3356564728823607,
+      "learning_rate": 9.566505388622412e-06,
+      "loss": 0.069,
+      "step": 1894
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.64209474126418,
+      "learning_rate": 9.565949435467233e-06,
+      "loss": 0.1152,
+      "step": 1895
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.46191438754969344,
+      "learning_rate": 9.565393142213543e-06,
+      "loss": 0.1188,
+      "step": 1896
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.46832712514110697,
+      "learning_rate": 9.564836508902776e-06,
+      "loss": 0.0937,
+      "step": 1897
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8101153380753563,
+      "learning_rate": 9.564279535576393e-06,
+      "loss": 0.1896,
+      "step": 1898
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.28854704042852003,
+      "learning_rate": 9.563722222275882e-06,
+      "loss": 0.0913,
+      "step": 1899
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.23398198139209972,
+      "learning_rate": 9.563164569042753e-06,
+      "loss": 0.0511,
+      "step": 1900
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.48133097737011343,
+      "learning_rate": 9.562606575918545e-06,
+      "loss": 0.0994,
+      "step": 1901
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3519487205930869,
+      "learning_rate": 9.562048242944822e-06,
+      "loss": 0.0399,
+      "step": 1902
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.31917410698880977,
+      "learning_rate": 9.561489570163167e-06,
+      "loss": 0.0576,
+      "step": 1903
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.21060618246657206,
+      "learning_rate": 9.560930557615198e-06,
+      "loss": 0.0641,
+      "step": 1904
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4187131422650143,
+      "learning_rate": 9.560371205342552e-06,
+      "loss": 0.1135,
+      "step": 1905
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.39822503095232115,
+      "learning_rate": 9.559811513386893e-06,
+      "loss": 0.0961,
+      "step": 1906
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5277395336641175,
+      "learning_rate": 9.559251481789905e-06,
+      "loss": 0.0432,
+      "step": 1907
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.32860558455906774,
+      "learning_rate": 9.558691110593312e-06,
+      "loss": 0.0545,
+      "step": 1908
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3986479527892182,
+      "learning_rate": 9.558130399838848e-06,
+      "loss": 0.1151,
+      "step": 1909
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.36150840517670046,
+      "learning_rate": 9.557569349568276e-06,
+      "loss": 0.0927,
+      "step": 1910
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.2997963789081962,
+      "learning_rate": 9.557007959823393e-06,
+      "loss": 0.0864,
+      "step": 1911
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.37580775723556203,
+      "learning_rate": 9.556446230646007e-06,
+      "loss": 0.1115,
+      "step": 1912
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.21565114468203345,
+      "learning_rate": 9.555884162077965e-06,
+      "loss": 0.0557,
+      "step": 1913
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5952209261857783,
+      "learning_rate": 9.555321754161128e-06,
+      "loss": 0.1081,
+      "step": 1914
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4633814326283105,
+      "learning_rate": 9.554759006937393e-06,
+      "loss": 0.1039,
+      "step": 1915
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.30483419406760776,
+      "learning_rate": 9.554195920448673e-06,
+      "loss": 0.0893,
+      "step": 1916
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4260064498029597,
+      "learning_rate": 9.553632494736908e-06,
+      "loss": 0.0993,
+      "step": 1917
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3865784602285552,
+      "learning_rate": 9.553068729844071e-06,
+      "loss": 0.0992,
+      "step": 1918
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.30712239111443296,
+      "learning_rate": 9.552504625812151e-06,
+      "loss": 0.0806,
+      "step": 1919
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3061093067852982,
+      "learning_rate": 9.551940182683165e-06,
+      "loss": 0.0976,
+      "step": 1920
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4536066329343321,
+      "learning_rate": 9.55137540049916e-06,
+      "loss": 0.0881,
+      "step": 1921
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6888589280724848,
+      "learning_rate": 9.550810279302197e-06,
+      "loss": 0.0632,
+      "step": 1922
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4881213863047809,
+      "learning_rate": 9.550244819134376e-06,
+      "loss": 0.1428,
+      "step": 1923
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4585627732004255,
+      "learning_rate": 9.549679020037815e-06,
+      "loss": 0.1209,
+      "step": 1924
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3796390823258864,
+      "learning_rate": 9.549112882054655e-06,
+      "loss": 0.0978,
+      "step": 1925
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.34617906720967245,
+      "learning_rate": 9.548546405227068e-06,
+      "loss": 0.1079,
+      "step": 1926
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3027132669964892,
+      "learning_rate": 9.547979589597247e-06,
+      "loss": 0.0877,
+      "step": 1927
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7228397512569966,
+      "learning_rate": 9.547412435207413e-06,
+      "loss": 0.0966,
+      "step": 1928
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.28552182170189494,
+      "learning_rate": 9.546844942099809e-06,
+      "loss": 0.0588,
+      "step": 1929
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.40501822776414936,
+      "learning_rate": 9.546277110316706e-06,
+      "loss": 0.0948,
+      "step": 1930
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.31485542662887833,
+      "learning_rate": 9.545708939900402e-06,
+      "loss": 0.0624,
+      "step": 1931
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.37669808553268835,
+      "learning_rate": 9.545140430893215e-06,
+      "loss": 0.1147,
+      "step": 1932
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4969298973553203,
+      "learning_rate": 9.54457158333749e-06,
+      "loss": 0.0838,
+      "step": 1933
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4469266156739036,
+      "learning_rate": 9.5440023972756e-06,
+      "loss": 0.1097,
+      "step": 1934
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.24151121098947906,
+      "learning_rate": 9.54343287274994e-06,
+      "loss": 0.0635,
+      "step": 1935
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.40705944128208127,
+      "learning_rate": 9.542863009802933e-06,
+      "loss": 0.1104,
+      "step": 1936
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.43116492972596615,
+      "learning_rate": 9.542292808477026e-06,
+      "loss": 0.0788,
+      "step": 1937
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.45539462383988244,
+      "learning_rate": 9.541722268814687e-06,
+      "loss": 0.1346,
+      "step": 1938
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3101595091055292,
+      "learning_rate": 9.541151390858417e-06,
+      "loss": 0.0832,
+      "step": 1939
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6876254577704088,
+      "learning_rate": 9.540580174650738e-06,
+      "loss": 0.1355,
+      "step": 1940
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3665868492527732,
+      "learning_rate": 9.540008620234197e-06,
+      "loss": 0.0872,
+      "step": 1941
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.29438406757155905,
+      "learning_rate": 9.539436727651365e-06,
+      "loss": 0.0869,
+      "step": 1942
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4029844747919658,
+      "learning_rate": 9.538864496944843e-06,
+      "loss": 0.0992,
+      "step": 1943
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.35002215559217426,
+      "learning_rate": 9.538291928157252e-06,
+      "loss": 0.0578,
+      "step": 1944
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.37963414284910696,
+      "learning_rate": 9.53771902133124e-06,
+      "loss": 0.0761,
+      "step": 1945
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5488268923899522,
+      "learning_rate": 9.537145776509482e-06,
+      "loss": 0.1256,
+      "step": 1946
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4027731444881979,
+      "learning_rate": 9.536572193734675e-06,
+      "loss": 0.1041,
+      "step": 1947
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5526595994132719,
+      "learning_rate": 9.535998273049542e-06,
+      "loss": 0.1167,
+      "step": 1948
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.2959930102857174,
+      "learning_rate": 9.535424014496838e-06,
+      "loss": 0.0584,
+      "step": 1949
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.49781507105261646,
+      "learning_rate": 9.534849418119328e-06,
+      "loss": 0.1113,
+      "step": 1950
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.2980499875388227,
+      "learning_rate": 9.534274483959818e-06,
+      "loss": 0.0812,
+      "step": 1951
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.45092725178630183,
+      "learning_rate": 9.533699212061131e-06,
+      "loss": 0.1071,
+      "step": 1952
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.291848327225128,
+      "learning_rate": 9.533123602466114e-06,
+      "loss": 0.0659,
+      "step": 1953
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4958560544525665,
+      "learning_rate": 9.532547655217645e-06,
+      "loss": 0.067,
+      "step": 1954
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.2990804492458282,
+      "learning_rate": 9.53197137035862e-06,
+      "loss": 0.0769,
+      "step": 1955
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.43936633789455953,
+      "learning_rate": 9.531394747931971e-06,
+      "loss": 0.0711,
+      "step": 1956
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.38520052624273077,
+      "learning_rate": 9.530817787980641e-06,
+      "loss": 0.1048,
+      "step": 1957
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5124849042811812,
+      "learning_rate": 9.53024049054761e-06,
+      "loss": 0.1564,
+      "step": 1958
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2429548537851146,
+      "learning_rate": 9.529662855675876e-06,
+      "loss": 0.0577,
+      "step": 1959
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.30580016502362695,
+      "learning_rate": 9.529084883408463e-06,
+      "loss": 0.0921,
+      "step": 1960
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3281601659052775,
+      "learning_rate": 9.528506573788428e-06,
+      "loss": 0.0743,
+      "step": 1961
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.37229018071737446,
+      "learning_rate": 9.52792792685884e-06,
+      "loss": 0.1052,
+      "step": 1962
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4109827953441356,
+      "learning_rate": 9.527348942662803e-06,
+      "loss": 0.0668,
+      "step": 1963
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3011620513730135,
+      "learning_rate": 9.526769621243444e-06,
+      "loss": 0.0719,
+      "step": 1964
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4098383268555689,
+      "learning_rate": 9.526189962643912e-06,
+      "loss": 0.1168,
+      "step": 1965
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2872030837061024,
+      "learning_rate": 9.525609966907386e-06,
+      "loss": 0.0579,
+      "step": 1966
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.28330110651646995,
+      "learning_rate": 9.525029634077067e-06,
+      "loss": 0.0593,
+      "step": 1967
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.341177430945532,
+      "learning_rate": 9.524448964196179e-06,
+      "loss": 0.0959,
+      "step": 1968
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4452198477193765,
+      "learning_rate": 9.523867957307975e-06,
+      "loss": 0.1281,
+      "step": 1969
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.20941745653167596,
+      "learning_rate": 9.523286613455732e-06,
+      "loss": 0.0541,
+      "step": 1970
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3149012028116164,
+      "learning_rate": 9.522704932682753e-06,
+      "loss": 0.0942,
+      "step": 1971
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.415166406816186,
+      "learning_rate": 9.522122915032364e-06,
+      "loss": 0.0888,
+      "step": 1972
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4143181662701593,
+      "learning_rate": 9.521540560547915e-06,
+      "loss": 0.1015,
+      "step": 1973
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6426815928836875,
+      "learning_rate": 9.520957869272786e-06,
+      "loss": 0.1438,
+      "step": 1974
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4888716888608612,
+      "learning_rate": 9.520374841250379e-06,
+      "loss": 0.1236,
+      "step": 1975
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3896014096527798,
+      "learning_rate": 9.519791476524121e-06,
+      "loss": 0.091,
+      "step": 1976
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6752570099183843,
+      "learning_rate": 9.519207775137463e-06,
+      "loss": 0.1422,
+      "step": 1977
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4218827812769126,
+      "learning_rate": 9.518623737133885e-06,
+      "loss": 0.0929,
+      "step": 1978
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2577582911623791,
+      "learning_rate": 9.518039362556887e-06,
+      "loss": 0.0536,
+      "step": 1979
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.40823862900848484,
+      "learning_rate": 9.517454651449998e-06,
+      "loss": 0.0575,
+      "step": 1980
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.32141817254171867,
+      "learning_rate": 9.516869603856772e-06,
+      "loss": 0.106,
+      "step": 1981
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.23420996589350715,
+      "learning_rate": 9.516284219820781e-06,
+      "loss": 0.0682,
+      "step": 1982
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.46985457661293323,
+      "learning_rate": 9.515698499385635e-06,
+      "loss": 0.1229,
+      "step": 1983
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6751457525658184,
+      "learning_rate": 9.51511244259496e-06,
+      "loss": 0.0951,
+      "step": 1984
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.23594047422484615,
+      "learning_rate": 9.514526049492406e-06,
+      "loss": 0.0588,
+      "step": 1985
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.25525652640834906,
+      "learning_rate": 9.513939320121653e-06,
+      "loss": 0.0499,
+      "step": 1986
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3048005503828047,
+      "learning_rate": 9.513352254526404e-06,
+      "loss": 0.0729,
+      "step": 1987
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.41088215119855725,
+      "learning_rate": 9.512764852750387e-06,
+      "loss": 0.074,
+      "step": 1988
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.41182032987326145,
+      "learning_rate": 9.512177114837355e-06,
+      "loss": 0.0938,
+      "step": 1989
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4364992103322481,
+      "learning_rate": 9.511589040831085e-06,
+      "loss": 0.1263,
+      "step": 1990
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6001211133422241,
+      "learning_rate": 9.511000630775382e-06,
+      "loss": 0.1272,
+      "step": 1991
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.38902850008709944,
+      "learning_rate": 9.510411884714073e-06,
+      "loss": 0.0937,
+      "step": 1992
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5857018855959341,
+      "learning_rate": 9.509822802691013e-06,
+      "loss": 0.118,
+      "step": 1993
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.45607906993765923,
+      "learning_rate": 9.509233384750079e-06,
+      "loss": 0.0999,
+      "step": 1994
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5838549497176535,
+      "learning_rate": 9.508643630935173e-06,
+      "loss": 0.1019,
+      "step": 1995
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.43132563951415875,
+      "learning_rate": 9.508053541290226e-06,
+      "loss": 0.1023,
+      "step": 1996
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.25768875678564085,
+      "learning_rate": 9.507463115859188e-06,
+      "loss": 0.0578,
+      "step": 1997
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4652198591159032,
+      "learning_rate": 9.506872354686041e-06,
+      "loss": 0.134,
+      "step": 1998
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3258001476598917,
+      "learning_rate": 9.506281257814786e-06,
+      "loss": 0.1066,
+      "step": 1999
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6777966835575625,
+      "learning_rate": 9.505689825289452e-06,
+      "loss": 0.1234,
+      "step": 2000
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5419415946372411,
+      "learning_rate": 9.505098057154091e-06,
+      "loss": 0.0992,
+      "step": 2001
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4182726820048844,
+      "learning_rate": 9.504505953452783e-06,
+      "loss": 0.1349,
+      "step": 2002
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2761378143920018,
+      "learning_rate": 9.503913514229633e-06,
+      "loss": 0.0878,
+      "step": 2003
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.35197591167277326,
+      "learning_rate": 9.503320739528765e-06,
+      "loss": 0.082,
+      "step": 2004
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9825059425492416,
+      "learning_rate": 9.502727629394335e-06,
+      "loss": 0.0923,
+      "step": 2005
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.34063089561454185,
+      "learning_rate": 9.502134183870522e-06,
+      "loss": 0.0879,
+      "step": 2006
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3159521140823055,
+      "learning_rate": 9.501540403001527e-06,
+      "loss": 0.1019,
+      "step": 2007
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.44742738624693507,
+      "learning_rate": 9.50094628683158e-06,
+      "loss": 0.1381,
+      "step": 2008
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.258281507717571,
+      "learning_rate": 9.500351835404932e-06,
+      "loss": 0.0893,
+      "step": 2009
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.35100134046858567,
+      "learning_rate": 9.499757048765864e-06,
+      "loss": 0.1014,
+      "step": 2010
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5145820463458471,
+      "learning_rate": 9.499161926958678e-06,
+      "loss": 0.0937,
+      "step": 2011
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.38285465738833085,
+      "learning_rate": 9.498566470027703e-06,
+      "loss": 0.0538,
+      "step": 2012
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.16053243785459514,
+      "learning_rate": 9.49797067801729e-06,
+      "loss": 0.0232,
+      "step": 2013
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.27437296355152263,
+      "learning_rate": 9.497374550971819e-06,
+      "loss": 0.0924,
+      "step": 2014
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2615857204563584,
+      "learning_rate": 9.496778088935691e-06,
+      "loss": 0.0762,
+      "step": 2015
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6125348829553893,
+      "learning_rate": 9.496181291953336e-06,
+      "loss": 0.1113,
+      "step": 2016
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6633824834058893,
+      "learning_rate": 9.495584160069207e-06,
+      "loss": 0.0922,
+      "step": 2017
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3447150786286894,
+      "learning_rate": 9.49498669332778e-06,
+      "loss": 0.0738,
+      "step": 2018
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.28013945144973423,
+      "learning_rate": 9.49438889177356e-06,
+      "loss": 0.0846,
+      "step": 2019
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2814589067238763,
+      "learning_rate": 9.493790755451072e-06,
+      "loss": 0.1006,
+      "step": 2020
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3037464454284003,
+      "learning_rate": 9.49319228440487e-06,
+      "loss": 0.1064,
+      "step": 2021
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.35575872226096344,
+      "learning_rate": 9.492593478679535e-06,
+      "loss": 0.1044,
+      "step": 2022
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7363800305196193,
+      "learning_rate": 9.491994338319664e-06,
+      "loss": 0.1356,
+      "step": 2023
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.27989258563291874,
+      "learning_rate": 9.491394863369887e-06,
+      "loss": 0.093,
+      "step": 2024
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.35861683058009075,
+      "learning_rate": 9.490795053874857e-06,
+      "loss": 0.089,
+      "step": 2025
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.30018812424838803,
+      "learning_rate": 9.49019490987925e-06,
+      "loss": 0.102,
+      "step": 2026
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4750617248573952,
+      "learning_rate": 9.489594431427768e-06,
+      "loss": 0.096,
+      "step": 2027
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4770863256516558,
+      "learning_rate": 9.48899361856514e-06,
+      "loss": 0.0759,
+      "step": 2028
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.45807891432183784,
+      "learning_rate": 9.488392471336117e-06,
+      "loss": 0.0962,
+      "step": 2029
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.35706347253363035,
+      "learning_rate": 9.487790989785475e-06,
+      "loss": 0.1141,
+      "step": 2030
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.44544491838767303,
+      "learning_rate": 9.487189173958019e-06,
+      "loss": 0.103,
+      "step": 2031
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.30005473159971874,
+      "learning_rate": 9.486587023898571e-06,
+      "loss": 0.0504,
+      "step": 2032
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5108988354208385,
+      "learning_rate": 9.485984539651987e-06,
+      "loss": 0.1225,
+      "step": 2033
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6548942786896536,
+      "learning_rate": 9.485381721263141e-06,
+      "loss": 0.118,
+      "step": 2034
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5959468979905713,
+      "learning_rate": 9.484778568776933e-06,
+      "loss": 0.1223,
+      "step": 2035
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4989012562735155,
+      "learning_rate": 9.484175082238293e-06,
+      "loss": 0.104,
+      "step": 2036
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2912968393800039,
+      "learning_rate": 9.483571261692173e-06,
+      "loss": 0.087,
+      "step": 2037
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.45542945615330743,
+      "learning_rate": 9.482967107183544e-06,
+      "loss": 0.1079,
+      "step": 2038
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5814941483700978,
+      "learning_rate": 9.48236261875741e-06,
+      "loss": 0.1483,
+      "step": 2039
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4678829560793343,
+      "learning_rate": 9.481757796458797e-06,
+      "loss": 0.1311,
+      "step": 2040
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2943329225561303,
+      "learning_rate": 9.481152640332756e-06,
+      "loss": 0.1014,
+      "step": 2041
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.41385069370464994,
+      "learning_rate": 9.48054715042436e-06,
+      "loss": 0.1079,
+      "step": 2042
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.332222148353211,
+      "learning_rate": 9.479941326778712e-06,
+      "loss": 0.1055,
+      "step": 2043
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.22114445308384245,
+      "learning_rate": 9.479335169440935e-06,
+      "loss": 0.0596,
+      "step": 2044
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.30042555756992606,
+      "learning_rate": 9.478728678456182e-06,
+      "loss": 0.0421,
+      "step": 2045
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.30598510548256924,
+      "learning_rate": 9.478121853869627e-06,
+      "loss": 0.0869,
+      "step": 2046
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5299024892029863,
+      "learning_rate": 9.477514695726468e-06,
+      "loss": 0.1297,
+      "step": 2047
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4489318886945348,
+      "learning_rate": 9.476907204071933e-06,
+      "loss": 0.0723,
+      "step": 2048
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4075155467965644,
+      "learning_rate": 9.476299378951267e-06,
+      "loss": 0.0857,
+      "step": 2049
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4863623130534948,
+      "learning_rate": 9.475691220409748e-06,
+      "loss": 0.1105,
+      "step": 2050
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5097727038303006,
+      "learning_rate": 9.475082728492674e-06,
+      "loss": 0.1235,
+      "step": 2051
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.25405081088878334,
+      "learning_rate": 9.474473903245369e-06,
+      "loss": 0.0473,
+      "step": 2052
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5517507443911186,
+      "learning_rate": 9.473864744713182e-06,
+      "loss": 0.1069,
+      "step": 2053
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.31697606388279304,
+      "learning_rate": 9.473255252941487e-06,
+      "loss": 0.0984,
+      "step": 2054
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4626150704732813,
+      "learning_rate": 9.472645427975681e-06,
+      "loss": 0.101,
+      "step": 2055
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3303079832175486,
+      "learning_rate": 9.472035269861191e-06,
+      "loss": 0.0878,
+      "step": 2056
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.21718309932001578,
+      "learning_rate": 9.471424778643459e-06,
+      "loss": 0.0563,
+      "step": 2057
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.34375817554153326,
+      "learning_rate": 9.470813954367964e-06,
+      "loss": 0.0607,
+      "step": 2058
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3203867115955599,
+      "learning_rate": 9.4702027970802e-06,
+      "loss": 0.0869,
+      "step": 2059
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.39817440859084197,
+      "learning_rate": 9.469591306825691e-06,
+      "loss": 0.1103,
+      "step": 2060
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5859653342600838,
+      "learning_rate": 9.468979483649987e-06,
+      "loss": 0.1542,
+      "step": 2061
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4474831365686752,
+      "learning_rate": 9.468367327598653e-06,
+      "loss": 0.1067,
+      "step": 2062
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5588590704569549,
+      "learning_rate": 9.467754838717293e-06,
+      "loss": 0.0693,
+      "step": 2063
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.29277718031774563,
+      "learning_rate": 9.467142017051525e-06,
+      "loss": 0.063,
+      "step": 2064
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4096378324409489,
+      "learning_rate": 9.466528862646998e-06,
+      "loss": 0.1117,
+      "step": 2065
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3556425694245333,
+      "learning_rate": 9.465915375549379e-06,
+      "loss": 0.0849,
+      "step": 2066
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3943073867480224,
+      "learning_rate": 9.46530155580437e-06,
+      "loss": 0.0856,
+      "step": 2067
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.32787111926269585,
+      "learning_rate": 9.464687403457687e-06,
+      "loss": 0.0907,
+      "step": 2068
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3656075721271882,
+      "learning_rate": 9.464072918555078e-06,
+      "loss": 0.1105,
+      "step": 2069
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.40992379207978313,
+      "learning_rate": 9.463458101142312e-06,
+      "loss": 0.0675,
+      "step": 2070
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.32100171900447483,
+      "learning_rate": 9.462842951265185e-06,
+      "loss": 0.0951,
+      "step": 2071
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3439146031528608,
+      "learning_rate": 9.462227468969518e-06,
+      "loss": 0.0632,
+      "step": 2072
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4536344153049165,
+      "learning_rate": 9.461611654301155e-06,
+      "loss": 0.1111,
+      "step": 2073
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3579289727457569,
+      "learning_rate": 9.460995507305965e-06,
+      "loss": 0.1062,
+      "step": 2074
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.30603795230627223,
+      "learning_rate": 9.460379028029842e-06,
+      "loss": 0.1014,
+      "step": 2075
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.24227467318400694,
+      "learning_rate": 9.459762216518706e-06,
+      "loss": 0.0597,
+      "step": 2076
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.302783728025485,
+      "learning_rate": 9.459145072818498e-06,
+      "loss": 0.1231,
+      "step": 2077
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.37431265541089237,
+      "learning_rate": 9.45852759697519e-06,
+      "loss": 0.0848,
+      "step": 2078
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3184615399757154,
+      "learning_rate": 9.457909789034774e-06,
+      "loss": 0.0909,
+      "step": 2079
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4409040223910068,
+      "learning_rate": 9.457291649043267e-06,
+      "loss": 0.1108,
+      "step": 2080
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4339535678649829,
+      "learning_rate": 9.456673177046713e-06,
+      "loss": 0.1097,
+      "step": 2081
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3612414208551163,
+      "learning_rate": 9.45605437309118e-06,
+      "loss": 0.0632,
+      "step": 2082
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3460709312038517,
+      "learning_rate": 9.455435237222756e-06,
+      "loss": 0.0821,
+      "step": 2083
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.2643084175150873,
+      "learning_rate": 9.454815769487563e-06,
+      "loss": 0.0527,
+      "step": 2084
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.20437615328576653,
+      "learning_rate": 9.454195969931739e-06,
+      "loss": 0.0323,
+      "step": 2085
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5603073807863361,
+      "learning_rate": 9.453575838601451e-06,
+      "loss": 0.1217,
+      "step": 2086
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.32887009037263376,
+      "learning_rate": 9.452955375542893e-06,
+      "loss": 0.0879,
+      "step": 2087
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.23916719362578137,
+      "learning_rate": 9.452334580802276e-06,
+      "loss": 0.0864,
+      "step": 2088
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.104995178205813,
+      "learning_rate": 9.451713454425845e-06,
+      "loss": 0.0859,
+      "step": 2089
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4746752900559491,
+      "learning_rate": 9.45109199645986e-06,
+      "loss": 0.1314,
+      "step": 2090
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5210535115740903,
+      "learning_rate": 9.450470206950615e-06,
+      "loss": 0.0816,
+      "step": 2091
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.30579030805112756,
+      "learning_rate": 9.449848085944422e-06,
+      "loss": 0.0983,
+      "step": 2092
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3743530382937883,
+      "learning_rate": 9.449225633487623e-06,
+      "loss": 0.1004,
+      "step": 2093
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.2927241158333802,
+      "learning_rate": 9.44860284962658e-06,
+      "loss": 0.0851,
+      "step": 2094
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3093823101893831,
+      "learning_rate": 9.44797973440768e-06,
+      "loss": 0.049,
+      "step": 2095
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.22866188925550068,
+      "learning_rate": 9.447356287877339e-06,
+      "loss": 0.0446,
+      "step": 2096
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.35844488092630356,
+      "learning_rate": 9.446732510081992e-06,
+      "loss": 0.1044,
+      "step": 2097
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5698069975196619,
+      "learning_rate": 9.446108401068108e-06,
+      "loss": 0.1238,
+      "step": 2098
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3582024936492228,
+      "learning_rate": 9.445483960882168e-06,
+      "loss": 0.0888,
+      "step": 2099
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6982166612907237,
+      "learning_rate": 9.444859189570683e-06,
+      "loss": 0.1523,
+      "step": 2100
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.2509513781254003,
+      "learning_rate": 9.444234087180195e-06,
+      "loss": 0.0559,
+      "step": 2101
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5946769788155676,
+      "learning_rate": 9.443608653757261e-06,
+      "loss": 0.099,
+      "step": 2102
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3495230953465962,
+      "learning_rate": 9.442982889348469e-06,
+      "loss": 0.1061,
+      "step": 2103
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5810342216470176,
+      "learning_rate": 9.44235679400043e-06,
+      "loss": 0.1336,
+      "step": 2104
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3876650337333317,
+      "learning_rate": 9.441730367759778e-06,
+      "loss": 0.0846,
+      "step": 2105
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3343822791123102,
+      "learning_rate": 9.441103610673172e-06,
+      "loss": 0.0916,
+      "step": 2106
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.36740235234976004,
+      "learning_rate": 9.4404765227873e-06,
+      "loss": 0.0378,
+      "step": 2107
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.2880713716427908,
+      "learning_rate": 9.439849104148869e-06,
+      "loss": 0.0688,
+      "step": 2108
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.2930059157292598,
+      "learning_rate": 9.439221354804609e-06,
+      "loss": 0.062,
+      "step": 2109
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4194614822083716,
+      "learning_rate": 9.438593274801285e-06,
+      "loss": 0.078,
+      "step": 2110
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.28625071666043683,
+      "learning_rate": 9.437964864185676e-06,
+      "loss": 0.0249,
+      "step": 2111
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.27928142027751585,
+      "learning_rate": 9.437336123004592e-06,
+      "loss": 0.0808,
+      "step": 2112
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4901406575441948,
+      "learning_rate": 9.436707051304865e-06,
+      "loss": 0.1345,
+      "step": 2113
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.28790817624002923,
+      "learning_rate": 9.436077649133348e-06,
+      "loss": 0.0712,
+      "step": 2114
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5233839900416991,
+      "learning_rate": 9.435447916536928e-06,
+      "loss": 0.1125,
+      "step": 2115
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.580184026875001,
+      "learning_rate": 9.434817853562507e-06,
+      "loss": 0.1174,
+      "step": 2116
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.28586132004515585,
+      "learning_rate": 9.43418746025702e-06,
+      "loss": 0.0971,
+      "step": 2117
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4285955082745664,
+      "learning_rate": 9.433556736667419e-06,
+      "loss": 0.1141,
+      "step": 2118
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.32944843354612124,
+      "learning_rate": 9.432925682840685e-06,
+      "loss": 0.0683,
+      "step": 2119
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5606661565661613,
+      "learning_rate": 9.432294298823821e-06,
+      "loss": 0.1292,
+      "step": 2120
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4655470066377842,
+      "learning_rate": 9.43166258466386e-06,
+      "loss": 0.0761,
+      "step": 2121
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.40724900851727097,
+      "learning_rate": 9.43103054040785e-06,
+      "loss": 0.0681,
+      "step": 2122
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5439863568660173,
+      "learning_rate": 9.430398166102875e-06,
+      "loss": 0.1128,
+      "step": 2123
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.40971313338262033,
+      "learning_rate": 9.429765461796037e-06,
+      "loss": 0.0997,
+      "step": 2124
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.46087660431247257,
+      "learning_rate": 9.42913242753446e-06,
+      "loss": 0.1073,
+      "step": 2125
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.2738214807812013,
+      "learning_rate": 9.428499063365297e-06,
+      "loss": 0.0545,
+      "step": 2126
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.427560980158112,
+      "learning_rate": 9.427865369335727e-06,
+      "loss": 0.1062,
+      "step": 2127
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5297915894354662,
+      "learning_rate": 9.42723134549295e-06,
+      "loss": 0.1403,
+      "step": 2128
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3145747221826406,
+      "learning_rate": 9.426596991884193e-06,
+      "loss": 0.0806,
+      "step": 2129
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4041049964939,
+      "learning_rate": 9.425962308556705e-06,
+      "loss": 0.128,
+      "step": 2130
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4590460480746908,
+      "learning_rate": 9.42532729555776e-06,
+      "loss": 0.1096,
+      "step": 2131
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3698802182995215,
+      "learning_rate": 9.42469195293466e-06,
+      "loss": 0.0737,
+      "step": 2132
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.31609869509130495,
+      "learning_rate": 9.424056280734726e-06,
+      "loss": 0.1015,
+      "step": 2133
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.31281062904517365,
+      "learning_rate": 9.423420279005309e-06,
+      "loss": 0.0427,
+      "step": 2134
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3358012374865554,
+      "learning_rate": 9.422783947793782e-06,
+      "loss": 0.0535,
+      "step": 2135
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3448611260191238,
+      "learning_rate": 9.422147287147541e-06,
+      "loss": 0.1163,
+      "step": 2136
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.2690747638099483,
+      "learning_rate": 9.42151029711401e-06,
+      "loss": 0.0481,
+      "step": 2137
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.29342023990224186,
+      "learning_rate": 9.420872977740634e-06,
+      "loss": 0.0729,
+      "step": 2138
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.41884751050214913,
+      "learning_rate": 9.420235329074884e-06,
+      "loss": 0.097,
+      "step": 2139
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6974565456357624,
+      "learning_rate": 9.41959735116426e-06,
+      "loss": 0.0754,
+      "step": 2140
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4781188006590574,
+      "learning_rate": 9.418959044056278e-06,
+      "loss": 0.0984,
+      "step": 2141
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3830671343176985,
+      "learning_rate": 9.418320407798482e-06,
+      "loss": 0.1028,
+      "step": 2142
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5419452405182396,
+      "learning_rate": 9.417681442438445e-06,
+      "loss": 0.1215,
+      "step": 2143
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.37285566721286717,
+      "learning_rate": 9.41704214802376e-06,
+      "loss": 0.115,
+      "step": 2144
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.29797407974493945,
+      "learning_rate": 9.416402524602044e-06,
+      "loss": 0.0718,
+      "step": 2145
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7029191866597169,
+      "learning_rate": 9.41576257222094e-06,
+      "loss": 0.1338,
+      "step": 2146
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.45343169034441955,
+      "learning_rate": 9.415122290928115e-06,
+      "loss": 0.1219,
+      "step": 2147
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.2935701162139371,
+      "learning_rate": 9.414481680771265e-06,
+      "loss": 0.0743,
+      "step": 2148
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5127436577194381,
+      "learning_rate": 9.413840741798101e-06,
+      "loss": 0.0998,
+      "step": 2149
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.35915615424616426,
+      "learning_rate": 9.413199474056365e-06,
+      "loss": 0.0828,
+      "step": 2150
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.30251352534766296,
+      "learning_rate": 9.412557877593826e-06,
+      "loss": 0.0759,
+      "step": 2151
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.33125740737477705,
+      "learning_rate": 9.411915952458271e-06,
+      "loss": 0.084,
+      "step": 2152
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.22244992737697733,
+      "learning_rate": 9.411273698697512e-06,
+      "loss": 0.0474,
+      "step": 2153
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.26695825821277774,
+      "learning_rate": 9.410631116359392e-06,
+      "loss": 0.0688,
+      "step": 2154
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.427976731321004,
+      "learning_rate": 9.409988205491772e-06,
+      "loss": 0.1381,
+      "step": 2155
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5136649457115914,
+      "learning_rate": 9.409344966142541e-06,
+      "loss": 0.1049,
+      "step": 2156
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.40478344377720044,
+      "learning_rate": 9.408701398359612e-06,
+      "loss": 0.0718,
+      "step": 2157
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.34759535441435085,
+      "learning_rate": 9.408057502190918e-06,
+      "loss": 0.0793,
+      "step": 2158
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5597623665737352,
+      "learning_rate": 9.407413277684423e-06,
+      "loss": 0.1169,
+      "step": 2159
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5577336513181574,
+      "learning_rate": 9.406768724888111e-06,
+      "loss": 0.0878,
+      "step": 2160
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3094706847580244,
+      "learning_rate": 9.406123843849995e-06,
+      "loss": 0.0789,
+      "step": 2161
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.365899021360758,
+      "learning_rate": 9.405478634618106e-06,
+      "loss": 0.0844,
+      "step": 2162
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.318880065901717,
+      "learning_rate": 9.404833097240506e-06,
+      "loss": 0.0936,
+      "step": 2163
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.22888144323069895,
+      "learning_rate": 9.404187231765275e-06,
+      "loss": 0.0501,
+      "step": 2164
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.35293310111611703,
+      "learning_rate": 9.403541038240525e-06,
+      "loss": 0.1079,
+      "step": 2165
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.532189552654461,
+      "learning_rate": 9.402894516714384e-06,
+      "loss": 0.1398,
+      "step": 2166
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3490138081760511,
+      "learning_rate": 9.402247667235012e-06,
+      "loss": 0.0966,
+      "step": 2167
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4630106728842395,
+      "learning_rate": 9.401600489850587e-06,
+      "loss": 0.1324,
+      "step": 2168
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.28697340323020426,
+      "learning_rate": 9.400952984609317e-06,
+      "loss": 0.0561,
+      "step": 2169
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.24536692584360947,
+      "learning_rate": 9.400305151559432e-06,
+      "loss": 0.076,
+      "step": 2170
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.27777638254115217,
+      "learning_rate": 9.399656990749185e-06,
+      "loss": 0.037,
+      "step": 2171
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.405952845587326,
+      "learning_rate": 9.399008502226856e-06,
+      "loss": 0.0888,
+      "step": 2172
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6072264127481234,
+      "learning_rate": 9.398359686040748e-06,
+      "loss": 0.1065,
+      "step": 2173
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.39351576737123956,
+      "learning_rate": 9.397710542239187e-06,
+      "loss": 0.0923,
+      "step": 2174
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3863214545580942,
+      "learning_rate": 9.397061070870525e-06,
+      "loss": 0.0926,
+      "step": 2175
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5129355088658487,
+      "learning_rate": 9.396411271983144e-06,
+      "loss": 0.1199,
+      "step": 2176
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.31038248872915813,
+      "learning_rate": 9.395761145625437e-06,
+      "loss": 0.0971,
+      "step": 2177
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.38879340596097023,
+      "learning_rate": 9.395110691845834e-06,
+      "loss": 0.0687,
+      "step": 2178
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.27890422321529085,
+      "learning_rate": 9.394459910692783e-06,
+      "loss": 0.0619,
+      "step": 2179
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5342788844276241,
+      "learning_rate": 9.393808802214757e-06,
+      "loss": 0.0944,
+      "step": 2180
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.34417901606388945,
+      "learning_rate": 9.393157366460257e-06,
+      "loss": 0.0533,
+      "step": 2181
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.31082763396869256,
+      "learning_rate": 9.392505603477804e-06,
+      "loss": 0.0614,
+      "step": 2182
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8692247839565632,
+      "learning_rate": 9.391853513315944e-06,
+      "loss": 0.0755,
+      "step": 2183
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4043925623698882,
+      "learning_rate": 9.391201096023253e-06,
+      "loss": 0.1359,
+      "step": 2184
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.40650029533689236,
+      "learning_rate": 9.390548351648322e-06,
+      "loss": 0.0636,
+      "step": 2185
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3076522315619764,
+      "learning_rate": 9.389895280239772e-06,
+      "loss": 0.0515,
+      "step": 2186
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.2831444067566987,
+      "learning_rate": 9.38924188184625e-06,
+      "loss": 0.0553,
+      "step": 2187
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4827459696495055,
+      "learning_rate": 9.388588156516422e-06,
+      "loss": 0.0715,
+      "step": 2188
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4365680729078727,
+      "learning_rate": 9.387934104298985e-06,
+      "loss": 0.1082,
+      "step": 2189
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3319271240524271,
+      "learning_rate": 9.387279725242654e-06,
+      "loss": 0.0771,
+      "step": 2190
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5746874041097297,
+      "learning_rate": 9.38662501939617e-06,
+      "loss": 0.1285,
+      "step": 2191
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.37256302255387674,
+      "learning_rate": 9.385969986808298e-06,
+      "loss": 0.0721,
+      "step": 2192
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4360181958704788,
+      "learning_rate": 9.385314627527835e-06,
+      "loss": 0.1054,
+      "step": 2193
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.33894115552721427,
+      "learning_rate": 9.38465894160359e-06,
+      "loss": 0.1179,
+      "step": 2194
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.333691616307029,
+      "learning_rate": 9.384002929084406e-06,
+      "loss": 0.0748,
+      "step": 2195
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.35609956030710144,
+      "learning_rate": 9.383346590019146e-06,
+      "loss": 0.0968,
+      "step": 2196
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.31760540487057437,
+      "learning_rate": 9.382689924456696e-06,
+      "loss": 0.1006,
+      "step": 2197
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3173277282721949,
+      "learning_rate": 9.38203293244597e-06,
+      "loss": 0.0655,
+      "step": 2198
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.298370438507746,
+      "learning_rate": 9.381375614035901e-06,
+      "loss": 0.0687,
+      "step": 2199
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8038513184861436,
+      "learning_rate": 9.380717969275456e-06,
+      "loss": 0.1364,
+      "step": 2200
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4106793509241525,
+      "learning_rate": 9.380059998213617e-06,
+      "loss": 0.1307,
+      "step": 2201
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3827998128832313,
+      "learning_rate": 9.379401700899392e-06,
+      "loss": 0.0971,
+      "step": 2202
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.33137795336936254,
+      "learning_rate": 9.378743077381818e-06,
+      "loss": 0.1062,
+      "step": 2203
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.49717090352775806,
+      "learning_rate": 9.37808412770995e-06,
+      "loss": 0.0937,
+      "step": 2204
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4191702221108936,
+      "learning_rate": 9.377424851932872e-06,
+      "loss": 0.1076,
+      "step": 2205
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2500759059903071,
+      "learning_rate": 9.376765250099692e-06,
+      "loss": 0.0596,
+      "step": 2206
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2270574154040705,
+      "learning_rate": 9.376105322259538e-06,
+      "loss": 0.0536,
+      "step": 2207
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4702194750647342,
+      "learning_rate": 9.375445068461568e-06,
+      "loss": 0.1214,
+      "step": 2208
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.40844161283483205,
+      "learning_rate": 9.37478448875496e-06,
+      "loss": 0.1211,
+      "step": 2209
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5057815197012512,
+      "learning_rate": 9.374123583188918e-06,
+      "loss": 0.1064,
+      "step": 2210
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.38530726005947286,
+      "learning_rate": 9.373462351812672e-06,
+      "loss": 0.0948,
+      "step": 2211
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.45419898384468865,
+      "learning_rate": 9.372800794675472e-06,
+      "loss": 0.1122,
+      "step": 2212
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.29468408769535404,
+      "learning_rate": 9.372138911826596e-06,
+      "loss": 0.0747,
+      "step": 2213
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4549855143651968,
+      "learning_rate": 9.371476703315342e-06,
+      "loss": 0.1069,
+      "step": 2214
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3598398049931699,
+      "learning_rate": 9.370814169191038e-06,
+      "loss": 0.1013,
+      "step": 2215
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2637270653793351,
+      "learning_rate": 9.370151309503033e-06,
+      "loss": 0.0759,
+      "step": 2216
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4325472240842392,
+      "learning_rate": 9.369488124300702e-06,
+      "loss": 0.1037,
+      "step": 2217
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.49140396441427825,
+      "learning_rate": 9.36882461363344e-06,
+      "loss": 0.1386,
+      "step": 2218
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.34060658528375604,
+      "learning_rate": 9.36816077755067e-06,
+      "loss": 0.0905,
+      "step": 2219
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.28820060315922585,
+      "learning_rate": 9.36749661610184e-06,
+      "loss": 0.082,
+      "step": 2220
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.21240068006372972,
+      "learning_rate": 9.366832129336421e-06,
+      "loss": 0.056,
+      "step": 2221
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.41145148918730623,
+      "learning_rate": 9.366167317303902e-06,
+      "loss": 0.1087,
+      "step": 2222
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.520918604172463,
+      "learning_rate": 9.36550218005381e-06,
+      "loss": 0.1533,
+      "step": 2223
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3312222826760618,
+      "learning_rate": 9.364836717635684e-06,
+      "loss": 0.0633,
+      "step": 2224
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.32281748387403,
+      "learning_rate": 9.364170930099092e-06,
+      "loss": 0.1257,
+      "step": 2225
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3238950495644765,
+      "learning_rate": 9.363504817493626e-06,
+      "loss": 0.0642,
+      "step": 2226
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.32683535644377365,
+      "learning_rate": 9.3628383798689e-06,
+      "loss": 0.1053,
+      "step": 2227
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.543738753720184,
+      "learning_rate": 9.362171617274558e-06,
+      "loss": 0.0928,
+      "step": 2228
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.33394596719448605,
+      "learning_rate": 9.361504529760261e-06,
+      "loss": 0.1135,
+      "step": 2229
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.38731728770733503,
+      "learning_rate": 9.3608371173757e-06,
+      "loss": 0.1317,
+      "step": 2230
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.493118648199292,
+      "learning_rate": 9.360169380170587e-06,
+      "loss": 0.1046,
+      "step": 2231
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.1734211831737521,
+      "learning_rate": 9.359501318194659e-06,
+      "loss": 0.0362,
+      "step": 2232
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.20783991923757028,
+      "learning_rate": 9.358832931497677e-06,
+      "loss": 0.0515,
+      "step": 2233
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.36125296433964915,
+      "learning_rate": 9.358164220129426e-06,
+      "loss": 0.1244,
+      "step": 2234
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.46045809876037175,
+      "learning_rate": 9.357495184139716e-06,
+      "loss": 0.1326,
+      "step": 2235
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.47959500646781406,
+      "learning_rate": 9.356825823578378e-06,
+      "loss": 0.1288,
+      "step": 2236
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.32732815716506336,
+      "learning_rate": 9.356156138495274e-06,
+      "loss": 0.077,
+      "step": 2237
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3337669782361158,
+      "learning_rate": 9.355486128940285e-06,
+      "loss": 0.0981,
+      "step": 2238
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.24849933527209633,
+      "learning_rate": 9.354815794963316e-06,
+      "loss": 0.0679,
+      "step": 2239
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.23056386406039184,
+      "learning_rate": 9.354145136614297e-06,
+      "loss": 0.0786,
+      "step": 2240
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4393384582260254,
+      "learning_rate": 9.353474153943183e-06,
+      "loss": 0.0706,
+      "step": 2241
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.65118502468742,
+      "learning_rate": 9.352802846999955e-06,
+      "loss": 0.0723,
+      "step": 2242
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3607947846547395,
+      "learning_rate": 9.352131215834613e-06,
+      "loss": 0.0468,
+      "step": 2243
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.587192983956507,
+      "learning_rate": 9.351459260497187e-06,
+      "loss": 0.1347,
+      "step": 2244
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3793657815751059,
+      "learning_rate": 9.350786981037722e-06,
+      "loss": 0.0855,
+      "step": 2245
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3195697837908121,
+      "learning_rate": 9.3501143775063e-06,
+      "loss": 0.0722,
+      "step": 2246
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.308546837720801,
+      "learning_rate": 9.349441449953018e-06,
+      "loss": 0.0956,
+      "step": 2247
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.45377656421589263,
+      "learning_rate": 9.348768198428e-06,
+      "loss": 0.0953,
+      "step": 2248
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3974460014355134,
+      "learning_rate": 9.348094622981392e-06,
+      "loss": 0.0889,
+      "step": 2249
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4431360828459046,
+      "learning_rate": 9.347420723663367e-06,
+      "loss": 0.0853,
+      "step": 2250
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5721494867356366,
+      "learning_rate": 9.346746500524122e-06,
+      "loss": 0.158,
+      "step": 2251
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.326486895017683,
+      "learning_rate": 9.346071953613876e-06,
+      "loss": 0.0853,
+      "step": 2252
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.32827540075127837,
+      "learning_rate": 9.345397082982873e-06,
+      "loss": 0.0774,
+      "step": 2253
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6338444646629003,
+      "learning_rate": 9.344721888681379e-06,
+      "loss": 0.0916,
+      "step": 2254
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.49580658565767105,
+      "learning_rate": 9.34404637075969e-06,
+      "loss": 0.0638,
+      "step": 2255
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2593705392274507,
+      "learning_rate": 9.343370529268123e-06,
+      "loss": 0.0651,
+      "step": 2256
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.31254673882322603,
+      "learning_rate": 9.342694364257015e-06,
+      "loss": 0.0625,
+      "step": 2257
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6255686129843026,
+      "learning_rate": 9.342017875776734e-06,
+      "loss": 0.1349,
+      "step": 2258
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.39507784829686776,
+      "learning_rate": 9.341341063877667e-06,
+      "loss": 0.115,
+      "step": 2259
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.21122245510312226,
+      "learning_rate": 9.340663928610227e-06,
+      "loss": 0.0624,
+      "step": 2260
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8570032934631387,
+      "learning_rate": 9.339986470024853e-06,
+      "loss": 0.1522,
+      "step": 2261
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3935293152076621,
+      "learning_rate": 9.339308688172003e-06,
+      "loss": 0.1046,
+      "step": 2262
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.41578165502469455,
+      "learning_rate": 9.338630583102164e-06,
+      "loss": 0.0802,
+      "step": 2263
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.42771859563311004,
+      "learning_rate": 9.337952154865844e-06,
+      "loss": 0.1275,
+      "step": 2264
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.33509295748410917,
+      "learning_rate": 9.337273403513578e-06,
+      "loss": 0.0916,
+      "step": 2265
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5729590793993866,
+      "learning_rate": 9.336594329095922e-06,
+      "loss": 0.0876,
+      "step": 2266
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.33791363124289925,
+      "learning_rate": 9.33591493166346e-06,
+      "loss": 0.1008,
+      "step": 2267
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.41375369908154275,
+      "learning_rate": 9.33523521126679e-06,
+      "loss": 0.0997,
+      "step": 2268
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3701664024187532,
+      "learning_rate": 9.334555167956551e-06,
+      "loss": 0.0898,
+      "step": 2269
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5660125964065214,
+      "learning_rate": 9.333874801783393e-06,
+      "loss": 0.0867,
+      "step": 2270
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3784728417953419,
+      "learning_rate": 9.333194112797991e-06,
+      "loss": 0.0732,
+      "step": 2271
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.47613957231436194,
+      "learning_rate": 9.332513101051049e-06,
+      "loss": 0.1201,
+      "step": 2272
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2936014193054595,
+      "learning_rate": 9.331831766593294e-06,
+      "loss": 0.0746,
+      "step": 2273
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.48955980433375457,
+      "learning_rate": 9.331150109475473e-06,
+      "loss": 0.1196,
+      "step": 2274
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8826617882758704,
+      "learning_rate": 9.33046812974836e-06,
+      "loss": 0.1255,
+      "step": 2275
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.30043859612719337,
+      "learning_rate": 9.329785827462757e-06,
+      "loss": 0.0644,
+      "step": 2276
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.41170057237837054,
+      "learning_rate": 9.32910320266948e-06,
+      "loss": 0.1034,
+      "step": 2277
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3056464923032891,
+      "learning_rate": 9.328420255419377e-06,
+      "loss": 0.0758,
+      "step": 2278
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.32642016227222376,
+      "learning_rate": 9.327736985763321e-06,
+      "loss": 0.0912,
+      "step": 2279
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.498843936151261,
+      "learning_rate": 9.3270533937522e-06,
+      "loss": 0.1431,
+      "step": 2280
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.29905728999205355,
+      "learning_rate": 9.326369479436938e-06,
+      "loss": 0.0705,
+      "step": 2281
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8265670552442287,
+      "learning_rate": 9.325685242868475e-06,
+      "loss": 0.1591,
+      "step": 2282
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.44808485307793977,
+      "learning_rate": 9.325000684097774e-06,
+      "loss": 0.1192,
+      "step": 2283
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.38484088495247626,
+      "learning_rate": 9.324315803175827e-06,
+      "loss": 0.0683,
+      "step": 2284
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3287297689946327,
+      "learning_rate": 9.32363060015365e-06,
+      "loss": 0.1304,
+      "step": 2285
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4806725005323175,
+      "learning_rate": 9.322945075082278e-06,
+      "loss": 0.0825,
+      "step": 2286
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5125632133098119,
+      "learning_rate": 9.322259228012774e-06,
+      "loss": 0.0768,
+      "step": 2287
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6962424315501092,
+      "learning_rate": 9.321573058996223e-06,
+      "loss": 0.1203,
+      "step": 2288
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7708997007821469,
+      "learning_rate": 9.320886568083736e-06,
+      "loss": 0.1216,
+      "step": 2289
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2948621485012947,
+      "learning_rate": 9.320199755326445e-06,
+      "loss": 0.0871,
+      "step": 2290
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.49689409162597564,
+      "learning_rate": 9.319512620775511e-06,
+      "loss": 0.1198,
+      "step": 2291
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.24977304550679208,
+      "learning_rate": 9.318825164482112e-06,
+      "loss": 0.055,
+      "step": 2292
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.30161144218119884,
+      "learning_rate": 9.318137386497457e-06,
+      "loss": 0.0359,
+      "step": 2293
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.42738038580433657,
+      "learning_rate": 9.317449286872775e-06,
+      "loss": 0.0986,
+      "step": 2294
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5937263243641031,
+      "learning_rate": 9.316760865659318e-06,
+      "loss": 0.1398,
+      "step": 2295
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.34344885761550314,
+      "learning_rate": 9.316072122908366e-06,
+      "loss": 0.0655,
+      "step": 2296
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4584300110460388,
+      "learning_rate": 9.315383058671219e-06,
+      "loss": 0.1029,
+      "step": 2297
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3314484366634366,
+      "learning_rate": 9.314693672999201e-06,
+      "loss": 0.0739,
+      "step": 2298
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.365909450893409,
+      "learning_rate": 9.314003965943665e-06,
+      "loss": 0.076,
+      "step": 2299
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2681066536098896,
+      "learning_rate": 9.313313937555982e-06,
+      "loss": 0.0517,
+      "step": 2300
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.37471136864222326,
+      "learning_rate": 9.31262358788755e-06,
+      "loss": 0.1204,
+      "step": 2301
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.40826293451460227,
+      "learning_rate": 9.311932916989792e-06,
+      "loss": 0.1175,
+      "step": 2302
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3441914664708332,
+      "learning_rate": 9.31124192491415e-06,
+      "loss": 0.0812,
+      "step": 2303
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.31759544099272835,
+      "learning_rate": 9.310550611712095e-06,
+      "loss": 0.1103,
+      "step": 2304
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2692933162350318,
+      "learning_rate": 9.309858977435118e-06,
+      "loss": 0.0604,
+      "step": 2305
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0808510236239905,
+      "learning_rate": 9.30916702213474e-06,
+      "loss": 0.0572,
+      "step": 2306
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.31330759661430957,
+      "learning_rate": 9.3084747458625e-06,
+      "loss": 0.0587,
+      "step": 2307
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2602518858215506,
+      "learning_rate": 9.307782148669959e-06,
+      "loss": 0.0736,
+      "step": 2308
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.41637649848546915,
+      "learning_rate": 9.30708923060871e-06,
+      "loss": 0.1,
+      "step": 2309
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4175821778769072,
+      "learning_rate": 9.306395991730365e-06,
+      "loss": 0.0896,
+      "step": 2310
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.46330499700208705,
+      "learning_rate": 9.30570243208656e-06,
+      "loss": 0.1124,
+      "step": 2311
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2974488155755075,
+      "learning_rate": 9.305008551728956e-06,
+      "loss": 0.0576,
+      "step": 2312
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.23501207289869444,
+      "learning_rate": 9.304314350709235e-06,
+      "loss": 0.0652,
+      "step": 2313
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.14831808420974502,
+      "learning_rate": 9.303619829079107e-06,
+      "loss": 0.0336,
+      "step": 2314
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.25836482441719333,
+      "learning_rate": 9.302924986890304e-06,
+      "loss": 0.0878,
+      "step": 2315
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6367297322592059,
+      "learning_rate": 9.302229824194582e-06,
+      "loss": 0.0912,
+      "step": 2316
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.2534834102230206,
+      "learning_rate": 9.30153434104372e-06,
+      "loss": 0.0913,
+      "step": 2317
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.30063392575062886,
+      "learning_rate": 9.300838537489522e-06,
+      "loss": 0.0779,
+      "step": 2318
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6081510267527095,
+      "learning_rate": 9.300142413583815e-06,
+      "loss": 0.1063,
+      "step": 2319
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.27676304477307284,
+      "learning_rate": 9.299445969378451e-06,
+      "loss": 0.0657,
+      "step": 2320
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.38594690112935753,
+      "learning_rate": 9.298749204925305e-06,
+      "loss": 0.0847,
+      "step": 2321
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3695285992143518,
+      "learning_rate": 9.298052120276277e-06,
+      "loss": 0.1073,
+      "step": 2322
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.45642639679703134,
+      "learning_rate": 9.297354715483288e-06,
+      "loss": 0.1376,
+      "step": 2323
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.20126860442180566,
+      "learning_rate": 9.296656990598288e-06,
+      "loss": 0.0356,
+      "step": 2324
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.46927238596496695,
+      "learning_rate": 9.295958945673243e-06,
+      "loss": 0.1077,
+      "step": 2325
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.29367986907341875,
+      "learning_rate": 9.29526058076015e-06,
+      "loss": 0.1075,
+      "step": 2326
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4288452773278697,
+      "learning_rate": 9.29456189591103e-06,
+      "loss": 0.0947,
+      "step": 2327
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.34003507946587996,
+      "learning_rate": 9.29386289117792e-06,
+      "loss": 0.1019,
+      "step": 2328
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.2659820539803868,
+      "learning_rate": 9.293163566612888e-06,
+      "loss": 0.0898,
+      "step": 2329
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.22683154753930837,
+      "learning_rate": 9.292463922268025e-06,
+      "loss": 0.059,
+      "step": 2330
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.24026047394087163,
+      "learning_rate": 9.291763958195444e-06,
+      "loss": 0.0666,
+      "step": 2331
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.25051432073764,
+      "learning_rate": 9.29106367444728e-06,
+      "loss": 0.1071,
+      "step": 2332
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.41609271981571083,
+      "learning_rate": 9.290363071075699e-06,
+      "loss": 0.0782,
+      "step": 2333
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.318479064236394,
+      "learning_rate": 9.28966214813288e-06,
+      "loss": 0.0829,
+      "step": 2334
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.2887733483241954,
+      "learning_rate": 9.288960905671038e-06,
+      "loss": 0.097,
+      "step": 2335
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5320088914392971,
+      "learning_rate": 9.2882593437424e-06,
+      "loss": 0.1005,
+      "step": 2336
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3139733696046921,
+      "learning_rate": 9.287557462399228e-06,
+      "loss": 0.069,
+      "step": 2337
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.41702533047427764,
+      "learning_rate": 9.286855261693798e-06,
+      "loss": 0.0985,
+      "step": 2338
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.32207282016958627,
+      "learning_rate": 9.286152741678416e-06,
+      "loss": 0.0986,
+      "step": 2339
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.41825571094765834,
+      "learning_rate": 9.285449902405409e-06,
+      "loss": 0.077,
+      "step": 2340
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.38126365470644813,
+      "learning_rate": 9.284746743927127e-06,
+      "loss": 0.0761,
+      "step": 2341
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5435371652562497,
+      "learning_rate": 9.284043266295948e-06,
+      "loss": 0.1306,
+      "step": 2342
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3030223415500926,
+      "learning_rate": 9.28333946956427e-06,
+      "loss": 0.0723,
+      "step": 2343
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.36309697990515394,
+      "learning_rate": 9.282635353784517e-06,
+      "loss": 0.0861,
+      "step": 2344
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5488755809777321,
+      "learning_rate": 9.281930919009134e-06,
+      "loss": 0.1391,
+      "step": 2345
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3449667837760911,
+      "learning_rate": 9.281226165290592e-06,
+      "loss": 0.0976,
+      "step": 2346
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5983205483149432,
+      "learning_rate": 9.280521092681386e-06,
+      "loss": 0.1305,
+      "step": 2347
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.314905034568646,
+      "learning_rate": 9.279815701234032e-06,
+      "loss": 0.072,
+      "step": 2348
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.44028354166455647,
+      "learning_rate": 9.279109991001073e-06,
+      "loss": 0.1186,
+      "step": 2349
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4038271835741766,
+      "learning_rate": 9.278403962035074e-06,
+      "loss": 0.0854,
+      "step": 2350
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3687027084033775,
+      "learning_rate": 9.277697614388624e-06,
+      "loss": 0.0964,
+      "step": 2351
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3238511718452813,
+      "learning_rate": 9.276990948114338e-06,
+      "loss": 0.0841,
+      "step": 2352
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.2705285581510575,
+      "learning_rate": 9.27628396326485e-06,
+      "loss": 0.0527,
+      "step": 2353
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.33347708661952113,
+      "learning_rate": 9.27557665989282e-06,
+      "loss": 0.0795,
+      "step": 2354
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5950398540763525,
+      "learning_rate": 9.274869038050936e-06,
+      "loss": 0.1286,
+      "step": 2355
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.22495757752564471,
+      "learning_rate": 9.2741610977919e-06,
+      "loss": 0.0693,
+      "step": 2356
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5407923898800172,
+      "learning_rate": 9.273452839168449e-06,
+      "loss": 0.1086,
+      "step": 2357
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.43297286812618463,
+      "learning_rate": 9.272744262233334e-06,
+      "loss": 0.0435,
+      "step": 2358
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.41466706197337605,
+      "learning_rate": 9.272035367039337e-06,
+      "loss": 0.097,
+      "step": 2359
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7195172691390104,
+      "learning_rate": 9.271326153639259e-06,
+      "loss": 0.0906,
+      "step": 2360
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5396883792642421,
+      "learning_rate": 9.270616622085926e-06,
+      "loss": 0.1367,
+      "step": 2361
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4296776492948369,
+      "learning_rate": 9.269906772432191e-06,
+      "loss": 0.0852,
+      "step": 2362
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6028479135930933,
+      "learning_rate": 9.269196604730923e-06,
+      "loss": 0.1211,
+      "step": 2363
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5700877451794317,
+      "learning_rate": 9.268486119035024e-06,
+      "loss": 0.1161,
+      "step": 2364
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.45844944809909693,
+      "learning_rate": 9.267775315397413e-06,
+      "loss": 0.1054,
+      "step": 2365
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.2358504304417916,
+      "learning_rate": 9.267064193871033e-06,
+      "loss": 0.0767,
+      "step": 2366
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4303326377577767,
+      "learning_rate": 9.266352754508855e-06,
+      "loss": 0.0775,
+      "step": 2367
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.48018137671701383,
+      "learning_rate": 9.265640997363871e-06,
+      "loss": 0.1213,
+      "step": 2368
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4550399629095118,
+      "learning_rate": 9.264928922489097e-06,
+      "loss": 0.0711,
+      "step": 2369
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6185696510012659,
+      "learning_rate": 9.26421652993757e-06,
+      "loss": 0.097,
+      "step": 2370
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.36518670265197134,
+      "learning_rate": 9.263503819762357e-06,
+      "loss": 0.0928,
+      "step": 2371
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5643621734805451,
+      "learning_rate": 9.262790792016543e-06,
+      "loss": 0.1034,
+      "step": 2372
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3684383768262814,
+      "learning_rate": 9.262077446753236e-06,
+      "loss": 0.0897,
+      "step": 2373
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.35139945998552913,
+      "learning_rate": 9.261363784025574e-06,
+      "loss": 0.0697,
+      "step": 2374
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.36686908286411946,
+      "learning_rate": 9.260649803886712e-06,
+      "loss": 0.1003,
+      "step": 2375
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.38824144474565186,
+      "learning_rate": 9.259935506389833e-06,
+      "loss": 0.1078,
+      "step": 2376
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.24022284061470212,
+      "learning_rate": 9.259220891588141e-06,
+      "loss": 0.0583,
+      "step": 2377
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.31705229857491707,
+      "learning_rate": 9.258505959534867e-06,
+      "loss": 0.0675,
+      "step": 2378
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3285945161504574,
+      "learning_rate": 9.257790710283258e-06,
+      "loss": 0.0715,
+      "step": 2379
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5356269063347308,
+      "learning_rate": 9.257075143886598e-06,
+      "loss": 0.1543,
+      "step": 2380
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3535088220286594,
+      "learning_rate": 9.256359260398178e-06,
+      "loss": 0.0745,
+      "step": 2381
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4514672660471702,
+      "learning_rate": 9.255643059871327e-06,
+      "loss": 0.1317,
+      "step": 2382
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.36110557820870454,
+      "learning_rate": 9.25492654235939e-06,
+      "loss": 0.0718,
+      "step": 2383
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.2564055114431298,
+      "learning_rate": 9.254209707915737e-06,
+      "loss": 0.0409,
+      "step": 2384
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4880596413762622,
+      "learning_rate": 9.253492556593763e-06,
+      "loss": 0.1038,
+      "step": 2385
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.26524893426896134,
+      "learning_rate": 9.252775088446884e-06,
+      "loss": 0.0576,
+      "step": 2386
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4003603107127384,
+      "learning_rate": 9.252057303528544e-06,
+      "loss": 0.0787,
+      "step": 2387
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3406245558755874,
+      "learning_rate": 9.251339201892203e-06,
+      "loss": 0.0813,
+      "step": 2388
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3261210254957526,
+      "learning_rate": 9.250620783591354e-06,
+      "loss": 0.0612,
+      "step": 2389
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.21396491269875514,
+      "learning_rate": 9.249902048679507e-06,
+      "loss": 0.0538,
+      "step": 2390
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.2796425878880943,
+      "learning_rate": 9.249182997210198e-06,
+      "loss": 0.0733,
+      "step": 2391
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5987007476421978,
+      "learning_rate": 9.248463629236987e-06,
+      "loss": 0.0689,
+      "step": 2392
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.26341944254403304,
+      "learning_rate": 9.247743944813454e-06,
+      "loss": 0.0443,
+      "step": 2393
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.29488704285765055,
+      "learning_rate": 9.247023943993208e-06,
+      "loss": 0.0729,
+      "step": 2394
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.649137868596947,
+      "learning_rate": 9.246303626829878e-06,
+      "loss": 0.145,
+      "step": 2395
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.36101111275762043,
+      "learning_rate": 9.245582993377117e-06,
+      "loss": 0.0702,
+      "step": 2396
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4622055912053914,
+      "learning_rate": 9.244862043688602e-06,
+      "loss": 0.0905,
+      "step": 2397
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.476919771871768,
+      "learning_rate": 9.244140777818036e-06,
+      "loss": 0.1108,
+      "step": 2398
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3624400139473859,
+      "learning_rate": 9.243419195819139e-06,
+      "loss": 0.0803,
+      "step": 2399
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.2591384784949351,
+      "learning_rate": 9.24269729774566e-06,
+      "loss": 0.0556,
+      "step": 2400
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6100973014729856,
+      "learning_rate": 9.241975083651372e-06,
+      "loss": 0.1099,
+      "step": 2401
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.33070414540174936,
+      "learning_rate": 9.241252553590068e-06,
+      "loss": 0.0579,
+      "step": 2402
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.45784867313565886,
+      "learning_rate": 9.240529707615566e-06,
+      "loss": 0.1064,
+      "step": 2403
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.47645895537709654,
+      "learning_rate": 9.23980654578171e-06,
+      "loss": 0.096,
+      "step": 2404
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.709165308299525,
+      "learning_rate": 9.239083068142362e-06,
+      "loss": 0.0844,
+      "step": 2405
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4169742623222363,
+      "learning_rate": 9.238359274751414e-06,
+      "loss": 0.1046,
+      "step": 2406
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3353419363137753,
+      "learning_rate": 9.237635165662777e-06,
+      "loss": 0.0742,
+      "step": 2407
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.30777185027966525,
+      "learning_rate": 9.236910740930385e-06,
+      "loss": 0.0772,
+      "step": 2408
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4025918348115341,
+      "learning_rate": 9.236186000608202e-06,
+      "loss": 0.0726,
+      "step": 2409
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.31819153065209155,
+      "learning_rate": 9.235460944750206e-06,
+      "loss": 0.0784,
+      "step": 2410
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.36466499319636253,
+      "learning_rate": 9.234735573410406e-06,
+      "loss": 0.1129,
+      "step": 2411
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.419476596826727,
+      "learning_rate": 9.234009886642832e-06,
+      "loss": 0.1039,
+      "step": 2412
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.43847129055073936,
+      "learning_rate": 9.233283884501535e-06,
+      "loss": 0.1252,
+      "step": 2413
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.24861610736164041,
+      "learning_rate": 9.232557567040593e-06,
+      "loss": 0.04,
+      "step": 2414
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3398709188441637,
+      "learning_rate": 9.23183093431411e-06,
+      "loss": 0.054,
+      "step": 2415
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4491830416661052,
+      "learning_rate": 9.231103986376207e-06,
+      "loss": 0.0919,
+      "step": 2416
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.29298627885131784,
+      "learning_rate": 9.230376723281027e-06,
+      "loss": 0.1054,
+      "step": 2417
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.41701764202936675,
+      "learning_rate": 9.229649145082749e-06,
+      "loss": 0.061,
+      "step": 2418
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4385706379841609,
+      "learning_rate": 9.228921251835562e-06,
+      "loss": 0.106,
+      "step": 2419
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.41182881312011,
+      "learning_rate": 9.228193043593682e-06,
+      "loss": 0.1181,
+      "step": 2420
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.44125008037721003,
+      "learning_rate": 9.227464520411355e-06,
+      "loss": 0.1028,
+      "step": 2421
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.34476323497136757,
+      "learning_rate": 9.226735682342846e-06,
+      "loss": 0.0996,
+      "step": 2422
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.32370349483702815,
+      "learning_rate": 9.226006529442439e-06,
+      "loss": 0.0564,
+      "step": 2423
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.639512593249112,
+      "learning_rate": 9.225277061764447e-06,
+      "loss": 0.0979,
+      "step": 2424
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.2591802918330556,
+      "learning_rate": 9.224547279363206e-06,
+      "loss": 0.072,
+      "step": 2425
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.25234330520223974,
+      "learning_rate": 9.223817182293074e-06,
+      "loss": 0.0726,
+      "step": 2426
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.39836143275817065,
+      "learning_rate": 9.223086770608432e-06,
+      "loss": 0.039,
+      "step": 2427
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.25288854154711077,
+      "learning_rate": 9.222356044363686e-06,
+      "loss": 0.0717,
+      "step": 2428
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.45690902308341924,
+      "learning_rate": 9.221625003613263e-06,
+      "loss": 0.1274,
+      "step": 2429
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.19947169844895804,
+      "learning_rate": 9.22089364841162e-06,
+      "loss": 0.057,
+      "step": 2430
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3574262981023576,
+      "learning_rate": 9.220161978813229e-06,
+      "loss": 0.0988,
+      "step": 2431
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4760489126589686,
+      "learning_rate": 9.219429994872588e-06,
+      "loss": 0.1129,
+      "step": 2432
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.35641492586824974,
+      "learning_rate": 9.218697696644223e-06,
+      "loss": 0.0794,
+      "step": 2433
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6287296604998953,
+      "learning_rate": 9.217965084182676e-06,
+      "loss": 0.1588,
+      "step": 2434
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4760913700629727,
+      "learning_rate": 9.217232157542519e-06,
+      "loss": 0.0825,
+      "step": 2435
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.42036001748068064,
+      "learning_rate": 9.216498916778345e-06,
+      "loss": 0.1032,
+      "step": 2436
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.40016852278390136,
+      "learning_rate": 9.215765361944766e-06,
+      "loss": 0.09,
+      "step": 2437
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3548804949368742,
+      "learning_rate": 9.215031493096428e-06,
+      "loss": 0.0833,
+      "step": 2438
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7681244631680875,
+      "learning_rate": 9.214297310287985e-06,
+      "loss": 0.1298,
+      "step": 2439
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.36731406638271036,
+      "learning_rate": 9.213562813574131e-06,
+      "loss": 0.0926,
+      "step": 2440
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.39382662804399066,
+      "learning_rate": 9.212828003009573e-06,
+      "loss": 0.0633,
+      "step": 2441
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3790270042765703,
+      "learning_rate": 9.212092878649044e-06,
+      "loss": 0.0772,
+      "step": 2442
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6067136613967411,
+      "learning_rate": 9.2113574405473e-06,
+      "loss": 0.1227,
+      "step": 2443
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4456505377332532,
+      "learning_rate": 9.21062168875912e-06,
+      "loss": 0.0958,
+      "step": 2444
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5324426399153132,
+      "learning_rate": 9.209885623339308e-06,
+      "loss": 0.1202,
+      "step": 2445
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3047861347538619,
+      "learning_rate": 9.209149244342692e-06,
+      "loss": 0.0797,
+      "step": 2446
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.32809350280302174,
+      "learning_rate": 9.208412551824117e-06,
+      "loss": 0.0891,
+      "step": 2447
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.25374701300360175,
+      "learning_rate": 9.207675545838463e-06,
+      "loss": 0.0654,
+      "step": 2448
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5349845644741081,
+      "learning_rate": 9.20693822644062e-06,
+      "loss": 0.0736,
+      "step": 2449
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.31489577087665815,
+      "learning_rate": 9.206200593685513e-06,
+      "loss": 0.0612,
+      "step": 2450
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.42791277062451033,
+      "learning_rate": 9.20546264762808e-06,
+      "loss": 0.113,
+      "step": 2451
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.38029799496525946,
+      "learning_rate": 9.204724388323292e-06,
+      "loss": 0.0876,
+      "step": 2452
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3190607566560223,
+      "learning_rate": 9.203985815826137e-06,
+      "loss": 0.1128,
+      "step": 2453
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.37339328848382636,
+      "learning_rate": 9.20324693019163e-06,
+      "loss": 0.1212,
+      "step": 2454
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6552134421042066,
+      "learning_rate": 9.202507731474803e-06,
+      "loss": 0.0428,
+      "step": 2455
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.48053717660764556,
+      "learning_rate": 9.201768219730722e-06,
+      "loss": 0.1225,
+      "step": 2456
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.2934395309830181,
+      "learning_rate": 9.201028395014464e-06,
+      "loss": 0.0913,
+      "step": 2457
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3017323536887283,
+      "learning_rate": 9.20028825738114e-06,
+      "loss": 0.083,
+      "step": 2458
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3035041835105427,
+      "learning_rate": 9.199547806885878e-06,
+      "loss": 0.0809,
+      "step": 2459
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.31855927492886943,
+      "learning_rate": 9.198807043583831e-06,
+      "loss": 0.0858,
+      "step": 2460
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.34542283170579696,
+      "learning_rate": 9.198065967530176e-06,
+      "loss": 0.0928,
+      "step": 2461
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5063082976442057,
+      "learning_rate": 9.197324578780111e-06,
+      "loss": 0.0975,
+      "step": 2462
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3680342856190144,
+      "learning_rate": 9.19658287738886e-06,
+      "loss": 0.0939,
+      "step": 2463
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.47664354574423795,
+      "learning_rate": 9.19584086341167e-06,
+      "loss": 0.1089,
+      "step": 2464
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.40856246493509385,
+      "learning_rate": 9.19509853690381e-06,
+      "loss": 0.0952,
+      "step": 2465
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.2896309161783139,
+      "learning_rate": 9.194355897920572e-06,
+      "loss": 0.0654,
+      "step": 2466
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3672838319742439,
+      "learning_rate": 9.193612946517274e-06,
+      "loss": 0.1358,
+      "step": 2467
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.380287434199588,
+      "learning_rate": 9.192869682749255e-06,
+      "loss": 0.1189,
+      "step": 2468
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.40946803912484275,
+      "learning_rate": 9.192126106671874e-06,
+      "loss": 0.1073,
+      "step": 2469
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.26418849420460605,
+      "learning_rate": 9.19138221834052e-06,
+      "loss": 0.0745,
+      "step": 2470
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.39109404299844663,
+      "learning_rate": 9.190638017810604e-06,
+      "loss": 0.1041,
+      "step": 2471
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4574658893406799,
+      "learning_rate": 9.189893505137555e-06,
+      "loss": 0.0931,
+      "step": 2472
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.24512730290637746,
+      "learning_rate": 9.189148680376829e-06,
+      "loss": 0.076,
+      "step": 2473
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5070097203437534,
+      "learning_rate": 9.188403543583909e-06,
+      "loss": 0.0952,
+      "step": 2474
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.25139712395579317,
+      "learning_rate": 9.187658094814288e-06,
+      "loss": 0.0603,
+      "step": 2475
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.48847416883191197,
+      "learning_rate": 9.186912334123503e-06,
+      "loss": 0.0929,
+      "step": 2476
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4937015884513572,
+      "learning_rate": 9.186166261567093e-06,
+      "loss": 0.0801,
+      "step": 2477
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3857002776517484,
+      "learning_rate": 9.185419877200636e-06,
+      "loss": 0.0735,
+      "step": 2478
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.268295717158448,
+      "learning_rate": 9.184673181079723e-06,
+      "loss": 0.0657,
+      "step": 2479
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.611197781324108,
+      "learning_rate": 9.183926173259974e-06,
+      "loss": 0.0999,
+      "step": 2480
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.35630875181460675,
+      "learning_rate": 9.18317885379703e-06,
+      "loss": 0.1128,
+      "step": 2481
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.30926034826405013,
+      "learning_rate": 9.182431222746557e-06,
+      "loss": 0.0943,
+      "step": 2482
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7206285124600662,
+      "learning_rate": 9.181683280164242e-06,
+      "loss": 0.1058,
+      "step": 2483
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.2983424124580018,
+      "learning_rate": 9.180935026105797e-06,
+      "loss": 0.0763,
+      "step": 2484
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4113553776440482,
+      "learning_rate": 9.180186460626955e-06,
+      "loss": 0.075,
+      "step": 2485
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4121105870105272,
+      "learning_rate": 9.179437583783474e-06,
+      "loss": 0.089,
+      "step": 2486
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.23062254175103497,
+      "learning_rate": 9.178688395631134e-06,
+      "loss": 0.0766,
+      "step": 2487
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4145102068514547,
+      "learning_rate": 9.17793889622574e-06,
+      "loss": 0.0978,
+      "step": 2488
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3219304579323048,
+      "learning_rate": 9.177189085623119e-06,
+      "loss": 0.1079,
+      "step": 2489
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.328843946556971,
+      "learning_rate": 9.176438963879121e-06,
+      "loss": 0.0803,
+      "step": 2490
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.24996337275980357,
+      "learning_rate": 9.17568853104962e-06,
+      "loss": 0.06,
+      "step": 2491
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4758088012089373,
+      "learning_rate": 9.174937787190512e-06,
+      "loss": 0.1465,
+      "step": 2492
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3223460355606129,
+      "learning_rate": 9.174186732357715e-06,
+      "loss": 0.0722,
+      "step": 2493
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.30139502075431784,
+      "learning_rate": 9.173435366607176e-06,
+      "loss": 0.0745,
+      "step": 2494
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4966418311455309,
+      "learning_rate": 9.172683689994856e-06,
+      "loss": 0.1067,
+      "step": 2495
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6473820198850618,
+      "learning_rate": 9.171931702576751e-06,
+      "loss": 0.1579,
+      "step": 2496
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4254062331229803,
+      "learning_rate": 9.171179404408866e-06,
+      "loss": 0.1003,
+      "step": 2497
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.2761746671294394,
+      "learning_rate": 9.170426795547241e-06,
+      "loss": 0.071,
+      "step": 2498
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4807357512511184,
+      "learning_rate": 9.169673876047935e-06,
+      "loss": 0.1233,
+      "step": 2499
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.37588415670573455,
+      "learning_rate": 9.168920645967027e-06,
+      "loss": 0.1085,
+      "step": 2500
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5007401461969425,
+      "learning_rate": 9.168167105360625e-06,
+      "loss": 0.0819,
+      "step": 2501
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4948651541518069,
+      "learning_rate": 9.167413254284854e-06,
+      "loss": 0.1089,
+      "step": 2502
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.48322159862417385,
+      "learning_rate": 9.166659092795869e-06,
+      "loss": 0.1079,
+      "step": 2503
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9899629578354386,
+      "learning_rate": 9.165904620949839e-06,
+      "loss": 0.1506,
+      "step": 2504
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.33693433977736276,
+      "learning_rate": 9.165149838802967e-06,
+      "loss": 0.062,
+      "step": 2505
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.36860823688203725,
+      "learning_rate": 9.164394746411471e-06,
+      "loss": 0.0961,
+      "step": 2506
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.24010900256428624,
+      "learning_rate": 9.163639343831595e-06,
+      "loss": 0.0742,
+      "step": 2507
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3240588744915587,
+      "learning_rate": 9.162883631119606e-06,
+      "loss": 0.0726,
+      "step": 2508
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.40624507683653166,
+      "learning_rate": 9.162127608331792e-06,
+      "loss": 0.0508,
+      "step": 2509
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.322478586487207,
+      "learning_rate": 9.161371275524469e-06,
+      "loss": 0.1069,
+      "step": 2510
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.45093396883388076,
+      "learning_rate": 9.160614632753971e-06,
+      "loss": 0.0892,
+      "step": 2511
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3815478254418809,
+      "learning_rate": 9.159857680076657e-06,
+      "loss": 0.0876,
+      "step": 2512
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6426436642489988,
+      "learning_rate": 9.159100417548911e-06,
+      "loss": 0.1224,
+      "step": 2513
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3331844016861563,
+      "learning_rate": 9.158342845227137e-06,
+      "loss": 0.0966,
+      "step": 2514
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.33857962471035274,
+      "learning_rate": 9.157584963167764e-06,
+      "loss": 0.108,
+      "step": 2515
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.29793596217942864,
+      "learning_rate": 9.156826771427243e-06,
+      "loss": 0.0993,
+      "step": 2516
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.37629183947643907,
+      "learning_rate": 9.156068270062048e-06,
+      "loss": 0.0911,
+      "step": 2517
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.445958521631865,
+      "learning_rate": 9.155309459128678e-06,
+      "loss": 0.107,
+      "step": 2518
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3023250317961179,
+      "learning_rate": 9.154550338683654e-06,
+      "loss": 0.0481,
+      "step": 2519
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.38060769590598115,
+      "learning_rate": 9.153790908783517e-06,
+      "loss": 0.0923,
+      "step": 2520
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5010774061190695,
+      "learning_rate": 9.153031169484838e-06,
+      "loss": 0.0957,
+      "step": 2521
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.45831047408042114,
+      "learning_rate": 9.152271120844203e-06,
+      "loss": 0.1121,
+      "step": 2522
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.23030855756275562,
+      "learning_rate": 9.151510762918225e-06,
+      "loss": 0.0397,
+      "step": 2523
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.24327685002587543,
+      "learning_rate": 9.150750095763543e-06,
+      "loss": 0.0632,
+      "step": 2524
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5162771619333781,
+      "learning_rate": 9.149989119436813e-06,
+      "loss": 0.16,
+      "step": 2525
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7383018806445086,
+      "learning_rate": 9.149227833994717e-06,
+      "loss": 0.0799,
+      "step": 2526
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4839474532854488,
+      "learning_rate": 9.148466239493962e-06,
+      "loss": 0.0994,
+      "step": 2527
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.22568671027585252,
+      "learning_rate": 9.147704335991275e-06,
+      "loss": 0.0487,
+      "step": 2528
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.35481442154136544,
+      "learning_rate": 9.146942123543406e-06,
+      "loss": 0.0866,
+      "step": 2529
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5514397097718858,
+      "learning_rate": 9.14617960220713e-06,
+      "loss": 0.1226,
+      "step": 2530
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5400627398502025,
+      "learning_rate": 9.145416772039245e-06,
+      "loss": 0.0827,
+      "step": 2531
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6267976470932751,
+      "learning_rate": 9.144653633096569e-06,
+      "loss": 0.1358,
+      "step": 2532
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.23745799704731144,
+      "learning_rate": 9.143890185435946e-06,
+      "loss": 0.0666,
+      "step": 2533
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.26523721134900063,
+      "learning_rate": 9.143126429114243e-06,
+      "loss": 0.0867,
+      "step": 2534
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4122795548462173,
+      "learning_rate": 9.142362364188348e-06,
+      "loss": 0.0945,
+      "step": 2535
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.48871316196481074,
+      "learning_rate": 9.141597990715172e-06,
+      "loss": 0.1086,
+      "step": 2536
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6270852697051894,
+      "learning_rate": 9.14083330875165e-06,
+      "loss": 0.1198,
+      "step": 2537
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.526201127494724,
+      "learning_rate": 9.140068318354745e-06,
+      "loss": 0.0831,
+      "step": 2538
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.1888757342131632,
+      "learning_rate": 9.139303019581432e-06,
+      "loss": 0.0625,
+      "step": 2539
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3612695813707224,
+      "learning_rate": 9.138537412488715e-06,
+      "loss": 0.0819,
+      "step": 2540
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.499989457781871,
+      "learning_rate": 9.137771497133625e-06,
+      "loss": 0.0911,
+      "step": 2541
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5989894805656822,
+      "learning_rate": 9.137005273573208e-06,
+      "loss": 0.1142,
+      "step": 2542
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.47768961548322536,
+      "learning_rate": 9.13623874186454e-06,
+      "loss": 0.1029,
+      "step": 2543
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5300784521691798,
+      "learning_rate": 9.135471902064715e-06,
+      "loss": 0.1319,
+      "step": 2544
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.2633714226731253,
+      "learning_rate": 9.13470475423085e-06,
+      "loss": 0.0633,
+      "step": 2545
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.24341585967279902,
+      "learning_rate": 9.13393729842009e-06,
+      "loss": 0.0708,
+      "step": 2546
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3549411553736421,
+      "learning_rate": 9.133169534689598e-06,
+      "loss": 0.0843,
+      "step": 2547
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.38410619968820586,
+      "learning_rate": 9.13240146309656e-06,
+      "loss": 0.0795,
+      "step": 2548
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5076940807071623,
+      "learning_rate": 9.13163308369819e-06,
+      "loss": 0.1438,
+      "step": 2549
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.40084210883167004,
+      "learning_rate": 9.13086439655172e-06,
+      "loss": 0.0806,
+      "step": 2550
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.526776219491348,
+      "learning_rate": 9.130095401714403e-06,
+      "loss": 0.1312,
+      "step": 2551
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.23052135332538107,
+      "learning_rate": 9.129326099243522e-06,
+      "loss": 0.0663,
+      "step": 2552
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.19425515962736847,
+      "learning_rate": 9.128556489196378e-06,
+      "loss": 0.0479,
+      "step": 2553
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.2823366831343273,
+      "learning_rate": 9.127786571630297e-06,
+      "loss": 0.0835,
+      "step": 2554
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.2608065392373284,
+      "learning_rate": 9.127016346602624e-06,
+      "loss": 0.0532,
+      "step": 2555
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.39571993060155386,
+      "learning_rate": 9.126245814170733e-06,
+      "loss": 0.1149,
+      "step": 2556
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9318515612608446,
+      "learning_rate": 9.125474974392018e-06,
+      "loss": 0.1649,
+      "step": 2557
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3081282224855193,
+      "learning_rate": 9.124703827323892e-06,
+      "loss": 0.0881,
+      "step": 2558
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5163842898771509,
+      "learning_rate": 9.123932373023798e-06,
+      "loss": 0.1211,
+      "step": 2559
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8761097371665139,
+      "learning_rate": 9.123160611549197e-06,
+      "loss": 0.1641,
+      "step": 2560
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.47144342685189805,
+      "learning_rate": 9.122388542957574e-06,
+      "loss": 0.0751,
+      "step": 2561
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3529385002467527,
+      "learning_rate": 9.12161616730644e-06,
+      "loss": 0.0974,
+      "step": 2562
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.20732876157411942,
+      "learning_rate": 9.120843484653321e-06,
+      "loss": 0.0543,
+      "step": 2563
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.41306360372722983,
+      "learning_rate": 9.120070495055774e-06,
+      "loss": 0.101,
+      "step": 2564
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7810116886043941,
+      "learning_rate": 9.119297198571375e-06,
+      "loss": 0.1893,
+      "step": 2565
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.257236972422567,
+      "learning_rate": 9.118523595257725e-06,
+      "loss": 0.084,
+      "step": 2566
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.437756239393404,
+      "learning_rate": 9.117749685172445e-06,
+      "loss": 0.1072,
+      "step": 2567
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.27137049235555033,
+      "learning_rate": 9.116975468373181e-06,
+      "loss": 0.0533,
+      "step": 2568
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8444980930223501,
+      "learning_rate": 9.116200944917602e-06,
+      "loss": 0.1487,
+      "step": 2569
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.328499656576064,
+      "learning_rate": 9.115426114863396e-06,
+      "loss": 0.1194,
+      "step": 2570
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.23009404115584572,
+      "learning_rate": 9.11465097826828e-06,
+      "loss": 0.0485,
+      "step": 2571
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.33661991461281293,
+      "learning_rate": 9.11387553518999e-06,
+      "loss": 0.0977,
+      "step": 2572
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.34755574168555936,
+      "learning_rate": 9.113099785686283e-06,
+      "loss": 0.0686,
+      "step": 2573
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5980102053817973,
+      "learning_rate": 9.112323729814945e-06,
+      "loss": 0.1274,
+      "step": 2574
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.2282413704088384,
+      "learning_rate": 9.11154736763378e-06,
+      "loss": 0.0529,
+      "step": 2575
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4160755343938698,
+      "learning_rate": 9.110770699200613e-06,
+      "loss": 0.1115,
+      "step": 2576
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.19579505862800656,
+      "learning_rate": 9.1099937245733e-06,
+      "loss": 0.0523,
+      "step": 2577
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.26222410254507084,
+      "learning_rate": 9.109216443809709e-06,
+      "loss": 0.065,
+      "step": 2578
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.37327179354221546,
+      "learning_rate": 9.108438856967742e-06,
+      "loss": 0.08,
+      "step": 2579
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6161828289850994,
+      "learning_rate": 9.107660964105314e-06,
+      "loss": 0.0905,
+      "step": 2580
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6065118804954795,
+      "learning_rate": 9.10688276528037e-06,
+      "loss": 0.1224,
+      "step": 2581
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.23746208395883547,
+      "learning_rate": 9.106104260550872e-06,
+      "loss": 0.0638,
+      "step": 2582
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.292860342825077,
+      "learning_rate": 9.105325449974809e-06,
+      "loss": 0.0532,
+      "step": 2583
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3676602802948076,
+      "learning_rate": 9.10454633361019e-06,
+      "loss": 0.0979,
+      "step": 2584
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.22188551180911154,
+      "learning_rate": 9.103766911515053e-06,
+      "loss": 0.0507,
+      "step": 2585
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4443835706268129,
+      "learning_rate": 9.102987183747449e-06,
+      "loss": 0.0876,
+      "step": 2586
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3464884058006399,
+      "learning_rate": 9.102207150365456e-06,
+      "loss": 0.0937,
+      "step": 2587
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.31448629595846794,
+      "learning_rate": 9.101426811427179e-06,
+      "loss": 0.067,
+      "step": 2588
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.43584081844549677,
+      "learning_rate": 9.10064616699074e-06,
+      "loss": 0.0942,
+      "step": 2589
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3930212409413606,
+      "learning_rate": 9.099865217114285e-06,
+      "loss": 0.0643,
+      "step": 2590
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.30530088815780404,
+      "learning_rate": 9.099083961855986e-06,
+      "loss": 0.1021,
+      "step": 2591
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5398788198959811,
+      "learning_rate": 9.098302401274035e-06,
+      "loss": 0.1436,
+      "step": 2592
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.20936585834393176,
+      "learning_rate": 9.097520535426646e-06,
+      "loss": 0.045,
+      "step": 2593
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.37911592986465925,
+      "learning_rate": 9.09673836437206e-06,
+      "loss": 0.0788,
+      "step": 2594
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.32963021373525403,
+      "learning_rate": 9.095955888168532e-06,
+      "loss": 0.0913,
+      "step": 2595
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5148699432817315,
+      "learning_rate": 9.095173106874352e-06,
+      "loss": 0.1048,
+      "step": 2596
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3907899794753764,
+      "learning_rate": 9.094390020547821e-06,
+      "loss": 0.0864,
+      "step": 2597
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.27910647194051497,
+      "learning_rate": 9.09360662924727e-06,
+      "loss": 0.0828,
+      "step": 2598
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.37978603053437976,
+      "learning_rate": 9.092822933031051e-06,
+      "loss": 0.0946,
+      "step": 2599
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3330646085488305,
+      "learning_rate": 9.092038931957537e-06,
+      "loss": 0.0999,
+      "step": 2600
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.25342448791173694,
+      "learning_rate": 9.091254626085124e-06,
+      "loss": 0.0645,
+      "step": 2601
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.36034819367439264,
+      "learning_rate": 9.090470015472234e-06,
+      "loss": 0.0904,
+      "step": 2602
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.35055986277243534,
+      "learning_rate": 9.08968510017731e-06,
+      "loss": 0.0925,
+      "step": 2603
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4599181961626947,
+      "learning_rate": 9.088899880258815e-06,
+      "loss": 0.0979,
+      "step": 2604
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.21793590238699348,
+      "learning_rate": 9.088114355775236e-06,
+      "loss": 0.0261,
+      "step": 2605
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.37647391968449523,
+      "learning_rate": 9.087328526785086e-06,
+      "loss": 0.1,
+      "step": 2606
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.34069634639924606,
+      "learning_rate": 9.086542393346895e-06,
+      "loss": 0.0861,
+      "step": 2607
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.37754994620913385,
+      "learning_rate": 9.085755955519222e-06,
+      "loss": 0.098,
+      "step": 2608
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4059475133425131,
+      "learning_rate": 9.084969213360644e-06,
+      "loss": 0.0658,
+      "step": 2609
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.30477118962372385,
+      "learning_rate": 9.084182166929764e-06,
+      "loss": 0.072,
+      "step": 2610
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5410236588885615,
+      "learning_rate": 9.083394816285203e-06,
+      "loss": 0.1093,
+      "step": 2611
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.2418257658008138,
+      "learning_rate": 9.082607161485607e-06,
+      "loss": 0.0736,
+      "step": 2612
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6436519610748231,
+      "learning_rate": 9.081819202589648e-06,
+      "loss": 0.122,
+      "step": 2613
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.25173956089787625,
+      "learning_rate": 9.081030939656017e-06,
+      "loss": 0.0556,
+      "step": 2614
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3106958839034273,
+      "learning_rate": 9.080242372743426e-06,
+      "loss": 0.0614,
+      "step": 2615
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.2914830713588264,
+      "learning_rate": 9.079453501910616e-06,
+      "loss": 0.0866,
+      "step": 2616
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5978257986526445,
+      "learning_rate": 9.078664327216344e-06,
+      "loss": 0.1264,
+      "step": 2617
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3629966725438927,
+      "learning_rate": 9.077874848719392e-06,
+      "loss": 0.1039,
+      "step": 2618
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3990503699568111,
+      "learning_rate": 9.077085066478567e-06,
+      "loss": 0.0985,
+      "step": 2619
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.32198765762253506,
+      "learning_rate": 9.076294980552694e-06,
+      "loss": 0.1053,
+      "step": 2620
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.43261813709377944,
+      "learning_rate": 9.075504591000626e-06,
+      "loss": 0.1194,
+      "step": 2621
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5412840291700423,
+      "learning_rate": 9.074713897881233e-06,
+      "loss": 0.1167,
+      "step": 2622
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3993745932464002,
+      "learning_rate": 9.073922901253413e-06,
+      "loss": 0.1104,
+      "step": 2623
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.2583907238984509,
+      "learning_rate": 9.073131601176084e-06,
+      "loss": 0.0766,
+      "step": 2624
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4320585829726278,
+      "learning_rate": 9.072339997708184e-06,
+      "loss": 0.1148,
+      "step": 2625
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3813778431770088,
+      "learning_rate": 9.07154809090868e-06,
+      "loss": 0.1141,
+      "step": 2626
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3473458305769543,
+      "learning_rate": 9.070755880836553e-06,
+      "loss": 0.0632,
+      "step": 2627
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5451070100529144,
+      "learning_rate": 9.069963367550815e-06,
+      "loss": 0.1182,
+      "step": 2628
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5404611314372313,
+      "learning_rate": 9.069170551110495e-06,
+      "loss": 0.1007,
+      "step": 2629
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.32558571550519905,
+      "learning_rate": 9.06837743157465e-06,
+      "loss": 0.0778,
+      "step": 2630
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.419590868998277,
+      "learning_rate": 9.067584009002353e-06,
+      "loss": 0.1082,
+      "step": 2631
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.23376487494108497,
+      "learning_rate": 9.066790283452702e-06,
+      "loss": 0.0621,
+      "step": 2632
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.39831459864013735,
+      "learning_rate": 9.065996254984822e-06,
+      "loss": 0.1086,
+      "step": 2633
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4948853966244402,
+      "learning_rate": 9.065201923657854e-06,
+      "loss": 0.1291,
+      "step": 2634
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.24317796996064392,
+      "learning_rate": 9.064407289530966e-06,
+      "loss": 0.0584,
+      "step": 2635
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.44600380841795656,
+      "learning_rate": 9.063612352663345e-06,
+      "loss": 0.0985,
+      "step": 2636
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.2316615180585202,
+      "learning_rate": 9.062817113114203e-06,
+      "loss": 0.0587,
+      "step": 2637
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4914966639986054,
+      "learning_rate": 9.062021570942776e-06,
+      "loss": 0.1163,
+      "step": 2638
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.2554579823720119,
+      "learning_rate": 9.06122572620832e-06,
+      "loss": 0.061,
+      "step": 2639
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.29633411946898025,
+      "learning_rate": 9.060429578970111e-06,
+      "loss": 0.0818,
+      "step": 2640
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3048364763696953,
+      "learning_rate": 9.059633129287456e-06,
+      "loss": 0.0904,
+      "step": 2641
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.280773975083365,
+      "learning_rate": 9.058836377219676e-06,
+      "loss": 0.0766,
+      "step": 2642
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.30711115012666435,
+      "learning_rate": 9.058039322826115e-06,
+      "loss": 0.0536,
+      "step": 2643
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.2757579844679646,
+      "learning_rate": 9.057241966166148e-06,
+      "loss": 0.0664,
+      "step": 2644
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.29503626421763296,
+      "learning_rate": 9.056444307299166e-06,
+      "loss": 0.0747,
+      "step": 2645
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4320680644659813,
+      "learning_rate": 9.05564634628458e-06,
+      "loss": 0.0937,
+      "step": 2646
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.252929903968619,
+      "learning_rate": 9.054848083181827e-06,
+      "loss": 0.0785,
+      "step": 2647
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.645218675775852,
+      "learning_rate": 9.054049518050367e-06,
+      "loss": 0.1047,
+      "step": 2648
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3629642421428972,
+      "learning_rate": 9.053250650949684e-06,
+      "loss": 0.0951,
+      "step": 2649
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3034759792994803,
+      "learning_rate": 9.05245148193928e-06,
+      "loss": 0.0848,
+      "step": 2650
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.33671694859370266,
+      "learning_rate": 9.051652011078681e-06,
+      "loss": 0.0816,
+      "step": 2651
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.23614811979000472,
+      "learning_rate": 9.05085223842744e-06,
+      "loss": 0.0517,
+      "step": 2652
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.37870939908090134,
+      "learning_rate": 9.050052164045125e-06,
+      "loss": 0.0741,
+      "step": 2653
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7553547903467922,
+      "learning_rate": 9.049251787991333e-06,
+      "loss": 0.0717,
+      "step": 2654
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.35198760650404876,
+      "learning_rate": 9.048451110325678e-06,
+      "loss": 0.0724,
+      "step": 2655
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4936877660182769,
+      "learning_rate": 9.0476501311078e-06,
+      "loss": 0.1379,
+      "step": 2656
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.44334452320078355,
+      "learning_rate": 9.046848850397361e-06,
+      "loss": 0.0914,
+      "step": 2657
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4532973311527628,
+      "learning_rate": 9.046047268254045e-06,
+      "loss": 0.074,
+      "step": 2658
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.42104868375019777,
+      "learning_rate": 9.04524538473756e-06,
+      "loss": 0.0851,
+      "step": 2659
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.21071206220743333,
+      "learning_rate": 9.044443199907633e-06,
+      "loss": 0.0621,
+      "step": 2660
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.28122218849144776,
+      "learning_rate": 9.043640713824013e-06,
+      "loss": 0.0655,
+      "step": 2661
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.39990627773414417,
+      "learning_rate": 9.04283792654648e-06,
+      "loss": 0.1149,
+      "step": 2662
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4726453379847842,
+      "learning_rate": 9.042034838134826e-06,
+      "loss": 0.1078,
+      "step": 2663
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.509778545882967,
+      "learning_rate": 9.04123144864887e-06,
+      "loss": 0.1044,
+      "step": 2664
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4869428672743166,
+      "learning_rate": 9.040427758148455e-06,
+      "loss": 0.0908,
+      "step": 2665
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.2764333055082348,
+      "learning_rate": 9.039623766693441e-06,
+      "loss": 0.0649,
+      "step": 2666
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3811368039556193,
+      "learning_rate": 9.038819474343718e-06,
+      "loss": 0.124,
+      "step": 2667
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5249246735671909,
+      "learning_rate": 9.038014881159194e-06,
+      "loss": 0.1068,
+      "step": 2668
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3905463546652125,
+      "learning_rate": 9.037209987199797e-06,
+      "loss": 0.0873,
+      "step": 2669
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1040854942502358,
+      "learning_rate": 9.036404792525482e-06,
+      "loss": 0.1815,
+      "step": 2670
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.32323871665779386,
+      "learning_rate": 9.035599297196222e-06,
+      "loss": 0.1055,
+      "step": 2671
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.262962789347326,
+      "learning_rate": 9.03479350127202e-06,
+      "loss": 0.0772,
+      "step": 2672
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4488840405858109,
+      "learning_rate": 9.033987404812893e-06,
+      "loss": 0.12,
+      "step": 2673
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.45378582246556004,
+      "learning_rate": 9.033181007878884e-06,
+      "loss": 0.0798,
+      "step": 2674
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5040432210327149,
+      "learning_rate": 9.032374310530058e-06,
+      "loss": 0.1244,
+      "step": 2675
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.2776149365654268,
+      "learning_rate": 9.031567312826504e-06,
+      "loss": 0.064,
+      "step": 2676
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4519826003045925,
+      "learning_rate": 9.030760014828332e-06,
+      "loss": 0.1114,
+      "step": 2677
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5554392327687911,
+      "learning_rate": 9.029952416595671e-06,
+      "loss": 0.1312,
+      "step": 2678
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5337728741387994,
+      "learning_rate": 9.029144518188679e-06,
+      "loss": 0.1304,
+      "step": 2679
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.30854457818498365,
+      "learning_rate": 9.028336319667532e-06,
+      "loss": 0.093,
+      "step": 2680
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4355621896484041,
+      "learning_rate": 9.027527821092428e-06,
+      "loss": 0.0791,
+      "step": 2681
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3658283551907798,
+      "learning_rate": 9.026719022523592e-06,
+      "loss": 0.0874,
+      "step": 2682
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.30816237347574826,
+      "learning_rate": 9.025909924021263e-06,
+      "loss": 0.051,
+      "step": 2683
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.2644213688472279,
+      "learning_rate": 9.025100525645712e-06,
+      "loss": 0.0661,
+      "step": 2684
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.33253195116671613,
+      "learning_rate": 9.024290827457225e-06,
+      "loss": 0.072,
+      "step": 2685
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.35030347374023424,
+      "learning_rate": 9.023480829516114e-06,
+      "loss": 0.1091,
+      "step": 2686
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3215639519320042,
+      "learning_rate": 9.022670531882712e-06,
+      "loss": 0.0816,
+      "step": 2687
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.597005491978897,
+      "learning_rate": 9.021859934617376e-06,
+      "loss": 0.0999,
+      "step": 2688
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.412522460686854,
+      "learning_rate": 9.021049037780483e-06,
+      "loss": 0.0651,
+      "step": 2689
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.2967607464327263,
+      "learning_rate": 9.020237841432431e-06,
+      "loss": 0.0745,
+      "step": 2690
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3353034253137983,
+      "learning_rate": 9.019426345633649e-06,
+      "loss": 0.0797,
+      "step": 2691
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.45998628652792406,
+      "learning_rate": 9.018614550444575e-06,
+      "loss": 0.1235,
+      "step": 2692
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.30943624871926556,
+      "learning_rate": 9.01780245592568e-06,
+      "loss": 0.0709,
+      "step": 2693
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.22998853597353697,
+      "learning_rate": 9.016990062137452e-06,
+      "loss": 0.0429,
+      "step": 2694
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5391389714112024,
+      "learning_rate": 9.016177369140405e-06,
+      "loss": 0.1276,
+      "step": 2695
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0545453687355895,
+      "learning_rate": 9.015364376995071e-06,
+      "loss": 0.1189,
+      "step": 2696
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5624534579982434,
+      "learning_rate": 9.014551085762006e-06,
+      "loss": 0.1066,
+      "step": 2697
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.43274477449888105,
+      "learning_rate": 9.013737495501788e-06,
+      "loss": 0.1229,
+      "step": 2698
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3110831120959545,
+      "learning_rate": 9.012923606275024e-06,
+      "loss": 0.1031,
+      "step": 2699
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.37218239130568853,
+      "learning_rate": 9.01210941814233e-06,
+      "loss": 0.0663,
+      "step": 2700
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4026313358909027,
+      "learning_rate": 9.011294931164354e-06,
+      "loss": 0.1389,
+      "step": 2701
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4559221185031558,
+      "learning_rate": 9.010480145401765e-06,
+      "loss": 0.1092,
+      "step": 2702
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5446599409991894,
+      "learning_rate": 9.009665060915252e-06,
+      "loss": 0.1224,
+      "step": 2703
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4158705364810137,
+      "learning_rate": 9.008849677765527e-06,
+      "loss": 0.076,
+      "step": 2704
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4381717523803752,
+      "learning_rate": 9.008033996013326e-06,
+      "loss": 0.0788,
+      "step": 2705
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.31993527647400266,
+      "learning_rate": 9.007218015719402e-06,
+      "loss": 0.0653,
+      "step": 2706
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3174546296094006,
+      "learning_rate": 9.006401736944538e-06,
+      "loss": 0.082,
+      "step": 2707
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.28381480585491237,
+      "learning_rate": 9.005585159749533e-06,
+      "loss": 0.0984,
+      "step": 2708
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5372730783968656,
+      "learning_rate": 9.004768284195212e-06,
+      "loss": 0.1011,
+      "step": 2709
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3294326952667977,
+      "learning_rate": 9.003951110342418e-06,
+      "loss": 0.0906,
+      "step": 2710
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.21896691538254417,
+      "learning_rate": 9.003133638252022e-06,
+      "loss": 0.059,
+      "step": 2711
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3326999807001258,
+      "learning_rate": 9.002315867984912e-06,
+      "loss": 0.0787,
+      "step": 2712
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.40439843280643584,
+      "learning_rate": 9.001497799602001e-06,
+      "loss": 0.0574,
+      "step": 2713
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4033431963451583,
+      "learning_rate": 9.000679433164224e-06,
+      "loss": 0.0951,
+      "step": 2714
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3559040485871925,
+      "learning_rate": 8.999860768732536e-06,
+      "loss": 0.0709,
+      "step": 2715
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3061154405179683,
+      "learning_rate": 8.999041806367919e-06,
+      "loss": 0.0756,
+      "step": 2716
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4525134213262327,
+      "learning_rate": 8.99822254613137e-06,
+      "loss": 0.104,
+      "step": 2717
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.30507556704077016,
+      "learning_rate": 8.997402988083916e-06,
+      "loss": 0.0988,
+      "step": 2718
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.345364014966951,
+      "learning_rate": 8.9965831322866e-06,
+      "loss": 0.0944,
+      "step": 2719
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5671088035707671,
+      "learning_rate": 8.995762978800492e-06,
+      "loss": 0.1199,
+      "step": 2720
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.27428010282051657,
+      "learning_rate": 8.994942527686677e-06,
+      "loss": 0.0731,
+      "step": 2721
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3704596620246462,
+      "learning_rate": 8.994121779006273e-06,
+      "loss": 0.0785,
+      "step": 2722
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.33800915646533625,
+      "learning_rate": 8.99330073282041e-06,
+      "loss": 0.0955,
+      "step": 2723
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4920755607101958,
+      "learning_rate": 8.992479389190247e-06,
+      "loss": 0.095,
+      "step": 2724
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.25387318046778673,
+      "learning_rate": 8.99165774817696e-06,
+      "loss": 0.0895,
+      "step": 2725
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3595736401784846,
+      "learning_rate": 8.990835809841751e-06,
+      "loss": 0.1049,
+      "step": 2726
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7542700445214279,
+      "learning_rate": 8.990013574245843e-06,
+      "loss": 0.1247,
+      "step": 2727
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3395562129942145,
+      "learning_rate": 8.98919104145048e-06,
+      "loss": 0.0727,
+      "step": 2728
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3963983887916889,
+      "learning_rate": 8.98836821151693e-06,
+      "loss": 0.1082,
+      "step": 2729
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.36094206402572404,
+      "learning_rate": 8.987545084506481e-06,
+      "loss": 0.0759,
+      "step": 2730
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.2660822618789366,
+      "learning_rate": 8.986721660480446e-06,
+      "loss": 0.0999,
+      "step": 2731
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.23723519517975003,
+      "learning_rate": 8.985897939500156e-06,
+      "loss": 0.0576,
+      "step": 2732
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3518410306265884,
+      "learning_rate": 8.98507392162697e-06,
+      "loss": 0.0905,
+      "step": 2733
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5090933764933581,
+      "learning_rate": 8.984249606922261e-06,
+      "loss": 0.1297,
+      "step": 2734
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.47058548146874457,
+      "learning_rate": 8.983424995447433e-06,
+      "loss": 0.0966,
+      "step": 2735
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.37973954061021137,
+      "learning_rate": 8.982600087263907e-06,
+      "loss": 0.1368,
+      "step": 2736
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.2521395228296642,
+      "learning_rate": 8.981774882433124e-06,
+      "loss": 0.0788,
+      "step": 2737
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3105013537155436,
+      "learning_rate": 8.980949381016553e-06,
+      "loss": 0.1077,
+      "step": 2738
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.33258028085388047,
+      "learning_rate": 8.98012358307568e-06,
+      "loss": 0.0673,
+      "step": 2739
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.30178482413869234,
+      "learning_rate": 8.979297488672019e-06,
+      "loss": 0.0741,
+      "step": 2740
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.571455642226894,
+      "learning_rate": 8.9784710978671e-06,
+      "loss": 0.1515,
+      "step": 2741
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.47980663204494933,
+      "learning_rate": 8.977644410722475e-06,
+      "loss": 0.0855,
+      "step": 2742
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.19803699107565897,
+      "learning_rate": 8.976817427299724e-06,
+      "loss": 0.0723,
+      "step": 2743
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7116909736110781,
+      "learning_rate": 8.975990147660444e-06,
+      "loss": 0.1062,
+      "step": 2744
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3031909469027027,
+      "learning_rate": 8.975162571866255e-06,
+      "loss": 0.0826,
+      "step": 2745
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.42399983097062166,
+      "learning_rate": 8.974334699978803e-06,
+      "loss": 0.0946,
+      "step": 2746
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6491029982640961,
+      "learning_rate": 8.973506532059749e-06,
+      "loss": 0.0727,
+      "step": 2747
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.2149978682935139,
+      "learning_rate": 8.972678068170781e-06,
+      "loss": 0.0631,
+      "step": 2748
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.450109538037787,
+      "learning_rate": 8.971849308373608e-06,
+      "loss": 0.1045,
+      "step": 2749
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.331023447233103,
+      "learning_rate": 8.971020252729961e-06,
+      "loss": 0.1064,
+      "step": 2750
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.20582904841185387,
+      "learning_rate": 8.970190901301595e-06,
+      "loss": 0.0605,
+      "step": 2751
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.24943963396730862,
+      "learning_rate": 8.96936125415028e-06,
+      "loss": 0.0661,
+      "step": 2752
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3130314373925038,
+      "learning_rate": 8.968531311337814e-06,
+      "loss": 0.0926,
+      "step": 2753
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.39420445098373563,
+      "learning_rate": 8.96770107292602e-06,
+      "loss": 0.0986,
+      "step": 2754
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4103619106541245,
+      "learning_rate": 8.966870538976736e-06,
+      "loss": 0.1,
+      "step": 2755
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3719851568235137,
+      "learning_rate": 8.966039709551826e-06,
+      "loss": 0.0982,
+      "step": 2756
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.2953239385620919,
+      "learning_rate": 8.965208584713173e-06,
+      "loss": 0.0757,
+      "step": 2757
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3406903095112857,
+      "learning_rate": 8.964377164522686e-06,
+      "loss": 0.0795,
+      "step": 2758
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3338676647667838,
+      "learning_rate": 8.963545449042294e-06,
+      "loss": 0.0789,
+      "step": 2759
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.30406041117699845,
+      "learning_rate": 8.962713438333948e-06,
+      "loss": 0.0602,
+      "step": 2760
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.42905916956599444,
+      "learning_rate": 8.96188113245962e-06,
+      "loss": 0.109,
+      "step": 2761
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7120739398509142,
+      "learning_rate": 8.961048531481306e-06,
+      "loss": 0.124,
+      "step": 2762
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.2887606799930288,
+      "learning_rate": 8.960215635461023e-06,
+      "loss": 0.0941,
+      "step": 2763
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.2647619570069839,
+      "learning_rate": 8.959382444460808e-06,
+      "loss": 0.05,
+      "step": 2764
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.2436400523389256,
+      "learning_rate": 8.958548958542726e-06,
+      "loss": 0.042,
+      "step": 2765
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.34521278704333364,
+      "learning_rate": 8.957715177768856e-06,
+      "loss": 0.0946,
+      "step": 2766
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9392575631150798,
+      "learning_rate": 8.956881102201303e-06,
+      "loss": 0.1483,
+      "step": 2767
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.40481432764914393,
+      "learning_rate": 8.956046731902196e-06,
+      "loss": 0.1014,
+      "step": 2768
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.2985600816022739,
+      "learning_rate": 8.955212066933683e-06,
+      "loss": 0.0796,
+      "step": 2769
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.24827947281851215,
+      "learning_rate": 8.954377107357935e-06,
+      "loss": 0.0674,
+      "step": 2770
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6783140624660977,
+      "learning_rate": 8.953541853237142e-06,
+      "loss": 0.1095,
+      "step": 2771
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.41651437942943265,
+      "learning_rate": 8.952706304633522e-06,
+      "loss": 0.109,
+      "step": 2772
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4421396954934899,
+      "learning_rate": 8.951870461609308e-06,
+      "loss": 0.1372,
+      "step": 2773
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.21061328471940802,
+      "learning_rate": 8.951034324226764e-06,
+      "loss": 0.0542,
+      "step": 2774
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4595805091546543,
+      "learning_rate": 8.950197892548166e-06,
+      "loss": 0.1173,
+      "step": 2775
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3471857963009139,
+      "learning_rate": 8.949361166635817e-06,
+      "loss": 0.0595,
+      "step": 2776
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.2798436353232557,
+      "learning_rate": 8.948524146552043e-06,
+      "loss": 0.0673,
+      "step": 2777
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.24665762314253364,
+      "learning_rate": 8.947686832359186e-06,
+      "loss": 0.0853,
+      "step": 2778
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.30843007646134885,
+      "learning_rate": 8.946849224119618e-06,
+      "loss": 0.0879,
+      "step": 2779
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4082239758110478,
+      "learning_rate": 8.946011321895727e-06,
+      "loss": 0.1236,
+      "step": 2780
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3782139103338437,
+      "learning_rate": 8.945173125749927e-06,
+      "loss": 0.1002,
+      "step": 2781
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4061858643167637,
+      "learning_rate": 8.94433463574465e-06,
+      "loss": 0.107,
+      "step": 2782
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.23856184381816875,
+      "learning_rate": 8.943495851942352e-06,
+      "loss": 0.0799,
+      "step": 2783
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.42964048314409087,
+      "learning_rate": 8.94265677440551e-06,
+      "loss": 0.0805,
+      "step": 2784
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3520249798638243,
+      "learning_rate": 8.941817403196624e-06,
+      "loss": 0.0751,
+      "step": 2785
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5090531036333168,
+      "learning_rate": 8.940977738378214e-06,
+      "loss": 0.118,
+      "step": 2786
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5029884313094897,
+      "learning_rate": 8.940137780012825e-06,
+      "loss": 0.1337,
+      "step": 2787
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.2897044651718947,
+      "learning_rate": 8.939297528163021e-06,
+      "loss": 0.0843,
+      "step": 2788
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.49933053200983224,
+      "learning_rate": 8.93845698289139e-06,
+      "loss": 0.114,
+      "step": 2789
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.2294182185806812,
+      "learning_rate": 8.937616144260537e-06,
+      "loss": 0.0725,
+      "step": 2790
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4258161410780781,
+      "learning_rate": 8.936775012333097e-06,
+      "loss": 0.1214,
+      "step": 2791
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.32472605810986227,
+      "learning_rate": 8.935933587171722e-06,
+      "loss": 0.115,
+      "step": 2792
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.47211953434842624,
+      "learning_rate": 8.935091868839083e-06,
+      "loss": 0.1195,
+      "step": 2793
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.255466045805625,
+      "learning_rate": 8.934249857397878e-06,
+      "loss": 0.0702,
+      "step": 2794
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3103380620273569,
+      "learning_rate": 8.933407552910825e-06,
+      "loss": 0.0755,
+      "step": 2795
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3137532550394677,
+      "learning_rate": 8.932564955440665e-06,
+      "loss": 0.0908,
+      "step": 2796
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.23521704931762324,
+      "learning_rate": 8.931722065050156e-06,
+      "loss": 0.0845,
+      "step": 2797
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.25882444005046445,
+      "learning_rate": 8.930878881802085e-06,
+      "loss": 0.0537,
+      "step": 2798
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.39272613559378483,
+      "learning_rate": 8.930035405759256e-06,
+      "loss": 0.1022,
+      "step": 2799
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.40178841740603577,
+      "learning_rate": 8.929191636984497e-06,
+      "loss": 0.0822,
+      "step": 2800
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.30141877929050653,
+      "learning_rate": 8.928347575540655e-06,
+      "loss": 0.0596,
+      "step": 2801
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3862418193709303,
+      "learning_rate": 8.927503221490598e-06,
+      "loss": 0.096,
+      "step": 2802
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3681834253637028,
+      "learning_rate": 8.926658574897225e-06,
+      "loss": 0.1129,
+      "step": 2803
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3268792490591226,
+      "learning_rate": 8.925813635823446e-06,
+      "loss": 0.0669,
+      "step": 2804
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.23037313144027546,
+      "learning_rate": 8.9249684043322e-06,
+      "loss": 0.0573,
+      "step": 2805
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.26532749631270036,
+      "learning_rate": 8.924122880486439e-06,
+      "loss": 0.0677,
+      "step": 2806
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.31522562941668775,
+      "learning_rate": 8.923277064349147e-06,
+      "loss": 0.0602,
+      "step": 2807
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5229278502205444,
+      "learning_rate": 8.922430955983326e-06,
+      "loss": 0.1347,
+      "step": 2808
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.37165257221447645,
+      "learning_rate": 8.921584555451997e-06,
+      "loss": 0.1056,
+      "step": 2809
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4678765455875168,
+      "learning_rate": 8.920737862818205e-06,
+      "loss": 0.0904,
+      "step": 2810
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3798818559732239,
+      "learning_rate": 8.919890878145016e-06,
+      "loss": 0.1146,
+      "step": 2811
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6224690370047794,
+      "learning_rate": 8.91904360149552e-06,
+      "loss": 0.1181,
+      "step": 2812
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4084461172779903,
+      "learning_rate": 8.918196032932827e-06,
+      "loss": 0.094,
+      "step": 2813
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.41239747915839853,
+      "learning_rate": 8.917348172520069e-06,
+      "loss": 0.1097,
+      "step": 2814
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.36825633041050204,
+      "learning_rate": 8.916500020320398e-06,
+      "loss": 0.0732,
+      "step": 2815
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5237830106558871,
+      "learning_rate": 8.91565157639699e-06,
+      "loss": 0.1306,
+      "step": 2816
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.350257487694403,
+      "learning_rate": 8.914802840813044e-06,
+      "loss": 0.0916,
+      "step": 2817
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.27900696208553033,
+      "learning_rate": 8.913953813631775e-06,
+      "loss": 0.0759,
+      "step": 2818
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3686581800894469,
+      "learning_rate": 8.913104494916425e-06,
+      "loss": 0.0577,
+      "step": 2819
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.618928945166433,
+      "learning_rate": 8.912254884730259e-06,
+      "loss": 0.1046,
+      "step": 2820
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.23629178657471436,
+      "learning_rate": 8.911404983136556e-06,
+      "loss": 0.0679,
+      "step": 2821
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.27315518329800387,
+      "learning_rate": 8.910554790198627e-06,
+      "loss": 0.0754,
+      "step": 2822
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5289553100616763,
+      "learning_rate": 8.909704305979795e-06,
+      "loss": 0.1315,
+      "step": 2823
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3553910233472979,
+      "learning_rate": 8.908853530543408e-06,
+      "loss": 0.11,
+      "step": 2824
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3311007901480106,
+      "learning_rate": 8.908002463952843e-06,
+      "loss": 0.0824,
+      "step": 2825
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.32371292554120545,
+      "learning_rate": 8.907151106271487e-06,
+      "loss": 0.0849,
+      "step": 2826
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5223889497269055,
+      "learning_rate": 8.906299457562757e-06,
+      "loss": 0.1099,
+      "step": 2827
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.23312336567390182,
+      "learning_rate": 8.905447517890087e-06,
+      "loss": 0.0428,
+      "step": 2828
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.37546354739097065,
+      "learning_rate": 8.904595287316936e-06,
+      "loss": 0.1063,
+      "step": 2829
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.37472073223353686,
+      "learning_rate": 8.90374276590678e-06,
+      "loss": 0.1065,
+      "step": 2830
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6325258861975911,
+      "learning_rate": 8.902889953723122e-06,
+      "loss": 0.143,
+      "step": 2831
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7134435322607835,
+      "learning_rate": 8.902036850829485e-06,
+      "loss": 0.1385,
+      "step": 2832
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.35354017725524867,
+      "learning_rate": 8.901183457289413e-06,
+      "loss": 0.0886,
+      "step": 2833
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5151913959789703,
+      "learning_rate": 8.900329773166471e-06,
+      "loss": 0.13,
+      "step": 2834
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.14351625360780815,
+      "learning_rate": 8.899475798524245e-06,
+      "loss": 0.0324,
+      "step": 2835
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.22916388682678246,
+      "learning_rate": 8.898621533426346e-06,
+      "loss": 0.0331,
+      "step": 2836
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4312486039375231,
+      "learning_rate": 8.897766977936404e-06,
+      "loss": 0.1088,
+      "step": 2837
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5615662450116726,
+      "learning_rate": 8.896912132118072e-06,
+      "loss": 0.1114,
+      "step": 2838
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.39168859423279895,
+      "learning_rate": 8.896056996035024e-06,
+      "loss": 0.1012,
+      "step": 2839
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4427772288729868,
+      "learning_rate": 8.895201569750952e-06,
+      "loss": 0.1285,
+      "step": 2840
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5172872940523813,
+      "learning_rate": 8.894345853329577e-06,
+      "loss": 0.066,
+      "step": 2841
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4320043699356264,
+      "learning_rate": 8.893489846834637e-06,
+      "loss": 0.1408,
+      "step": 2842
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.21826385738217519,
+      "learning_rate": 8.892633550329892e-06,
+      "loss": 0.0451,
+      "step": 2843
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.30536975239453507,
+      "learning_rate": 8.891776963879124e-06,
+      "loss": 0.0987,
+      "step": 2844
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.41305233935395874,
+      "learning_rate": 8.890920087546135e-06,
+      "loss": 0.1219,
+      "step": 2845
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.29496752962675066,
+      "learning_rate": 8.890062921394753e-06,
+      "loss": 0.0659,
+      "step": 2846
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.27501347898561496,
+      "learning_rate": 8.889205465488823e-06,
+      "loss": 0.0986,
+      "step": 2847
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.33111899225956937,
+      "learning_rate": 8.888347719892214e-06,
+      "loss": 0.0849,
+      "step": 2848
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.27545672599575527,
+      "learning_rate": 8.887489684668816e-06,
+      "loss": 0.0538,
+      "step": 2849
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.25472040155819675,
+      "learning_rate": 8.886631359882538e-06,
+      "loss": 0.0735,
+      "step": 2850
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.37903050561502716,
+      "learning_rate": 8.885772745597317e-06,
+      "loss": 0.0884,
+      "step": 2851
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5115405883212079,
+      "learning_rate": 8.884913841877106e-06,
+      "loss": 0.0863,
+      "step": 2852
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.40063681772782056,
+      "learning_rate": 8.884054648785878e-06,
+      "loss": 0.0942,
+      "step": 2853
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.41657987661935486,
+      "learning_rate": 8.883195166387635e-06,
+      "loss": 0.0624,
+      "step": 2854
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.43316848546427505,
+      "learning_rate": 8.882335394746395e-06,
+      "loss": 0.1244,
+      "step": 2855
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3297483499339481,
+      "learning_rate": 8.881475333926198e-06,
+      "loss": 0.1023,
+      "step": 2856
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.279182412492037,
+      "learning_rate": 8.880614983991106e-06,
+      "loss": 0.0763,
+      "step": 2857
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3678743158692316,
+      "learning_rate": 8.879754345005207e-06,
+      "loss": 0.1089,
+      "step": 2858
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3975343197899079,
+      "learning_rate": 8.8788934170326e-06,
+      "loss": 0.1185,
+      "step": 2859
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3612735218594484,
+      "learning_rate": 8.878032200137417e-06,
+      "loss": 0.0608,
+      "step": 2860
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.36003023521498584,
+      "learning_rate": 8.877170694383803e-06,
+      "loss": 0.0891,
+      "step": 2861
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4427646180924044,
+      "learning_rate": 8.876308899835931e-06,
+      "loss": 0.1154,
+      "step": 2862
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.37212812679235696,
+      "learning_rate": 8.875446816557988e-06,
+      "loss": 0.0689,
+      "step": 2863
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.23265850182196846,
+      "learning_rate": 8.874584444614193e-06,
+      "loss": 0.0649,
+      "step": 2864
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.26637280780579425,
+      "learning_rate": 8.873721784068775e-06,
+      "loss": 0.0493,
+      "step": 2865
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.2719625337348453,
+      "learning_rate": 8.872858834985995e-06,
+      "loss": 0.073,
+      "step": 2866
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3883071654519712,
+      "learning_rate": 8.871995597430128e-06,
+      "loss": 0.0679,
+      "step": 2867
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.273445482848648,
+      "learning_rate": 8.871132071465472e-06,
+      "loss": 0.0611,
+      "step": 2868
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.14251544165414695,
+      "learning_rate": 8.870268257156346e-06,
+      "loss": 0.0356,
+      "step": 2869
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.36282991457711605,
+      "learning_rate": 8.869404154567097e-06,
+      "loss": 0.0564,
+      "step": 2870
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4724302640022304,
+      "learning_rate": 8.868539763762085e-06,
+      "loss": 0.095,
+      "step": 2871
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.40559956579385575,
+      "learning_rate": 8.867675084805696e-06,
+      "loss": 0.1012,
+      "step": 2872
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.42567031255506615,
+      "learning_rate": 8.866810117762335e-06,
+      "loss": 0.1294,
+      "step": 2873
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.30781394833221576,
+      "learning_rate": 8.865944862696431e-06,
+      "loss": 0.0545,
+      "step": 2874
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.31404349108611535,
+      "learning_rate": 8.865079319672432e-06,
+      "loss": 0.0988,
+      "step": 2875
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3086156659255089,
+      "learning_rate": 8.864213488754808e-06,
+      "loss": 0.0929,
+      "step": 2876
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.29320734567447476,
+      "learning_rate": 8.863347370008058e-06,
+      "loss": 0.0864,
+      "step": 2877
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3587082246279781,
+      "learning_rate": 8.862480963496687e-06,
+      "loss": 0.064,
+      "step": 2878
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.39194698361953034,
+      "learning_rate": 8.861614269285233e-06,
+      "loss": 0.0586,
+      "step": 2879
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.38667544477744226,
+      "learning_rate": 8.860747287438252e-06,
+      "loss": 0.1191,
+      "step": 2880
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.2874327132378102,
+      "learning_rate": 8.859880018020323e-06,
+      "loss": 0.0775,
+      "step": 2881
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.34855671160765467,
+      "learning_rate": 8.859012461096045e-06,
+      "loss": 0.0966,
+      "step": 2882
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4004832468044629,
+      "learning_rate": 8.858144616730038e-06,
+      "loss": 0.0995,
+      "step": 2883
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.29411404420807263,
+      "learning_rate": 8.857276484986946e-06,
+      "loss": 0.1008,
+      "step": 2884
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6280329599597151,
+      "learning_rate": 8.856408065931429e-06,
+      "loss": 0.1251,
+      "step": 2885
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.25545803823799185,
+      "learning_rate": 8.855539359628177e-06,
+      "loss": 0.0622,
+      "step": 2886
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3207803412588621,
+      "learning_rate": 8.85467036614189e-06,
+      "loss": 0.0759,
+      "step": 2887
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3119671468797237,
+      "learning_rate": 8.853801085537303e-06,
+      "loss": 0.0648,
+      "step": 2888
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5293291658847464,
+      "learning_rate": 8.852931517879157e-06,
+      "loss": 0.1208,
+      "step": 2889
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3874955631323083,
+      "learning_rate": 8.85206166323223e-06,
+      "loss": 0.11,
+      "step": 2890
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.2395488291426755,
+      "learning_rate": 8.851191521661308e-06,
+      "loss": 0.0811,
+      "step": 2891
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4209398389180246,
+      "learning_rate": 8.850321093231209e-06,
+      "loss": 0.0567,
+      "step": 2892
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.2421634663167766,
+      "learning_rate": 8.849450378006763e-06,
+      "loss": 0.0745,
+      "step": 2893
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.2208218228211407,
+      "learning_rate": 8.848579376052828e-06,
+      "loss": 0.0574,
+      "step": 2894
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5694713614156356,
+      "learning_rate": 8.847708087434282e-06,
+      "loss": 0.095,
+      "step": 2895
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3097219552399733,
+      "learning_rate": 8.846836512216022e-06,
+      "loss": 0.0514,
+      "step": 2896
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.21961186232410493,
+      "learning_rate": 8.845964650462971e-06,
+      "loss": 0.0611,
+      "step": 2897
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4584144280189024,
+      "learning_rate": 8.845092502240067e-06,
+      "loss": 0.1088,
+      "step": 2898
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4130477547080702,
+      "learning_rate": 8.844220067612276e-06,
+      "loss": 0.1015,
+      "step": 2899
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.2802338282277875,
+      "learning_rate": 8.843347346644579e-06,
+      "loss": 0.0721,
+      "step": 2900
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3561563520956477,
+      "learning_rate": 8.842474339401983e-06,
+      "loss": 0.0906,
+      "step": 2901
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.2778591764946319,
+      "learning_rate": 8.841601045949513e-06,
+      "loss": 0.0749,
+      "step": 2902
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5255139502632714,
+      "learning_rate": 8.84072746635222e-06,
+      "loss": 0.1101,
+      "step": 2903
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5799560394608094,
+      "learning_rate": 8.83985360067517e-06,
+      "loss": 0.1276,
+      "step": 2904
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.2367330495962254,
+      "learning_rate": 8.838979448983453e-06,
+      "loss": 0.0405,
+      "step": 2905
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5101877524234021,
+      "learning_rate": 8.838105011342186e-06,
+      "loss": 0.1049,
+      "step": 2906
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.24916647818240295,
+      "learning_rate": 8.837230287816498e-06,
+      "loss": 0.0685,
+      "step": 2907
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.24931999935395047,
+      "learning_rate": 8.836355278471545e-06,
+      "loss": 0.0685,
+      "step": 2908
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3354044661605128,
+      "learning_rate": 8.835479983372502e-06,
+      "loss": 0.0881,
+      "step": 2909
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3390664443223106,
+      "learning_rate": 8.834604402584566e-06,
+      "loss": 0.0814,
+      "step": 2910
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3003475997487348,
+      "learning_rate": 8.833728536172958e-06,
+      "loss": 0.0739,
+      "step": 2911
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3181003839413497,
+      "learning_rate": 8.832852384202915e-06,
+      "loss": 0.0978,
+      "step": 2912
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.19012958822881082,
+      "learning_rate": 8.831975946739697e-06,
+      "loss": 0.0466,
+      "step": 2913
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3871243806789563,
+      "learning_rate": 8.831099223848587e-06,
+      "loss": 0.0927,
+      "step": 2914
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.47936710351009654,
+      "learning_rate": 8.83022221559489e-06,
+      "loss": 0.15,
+      "step": 2915
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4124137259689811,
+      "learning_rate": 8.82934492204393e-06,
+      "loss": 0.1078,
+      "step": 2916
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.25830085286326315,
+      "learning_rate": 8.828467343261054e-06,
+      "loss": 0.0748,
+      "step": 2917
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.33899724041855966,
+      "learning_rate": 8.827589479311626e-06,
+      "loss": 0.0551,
+      "step": 2918
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6024074170621591,
+      "learning_rate": 8.826711330261036e-06,
+      "loss": 0.1035,
+      "step": 2919
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.2353928638402001,
+      "learning_rate": 8.825832896174696e-06,
+      "loss": 0.0552,
+      "step": 2920
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.429466880403828,
+      "learning_rate": 8.824954177118032e-06,
+      "loss": 0.1235,
+      "step": 2921
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4304723017926431,
+      "learning_rate": 8.8240751731565e-06,
+      "loss": 0.0806,
+      "step": 2922
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4218768100065337,
+      "learning_rate": 8.823195884355572e-06,
+      "loss": 0.1126,
+      "step": 2923
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.2872072748294613,
+      "learning_rate": 8.822316310780743e-06,
+      "loss": 0.0794,
+      "step": 2924
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.29955523937796386,
+      "learning_rate": 8.82143645249753e-06,
+      "loss": 0.0785,
+      "step": 2925
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3581329494392368,
+      "learning_rate": 8.820556309571467e-06,
+      "loss": 0.0873,
+      "step": 2926
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4154836976986268,
+      "learning_rate": 8.819675882068113e-06,
+      "loss": 0.0443,
+      "step": 2927
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.2929112583999319,
+      "learning_rate": 8.818795170053049e-06,
+      "loss": 0.0679,
+      "step": 2928
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.2283100763348576,
+      "learning_rate": 8.817914173591873e-06,
+      "loss": 0.088,
+      "step": 2929
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5825168582646038,
+      "learning_rate": 8.81703289275021e-06,
+      "loss": 0.118,
+      "step": 2930
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3870929273539651,
+      "learning_rate": 8.816151327593702e-06,
+      "loss": 0.1041,
+      "step": 2931
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.40418358212486455,
+      "learning_rate": 8.81526947818801e-06,
+      "loss": 0.0872,
+      "step": 2932
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.43715991417852335,
+      "learning_rate": 8.814387344598823e-06,
+      "loss": 0.1258,
+      "step": 2933
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.396511259850949,
+      "learning_rate": 8.813504926891848e-06,
+      "loss": 0.0837,
+      "step": 2934
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3998401387987405,
+      "learning_rate": 8.812622225132809e-06,
+      "loss": 0.1155,
+      "step": 2935
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.229249685559685,
+      "learning_rate": 8.811739239387456e-06,
+      "loss": 0.0659,
+      "step": 2936
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.31346605930253274,
+      "learning_rate": 8.81085596972156e-06,
+      "loss": 0.0862,
+      "step": 2937
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.26630506028654455,
+      "learning_rate": 8.809972416200911e-06,
+      "loss": 0.0759,
+      "step": 2938
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.47053928460776173,
+      "learning_rate": 8.809088578891323e-06,
+      "loss": 0.1145,
+      "step": 2939
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3856814689919608,
+      "learning_rate": 8.808204457858627e-06,
+      "loss": 0.1078,
+      "step": 2940
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4581838423026915,
+      "learning_rate": 8.807320053168677e-06,
+      "loss": 0.0998,
+      "step": 2941
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.33724341757192344,
+      "learning_rate": 8.806435364887353e-06,
+      "loss": 0.0791,
+      "step": 2942
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.24557272507719297,
+      "learning_rate": 8.80555039308055e-06,
+      "loss": 0.0522,
+      "step": 2943
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3706194307175531,
+      "learning_rate": 8.80466513781418e-06,
+      "loss": 0.1475,
+      "step": 2944
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.23770944744467992,
+      "learning_rate": 8.80377959915419e-06,
+      "loss": 0.0312,
+      "step": 2945
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4996902536641272,
+      "learning_rate": 8.802893777166535e-06,
+      "loss": 0.1042,
+      "step": 2946
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3937901437884222,
+      "learning_rate": 8.8020076719172e-06,
+      "loss": 0.1113,
+      "step": 2947
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4067606665286067,
+      "learning_rate": 8.801121283472185e-06,
+      "loss": 0.0884,
+      "step": 2948
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.2721738940222727,
+      "learning_rate": 8.800234611897513e-06,
+      "loss": 0.0916,
+      "step": 2949
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.41975841466611485,
+      "learning_rate": 8.799347657259228e-06,
+      "loss": 0.0866,
+      "step": 2950
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3128445145052996,
+      "learning_rate": 8.798460419623398e-06,
+      "loss": 0.0761,
+      "step": 2951
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.20865477733894178,
+      "learning_rate": 8.797572899056109e-06,
+      "loss": 0.0371,
+      "step": 2952
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.35299886958277493,
+      "learning_rate": 8.796685095623466e-06,
+      "loss": 0.0769,
+      "step": 2953
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.36173531563296485,
+      "learning_rate": 8.7957970093916e-06,
+      "loss": 0.0921,
+      "step": 2954
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4767283692558023,
+      "learning_rate": 8.794908640426663e-06,
+      "loss": 0.1292,
+      "step": 2955
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.2861666948176078,
+      "learning_rate": 8.79401998879482e-06,
+      "loss": 0.0783,
+      "step": 2956
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.26759349085206163,
+      "learning_rate": 8.79313105456227e-06,
+      "loss": 0.0901,
+      "step": 2957
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.32168013246623794,
+      "learning_rate": 8.792241837795219e-06,
+      "loss": 0.0719,
+      "step": 2958
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3334721712264962,
+      "learning_rate": 8.791352338559906e-06,
+      "loss": 0.0653,
+      "step": 2959
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3429156405548126,
+      "learning_rate": 8.790462556922585e-06,
+      "loss": 0.0619,
+      "step": 2960
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4742728942252845,
+      "learning_rate": 8.789572492949533e-06,
+      "loss": 0.134,
+      "step": 2961
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3652179769118772,
+      "learning_rate": 8.788682146707043e-06,
+      "loss": 0.1204,
+      "step": 2962
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3381616096317784,
+      "learning_rate": 8.787791518261437e-06,
+      "loss": 0.0656,
+      "step": 2963
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3457025712510556,
+      "learning_rate": 8.786900607679053e-06,
+      "loss": 0.0548,
+      "step": 2964
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4812280973689279,
+      "learning_rate": 8.786009415026254e-06,
+      "loss": 0.1399,
+      "step": 2965
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7052403817332502,
+      "learning_rate": 8.785117940369415e-06,
+      "loss": 0.1001,
+      "step": 2966
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.2287226237926877,
+      "learning_rate": 8.784226183774944e-06,
+      "loss": 0.0715,
+      "step": 2967
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7434693176748717,
+      "learning_rate": 8.783334145309262e-06,
+      "loss": 0.0882,
+      "step": 2968
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.45715312805054975,
+      "learning_rate": 8.782441825038811e-06,
+      "loss": 0.0771,
+      "step": 2969
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.304969482041532,
+      "learning_rate": 8.781549223030061e-06,
+      "loss": 0.0653,
+      "step": 2970
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.34863359697087204,
+      "learning_rate": 8.780656339349495e-06,
+      "loss": 0.0753,
+      "step": 2971
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.24027347350793954,
+      "learning_rate": 8.77976317406362e-06,
+      "loss": 0.0493,
+      "step": 2972
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.40889968863113907,
+      "learning_rate": 8.778869727238967e-06,
+      "loss": 0.1111,
+      "step": 2973
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.18882107639783857,
+      "learning_rate": 8.77797599894208e-06,
+      "loss": 0.0435,
+      "step": 2974
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3838513400340036,
+      "learning_rate": 8.777081989239536e-06,
+      "loss": 0.0999,
+      "step": 2975
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.45983458419741485,
+      "learning_rate": 8.77618769819792e-06,
+      "loss": 0.1147,
+      "step": 2976
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5124004979419058,
+      "learning_rate": 8.775293125883848e-06,
+      "loss": 0.0741,
+      "step": 2977
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.39616417187255293,
+      "learning_rate": 8.77439827236395e-06,
+      "loss": 0.0732,
+      "step": 2978
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4479651846763837,
+      "learning_rate": 8.773503137704882e-06,
+      "loss": 0.086,
+      "step": 2979
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.39806397267298743,
+      "learning_rate": 8.772607721973317e-06,
+      "loss": 0.098,
+      "step": 2980
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.34899815414181645,
+      "learning_rate": 8.771712025235953e-06,
+      "loss": 0.0717,
+      "step": 2981
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.23439161833035924,
+      "learning_rate": 8.770816047559504e-06,
+      "loss": 0.0622,
+      "step": 2982
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.36554932962078446,
+      "learning_rate": 8.76991978901071e-06,
+      "loss": 0.0747,
+      "step": 2983
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3532014102586062,
+      "learning_rate": 8.769023249656327e-06,
+      "loss": 0.0774,
+      "step": 2984
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.345853014483529,
+      "learning_rate": 8.768126429563138e-06,
+      "loss": 0.0573,
+      "step": 2985
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.35310386537462773,
+      "learning_rate": 8.767229328797943e-06,
+      "loss": 0.0952,
+      "step": 2986
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.37319853034545414,
+      "learning_rate": 8.76633194742756e-06,
+      "loss": 0.0684,
+      "step": 2987
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.23295897174057056,
+      "learning_rate": 8.765434285518832e-06,
+      "loss": 0.0518,
+      "step": 2988
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.23156680819498554,
+      "learning_rate": 8.764536343138626e-06,
+      "loss": 0.0718,
+      "step": 2989
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3775996706689483,
+      "learning_rate": 8.763638120353821e-06,
+      "loss": 0.101,
+      "step": 2990
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5060511053304001,
+      "learning_rate": 8.762739617231325e-06,
+      "loss": 0.122,
+      "step": 2991
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3203705557553357,
+      "learning_rate": 8.761840833838062e-06,
+      "loss": 0.0664,
+      "step": 2992
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4165276899833806,
+      "learning_rate": 8.760941770240979e-06,
+      "loss": 0.0926,
+      "step": 2993
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.30619266291650543,
+      "learning_rate": 8.760042426507044e-06,
+      "loss": 0.0866,
+      "step": 2994
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.44687678887352683,
+      "learning_rate": 8.759142802703246e-06,
+      "loss": 0.1012,
+      "step": 2995
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.42565070428799945,
+      "learning_rate": 8.758242898896594e-06,
+      "loss": 0.0941,
+      "step": 2996
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3334117676579736,
+      "learning_rate": 8.757342715154115e-06,
+      "loss": 0.07,
+      "step": 2997
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5134372633258016,
+      "learning_rate": 8.756442251542864e-06,
+      "loss": 0.1225,
+      "step": 2998
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.29412708525730175,
+      "learning_rate": 8.755541508129912e-06,
+      "loss": 0.071,
+      "step": 2999
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5219701811151635,
+      "learning_rate": 8.754640484982351e-06,
+      "loss": 0.111,
+      "step": 3000
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.46805525685333077,
+      "learning_rate": 8.753739182167294e-06,
+      "loss": 0.0612,
+      "step": 3001
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.29479732104974793,
+      "learning_rate": 8.752837599751873e-06,
+      "loss": 0.0618,
+      "step": 3002
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.27199341588978804,
+      "learning_rate": 8.751935737803248e-06,
+      "loss": 0.0879,
+      "step": 3003
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3978160169497034,
+      "learning_rate": 8.751033596388593e-06,
+      "loss": 0.0744,
+      "step": 3004
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.34080311756766973,
+      "learning_rate": 8.750131175575103e-06,
+      "loss": 0.0733,
+      "step": 3005
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3631759022776894,
+      "learning_rate": 8.749228475429999e-06,
+      "loss": 0.1012,
+      "step": 3006
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.32774359259040653,
+      "learning_rate": 8.748325496020512e-06,
+      "loss": 0.1149,
+      "step": 3007
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.2992196278523508,
+      "learning_rate": 8.747422237413912e-06,
+      "loss": 0.0657,
+      "step": 3008
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.34297064806118344,
+      "learning_rate": 8.746518699677472e-06,
+      "loss": 0.0974,
+      "step": 3009
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.1873070952707471,
+      "learning_rate": 8.745614882878494e-06,
+      "loss": 0.0338,
+      "step": 3010
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.307782549440564,
+      "learning_rate": 8.744710787084299e-06,
+      "loss": 0.0665,
+      "step": 3011
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.19646688663582387,
+      "learning_rate": 8.74380641236223e-06,
+      "loss": 0.0399,
+      "step": 3012
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.29279733842085304,
+      "learning_rate": 8.742901758779651e-06,
+      "loss": 0.0792,
+      "step": 3013
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.2263859255474588,
+      "learning_rate": 8.741996826403944e-06,
+      "loss": 0.0619,
+      "step": 3014
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.20951131458836975,
+      "learning_rate": 8.741091615302516e-06,
+      "loss": 0.0537,
+      "step": 3015
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.2790485558260262,
+      "learning_rate": 8.74018612554279e-06,
+      "loss": 0.0686,
+      "step": 3016
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.2848789155589141,
+      "learning_rate": 8.739280357192213e-06,
+      "loss": 0.0794,
+      "step": 3017
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.26033986062806086,
+      "learning_rate": 8.738374310318253e-06,
+      "loss": 0.0594,
+      "step": 3018
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4880779228685505,
+      "learning_rate": 8.737467984988397e-06,
+      "loss": 0.1166,
+      "step": 3019
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4776696157447335,
+      "learning_rate": 8.736561381270152e-06,
+      "loss": 0.1388,
+      "step": 3020
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4597053267578651,
+      "learning_rate": 8.735654499231049e-06,
+      "loss": 0.1015,
+      "step": 3021
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3739923924755732,
+      "learning_rate": 8.734747338938636e-06,
+      "loss": 0.0821,
+      "step": 3022
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.19204212567712858,
+      "learning_rate": 8.733839900460487e-06,
+      "loss": 0.0588,
+      "step": 3023
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.49656902512924556,
+      "learning_rate": 8.732932183864188e-06,
+      "loss": 0.0884,
+      "step": 3024
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.30436965255209997,
+      "learning_rate": 8.732024189217355e-06,
+      "loss": 0.0852,
+      "step": 3025
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.37335454443625116,
+      "learning_rate": 8.73111591658762e-06,
+      "loss": 0.0584,
+      "step": 3026
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3266896381540124,
+      "learning_rate": 8.730207366042636e-06,
+      "loss": 0.0936,
+      "step": 3027
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4783785021587216,
+      "learning_rate": 8.729298537650078e-06,
+      "loss": 0.1237,
+      "step": 3028
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.333226837475945,
+      "learning_rate": 8.728389431477639e-06,
+      "loss": 0.0825,
+      "step": 3029
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.25996530304656607,
+      "learning_rate": 8.727480047593037e-06,
+      "loss": 0.0897,
+      "step": 3030
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2920157497176272,
+      "learning_rate": 8.726570386064006e-06,
+      "loss": 0.0521,
+      "step": 3031
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.42987280768225616,
+      "learning_rate": 8.725660446958304e-06,
+      "loss": 0.1024,
+      "step": 3032
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.308683485152685,
+      "learning_rate": 8.72475023034371e-06,
+      "loss": 0.0818,
+      "step": 3033
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.294305976354729,
+      "learning_rate": 8.723839736288017e-06,
+      "loss": 0.0682,
+      "step": 3034
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.23245477279577353,
+      "learning_rate": 8.72292896485905e-06,
+      "loss": 0.0648,
+      "step": 3035
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2762312648125152,
+      "learning_rate": 8.722017916124646e-06,
+      "loss": 0.0466,
+      "step": 3036
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.1544132769203554,
+      "learning_rate": 8.721106590152666e-06,
+      "loss": 0.0322,
+      "step": 3037
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7385130726875857,
+      "learning_rate": 8.720194987010988e-06,
+      "loss": 0.1177,
+      "step": 3038
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3876200212981495,
+      "learning_rate": 8.719283106767515e-06,
+      "loss": 0.0681,
+      "step": 3039
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2864410882581645,
+      "learning_rate": 8.718370949490172e-06,
+      "loss": 0.0825,
+      "step": 3040
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.44777902534983166,
+      "learning_rate": 8.717458515246899e-06,
+      "loss": 0.0847,
+      "step": 3041
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4304585180889677,
+      "learning_rate": 8.71654580410566e-06,
+      "loss": 0.0967,
+      "step": 3042
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.444879683370426,
+      "learning_rate": 8.715632816134438e-06,
+      "loss": 0.0779,
+      "step": 3043
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.24787005757509195,
+      "learning_rate": 8.71471955140124e-06,
+      "loss": 0.0608,
+      "step": 3044
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3237969194907698,
+      "learning_rate": 8.71380600997409e-06,
+      "loss": 0.089,
+      "step": 3045
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3078945057934309,
+      "learning_rate": 8.712892191921032e-06,
+      "loss": 0.0813,
+      "step": 3046
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5526745925550122,
+      "learning_rate": 8.711978097310137e-06,
+      "loss": 0.1168,
+      "step": 3047
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.26734762486589675,
+      "learning_rate": 8.711063726209488e-06,
+      "loss": 0.0695,
+      "step": 3048
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4340663189824596,
+      "learning_rate": 8.710149078687193e-06,
+      "loss": 0.109,
+      "step": 3049
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.26267975452279574,
+      "learning_rate": 8.709234154811384e-06,
+      "loss": 0.06,
+      "step": 3050
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.26464568140649297,
+      "learning_rate": 8.708318954650206e-06,
+      "loss": 0.0305,
+      "step": 3051
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.20673229308521307,
+      "learning_rate": 8.70740347827183e-06,
+      "loss": 0.0543,
+      "step": 3052
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.42673339340516275,
+      "learning_rate": 8.706487725744444e-06,
+      "loss": 0.1238,
+      "step": 3053
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3143014509646133,
+      "learning_rate": 8.705571697136263e-06,
+      "loss": 0.0968,
+      "step": 3054
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.30545729716000103,
+      "learning_rate": 8.704655392515513e-06,
+      "loss": 0.0945,
+      "step": 3055
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2560163250501721,
+      "learning_rate": 8.70373881195045e-06,
+      "loss": 0.0463,
+      "step": 3056
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2537603282629021,
+      "learning_rate": 8.702821955509345e-06,
+      "loss": 0.0891,
+      "step": 3057
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.39570207196513024,
+      "learning_rate": 8.701904823260488e-06,
+      "loss": 0.083,
+      "step": 3058
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6619417778811499,
+      "learning_rate": 8.700987415272195e-06,
+      "loss": 0.0507,
+      "step": 3059
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4229151948048635,
+      "learning_rate": 8.700069731612799e-06,
+      "loss": 0.0992,
+      "step": 3060
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.31991536612454646,
+      "learning_rate": 8.699151772350656e-06,
+      "loss": 0.0978,
+      "step": 3061
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.24058634715423152,
+      "learning_rate": 8.69823353755414e-06,
+      "loss": 0.0626,
+      "step": 3062
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.40669894172985,
+      "learning_rate": 8.697315027291645e-06,
+      "loss": 0.1698,
+      "step": 3063
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3774105248406947,
+      "learning_rate": 8.69639624163159e-06,
+      "loss": 0.1113,
+      "step": 3064
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3624513429718424,
+      "learning_rate": 8.695477180642409e-06,
+      "loss": 0.1022,
+      "step": 3065
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.37015760486948013,
+      "learning_rate": 8.69455784439256e-06,
+      "loss": 0.0826,
+      "step": 3066
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2556386154762862,
+      "learning_rate": 8.693638232950522e-06,
+      "loss": 0.0761,
+      "step": 3067
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.22507158405401548,
+      "learning_rate": 8.69271834638479e-06,
+      "loss": 0.0728,
+      "step": 3068
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.37669699910002513,
+      "learning_rate": 8.691798184763883e-06,
+      "loss": 0.1068,
+      "step": 3069
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4168428571113629,
+      "learning_rate": 8.69087774815634e-06,
+      "loss": 0.1006,
+      "step": 3070
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3939861919101858,
+      "learning_rate": 8.689957036630723e-06,
+      "loss": 0.0905,
+      "step": 3071
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6139665360505826,
+      "learning_rate": 8.689036050255611e-06,
+      "loss": 0.1464,
+      "step": 3072
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3155217628362793,
+      "learning_rate": 8.688114789099603e-06,
+      "loss": 0.0982,
+      "step": 3073
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6456320150900486,
+      "learning_rate": 8.687193253231321e-06,
+      "loss": 0.1239,
+      "step": 3074
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4089570285304536,
+      "learning_rate": 8.686271442719404e-06,
+      "loss": 0.1085,
+      "step": 3075
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.31931112691738567,
+      "learning_rate": 8.685349357632518e-06,
+      "loss": 0.0804,
+      "step": 3076
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5620520747543952,
+      "learning_rate": 8.684426998039342e-06,
+      "loss": 0.0923,
+      "step": 3077
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2980340551852319,
+      "learning_rate": 8.683504364008582e-06,
+      "loss": 0.0499,
+      "step": 3078
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5172604670301609,
+      "learning_rate": 8.682581455608956e-06,
+      "loss": 0.1117,
+      "step": 3079
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.44942208783293075,
+      "learning_rate": 8.681658272909213e-06,
+      "loss": 0.1105,
+      "step": 3080
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.645272144801532,
+      "learning_rate": 8.680734815978112e-06,
+      "loss": 0.0833,
+      "step": 3081
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3886851201409186,
+      "learning_rate": 8.679811084884442e-06,
+      "loss": 0.0907,
+      "step": 3082
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2926562268833976,
+      "learning_rate": 8.678887079697005e-06,
+      "loss": 0.1033,
+      "step": 3083
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2592224304525484,
+      "learning_rate": 8.677962800484628e-06,
+      "loss": 0.0809,
+      "step": 3084
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5386606562840128,
+      "learning_rate": 8.677038247316156e-06,
+      "loss": 0.0675,
+      "step": 3085
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.29711257637923505,
+      "learning_rate": 8.676113420260454e-06,
+      "loss": 0.0643,
+      "step": 3086
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.23687841914215468,
+      "learning_rate": 8.675188319386409e-06,
+      "loss": 0.0692,
+      "step": 3087
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.40167497144295966,
+      "learning_rate": 8.67426294476293e-06,
+      "loss": 0.1198,
+      "step": 3088
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2667791449724475,
+      "learning_rate": 8.673337296458943e-06,
+      "loss": 0.0593,
+      "step": 3089
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.30843353990812183,
+      "learning_rate": 8.672411374543395e-06,
+      "loss": 0.0729,
+      "step": 3090
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3674815263894456,
+      "learning_rate": 8.671485179085253e-06,
+      "loss": 0.0783,
+      "step": 3091
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.37472969635240316,
+      "learning_rate": 8.670558710153507e-06,
+      "loss": 0.1174,
+      "step": 3092
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3503859966142475,
+      "learning_rate": 8.669631967817168e-06,
+      "loss": 0.0966,
+      "step": 3093
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.37443217149266933,
+      "learning_rate": 8.66870495214526e-06,
+      "loss": 0.1056,
+      "step": 3094
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4196137241317033,
+      "learning_rate": 8.667777663206837e-06,
+      "loss": 0.1063,
+      "step": 3095
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.35339454112322427,
+      "learning_rate": 8.666850101070968e-06,
+      "loss": 0.0552,
+      "step": 3096
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.17312533850363393,
+      "learning_rate": 8.665922265806741e-06,
+      "loss": 0.0534,
+      "step": 3097
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.25386378969272116,
+      "learning_rate": 8.664994157483269e-06,
+      "loss": 0.0542,
+      "step": 3098
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6522153747464053,
+      "learning_rate": 8.664065776169681e-06,
+      "loss": 0.1661,
+      "step": 3099
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.33530007524378136,
+      "learning_rate": 8.66313712193513e-06,
+      "loss": 0.0748,
+      "step": 3100
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.530169144597035,
+      "learning_rate": 8.662208194848786e-06,
+      "loss": 0.1216,
+      "step": 3101
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5103953536176062,
+      "learning_rate": 8.661278994979843e-06,
+      "loss": 0.1227,
+      "step": 3102
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3606476525458075,
+      "learning_rate": 8.66034952239751e-06,
+      "loss": 0.0826,
+      "step": 3103
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2785053445972938,
+      "learning_rate": 8.659419777171022e-06,
+      "loss": 0.076,
+      "step": 3104
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2687087666482121,
+      "learning_rate": 8.658489759369633e-06,
+      "loss": 0.0799,
+      "step": 3105
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3410263948835231,
+      "learning_rate": 8.657559469062614e-06,
+      "loss": 0.0706,
+      "step": 3106
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3079909184001713,
+      "learning_rate": 8.656628906319258e-06,
+      "loss": 0.0687,
+      "step": 3107
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.35143030901980515,
+      "learning_rate": 8.65569807120888e-06,
+      "loss": 0.103,
+      "step": 3108
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.22985844520561322,
+      "learning_rate": 8.654766963800814e-06,
+      "loss": 0.0707,
+      "step": 3109
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2158617418890222,
+      "learning_rate": 8.653835584164415e-06,
+      "loss": 0.0395,
+      "step": 3110
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.289186626036599,
+      "learning_rate": 8.652903932369053e-06,
+      "loss": 0.0809,
+      "step": 3111
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.25289569586848415,
+      "learning_rate": 8.651972008484129e-06,
+      "loss": 0.0998,
+      "step": 3112
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5648942626602366,
+      "learning_rate": 8.651039812579054e-06,
+      "loss": 0.1131,
+      "step": 3113
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3601583992794044,
+      "learning_rate": 8.650107344723266e-06,
+      "loss": 0.1083,
+      "step": 3114
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3309370885185487,
+      "learning_rate": 8.649174604986219e-06,
+      "loss": 0.0799,
+      "step": 3115
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.32542710895638866,
+      "learning_rate": 8.64824159343739e-06,
+      "loss": 0.0946,
+      "step": 3116
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3459152992662643,
+      "learning_rate": 8.647308310146273e-06,
+      "loss": 0.0936,
+      "step": 3117
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5691984017877604,
+      "learning_rate": 8.646374755182388e-06,
+      "loss": 0.1248,
+      "step": 3118
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4569362608514308,
+      "learning_rate": 8.645440928615269e-06,
+      "loss": 0.1279,
+      "step": 3119
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4338039577266958,
+      "learning_rate": 8.644506830514475e-06,
+      "loss": 0.1007,
+      "step": 3120
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3044495409370965,
+      "learning_rate": 8.643572460949578e-06,
+      "loss": 0.0947,
+      "step": 3121
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.29216930135792835,
+      "learning_rate": 8.64263781999018e-06,
+      "loss": 0.0732,
+      "step": 3122
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5215553813165708,
+      "learning_rate": 8.641702907705898e-06,
+      "loss": 0.1351,
+      "step": 3123
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4176139003627783,
+      "learning_rate": 8.640767724166368e-06,
+      "loss": 0.1094,
+      "step": 3124
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3078718509833029,
+      "learning_rate": 8.63983226944125e-06,
+      "loss": 0.0566,
+      "step": 3125
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.38293845412458544,
+      "learning_rate": 8.63889654360022e-06,
+      "loss": 0.0794,
+      "step": 3126
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.49416550841508555,
+      "learning_rate": 8.637960546712977e-06,
+      "loss": 0.1037,
+      "step": 3127
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5180489303394368,
+      "learning_rate": 8.637024278849241e-06,
+      "loss": 0.1162,
+      "step": 3128
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.30283153033511245,
+      "learning_rate": 8.636087740078749e-06,
+      "loss": 0.0492,
+      "step": 3129
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.24541939070108781,
+      "learning_rate": 8.63515093047126e-06,
+      "loss": 0.0475,
+      "step": 3130
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3284929833474759,
+      "learning_rate": 8.634213850096554e-06,
+      "loss": 0.1035,
+      "step": 3131
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2690705848976222,
+      "learning_rate": 8.63327649902443e-06,
+      "loss": 0.0765,
+      "step": 3132
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.36139522796423673,
+      "learning_rate": 8.632338877324705e-06,
+      "loss": 0.1057,
+      "step": 3133
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.354741597268939,
+      "learning_rate": 8.63140098506722e-06,
+      "loss": 0.1171,
+      "step": 3134
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3939376050659103,
+      "learning_rate": 8.630462822321838e-06,
+      "loss": 0.0914,
+      "step": 3135
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.37242398320385267,
+      "learning_rate": 8.629524389158435e-06,
+      "loss": 0.1162,
+      "step": 3136
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6210828343589856,
+      "learning_rate": 8.62858568564691e-06,
+      "loss": 0.1042,
+      "step": 3137
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.44341954276916,
+      "learning_rate": 8.627646711857188e-06,
+      "loss": 0.1074,
+      "step": 3138
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.40703856623009427,
+      "learning_rate": 8.626707467859205e-06,
+      "loss": 0.0809,
+      "step": 3139
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5131052633462835,
+      "learning_rate": 8.625767953722923e-06,
+      "loss": 0.1299,
+      "step": 3140
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.40440762471270764,
+      "learning_rate": 8.624828169518322e-06,
+      "loss": 0.0673,
+      "step": 3141
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4026228171449898,
+      "learning_rate": 8.623888115315402e-06,
+      "loss": 0.1242,
+      "step": 3142
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0405054218533596,
+      "learning_rate": 8.622947791184186e-06,
+      "loss": 0.1539,
+      "step": 3143
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.45489837857045895,
+      "learning_rate": 8.622007197194712e-06,
+      "loss": 0.1188,
+      "step": 3144
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.458041621658602,
+      "learning_rate": 8.621066333417044e-06,
+      "loss": 0.0983,
+      "step": 3145
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5868371905219253,
+      "learning_rate": 8.620125199921259e-06,
+      "loss": 0.0838,
+      "step": 3146
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4996584318164899,
+      "learning_rate": 8.61918379677746e-06,
+      "loss": 0.1522,
+      "step": 3147
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6631142003833113,
+      "learning_rate": 8.61824212405577e-06,
+      "loss": 0.1108,
+      "step": 3148
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4410454527504012,
+      "learning_rate": 8.617300181826328e-06,
+      "loss": 0.1194,
+      "step": 3149
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.27908004179805007,
+      "learning_rate": 8.616357970159296e-06,
+      "loss": 0.07,
+      "step": 3150
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.34609408537989883,
+      "learning_rate": 8.615415489124857e-06,
+      "loss": 0.0669,
+      "step": 3151
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.419953704372675,
+      "learning_rate": 8.61447273879321e-06,
+      "loss": 0.1108,
+      "step": 3152
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.40446890730651647,
+      "learning_rate": 8.613529719234577e-06,
+      "loss": 0.1217,
+      "step": 3153
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.28617730337793196,
+      "learning_rate": 8.612586430519201e-06,
+      "loss": 0.092,
+      "step": 3154
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.28130652207297047,
+      "learning_rate": 8.611642872717343e-06,
+      "loss": 0.0772,
+      "step": 3155
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5650392784552916,
+      "learning_rate": 8.610699045899283e-06,
+      "loss": 0.1281,
+      "step": 3156
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.17599631933648477,
+      "learning_rate": 8.609754950135325e-06,
+      "loss": 0.0444,
+      "step": 3157
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.33425221227894586,
+      "learning_rate": 8.60881058549579e-06,
+      "loss": 0.0709,
+      "step": 3158
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.18045066763482817,
+      "learning_rate": 8.607865952051021e-06,
+      "loss": 0.0299,
+      "step": 3159
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.42381083368727096,
+      "learning_rate": 8.606921049871377e-06,
+      "loss": 0.0909,
+      "step": 3160
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4385859831444366,
+      "learning_rate": 8.605975879027242e-06,
+      "loss": 0.0952,
+      "step": 3161
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3716520722317863,
+      "learning_rate": 8.605030439589017e-06,
+      "loss": 0.1081,
+      "step": 3162
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2660368474783016,
+      "learning_rate": 8.604084731627123e-06,
+      "loss": 0.053,
+      "step": 3163
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.32090225166767555,
+      "learning_rate": 8.603138755212004e-06,
+      "loss": 0.0668,
+      "step": 3164
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4790531991050869,
+      "learning_rate": 8.602192510414121e-06,
+      "loss": 0.0858,
+      "step": 3165
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2840525100702858,
+      "learning_rate": 8.601245997303954e-06,
+      "loss": 0.0673,
+      "step": 3166
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6141098224050812,
+      "learning_rate": 8.600299215952009e-06,
+      "loss": 0.1266,
+      "step": 3167
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2597524150143281,
+      "learning_rate": 8.5993521664288e-06,
+      "loss": 0.0653,
+      "step": 3168
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2904008801759814,
+      "learning_rate": 8.598404848804879e-06,
+      "loss": 0.0612,
+      "step": 3169
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.21435316282959738,
+      "learning_rate": 8.597457263150801e-06,
+      "loss": 0.0703,
+      "step": 3170
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.43803103169438823,
+      "learning_rate": 8.596509409537148e-06,
+      "loss": 0.103,
+      "step": 3171
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.40892091518874096,
+      "learning_rate": 8.595561288034524e-06,
+      "loss": 0.0771,
+      "step": 3172
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3598338325366773,
+      "learning_rate": 8.59461289871355e-06,
+      "loss": 0.076,
+      "step": 3173
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.281174090911734,
+      "learning_rate": 8.593664241644868e-06,
+      "loss": 0.0543,
+      "step": 3174
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8070277444030411,
+      "learning_rate": 8.592715316899136e-06,
+      "loss": 0.0899,
+      "step": 3175
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4905251976539588,
+      "learning_rate": 8.59176612454704e-06,
+      "loss": 0.0813,
+      "step": 3176
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.27669804600285297,
+      "learning_rate": 8.590816664659279e-06,
+      "loss": 0.097,
+      "step": 3177
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.22095066131920046,
+      "learning_rate": 8.589866937306574e-06,
+      "loss": 0.0707,
+      "step": 3178
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3147108398502547,
+      "learning_rate": 8.588916942559668e-06,
+      "loss": 0.0778,
+      "step": 3179
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.46495837026485753,
+      "learning_rate": 8.58796668048932e-06,
+      "loss": 0.0704,
+      "step": 3180
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.47869112909600686,
+      "learning_rate": 8.587016151166314e-06,
+      "loss": 0.0922,
+      "step": 3181
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.39161912982290736,
+      "learning_rate": 8.586065354661448e-06,
+      "loss": 0.1078,
+      "step": 3182
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2928832466448882,
+      "learning_rate": 8.585114291045544e-06,
+      "loss": 0.0767,
+      "step": 3183
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4152114814618614,
+      "learning_rate": 8.584162960389443e-06,
+      "loss": 0.103,
+      "step": 3184
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.34526393663227817,
+      "learning_rate": 8.583211362764006e-06,
+      "loss": 0.0696,
+      "step": 3185
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4783039155017807,
+      "learning_rate": 8.582259498240111e-06,
+      "loss": 0.1255,
+      "step": 3186
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2983994278844034,
+      "learning_rate": 8.581307366888663e-06,
+      "loss": 0.0953,
+      "step": 3187
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4297595322064294,
+      "learning_rate": 8.580354968780578e-06,
+      "loss": 0.087,
+      "step": 3188
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2702825337725005,
+      "learning_rate": 8.579402303986798e-06,
+      "loss": 0.064,
+      "step": 3189
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.25057890993016113,
+      "learning_rate": 8.578449372578284e-06,
+      "loss": 0.0536,
+      "step": 3190
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4997154962747346,
+      "learning_rate": 8.577496174626013e-06,
+      "loss": 0.1462,
+      "step": 3191
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5484457245657279,
+      "learning_rate": 8.576542710200987e-06,
+      "loss": 0.1782,
+      "step": 3192
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7553463793023507,
+      "learning_rate": 8.575588979374226e-06,
+      "loss": 0.1222,
+      "step": 3193
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.21169058914202843,
+      "learning_rate": 8.574634982216767e-06,
+      "loss": 0.0697,
+      "step": 3194
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4280841174287749,
+      "learning_rate": 8.573680718799671e-06,
+      "loss": 0.0999,
+      "step": 3195
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5388658852358397,
+      "learning_rate": 8.572726189194019e-06,
+      "loss": 0.1348,
+      "step": 3196
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3920844544821534,
+      "learning_rate": 8.571771393470904e-06,
+      "loss": 0.1133,
+      "step": 3197
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.27407678053272166,
+      "learning_rate": 8.570816331701451e-06,
+      "loss": 0.0695,
+      "step": 3198
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3430960494607401,
+      "learning_rate": 8.569861003956795e-06,
+      "loss": 0.0869,
+      "step": 3199
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2686785339460621,
+      "learning_rate": 8.568905410308097e-06,
+      "loss": 0.0621,
+      "step": 3200
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.26427304549817626,
+      "learning_rate": 8.567949550826532e-06,
+      "loss": 0.0731,
+      "step": 3201
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3711964973476967,
+      "learning_rate": 8.5669934255833e-06,
+      "loss": 0.0971,
+      "step": 3202
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5404813466275674,
+      "learning_rate": 8.56603703464962e-06,
+      "loss": 0.0863,
+      "step": 3203
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8233651893174969,
+      "learning_rate": 8.565080378096725e-06,
+      "loss": 0.116,
+      "step": 3204
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3238031058502858,
+      "learning_rate": 8.564123455995877e-06,
+      "loss": 0.0896,
+      "step": 3205
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.376880042601365,
+      "learning_rate": 8.563166268418351e-06,
+      "loss": 0.0862,
+      "step": 3206
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2526099829136848,
+      "learning_rate": 8.562208815435445e-06,
+      "loss": 0.082,
+      "step": 3207
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.36082686878299813,
+      "learning_rate": 8.561251097118475e-06,
+      "loss": 0.1148,
+      "step": 3208
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.32571027721387186,
+      "learning_rate": 8.560293113538777e-06,
+      "loss": 0.0968,
+      "step": 3209
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3023408386922019,
+      "learning_rate": 8.559334864767707e-06,
+      "loss": 0.0466,
+      "step": 3210
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.1878076277003008,
+      "learning_rate": 8.558376350876644e-06,
+      "loss": 0.0481,
+      "step": 3211
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3410576094691958,
+      "learning_rate": 8.557417571936978e-06,
+      "loss": 0.1131,
+      "step": 3212
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.49962613976695847,
+      "learning_rate": 8.556458528020131e-06,
+      "loss": 0.1155,
+      "step": 3213
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.34573822382790637,
+      "learning_rate": 8.555499219197534e-06,
+      "loss": 0.1105,
+      "step": 3214
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2873693084160559,
+      "learning_rate": 8.554539645540642e-06,
+      "loss": 0.0796,
+      "step": 3215
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2675779164142398,
+      "learning_rate": 8.55357980712093e-06,
+      "loss": 0.0512,
+      "step": 3216
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.40171732986158,
+      "learning_rate": 8.552619704009895e-06,
+      "loss": 0.1018,
+      "step": 3217
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3130559161605051,
+      "learning_rate": 8.551659336279047e-06,
+      "loss": 0.0994,
+      "step": 3218
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.35490154059866313,
+      "learning_rate": 8.550698703999922e-06,
+      "loss": 0.0673,
+      "step": 3219
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5070013958030166,
+      "learning_rate": 8.549737807244073e-06,
+      "loss": 0.0994,
+      "step": 3220
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3087753827818303,
+      "learning_rate": 8.548776646083074e-06,
+      "loss": 0.0595,
+      "step": 3221
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.32824077294457665,
+      "learning_rate": 8.547815220588517e-06,
+      "loss": 0.0863,
+      "step": 3222
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3743889538050714,
+      "learning_rate": 8.546853530832015e-06,
+      "loss": 0.1064,
+      "step": 3223
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2801371714876428,
+      "learning_rate": 8.545891576885198e-06,
+      "loss": 0.0566,
+      "step": 3224
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5621597618454878,
+      "learning_rate": 8.544929358819724e-06,
+      "loss": 0.0845,
+      "step": 3225
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.34000926095524525,
+      "learning_rate": 8.543966876707258e-06,
+      "loss": 0.0973,
+      "step": 3226
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.537761908965727,
+      "learning_rate": 8.543004130619494e-06,
+      "loss": 0.1599,
+      "step": 3227
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4273946410502113,
+      "learning_rate": 8.542041120628143e-06,
+      "loss": 0.0933,
+      "step": 3228
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2548743596543969,
+      "learning_rate": 8.541077846804937e-06,
+      "loss": 0.0593,
+      "step": 3229
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5140422198652104,
+      "learning_rate": 8.540114309221624e-06,
+      "loss": 0.0918,
+      "step": 3230
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.314189752415136,
+      "learning_rate": 8.539150507949972e-06,
+      "loss": 0.0877,
+      "step": 3231
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.42965783813692837,
+      "learning_rate": 8.538186443061775e-06,
+      "loss": 0.12,
+      "step": 3232
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3936450967581396,
+      "learning_rate": 8.53722211462884e-06,
+      "loss": 0.0916,
+      "step": 3233
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.36522016352049114,
+      "learning_rate": 8.536257522722997e-06,
+      "loss": 0.0982,
+      "step": 3234
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.41862391248228376,
+      "learning_rate": 8.535292667416093e-06,
+      "loss": 0.1167,
+      "step": 3235
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.22624187113222125,
+      "learning_rate": 8.534327548779996e-06,
+      "loss": 0.0608,
+      "step": 3236
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.22955399230281884,
+      "learning_rate": 8.533362166886595e-06,
+      "loss": 0.0576,
+      "step": 3237
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.34469200880000195,
+      "learning_rate": 8.532396521807795e-06,
+      "loss": 0.0435,
+      "step": 3238
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2884067616331789,
+      "learning_rate": 8.531430613615525e-06,
+      "loss": 0.1152,
+      "step": 3239
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.732762773316448,
+      "learning_rate": 8.530464442381731e-06,
+      "loss": 0.1002,
+      "step": 3240
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2891667731297482,
+      "learning_rate": 8.52949800817838e-06,
+      "loss": 0.0744,
+      "step": 3241
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4674154182735356,
+      "learning_rate": 8.528531311077456e-06,
+      "loss": 0.113,
+      "step": 3242
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3753286632948059,
+      "learning_rate": 8.527564351150964e-06,
+      "loss": 0.076,
+      "step": 3243
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5524660177900298,
+      "learning_rate": 8.52659712847093e-06,
+      "loss": 0.0979,
+      "step": 3244
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3327205218749006,
+      "learning_rate": 8.525629643109397e-06,
+      "loss": 0.0762,
+      "step": 3245
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.34950465810645814,
+      "learning_rate": 8.52466189513843e-06,
+      "loss": 0.1137,
+      "step": 3246
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4291830250885162,
+      "learning_rate": 8.523693884630113e-06,
+      "loss": 0.1211,
+      "step": 3247
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.391255268613478,
+      "learning_rate": 8.522725611656548e-06,
+      "loss": 0.0879,
+      "step": 3248
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.33353056761724076,
+      "learning_rate": 8.52175707628986e-06,
+      "loss": 0.0845,
+      "step": 3249
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.403392058994754,
+      "learning_rate": 8.520788278602186e-06,
+      "loss": 0.0961,
+      "step": 3250
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.39243587519409867,
+      "learning_rate": 8.519819218665692e-06,
+      "loss": 0.1291,
+      "step": 3251
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.32935199675208676,
+      "learning_rate": 8.51884989655256e-06,
+      "loss": 0.0543,
+      "step": 3252
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.40585213969785006,
+      "learning_rate": 8.517880312334987e-06,
+      "loss": 0.1092,
+      "step": 3253
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3890174081017157,
+      "learning_rate": 8.516910466085196e-06,
+      "loss": 0.1074,
+      "step": 3254
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2484434455016351,
+      "learning_rate": 8.515940357875425e-06,
+      "loss": 0.0791,
+      "step": 3255
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2658535939025943,
+      "learning_rate": 8.514969987777935e-06,
+      "loss": 0.0874,
+      "step": 3256
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.43359056993230394,
+      "learning_rate": 8.513999355865003e-06,
+      "loss": 0.1033,
+      "step": 3257
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.33723946502703545,
+      "learning_rate": 8.51302846220893e-06,
+      "loss": 0.0714,
+      "step": 3258
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.34742262264342544,
+      "learning_rate": 8.512057306882032e-06,
+      "loss": 0.1103,
+      "step": 3259
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.37170281098989383,
+      "learning_rate": 8.511085889956646e-06,
+      "loss": 0.0966,
+      "step": 3260
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.22499351619756636,
+      "learning_rate": 8.51011421150513e-06,
+      "loss": 0.0859,
+      "step": 3261
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.466483122628316,
+      "learning_rate": 8.509142271599859e-06,
+      "loss": 0.0821,
+      "step": 3262
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.42798305368317,
+      "learning_rate": 8.508170070313229e-06,
+      "loss": 0.1029,
+      "step": 3263
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2784497769160566,
+      "learning_rate": 8.507197607717656e-06,
+      "loss": 0.0548,
+      "step": 3264
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5581748185905202,
+      "learning_rate": 8.506224883885574e-06,
+      "loss": 0.1074,
+      "step": 3265
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3189553204847817,
+      "learning_rate": 8.50525189888944e-06,
+      "loss": 0.1005,
+      "step": 3266
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.28504913214545835,
+      "learning_rate": 8.50427865280172e-06,
+      "loss": 0.0582,
+      "step": 3267
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.49875898758137166,
+      "learning_rate": 8.503305145694916e-06,
+      "loss": 0.1195,
+      "step": 3268
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2315835794495503,
+      "learning_rate": 8.502331377641535e-06,
+      "loss": 0.0555,
+      "step": 3269
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.30933923842008554,
+      "learning_rate": 8.50135734871411e-06,
+      "loss": 0.0939,
+      "step": 3270
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5624360134399001,
+      "learning_rate": 8.500383058985193e-06,
+      "loss": 0.1329,
+      "step": 3271
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.295397970556791,
+      "learning_rate": 8.499408508527355e-06,
+      "loss": 0.0851,
+      "step": 3272
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.1675894245864827,
+      "learning_rate": 8.498433697413186e-06,
+      "loss": 0.0307,
+      "step": 3273
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5132502603384155,
+      "learning_rate": 8.497458625715294e-06,
+      "loss": 0.0918,
+      "step": 3274
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4733679774538817,
+      "learning_rate": 8.49648329350631e-06,
+      "loss": 0.1027,
+      "step": 3275
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.27256448915757814,
+      "learning_rate": 8.495507700858882e-06,
+      "loss": 0.0805,
+      "step": 3276
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.37408728493161036,
+      "learning_rate": 8.494531847845678e-06,
+      "loss": 0.1008,
+      "step": 3277
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.39925872046497757,
+      "learning_rate": 8.493555734539386e-06,
+      "loss": 0.0921,
+      "step": 3278
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.35720907920386613,
+      "learning_rate": 8.49257936101271e-06,
+      "loss": 0.0671,
+      "step": 3279
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2864578666391323,
+      "learning_rate": 8.491602727338375e-06,
+      "loss": 0.0878,
+      "step": 3280
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.39967924543228067,
+      "learning_rate": 8.490625833589133e-06,
+      "loss": 0.1144,
+      "step": 3281
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.41537538493289095,
+      "learning_rate": 8.489648679837741e-06,
+      "loss": 0.1004,
+      "step": 3282
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.47951841674767526,
+      "learning_rate": 8.48867126615699e-06,
+      "loss": 0.0895,
+      "step": 3283
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.38633985217669187,
+      "learning_rate": 8.487693592619677e-06,
+      "loss": 0.0836,
+      "step": 3284
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.40296827372207644,
+      "learning_rate": 8.486715659298627e-06,
+      "loss": 0.104,
+      "step": 3285
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.38767267164524444,
+      "learning_rate": 8.485737466266688e-06,
+      "loss": 0.1304,
+      "step": 3286
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.35823461375247706,
+      "learning_rate": 8.484759013596713e-06,
+      "loss": 0.0967,
+      "step": 3287
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.35578414199319686,
+      "learning_rate": 8.483780301361587e-06,
+      "loss": 0.0866,
+      "step": 3288
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3968411679378174,
+      "learning_rate": 8.482801329634209e-06,
+      "loss": 0.116,
+      "step": 3289
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.44163512755169076,
+      "learning_rate": 8.481822098487502e-06,
+      "loss": 0.0716,
+      "step": 3290
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3954139121320707,
+      "learning_rate": 8.4808426079944e-06,
+      "loss": 0.0712,
+      "step": 3291
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.44121891200486024,
+      "learning_rate": 8.479862858227863e-06,
+      "loss": 0.1385,
+      "step": 3292
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.356646533576252,
+      "learning_rate": 8.478882849260868e-06,
+      "loss": 0.094,
+      "step": 3293
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3908175516053105,
+      "learning_rate": 8.477902581166413e-06,
+      "loss": 0.0981,
+      "step": 3294
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.41678076025180893,
+      "learning_rate": 8.476922054017514e-06,
+      "loss": 0.0978,
+      "step": 3295
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2448513105698184,
+      "learning_rate": 8.475941267887206e-06,
+      "loss": 0.0542,
+      "step": 3296
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3616055497447477,
+      "learning_rate": 8.474960222848542e-06,
+      "loss": 0.1087,
+      "step": 3297
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.33022684331591606,
+      "learning_rate": 8.473978918974598e-06,
+      "loss": 0.0679,
+      "step": 3298
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.40708789926489036,
+      "learning_rate": 8.472997356338469e-06,
+      "loss": 0.0677,
+      "step": 3299
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.30322614204820136,
+      "learning_rate": 8.472015535013262e-06,
+      "loss": 0.0914,
+      "step": 3300
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5341928224666488,
+      "learning_rate": 8.471033455072114e-06,
+      "loss": 0.1415,
+      "step": 3301
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4299489974869717,
+      "learning_rate": 8.470051116588174e-06,
+      "loss": 0.119,
+      "step": 3302
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7760515560778837,
+      "learning_rate": 8.46906851963461e-06,
+      "loss": 0.1193,
+      "step": 3303
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.43172189096797997,
+      "learning_rate": 8.468085664284615e-06,
+      "loss": 0.1018,
+      "step": 3304
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.36875631357121696,
+      "learning_rate": 8.467102550611398e-06,
+      "loss": 0.082,
+      "step": 3305
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.43235261026342625,
+      "learning_rate": 8.466119178688186e-06,
+      "loss": 0.0841,
+      "step": 3306
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3569165129764003,
+      "learning_rate": 8.465135548588224e-06,
+      "loss": 0.0472,
+      "step": 3307
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.31461599294489223,
+      "learning_rate": 8.46415166038478e-06,
+      "loss": 0.0759,
+      "step": 3308
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4468755758505138,
+      "learning_rate": 8.463167514151142e-06,
+      "loss": 0.1099,
+      "step": 3309
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.21892813775295478,
+      "learning_rate": 8.462183109960613e-06,
+      "loss": 0.0521,
+      "step": 3310
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2304771767209447,
+      "learning_rate": 8.461198447886517e-06,
+      "loss": 0.0693,
+      "step": 3311
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.41580216945953946,
+      "learning_rate": 8.460213528002197e-06,
+      "loss": 0.0788,
+      "step": 3312
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3435692167999843,
+      "learning_rate": 8.459228350381017e-06,
+      "loss": 0.0884,
+      "step": 3313
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3845269099039142,
+      "learning_rate": 8.458242915096356e-06,
+      "loss": 0.088,
+      "step": 3314
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.37064353233896713,
+      "learning_rate": 8.45725722222162e-06,
+      "loss": 0.0855,
+      "step": 3315
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4662331189380145,
+      "learning_rate": 8.456271271830223e-06,
+      "loss": 0.1055,
+      "step": 3316
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2130347691131314,
+      "learning_rate": 8.45528506399561e-06,
+      "loss": 0.0627,
+      "step": 3317
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3033374201811118,
+      "learning_rate": 8.454298598791235e-06,
+      "loss": 0.0567,
+      "step": 3318
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.28753764074345245,
+      "learning_rate": 8.45331187629058e-06,
+      "loss": 0.0688,
+      "step": 3319
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4298701532828766,
+      "learning_rate": 8.452324896567137e-06,
+      "loss": 0.1134,
+      "step": 3320
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.24269968958832125,
+      "learning_rate": 8.451337659694424e-06,
+      "loss": 0.0547,
+      "step": 3321
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.45275274014640904,
+      "learning_rate": 8.450350165745979e-06,
+      "loss": 0.1158,
+      "step": 3322
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.33374972792258467,
+      "learning_rate": 8.449362414795353e-06,
+      "loss": 0.0934,
+      "step": 3323
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3141848919402492,
+      "learning_rate": 8.44837440691612e-06,
+      "loss": 0.0724,
+      "step": 3324
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5028623126882972,
+      "learning_rate": 8.447386142181873e-06,
+      "loss": 0.0945,
+      "step": 3325
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.40024990555539547,
+      "learning_rate": 8.446397620666222e-06,
+      "loss": 0.1056,
+      "step": 3326
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2410627703341638,
+      "learning_rate": 8.445408842442802e-06,
+      "loss": 0.0929,
+      "step": 3327
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.32016105033735426,
+      "learning_rate": 8.44441980758526e-06,
+      "loss": 0.096,
+      "step": 3328
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.38275698114661477,
+      "learning_rate": 8.443430516167263e-06,
+      "loss": 0.1167,
+      "step": 3329
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.23848083978635368,
+      "learning_rate": 8.442440968262503e-06,
+      "loss": 0.0758,
+      "step": 3330
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3357317616355323,
+      "learning_rate": 8.441451163944687e-06,
+      "loss": 0.0666,
+      "step": 3331
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4315098014956081,
+      "learning_rate": 8.440461103287541e-06,
+      "loss": 0.1221,
+      "step": 3332
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.22908625309514344,
+      "learning_rate": 8.439470786364808e-06,
+      "loss": 0.0573,
+      "step": 3333
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.19572494488928305,
+      "learning_rate": 8.438480213250256e-06,
+      "loss": 0.0467,
+      "step": 3334
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.32482667806699966,
+      "learning_rate": 8.437489384017667e-06,
+      "loss": 0.1303,
+      "step": 3335
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4469492501081355,
+      "learning_rate": 8.436498298740842e-06,
+      "loss": 0.1065,
+      "step": 3336
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.46064522623430737,
+      "learning_rate": 8.435506957493606e-06,
+      "loss": 0.1077,
+      "step": 3337
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.267567471283384,
+      "learning_rate": 8.434515360349798e-06,
+      "loss": 0.0619,
+      "step": 3338
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.30638934078597135,
+      "learning_rate": 8.433523507383279e-06,
+      "loss": 0.0894,
+      "step": 3339
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.40060837399836735,
+      "learning_rate": 8.432531398667928e-06,
+      "loss": 0.1157,
+      "step": 3340
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.324327732462354,
+      "learning_rate": 8.431539034277642e-06,
+      "loss": 0.1002,
+      "step": 3341
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.39166590891194175,
+      "learning_rate": 8.430546414286336e-06,
+      "loss": 0.0946,
+      "step": 3342
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6252142762430505,
+      "learning_rate": 8.429553538767952e-06,
+      "loss": 0.0928,
+      "step": 3343
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6107044291807683,
+      "learning_rate": 8.42856040779644e-06,
+      "loss": 0.0728,
+      "step": 3344
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4558872260103241,
+      "learning_rate": 8.427567021445777e-06,
+      "loss": 0.1205,
+      "step": 3345
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3875100230588183,
+      "learning_rate": 8.426573379789956e-06,
+      "loss": 0.0436,
+      "step": 3346
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3771494326187267,
+      "learning_rate": 8.425579482902986e-06,
+      "loss": 0.0913,
+      "step": 3347
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5725745290525726,
+      "learning_rate": 8.4245853308589e-06,
+      "loss": 0.1401,
+      "step": 3348
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3774470102625326,
+      "learning_rate": 8.423590923731753e-06,
+      "loss": 0.0682,
+      "step": 3349
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.562252413355768,
+      "learning_rate": 8.422596261595608e-06,
+      "loss": 0.1251,
+      "step": 3350
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.31680553855293964,
+      "learning_rate": 8.421601344524555e-06,
+      "loss": 0.1092,
+      "step": 3351
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.38373734913665547,
+      "learning_rate": 8.4206061725927e-06,
+      "loss": 0.1094,
+      "step": 3352
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.20837822263642564,
+      "learning_rate": 8.419610745874175e-06,
+      "loss": 0.0621,
+      "step": 3353
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3703888947655228,
+      "learning_rate": 8.418615064443116e-06,
+      "loss": 0.0915,
+      "step": 3354
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6442335432160816,
+      "learning_rate": 8.417619128373695e-06,
+      "loss": 0.1201,
+      "step": 3355
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4260252350601668,
+      "learning_rate": 8.416622937740091e-06,
+      "loss": 0.0503,
+      "step": 3356
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5819320198359181,
+      "learning_rate": 8.415626492616509e-06,
+      "loss": 0.0822,
+      "step": 3357
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4235205358802179,
+      "learning_rate": 8.414629793077167e-06,
+      "loss": 0.1046,
+      "step": 3358
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2910420856174988,
+      "learning_rate": 8.413632839196306e-06,
+      "loss": 0.0737,
+      "step": 3359
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.406100722013346,
+      "learning_rate": 8.412635631048188e-06,
+      "loss": 0.1004,
+      "step": 3360
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7850592134220534,
+      "learning_rate": 8.411638168707086e-06,
+      "loss": 0.1885,
+      "step": 3361
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.39435603762595733,
+      "learning_rate": 8.410640452247299e-06,
+      "loss": 0.0798,
+      "step": 3362
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.25497441405730864,
+      "learning_rate": 8.409642481743141e-06,
+      "loss": 0.0436,
+      "step": 3363
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2379903396782314,
+      "learning_rate": 8.408644257268951e-06,
+      "loss": 0.0468,
+      "step": 3364
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5142612670676274,
+      "learning_rate": 8.407645778899078e-06,
+      "loss": 0.0838,
+      "step": 3365
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3831801946782589,
+      "learning_rate": 8.4066470467079e-06,
+      "loss": 0.0887,
+      "step": 3366
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.29258461204526076,
+      "learning_rate": 8.4056480607698e-06,
+      "loss": 0.0829,
+      "step": 3367
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.20288648934254114,
+      "learning_rate": 8.404648821159196e-06,
+      "loss": 0.0543,
+      "step": 3368
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4027044736075758,
+      "learning_rate": 8.403649327950511e-06,
+      "loss": 0.125,
+      "step": 3369
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6729101275743553,
+      "learning_rate": 8.4026495812182e-06,
+      "loss": 0.1429,
+      "step": 3370
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2694361518728078,
+      "learning_rate": 8.401649581036724e-06,
+      "loss": 0.0842,
+      "step": 3371
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2010541498615636,
+      "learning_rate": 8.400649327480572e-06,
+      "loss": 0.0416,
+      "step": 3372
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.38034497185975724,
+      "learning_rate": 8.399648820624249e-06,
+      "loss": 0.0667,
+      "step": 3373
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2916757737457415,
+      "learning_rate": 8.398648060542275e-06,
+      "loss": 0.0696,
+      "step": 3374
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2542035852949873,
+      "learning_rate": 8.397647047309198e-06,
+      "loss": 0.042,
+      "step": 3375
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.19226230547334297,
+      "learning_rate": 8.396645780999573e-06,
+      "loss": 0.0291,
+      "step": 3376
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.27866466288238123,
+      "learning_rate": 8.395644261687985e-06,
+      "loss": 0.0645,
+      "step": 3377
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.43291053893351605,
+      "learning_rate": 8.394642489449031e-06,
+      "loss": 0.0949,
+      "step": 3378
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2437447326906455,
+      "learning_rate": 8.393640464357329e-06,
+      "loss": 0.051,
+      "step": 3379
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.40948147827448067,
+      "learning_rate": 8.392638186487516e-06,
+      "loss": 0.119,
+      "step": 3380
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3842415440354286,
+      "learning_rate": 8.391635655914247e-06,
+      "loss": 0.0994,
+      "step": 3381
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3067628130372608,
+      "learning_rate": 8.390632872712198e-06,
+      "loss": 0.1152,
+      "step": 3382
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.24297628585068776,
+      "learning_rate": 8.38962983695606e-06,
+      "loss": 0.0581,
+      "step": 3383
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.46841119776419604,
+      "learning_rate": 8.388626548720544e-06,
+      "loss": 0.0951,
+      "step": 3384
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.32623700397439986,
+      "learning_rate": 8.387623008080385e-06,
+      "loss": 0.092,
+      "step": 3385
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3053157098640161,
+      "learning_rate": 8.386619215110328e-06,
+      "loss": 0.0773,
+      "step": 3386
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.46665660376585216,
+      "learning_rate": 8.385615169885144e-06,
+      "loss": 0.0956,
+      "step": 3387
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4055192819865835,
+      "learning_rate": 8.38461087247962e-06,
+      "loss": 0.1216,
+      "step": 3388
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2561331108094009,
+      "learning_rate": 8.38360632296856e-06,
+      "loss": 0.092,
+      "step": 3389
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.32356700897078516,
+      "learning_rate": 8.382601521426792e-06,
+      "loss": 0.0752,
+      "step": 3390
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.30851726092992915,
+      "learning_rate": 8.381596467929156e-06,
+      "loss": 0.0639,
+      "step": 3391
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2730356500934993,
+      "learning_rate": 8.380591162550516e-06,
+      "loss": 0.0625,
+      "step": 3392
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.35882615126242606,
+      "learning_rate": 8.379585605365753e-06,
+      "loss": 0.0856,
+      "step": 3393
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3049135956875774,
+      "learning_rate": 8.378579796449767e-06,
+      "loss": 0.0467,
+      "step": 3394
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7393813429775998,
+      "learning_rate": 8.377573735877476e-06,
+      "loss": 0.1475,
+      "step": 3395
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.38477887793628174,
+      "learning_rate": 8.376567423723817e-06,
+      "loss": 0.0804,
+      "step": 3396
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3014155759279757,
+      "learning_rate": 8.375560860063747e-06,
+      "loss": 0.0525,
+      "step": 3397
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4305428006655133,
+      "learning_rate": 8.374554044972239e-06,
+      "loss": 0.1105,
+      "step": 3398
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.20110282404068208,
+      "learning_rate": 8.373546978524288e-06,
+      "loss": 0.0358,
+      "step": 3399
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.24467207764729795,
+      "learning_rate": 8.372539660794907e-06,
+      "loss": 0.0649,
+      "step": 3400
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2879025959924376,
+      "learning_rate": 8.371532091859123e-06,
+      "loss": 0.0589,
+      "step": 3401
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.39648550462949134,
+      "learning_rate": 8.37052427179199e-06,
+      "loss": 0.1069,
+      "step": 3402
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.39206689623802476,
+      "learning_rate": 8.369516200668574e-06,
+      "loss": 0.0887,
+      "step": 3403
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3661587965246451,
+      "learning_rate": 8.368507878563963e-06,
+      "loss": 0.0656,
+      "step": 3404
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8021457762769479,
+      "learning_rate": 8.367499305553261e-06,
+      "loss": 0.1068,
+      "step": 3405
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5576168217332138,
+      "learning_rate": 8.366490481711594e-06,
+      "loss": 0.0585,
+      "step": 3406
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.42226592023703713,
+      "learning_rate": 8.365481407114104e-06,
+      "loss": 0.1144,
+      "step": 3407
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.38381436000033164,
+      "learning_rate": 8.364472081835955e-06,
+      "loss": 0.1067,
+      "step": 3408
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.42376842507324347,
+      "learning_rate": 8.363462505952325e-06,
+      "loss": 0.0711,
+      "step": 3409
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3744793545771979,
+      "learning_rate": 8.362452679538415e-06,
+      "loss": 0.1398,
+      "step": 3410
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.424122164805052,
+      "learning_rate": 8.36144260266944e-06,
+      "loss": 0.119,
+      "step": 3411
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.40064419966549436,
+      "learning_rate": 8.36043227542064e-06,
+      "loss": 0.0956,
+      "step": 3412
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5196394155545212,
+      "learning_rate": 8.359421697867266e-06,
+      "loss": 0.1166,
+      "step": 3413
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3187101905279199,
+      "learning_rate": 8.358410870084595e-06,
+      "loss": 0.0723,
+      "step": 3414
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.38243723969580906,
+      "learning_rate": 8.357399792147917e-06,
+      "loss": 0.1356,
+      "step": 3415
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.28558225328996106,
+      "learning_rate": 8.356388464132546e-06,
+      "loss": 0.0936,
+      "step": 3416
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.268210303968939,
+      "learning_rate": 8.35537688611381e-06,
+      "loss": 0.0905,
+      "step": 3417
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3855491479870864,
+      "learning_rate": 8.354365058167055e-06,
+      "loss": 0.1107,
+      "step": 3418
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6535339478613275,
+      "learning_rate": 8.353352980367653e-06,
+      "loss": 0.1088,
+      "step": 3419
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.32860708361758817,
+      "learning_rate": 8.352340652790982e-06,
+      "loss": 0.0565,
+      "step": 3420
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6543902840969773,
+      "learning_rate": 8.351328075512454e-06,
+      "loss": 0.1378,
+      "step": 3421
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.30560326658256287,
+      "learning_rate": 8.350315248607485e-06,
+      "loss": 0.0918,
+      "step": 3422
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3127499532125756,
+      "learning_rate": 8.34930217215152e-06,
+      "loss": 0.0786,
+      "step": 3423
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.34923588434775715,
+      "learning_rate": 8.34828884622002e-06,
+      "loss": 0.0773,
+      "step": 3424
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.36858274709077565,
+      "learning_rate": 8.347275270888462e-06,
+      "loss": 0.0954,
+      "step": 3425
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.20811523959502642,
+      "learning_rate": 8.34626144623234e-06,
+      "loss": 0.0496,
+      "step": 3426
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3158100457855755,
+      "learning_rate": 8.345247372327174e-06,
+      "loss": 0.0809,
+      "step": 3427
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.27452373484084475,
+      "learning_rate": 8.344233049248495e-06,
+      "loss": 0.0637,
+      "step": 3428
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2608608757382667,
+      "learning_rate": 8.343218477071857e-06,
+      "loss": 0.0793,
+      "step": 3429
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2680826696313099,
+      "learning_rate": 8.342203655872832e-06,
+      "loss": 0.0852,
+      "step": 3430
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.32596781746935116,
+      "learning_rate": 8.341188585727006e-06,
+      "loss": 0.0928,
+      "step": 3431
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.37439402604427874,
+      "learning_rate": 8.340173266709994e-06,
+      "loss": 0.0581,
+      "step": 3432
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3936628481772283,
+      "learning_rate": 8.339157698897417e-06,
+      "loss": 0.077,
+      "step": 3433
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.34204410315491285,
+      "learning_rate": 8.338141882364924e-06,
+      "loss": 0.1201,
+      "step": 3434
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2479212433040463,
+      "learning_rate": 8.337125817188177e-06,
+      "loss": 0.0652,
+      "step": 3435
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.31367139096097785,
+      "learning_rate": 8.336109503442858e-06,
+      "loss": 0.0941,
+      "step": 3436
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3669283433980056,
+      "learning_rate": 8.335092941204668e-06,
+      "loss": 0.1177,
+      "step": 3437
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.30490197760017007,
+      "learning_rate": 8.33407613054933e-06,
+      "loss": 0.0925,
+      "step": 3438
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.43609424706837446,
+      "learning_rate": 8.333059071552579e-06,
+      "loss": 0.0962,
+      "step": 3439
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3523186587277727,
+      "learning_rate": 8.332041764290171e-06,
+      "loss": 0.0867,
+      "step": 3440
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.25415591107423124,
+      "learning_rate": 8.331024208837882e-06,
+      "loss": 0.0555,
+      "step": 3441
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.36675167917227486,
+      "learning_rate": 8.330006405271504e-06,
+      "loss": 0.043,
+      "step": 3442
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5277467307141047,
+      "learning_rate": 8.328988353666852e-06,
+      "loss": 0.1209,
+      "step": 3443
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.21321544284063657,
+      "learning_rate": 8.327970054099754e-06,
+      "loss": 0.0516,
+      "step": 3444
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.35728047458352996,
+      "learning_rate": 8.32695150664606e-06,
+      "loss": 0.1053,
+      "step": 3445
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.32190113084790645,
+      "learning_rate": 8.325932711381636e-06,
+      "loss": 0.1214,
+      "step": 3446
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.27364183733812764,
+      "learning_rate": 8.324913668382368e-06,
+      "loss": 0.0831,
+      "step": 3447
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3117275454118977,
+      "learning_rate": 8.323894377724163e-06,
+      "loss": 0.1122,
+      "step": 3448
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.40745333124696304,
+      "learning_rate": 8.322874839482941e-06,
+      "loss": 0.1058,
+      "step": 3449
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3805713969100236,
+      "learning_rate": 8.321855053734643e-06,
+      "loss": 0.1098,
+      "step": 3450
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.432110539569376,
+      "learning_rate": 8.32083502055523e-06,
+      "loss": 0.0969,
+      "step": 3451
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5270207021297184,
+      "learning_rate": 8.319814740020678e-06,
+      "loss": 0.1184,
+      "step": 3452
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4031304466773929,
+      "learning_rate": 8.318794212206986e-06,
+      "loss": 0.0474,
+      "step": 3453
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6952881550734994,
+      "learning_rate": 8.317773437190169e-06,
+      "loss": 0.1224,
+      "step": 3454
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.39883864374984407,
+      "learning_rate": 8.316752415046258e-06,
+      "loss": 0.1101,
+      "step": 3455
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.20566967699171593,
+      "learning_rate": 8.315731145851305e-06,
+      "loss": 0.0626,
+      "step": 3456
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3322982673747095,
+      "learning_rate": 8.31470962968138e-06,
+      "loss": 0.0628,
+      "step": 3457
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2523443883086348,
+      "learning_rate": 8.313687866612574e-06,
+      "loss": 0.094,
+      "step": 3458
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.40750634059975216,
+      "learning_rate": 8.31266585672099e-06,
+      "loss": 0.1194,
+      "step": 3459
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.21047445463151845,
+      "learning_rate": 8.311643600082758e-06,
+      "loss": 0.069,
+      "step": 3460
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.18731140143254316,
+      "learning_rate": 8.310621096774016e-06,
+      "loss": 0.0535,
+      "step": 3461
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5631463785512291,
+      "learning_rate": 8.309598346870931e-06,
+      "loss": 0.0901,
+      "step": 3462
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.48073372924624547,
+      "learning_rate": 8.308575350449679e-06,
+      "loss": 0.1201,
+      "step": 3463
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5515112312867327,
+      "learning_rate": 8.307552107586463e-06,
+      "loss": 0.1343,
+      "step": 3464
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.30658532689762324,
+      "learning_rate": 8.306528618357498e-06,
+      "loss": 0.0635,
+      "step": 3465
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.26958992981370533,
+      "learning_rate": 8.30550488283902e-06,
+      "loss": 0.0983,
+      "step": 3466
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.24971534817696114,
+      "learning_rate": 8.304480901107282e-06,
+      "loss": 0.0463,
+      "step": 3467
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.36192954983787184,
+      "learning_rate": 8.303456673238555e-06,
+      "loss": 0.1318,
+      "step": 3468
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.26012356317526036,
+      "learning_rate": 8.302432199309132e-06,
+      "loss": 0.0964,
+      "step": 3469
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.22578942833083543,
+      "learning_rate": 8.30140747939532e-06,
+      "loss": 0.0751,
+      "step": 3470
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3906385308712652,
+      "learning_rate": 8.300382513573447e-06,
+      "loss": 0.0834,
+      "step": 3471
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.25919778751735095,
+      "learning_rate": 8.299357301919859e-06,
+      "loss": 0.056,
+      "step": 3472
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3045165011962118,
+      "learning_rate": 8.298331844510918e-06,
+      "loss": 0.0811,
+      "step": 3473
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.25508796042251275,
+      "learning_rate": 8.297306141423009e-06,
+      "loss": 0.0684,
+      "step": 3474
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3406187460622309,
+      "learning_rate": 8.296280192732529e-06,
+      "loss": 0.123,
+      "step": 3475
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.39811895158647004,
+      "learning_rate": 8.295253998515897e-06,
+      "loss": 0.0992,
+      "step": 3476
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.28772373208319796,
+      "learning_rate": 8.294227558849553e-06,
+      "loss": 0.0981,
+      "step": 3477
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.43034726107327487,
+      "learning_rate": 8.293200873809948e-06,
+      "loss": 0.1163,
+      "step": 3478
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3795166731517897,
+      "learning_rate": 8.29217394347356e-06,
+      "loss": 0.0828,
+      "step": 3479
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.26059099469733243,
+      "learning_rate": 8.291146767916875e-06,
+      "loss": 0.0819,
+      "step": 3480
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.26233130541654637,
+      "learning_rate": 8.29011934721641e-06,
+      "loss": 0.0412,
+      "step": 3481
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3730884760728986,
+      "learning_rate": 8.289091681448688e-06,
+      "loss": 0.0904,
+      "step": 3482
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.35183852015895173,
+      "learning_rate": 8.288063770690257e-06,
+      "loss": 0.0987,
+      "step": 3483
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6369365079551031,
+      "learning_rate": 8.287035615017682e-06,
+      "loss": 0.1072,
+      "step": 3484
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2618042624267544,
+      "learning_rate": 8.286007214507547e-06,
+      "loss": 0.0632,
+      "step": 3485
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.24961948529249958,
+      "learning_rate": 8.28497856923645e-06,
+      "loss": 0.0738,
+      "step": 3486
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.35731559779365896,
+      "learning_rate": 8.283949679281014e-06,
+      "loss": 0.1242,
+      "step": 3487
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3841415219010094,
+      "learning_rate": 8.282920544717876e-06,
+      "loss": 0.1144,
+      "step": 3488
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3050665812414802,
+      "learning_rate": 8.281891165623693e-06,
+      "loss": 0.0651,
+      "step": 3489
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.34841682970388693,
+      "learning_rate": 8.280861542075134e-06,
+      "loss": 0.0901,
+      "step": 3490
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3746083592924003,
+      "learning_rate": 8.279831674148898e-06,
+      "loss": 0.1214,
+      "step": 3491
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.28994739762570476,
+      "learning_rate": 8.27880156192169e-06,
+      "loss": 0.0745,
+      "step": 3492
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.40156534165780644,
+      "learning_rate": 8.277771205470242e-06,
+      "loss": 0.0965,
+      "step": 3493
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4694312903797416,
+      "learning_rate": 8.2767406048713e-06,
+      "loss": 0.1042,
+      "step": 3494
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2651029565468178,
+      "learning_rate": 8.27570976020163e-06,
+      "loss": 0.0955,
+      "step": 3495
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.36348285242495476,
+      "learning_rate": 8.274678671538014e-06,
+      "loss": 0.078,
+      "step": 3496
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.35051604730621977,
+      "learning_rate": 8.273647338957256e-06,
+      "loss": 0.0678,
+      "step": 3497
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.36564406133536825,
+      "learning_rate": 8.272615762536171e-06,
+      "loss": 0.0883,
+      "step": 3498
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.25322654230355174,
+      "learning_rate": 8.271583942351602e-06,
+      "loss": 0.0714,
+      "step": 3499
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3586927081095055,
+      "learning_rate": 8.270551878480402e-06,
+      "loss": 0.0849,
+      "step": 3500
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3027782906153767,
+      "learning_rate": 8.269519570999444e-06,
+      "loss": 0.0771,
+      "step": 3501
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4608092505711324,
+      "learning_rate": 8.268487019985624e-06,
+      "loss": 0.0808,
+      "step": 3502
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2735354911387808,
+      "learning_rate": 8.267454225515848e-06,
+      "loss": 0.0994,
+      "step": 3503
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.304705067909962,
+      "learning_rate": 8.26642118766705e-06,
+      "loss": 0.0794,
+      "step": 3504
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3965411136038813,
+      "learning_rate": 8.265387906516171e-06,
+      "loss": 0.0676,
+      "step": 3505
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4386935815845951,
+      "learning_rate": 8.26435438214018e-06,
+      "loss": 0.1038,
+      "step": 3506
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.25239037871008524,
+      "learning_rate": 8.263320614616059e-06,
+      "loss": 0.0731,
+      "step": 3507
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.30296284753705743,
+      "learning_rate": 8.262286604020808e-06,
+      "loss": 0.0811,
+      "step": 3508
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3138322538743376,
+      "learning_rate": 8.261252350431446e-06,
+      "loss": 0.0718,
+      "step": 3509
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.19164499727840723,
+      "learning_rate": 8.26021785392501e-06,
+      "loss": 0.0447,
+      "step": 3510
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.34125096007889766,
+      "learning_rate": 8.259183114578556e-06,
+      "loss": 0.0892,
+      "step": 3511
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3341500671102026,
+      "learning_rate": 8.258148132469157e-06,
+      "loss": 0.082,
+      "step": 3512
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.46559353871791576,
+      "learning_rate": 8.257112907673908e-06,
+      "loss": 0.1557,
+      "step": 3513
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.22963138197462746,
+      "learning_rate": 8.256077440269912e-06,
+      "loss": 0.0549,
+      "step": 3514
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3097432695846166,
+      "learning_rate": 8.255041730334303e-06,
+      "loss": 0.0907,
+      "step": 3515
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2966561078055357,
+      "learning_rate": 8.254005777944223e-06,
+      "loss": 0.1031,
+      "step": 3516
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.20922882769907816,
+      "learning_rate": 8.252969583176837e-06,
+      "loss": 0.0647,
+      "step": 3517
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2843670838712974,
+      "learning_rate": 8.251933146109326e-06,
+      "loss": 0.0813,
+      "step": 3518
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.35062437946849434,
+      "learning_rate": 8.25089646681889e-06,
+      "loss": 0.054,
+      "step": 3519
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.28324491593157763,
+      "learning_rate": 8.249859545382748e-06,
+      "loss": 0.0886,
+      "step": 3520
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.32046533788374665,
+      "learning_rate": 8.248822381878135e-06,
+      "loss": 0.0854,
+      "step": 3521
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4422833536255353,
+      "learning_rate": 8.247784976382306e-06,
+      "loss": 0.0711,
+      "step": 3522
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.19907422451738827,
+      "learning_rate": 8.24674732897253e-06,
+      "loss": 0.0651,
+      "step": 3523
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.23001856860185055,
+      "learning_rate": 8.2457094397261e-06,
+      "loss": 0.0659,
+      "step": 3524
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.400343839236898,
+      "learning_rate": 8.244671308720324e-06,
+      "loss": 0.0754,
+      "step": 3525
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.41407416669775604,
+      "learning_rate": 8.243632936032526e-06,
+      "loss": 0.0657,
+      "step": 3526
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4794970798554485,
+      "learning_rate": 8.242594321740052e-06,
+      "loss": 0.0884,
+      "step": 3527
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.562444624737108,
+      "learning_rate": 8.241555465920265e-06,
+      "loss": 0.1285,
+      "step": 3528
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.242208397891645,
+      "learning_rate": 8.240516368650542e-06,
+      "loss": 0.0535,
+      "step": 3529
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.31719335091294476,
+      "learning_rate": 8.239477030008284e-06,
+      "loss": 0.0807,
+      "step": 3530
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.40360204486442014,
+      "learning_rate": 8.238437450070904e-06,
+      "loss": 0.0742,
+      "step": 3531
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.23549210019814987,
+      "learning_rate": 8.237397628915838e-06,
+      "loss": 0.0627,
+      "step": 3532
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.27018376168501396,
+      "learning_rate": 8.236357566620537e-06,
+      "loss": 0.0544,
+      "step": 3533
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.36156645275346017,
+      "learning_rate": 8.23531726326247e-06,
+      "loss": 0.0974,
+      "step": 3534
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3345352689895445,
+      "learning_rate": 8.234276718919127e-06,
+      "loss": 0.0818,
+      "step": 3535
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2747257354504063,
+      "learning_rate": 8.233235933668014e-06,
+      "loss": 0.0488,
+      "step": 3536
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.27251394631729614,
+      "learning_rate": 8.232194907586652e-06,
+      "loss": 0.087,
+      "step": 3537
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.24365137952845387,
+      "learning_rate": 8.231153640752588e-06,
+      "loss": 0.0876,
+      "step": 3538
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.34021744218146943,
+      "learning_rate": 8.230112133243374e-06,
+      "loss": 0.0611,
+      "step": 3539
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5712587597484108,
+      "learning_rate": 8.229070385136593e-06,
+      "loss": 0.0825,
+      "step": 3540
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2661618508196568,
+      "learning_rate": 8.22802839650984e-06,
+      "loss": 0.0883,
+      "step": 3541
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.22350074367142747,
+      "learning_rate": 8.226986167440725e-06,
+      "loss": 0.0521,
+      "step": 3542
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3454542046527263,
+      "learning_rate": 8.225943698006883e-06,
+      "loss": 0.1,
+      "step": 3543
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.23772672229894337,
+      "learning_rate": 8.22490098828596e-06,
+      "loss": 0.0603,
+      "step": 3544
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2369100399154882,
+      "learning_rate": 8.223858038355628e-06,
+      "loss": 0.0571,
+      "step": 3545
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8091001582078899,
+      "learning_rate": 8.222814848293568e-06,
+      "loss": 0.1457,
+      "step": 3546
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.665874188650964,
+      "learning_rate": 8.221771418177483e-06,
+      "loss": 0.1688,
+      "step": 3547
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.41061806825129316,
+      "learning_rate": 8.220727748085096e-06,
+      "loss": 0.0957,
+      "step": 3548
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2206177184196639,
+      "learning_rate": 8.219683838094143e-06,
+      "loss": 0.0595,
+      "step": 3549
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4672874731214148,
+      "learning_rate": 8.21863968828238e-06,
+      "loss": 0.1043,
+      "step": 3550
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6024646838735105,
+      "learning_rate": 8.217595298727586e-06,
+      "loss": 0.123,
+      "step": 3551
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.30994340675601734,
+      "learning_rate": 8.216550669507549e-06,
+      "loss": 0.0525,
+      "step": 3552
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.33669491623773484,
+      "learning_rate": 8.21550580070008e-06,
+      "loss": 0.0951,
+      "step": 3553
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.23209858450625856,
+      "learning_rate": 8.214460692383008e-06,
+      "loss": 0.067,
+      "step": 3554
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.30621287286132703,
+      "learning_rate": 8.213415344634178e-06,
+      "loss": 0.0846,
+      "step": 3555
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2740132163539907,
+      "learning_rate": 8.212369757531453e-06,
+      "loss": 0.0746,
+      "step": 3556
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3725701253001791,
+      "learning_rate": 8.211323931152718e-06,
+      "loss": 0.0817,
+      "step": 3557
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3150927742638896,
+      "learning_rate": 8.210277865575866e-06,
+      "loss": 0.0626,
+      "step": 3558
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.25094659183947016,
+      "learning_rate": 8.209231560878818e-06,
+      "loss": 0.0777,
+      "step": 3559
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.43274462771994365,
+      "learning_rate": 8.208185017139508e-06,
+      "loss": 0.1107,
+      "step": 3560
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3118682713527896,
+      "learning_rate": 8.20713823443589e-06,
+      "loss": 0.0652,
+      "step": 3561
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.21452713826020006,
+      "learning_rate": 8.206091212845932e-06,
+      "loss": 0.047,
+      "step": 3562
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.19473713711449442,
+      "learning_rate": 8.205043952447622e-06,
+      "loss": 0.0601,
+      "step": 3563
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5717441589621339,
+      "learning_rate": 8.203996453318971e-06,
+      "loss": 0.0765,
+      "step": 3564
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.23739584595384405,
+      "learning_rate": 8.202948715537998e-06,
+      "loss": 0.0574,
+      "step": 3565
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.33774805467086577,
+      "learning_rate": 8.201900739182746e-06,
+      "loss": 0.0985,
+      "step": 3566
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.23469365327432187,
+      "learning_rate": 8.200852524331275e-06,
+      "loss": 0.0811,
+      "step": 3567
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.704950679816774,
+      "learning_rate": 8.199804071061661e-06,
+      "loss": 0.0822,
+      "step": 3568
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2774618020586867,
+      "learning_rate": 8.198755379452e-06,
+      "loss": 0.0867,
+      "step": 3569
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3011726794268524,
+      "learning_rate": 8.197706449580405e-06,
+      "loss": 0.1231,
+      "step": 3570
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.26096145580664165,
+      "learning_rate": 8.196657281525005e-06,
+      "loss": 0.0468,
+      "step": 3571
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.78419528255088,
+      "learning_rate": 8.195607875363949e-06,
+      "loss": 0.0874,
+      "step": 3572
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.29915199365948836,
+      "learning_rate": 8.194558231175401e-06,
+      "loss": 0.096,
+      "step": 3573
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.19783771611164497,
+      "learning_rate": 8.193508349037548e-06,
+      "loss": 0.0611,
+      "step": 3574
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.21091219727055546,
+      "learning_rate": 8.19245822902859e-06,
+      "loss": 0.0659,
+      "step": 3575
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.22506974299469842,
+      "learning_rate": 8.191407871226744e-06,
+      "loss": 0.0512,
+      "step": 3576
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.21873567408392205,
+      "learning_rate": 8.190357275710249e-06,
+      "loss": 0.048,
+      "step": 3577
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.41682267051210686,
+      "learning_rate": 8.189306442557359e-06,
+      "loss": 0.0959,
+      "step": 3578
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.391513123048695,
+      "learning_rate": 8.188255371846347e-06,
+      "loss": 0.0933,
+      "step": 3579
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3307180489648822,
+      "learning_rate": 8.187204063655501e-06,
+      "loss": 0.1214,
+      "step": 3580
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3335712656891786,
+      "learning_rate": 8.186152518063128e-06,
+      "loss": 0.0626,
+      "step": 3581
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2570347551445168,
+      "learning_rate": 8.185100735147558e-06,
+      "loss": 0.0808,
+      "step": 3582
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.27483236727369664,
+      "learning_rate": 8.184048714987129e-06,
+      "loss": 0.0829,
+      "step": 3583
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3363792797239031,
+      "learning_rate": 8.182996457660202e-06,
+      "loss": 0.0486,
+      "step": 3584
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3243358621086891,
+      "learning_rate": 8.181943963245157e-06,
+      "loss": 0.0753,
+      "step": 3585
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.20914061746171497,
+      "learning_rate": 8.180891231820392e-06,
+      "loss": 0.0506,
+      "step": 3586
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.23723633120098409,
+      "learning_rate": 8.179838263464316e-06,
+      "loss": 0.0882,
+      "step": 3587
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.37242300880644585,
+      "learning_rate": 8.178785058255363e-06,
+      "loss": 0.0885,
+      "step": 3588
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5209478834628155,
+      "learning_rate": 8.17773161627198e-06,
+      "loss": 0.1358,
+      "step": 3589
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4079965886472989,
+      "learning_rate": 8.176677937592634e-06,
+      "loss": 0.0767,
+      "step": 3590
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.20742024159347838,
+      "learning_rate": 8.175624022295812e-06,
+      "loss": 0.0457,
+      "step": 3591
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.393137049734039,
+      "learning_rate": 8.174569870460012e-06,
+      "loss": 0.0696,
+      "step": 3592
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.47523286203037013,
+      "learning_rate": 8.173515482163756e-06,
+      "loss": 0.1363,
+      "step": 3593
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.26323421201168046,
+      "learning_rate": 8.172460857485578e-06,
+      "loss": 0.0739,
+      "step": 3594
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6857273593503733,
+      "learning_rate": 8.171405996504033e-06,
+      "loss": 0.1454,
+      "step": 3595
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3201587353551199,
+      "learning_rate": 8.1703508992977e-06,
+      "loss": 0.0606,
+      "step": 3596
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3144356367893049,
+      "learning_rate": 8.169295565945158e-06,
+      "loss": 0.0935,
+      "step": 3597
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4174814452084571,
+      "learning_rate": 8.168239996525022e-06,
+      "loss": 0.1347,
+      "step": 3598
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3661488273990816,
+      "learning_rate": 8.167184191115915e-06,
+      "loss": 0.0919,
+      "step": 3599
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.38804316550183937,
+      "learning_rate": 8.16612814979648e-06,
+      "loss": 0.0923,
+      "step": 3600
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3199892317272469,
+      "learning_rate": 8.165071872645373e-06,
+      "loss": 0.0927,
+      "step": 3601
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2893005194217812,
+      "learning_rate": 8.164015359741277e-06,
+      "loss": 0.0981,
+      "step": 3602
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3129566487938126,
+      "learning_rate": 8.162958611162886e-06,
+      "loss": 0.0413,
+      "step": 3603
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2835923847409358,
+      "learning_rate": 8.161901626988912e-06,
+      "loss": 0.0681,
+      "step": 3604
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3643393212368736,
+      "learning_rate": 8.160844407298085e-06,
+      "loss": 0.0784,
+      "step": 3605
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2981156765135449,
+      "learning_rate": 8.159786952169153e-06,
+      "loss": 0.086,
+      "step": 3606
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.37979050067759085,
+      "learning_rate": 8.158729261680881e-06,
+      "loss": 0.0879,
+      "step": 3607
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.19201077150890355,
+      "learning_rate": 8.157671335912054e-06,
+      "loss": 0.0425,
+      "step": 3608
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5772767501716746,
+      "learning_rate": 8.15661317494147e-06,
+      "loss": 0.1064,
+      "step": 3609
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6341318181437655,
+      "learning_rate": 8.15555477884795e-06,
+      "loss": 0.076,
+      "step": 3610
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.47905500146513535,
+      "learning_rate": 8.154496147710325e-06,
+      "loss": 0.1207,
+      "step": 3611
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4694424233853518,
+      "learning_rate": 8.153437281607452e-06,
+      "loss": 0.1391,
+      "step": 3612
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.28938582919678285,
+      "learning_rate": 8.1523781806182e-06,
+      "loss": 0.06,
+      "step": 3613
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4116984259175424,
+      "learning_rate": 8.151318844821456e-06,
+      "loss": 0.0856,
+      "step": 3614
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.23732564409107743,
+      "learning_rate": 8.150259274296128e-06,
+      "loss": 0.0777,
+      "step": 3615
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4813275978794444,
+      "learning_rate": 8.149199469121138e-06,
+      "loss": 0.0826,
+      "step": 3616
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4143177241081707,
+      "learning_rate": 8.148139429375425e-06,
+      "loss": 0.0945,
+      "step": 3617
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3283683014729715,
+      "learning_rate": 8.14707915513795e-06,
+      "loss": 0.0573,
+      "step": 3618
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.22316466148213573,
+      "learning_rate": 8.146018646487684e-06,
+      "loss": 0.0755,
+      "step": 3619
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3678202858659283,
+      "learning_rate": 8.144957903503622e-06,
+      "loss": 0.0887,
+      "step": 3620
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1949478289167243,
+      "learning_rate": 8.143896926264778e-06,
+      "loss": 0.0538,
+      "step": 3621
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2561366638630743,
+      "learning_rate": 8.142835714850173e-06,
+      "loss": 0.0721,
+      "step": 3622
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3465086359544568,
+      "learning_rate": 8.141774269338858e-06,
+      "loss": 0.1099,
+      "step": 3623
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3747047550201552,
+      "learning_rate": 8.140712589809891e-06,
+      "loss": 0.0824,
+      "step": 3624
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2795556229996969,
+      "learning_rate": 8.139650676342357e-06,
+      "loss": 0.0865,
+      "step": 3625
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3336606853708106,
+      "learning_rate": 8.138588529015348e-06,
+      "loss": 0.0932,
+      "step": 3626
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.24890182717019194,
+      "learning_rate": 8.137526147907984e-06,
+      "loss": 0.0474,
+      "step": 3627
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5590284713264075,
+      "learning_rate": 8.136463533099392e-06,
+      "loss": 0.1092,
+      "step": 3628
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.283909236953108,
+      "learning_rate": 8.135400684668727e-06,
+      "loss": 0.078,
+      "step": 3629
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.22304138643966928,
+      "learning_rate": 8.134337602695154e-06,
+      "loss": 0.0485,
+      "step": 3630
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5463154782975195,
+      "learning_rate": 8.133274287257857e-06,
+      "loss": 0.0474,
+      "step": 3631
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.33949702524007497,
+      "learning_rate": 8.13221073843604e-06,
+      "loss": 0.0753,
+      "step": 3632
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2104648296125629,
+      "learning_rate": 8.131146956308917e-06,
+      "loss": 0.0494,
+      "step": 3633
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.34143602421276253,
+      "learning_rate": 8.130082940955733e-06,
+      "loss": 0.0876,
+      "step": 3634
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2534738497041492,
+      "learning_rate": 8.129018692455735e-06,
+      "loss": 0.0637,
+      "step": 3635
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3284661252711693,
+      "learning_rate": 8.127954210888197e-06,
+      "loss": 0.1019,
+      "step": 3636
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.26153305852086406,
+      "learning_rate": 8.126889496332408e-06,
+      "loss": 0.0668,
+      "step": 3637
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.48344683443548797,
+      "learning_rate": 8.125824548867673e-06,
+      "loss": 0.0757,
+      "step": 3638
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3629083209154181,
+      "learning_rate": 8.124759368573316e-06,
+      "loss": 0.1273,
+      "step": 3639
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.27587691975975437,
+      "learning_rate": 8.12369395552868e-06,
+      "loss": 0.0627,
+      "step": 3640
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2409562128772796,
+      "learning_rate": 8.122628309813121e-06,
+      "loss": 0.0748,
+      "step": 3641
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.27847207017261966,
+      "learning_rate": 8.121562431506013e-06,
+      "loss": 0.0678,
+      "step": 3642
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.24451562099210608,
+      "learning_rate": 8.120496320686752e-06,
+      "loss": 0.0614,
+      "step": 3643
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.31305820048668476,
+      "learning_rate": 8.119429977434747e-06,
+      "loss": 0.1057,
+      "step": 3644
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.30697942267223816,
+      "learning_rate": 8.118363401829425e-06,
+      "loss": 0.0781,
+      "step": 3645
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.20905448528153592,
+      "learning_rate": 8.117296593950231e-06,
+      "loss": 0.047,
+      "step": 3646
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3511743706260753,
+      "learning_rate": 8.116229553876628e-06,
+      "loss": 0.0664,
+      "step": 3647
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3776031318940993,
+      "learning_rate": 8.115162281688092e-06,
+      "loss": 0.1237,
+      "step": 3648
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5453083053449695,
+      "learning_rate": 8.114094777464125e-06,
+      "loss": 0.1156,
+      "step": 3649
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.24765382057948584,
+      "learning_rate": 8.113027041284236e-06,
+      "loss": 0.0866,
+      "step": 3650
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.27328041672248937,
+      "learning_rate": 8.111959073227959e-06,
+      "loss": 0.0499,
+      "step": 3651
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.30793815360314775,
+      "learning_rate": 8.110890873374842e-06,
+      "loss": 0.0507,
+      "step": 3652
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.24849401989727152,
+      "learning_rate": 8.10982244180445e-06,
+      "loss": 0.1009,
+      "step": 3653
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.30482750278982224,
+      "learning_rate": 8.108753778596367e-06,
+      "loss": 0.0795,
+      "step": 3654
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.17556756459036374,
+      "learning_rate": 8.107684883830193e-06,
+      "loss": 0.0798,
+      "step": 3655
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.18133493124607133,
+      "learning_rate": 8.106615757585544e-06,
+      "loss": 0.058,
+      "step": 3656
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.45069641238224784,
+      "learning_rate": 8.105546399942057e-06,
+      "loss": 0.0814,
+      "step": 3657
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.22320412028183,
+      "learning_rate": 8.104476810979382e-06,
+      "loss": 0.069,
+      "step": 3658
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.20819664079902336,
+      "learning_rate": 8.103406990777189e-06,
+      "loss": 0.0455,
+      "step": 3659
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.48140303058035505,
+      "learning_rate": 8.102336939415167e-06,
+      "loss": 0.0897,
+      "step": 3660
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4209688726266939,
+      "learning_rate": 8.101266656973015e-06,
+      "loss": 0.1058,
+      "step": 3661
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3292900339491034,
+      "learning_rate": 8.100196143530456e-06,
+      "loss": 0.0818,
+      "step": 3662
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3747237462560882,
+      "learning_rate": 8.099125399167231e-06,
+      "loss": 0.1113,
+      "step": 3663
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4722430355332535,
+      "learning_rate": 8.098054423963089e-06,
+      "loss": 0.1183,
+      "step": 3664
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.20359783999502326,
+      "learning_rate": 8.096983217997808e-06,
+      "loss": 0.0336,
+      "step": 3665
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.40177786233970975,
+      "learning_rate": 8.095911781351175e-06,
+      "loss": 0.084,
+      "step": 3666
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2807103633055263,
+      "learning_rate": 8.094840114102999e-06,
+      "loss": 0.0695,
+      "step": 3667
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6461125023952256,
+      "learning_rate": 8.093768216333102e-06,
+      "loss": 0.1055,
+      "step": 3668
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.38947143382002863,
+      "learning_rate": 8.092696088121324e-06,
+      "loss": 0.0941,
+      "step": 3669
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3333353050447398,
+      "learning_rate": 8.091623729547526e-06,
+      "loss": 0.0494,
+      "step": 3670
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2086566818308642,
+      "learning_rate": 8.090551140691582e-06,
+      "loss": 0.0512,
+      "step": 3671
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4242079701764188,
+      "learning_rate": 8.089478321633386e-06,
+      "loss": 0.1267,
+      "step": 3672
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5420750095291259,
+      "learning_rate": 8.088405272452848e-06,
+      "loss": 0.1095,
+      "step": 3673
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4147364802054457,
+      "learning_rate": 8.087331993229895e-06,
+      "loss": 0.0875,
+      "step": 3674
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.39723798537863864,
+      "learning_rate": 8.086258484044466e-06,
+      "loss": 0.1235,
+      "step": 3675
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.40117143883772277,
+      "learning_rate": 8.08518474497653e-06,
+      "loss": 0.0873,
+      "step": 3676
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3959386515439047,
+      "learning_rate": 8.08411077610606e-06,
+      "loss": 0.063,
+      "step": 3677
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.27169115320492493,
+      "learning_rate": 8.083036577513052e-06,
+      "loss": 0.0541,
+      "step": 3678
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2926125995064389,
+      "learning_rate": 8.081962149277523e-06,
+      "loss": 0.0968,
+      "step": 3679
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.43539065859844706,
+      "learning_rate": 8.080887491479497e-06,
+      "loss": 0.0914,
+      "step": 3680
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.31477202594534875,
+      "learning_rate": 8.079812604199023e-06,
+      "loss": 0.0933,
+      "step": 3681
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2560898443670172,
+      "learning_rate": 8.078737487516164e-06,
+      "loss": 0.1258,
+      "step": 3682
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.295743862308518,
+      "learning_rate": 8.077662141511004e-06,
+      "loss": 0.0515,
+      "step": 3683
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.48089399395202886,
+      "learning_rate": 8.07658656626364e-06,
+      "loss": 0.1035,
+      "step": 3684
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.40529573665224317,
+      "learning_rate": 8.075510761854184e-06,
+      "loss": 0.0883,
+      "step": 3685
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.24327707616647226,
+      "learning_rate": 8.07443472836277e-06,
+      "loss": 0.0627,
+      "step": 3686
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3328996560533555,
+      "learning_rate": 8.073358465869549e-06,
+      "loss": 0.1126,
+      "step": 3687
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6477096064387154,
+      "learning_rate": 8.072281974454685e-06,
+      "loss": 0.1033,
+      "step": 3688
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.35847201575424137,
+      "learning_rate": 8.071205254198362e-06,
+      "loss": 0.0729,
+      "step": 3689
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.26547948699896057,
+      "learning_rate": 8.070128305180783e-06,
+      "loss": 0.0845,
+      "step": 3690
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.35094222777145845,
+      "learning_rate": 8.06905112748216e-06,
+      "loss": 0.0799,
+      "step": 3691
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6214925157759148,
+      "learning_rate": 8.06797372118273e-06,
+      "loss": 0.0904,
+      "step": 3692
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2412061053804197,
+      "learning_rate": 8.066896086362747e-06,
+      "loss": 0.0483,
+      "step": 3693
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5907969227458691,
+      "learning_rate": 8.065818223102477e-06,
+      "loss": 0.115,
+      "step": 3694
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.26826139996480475,
+      "learning_rate": 8.064740131482207e-06,
+      "loss": 0.0751,
+      "step": 3695
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4743967680165217,
+      "learning_rate": 8.063661811582237e-06,
+      "loss": 0.1316,
+      "step": 3696
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.493772496394193,
+      "learning_rate": 8.062583263482888e-06,
+      "loss": 0.0711,
+      "step": 3697
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2779520010329725,
+      "learning_rate": 8.061504487264498e-06,
+      "loss": 0.0748,
+      "step": 3698
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3040146023723784,
+      "learning_rate": 8.060425483007418e-06,
+      "loss": 0.0361,
+      "step": 3699
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.318086531917773,
+      "learning_rate": 8.05934625079202e-06,
+      "loss": 0.0767,
+      "step": 3700
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.28112501659550976,
+      "learning_rate": 8.058266790698693e-06,
+      "loss": 0.0712,
+      "step": 3701
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4194752463188414,
+      "learning_rate": 8.057187102807837e-06,
+      "loss": 0.0841,
+      "step": 3702
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2745856466850972,
+      "learning_rate": 8.056107187199877e-06,
+      "loss": 0.0683,
+      "step": 3703
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2939525718786237,
+      "learning_rate": 8.05502704395525e-06,
+      "loss": 0.0899,
+      "step": 3704
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.24799364574236718,
+      "learning_rate": 8.053946673154414e-06,
+      "loss": 0.0711,
+      "step": 3705
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.354084552460418,
+      "learning_rate": 8.052866074877839e-06,
+      "loss": 0.11,
+      "step": 3706
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4658330039787627,
+      "learning_rate": 8.051785249206013e-06,
+      "loss": 0.059,
+      "step": 3707
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.37496933240899205,
+      "learning_rate": 8.050704196219443e-06,
+      "loss": 0.0843,
+      "step": 3708
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.291499464635093,
+      "learning_rate": 8.049622915998655e-06,
+      "loss": 0.0558,
+      "step": 3709
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.33919935477158353,
+      "learning_rate": 8.048541408624186e-06,
+      "loss": 0.0856,
+      "step": 3710
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.21457262177588945,
+      "learning_rate": 8.047459674176592e-06,
+      "loss": 0.0767,
+      "step": 3711
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.22744239761159804,
+      "learning_rate": 8.046377712736453e-06,
+      "loss": 0.0838,
+      "step": 3712
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3779288965197235,
+      "learning_rate": 8.04529552438435e-06,
+      "loss": 0.1181,
+      "step": 3713
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.33942952897197093,
+      "learning_rate": 8.044213109200901e-06,
+      "loss": 0.0582,
+      "step": 3714
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.37548894951209105,
+      "learning_rate": 8.043130467266725e-06,
+      "loss": 0.0824,
+      "step": 3715
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2819441628175042,
+      "learning_rate": 8.04204759866246e-06,
+      "loss": 0.0805,
+      "step": 3716
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3187724688118607,
+      "learning_rate": 8.040964503468774e-06,
+      "loss": 0.1112,
+      "step": 3717
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4284914009235011,
+      "learning_rate": 8.039881181766334e-06,
+      "loss": 0.067,
+      "step": 3718
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5179409049099479,
+      "learning_rate": 8.038797633635836e-06,
+      "loss": 0.0994,
+      "step": 3719
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3073693328207546,
+      "learning_rate": 8.037713859157987e-06,
+      "loss": 0.0832,
+      "step": 3720
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4280695536533562,
+      "learning_rate": 8.036629858413514e-06,
+      "loss": 0.1194,
+      "step": 3721
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.32675315090568596,
+      "learning_rate": 8.03554563148316e-06,
+      "loss": 0.1209,
+      "step": 3722
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.31071896612432215,
+      "learning_rate": 8.034461178447683e-06,
+      "loss": 0.0868,
+      "step": 3723
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.40817108240019284,
+      "learning_rate": 8.033376499387862e-06,
+      "loss": 0.1357,
+      "step": 3724
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6003230305454093,
+      "learning_rate": 8.032291594384488e-06,
+      "loss": 0.1261,
+      "step": 3725
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.25260403102843826,
+      "learning_rate": 8.03120646351837e-06,
+      "loss": 0.0937,
+      "step": 3726
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.249582174415589,
+      "learning_rate": 8.030121106870339e-06,
+      "loss": 0.0872,
+      "step": 3727
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.26247104330602844,
+      "learning_rate": 8.029035524521234e-06,
+      "loss": 0.0811,
+      "step": 3728
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.43865107517533963,
+      "learning_rate": 8.02794971655192e-06,
+      "loss": 0.1098,
+      "step": 3729
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2891859854900654,
+      "learning_rate": 8.02686368304327e-06,
+      "loss": 0.0842,
+      "step": 3730
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4965339159025126,
+      "learning_rate": 8.02577742407618e-06,
+      "loss": 0.1356,
+      "step": 3731
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.29430826029829954,
+      "learning_rate": 8.024690939731563e-06,
+      "loss": 0.0832,
+      "step": 3732
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.16406307966210348,
+      "learning_rate": 8.023604230090342e-06,
+      "loss": 0.0339,
+      "step": 3733
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.40968631719028714,
+      "learning_rate": 8.022517295233467e-06,
+      "loss": 0.0925,
+      "step": 3734
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.33663372437045724,
+      "learning_rate": 8.021430135241895e-06,
+      "loss": 0.0763,
+      "step": 3735
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.32420294957211787,
+      "learning_rate": 8.020342750196605e-06,
+      "loss": 0.0731,
+      "step": 3736
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6422374175470135,
+      "learning_rate": 8.019255140178593e-06,
+      "loss": 0.107,
+      "step": 3737
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.41598252984729867,
+      "learning_rate": 8.018167305268869e-06,
+      "loss": 0.0766,
+      "step": 3738
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.32295314016272697,
+      "learning_rate": 8.017079245548462e-06,
+      "loss": 0.0703,
+      "step": 3739
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.26044425627338225,
+      "learning_rate": 8.015990961098419e-06,
+      "loss": 0.0538,
+      "step": 3740
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4061907312453345,
+      "learning_rate": 8.014902451999797e-06,
+      "loss": 0.1057,
+      "step": 3741
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6102667859978426,
+      "learning_rate": 8.013813718333679e-06,
+      "loss": 0.1603,
+      "step": 3742
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.46677668651824966,
+      "learning_rate": 8.012724760181158e-06,
+      "loss": 0.14,
+      "step": 3743
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5604786397016787,
+      "learning_rate": 8.011635577623346e-06,
+      "loss": 0.1216,
+      "step": 3744
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2799842894426531,
+      "learning_rate": 8.010546170741371e-06,
+      "loss": 0.0589,
+      "step": 3745
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.37356984686095,
+      "learning_rate": 8.009456539616383e-06,
+      "loss": 0.1057,
+      "step": 3746
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.33033500975523866,
+      "learning_rate": 8.008366684329537e-06,
+      "loss": 0.0874,
+      "step": 3747
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.23759675904431252,
+      "learning_rate": 8.007276604962018e-06,
+      "loss": 0.0525,
+      "step": 3748
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.28392591091699076,
+      "learning_rate": 8.006186301595015e-06,
+      "loss": 0.0888,
+      "step": 3749
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4213891765355426,
+      "learning_rate": 8.005095774309748e-06,
+      "loss": 0.0911,
+      "step": 3750
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.44643507668545834,
+      "learning_rate": 8.00400502318744e-06,
+      "loss": 0.0898,
+      "step": 3751
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.38592621563839047,
+      "learning_rate": 8.002914048309337e-06,
+      "loss": 0.1081,
+      "step": 3752
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2819134600796379,
+      "learning_rate": 8.001822849756706e-06,
+      "loss": 0.092,
+      "step": 3753
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2737352135456345,
+      "learning_rate": 8.00073142761082e-06,
+      "loss": 0.0697,
+      "step": 3754
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2904980900951311,
+      "learning_rate": 7.999639781952977e-06,
+      "loss": 0.1116,
+      "step": 3755
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3701877233316966,
+      "learning_rate": 7.998547912864487e-06,
+      "loss": 0.1094,
+      "step": 3756
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3663783360100223,
+      "learning_rate": 7.997455820426683e-06,
+      "loss": 0.0544,
+      "step": 3757
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3242944619684403,
+      "learning_rate": 7.996363504720908e-06,
+      "loss": 0.065,
+      "step": 3758
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.34193773516066184,
+      "learning_rate": 7.995270965828523e-06,
+      "loss": 0.0937,
+      "step": 3759
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.30047277951887874,
+      "learning_rate": 7.994178203830908e-06,
+      "loss": 0.0931,
+      "step": 3760
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2954148708980294,
+      "learning_rate": 7.993085218809458e-06,
+      "loss": 0.0752,
+      "step": 3761
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.43071383283306847,
+      "learning_rate": 7.991992010845585e-06,
+      "loss": 0.0976,
+      "step": 3762
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.27370587956260023,
+      "learning_rate": 7.990898580020718e-06,
+      "loss": 0.0892,
+      "step": 3763
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3618802701649216,
+      "learning_rate": 7.989804926416301e-06,
+      "loss": 0.1002,
+      "step": 3764
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.33916408424279826,
+      "learning_rate": 7.988711050113795e-06,
+      "loss": 0.0723,
+      "step": 3765
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.32785999220023543,
+      "learning_rate": 7.987616951194681e-06,
+      "loss": 0.0826,
+      "step": 3766
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2811981708809026,
+      "learning_rate": 7.986522629740452e-06,
+      "loss": 0.0784,
+      "step": 3767
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2751896598181098,
+      "learning_rate": 7.98542808583262e-06,
+      "loss": 0.0913,
+      "step": 3768
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.23900676243501434,
+      "learning_rate": 7.984333319552711e-06,
+      "loss": 0.0588,
+      "step": 3769
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3532744854124887,
+      "learning_rate": 7.983238330982274e-06,
+      "loss": 0.0991,
+      "step": 3770
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3650630600543345,
+      "learning_rate": 7.982143120202866e-06,
+      "loss": 0.0671,
+      "step": 3771
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.29389219374624903,
+      "learning_rate": 7.981047687296066e-06,
+      "loss": 0.0588,
+      "step": 3772
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.22012455743083292,
+      "learning_rate": 7.979952032343467e-06,
+      "loss": 0.0593,
+      "step": 3773
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2491490718056531,
+      "learning_rate": 7.978856155426684e-06,
+      "loss": 0.0586,
+      "step": 3774
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7935956973755018,
+      "learning_rate": 7.97776005662734e-06,
+      "loss": 0.1274,
+      "step": 3775
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4266731885008378,
+      "learning_rate": 7.976663736027082e-06,
+      "loss": 0.0967,
+      "step": 3776
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.35885999500722937,
+      "learning_rate": 7.975567193707565e-06,
+      "loss": 0.0793,
+      "step": 3777
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.33575430918281457,
+      "learning_rate": 7.974470429750473e-06,
+      "loss": 0.0761,
+      "step": 3778
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.27771340377330506,
+      "learning_rate": 7.973373444237493e-06,
+      "loss": 0.0862,
+      "step": 3779
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.30167685911801534,
+      "learning_rate": 7.97227623725034e-06,
+      "loss": 0.0695,
+      "step": 3780
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5315730211112671,
+      "learning_rate": 7.971178808870736e-06,
+      "loss": 0.0851,
+      "step": 3781
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.46850194183641813,
+      "learning_rate": 7.970081159180427e-06,
+      "loss": 0.1096,
+      "step": 3782
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3101169668685658,
+      "learning_rate": 7.968983288261171e-06,
+      "loss": 0.0978,
+      "step": 3783
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.36512434846685166,
+      "learning_rate": 7.967885196194744e-06,
+      "loss": 0.0675,
+      "step": 3784
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4697088769772414,
+      "learning_rate": 7.96678688306294e-06,
+      "loss": 0.0859,
+      "step": 3785
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.85916001771976,
+      "learning_rate": 7.965688348947565e-06,
+      "loss": 0.2002,
+      "step": 3786
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.42580393030440017,
+      "learning_rate": 7.964589593930446e-06,
+      "loss": 0.0867,
+      "step": 3787
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.27080932444298184,
+      "learning_rate": 7.963490618093422e-06,
+      "loss": 0.0714,
+      "step": 3788
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.45194285866952794,
+      "learning_rate": 7.962391421518354e-06,
+      "loss": 0.0931,
+      "step": 3789
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7537514573794841,
+      "learning_rate": 7.961292004287116e-06,
+      "loss": 0.1332,
+      "step": 3790
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6315922545752019,
+      "learning_rate": 7.9601923664816e-06,
+      "loss": 0.1204,
+      "step": 3791
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.20256743726844248,
+      "learning_rate": 7.95909250818371e-06,
+      "loss": 0.0502,
+      "step": 3792
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3088851271048101,
+      "learning_rate": 7.957992429475375e-06,
+      "loss": 0.0736,
+      "step": 3793
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.44718490779769315,
+      "learning_rate": 7.95689213043853e-06,
+      "loss": 0.0898,
+      "step": 3794
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3308718439606947,
+      "learning_rate": 7.955791611155132e-06,
+      "loss": 0.0962,
+      "step": 3795
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5377240640959181,
+      "learning_rate": 7.954690871707159e-06,
+      "loss": 0.0767,
+      "step": 3796
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.46453737444210064,
+      "learning_rate": 7.953589912176597e-06,
+      "loss": 0.1287,
+      "step": 3797
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.22360239110456812,
+      "learning_rate": 7.952488732645452e-06,
+      "loss": 0.0551,
+      "step": 3798
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4090020753345458,
+      "learning_rate": 7.951387333195747e-06,
+      "loss": 0.1242,
+      "step": 3799
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.43960818987052247,
+      "learning_rate": 7.95028571390952e-06,
+      "loss": 0.133,
+      "step": 3800
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4003211363864356,
+      "learning_rate": 7.949183874868827e-06,
+      "loss": 0.0825,
+      "step": 3801
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.22981992857338113,
+      "learning_rate": 7.948081816155737e-06,
+      "loss": 0.0563,
+      "step": 3802
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5372224592880047,
+      "learning_rate": 7.946979537852341e-06,
+      "loss": 0.1108,
+      "step": 3803
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.22713161151803155,
+      "learning_rate": 7.945877040040742e-06,
+      "loss": 0.0546,
+      "step": 3804
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.24759246726908532,
+      "learning_rate": 7.94477432280306e-06,
+      "loss": 0.0884,
+      "step": 3805
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.34277318370418386,
+      "learning_rate": 7.94367138622143e-06,
+      "loss": 0.0397,
+      "step": 3806
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.25914238919115246,
+      "learning_rate": 7.942568230378008e-06,
+      "loss": 0.0792,
+      "step": 3807
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5739759948778025,
+      "learning_rate": 7.941464855354963e-06,
+      "loss": 0.1555,
+      "step": 3808
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3415678364698252,
+      "learning_rate": 7.940361261234481e-06,
+      "loss": 0.0905,
+      "step": 3809
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5800951868416174,
+      "learning_rate": 7.939257448098762e-06,
+      "loss": 0.0822,
+      "step": 3810
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.33258165736331224,
+      "learning_rate": 7.938153416030025e-06,
+      "loss": 0.0919,
+      "step": 3811
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.513736024388914,
+      "learning_rate": 7.937049165110509e-06,
+      "loss": 0.1251,
+      "step": 3812
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4315277977310471,
+      "learning_rate": 7.935944695422457e-06,
+      "loss": 0.1006,
+      "step": 3813
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7341287051289989,
+      "learning_rate": 7.934840007048144e-06,
+      "loss": 0.1098,
+      "step": 3814
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2788014270167921,
+      "learning_rate": 7.93373510006985e-06,
+      "loss": 0.0808,
+      "step": 3815
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.22348277967031732,
+      "learning_rate": 7.932629974569877e-06,
+      "loss": 0.0647,
+      "step": 3816
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0739938558570539,
+      "learning_rate": 7.931524630630536e-06,
+      "loss": 0.0765,
+      "step": 3817
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2860032630299111,
+      "learning_rate": 7.930419068334166e-06,
+      "loss": 0.0449,
+      "step": 3818
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.32122209962220377,
+      "learning_rate": 7.929313287763112e-06,
+      "loss": 0.0429,
+      "step": 3819
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3780656956081498,
+      "learning_rate": 7.92820728899974e-06,
+      "loss": 0.1374,
+      "step": 3820
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4086574044775283,
+      "learning_rate": 7.927101072126431e-06,
+      "loss": 0.1014,
+      "step": 3821
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.35111464101065787,
+      "learning_rate": 7.925994637225582e-06,
+      "loss": 0.0931,
+      "step": 3822
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5400846971606367,
+      "learning_rate": 7.924887984379608e-06,
+      "loss": 0.1496,
+      "step": 3823
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4326758984865909,
+      "learning_rate": 7.923781113670937e-06,
+      "loss": 0.1256,
+      "step": 3824
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2422839778474585,
+      "learning_rate": 7.922674025182019e-06,
+      "loss": 0.0645,
+      "step": 3825
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.21447927382797952,
+      "learning_rate": 7.92156671899531e-06,
+      "loss": 0.046,
+      "step": 3826
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.46684582091269483,
+      "learning_rate": 7.920459195193296e-06,
+      "loss": 0.1158,
+      "step": 3827
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.30989675276313333,
+      "learning_rate": 7.919351453858466e-06,
+      "loss": 0.0845,
+      "step": 3828
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3688883811791314,
+      "learning_rate": 7.918243495073333e-06,
+      "loss": 0.1002,
+      "step": 3829
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.234582937182848,
+      "learning_rate": 7.917135318920425e-06,
+      "loss": 0.0906,
+      "step": 3830
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3188407941578234,
+      "learning_rate": 7.916026925482286e-06,
+      "loss": 0.0826,
+      "step": 3831
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.18750359060658844,
+      "learning_rate": 7.914918314841473e-06,
+      "loss": 0.0569,
+      "step": 3832
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.24653489913298587,
+      "learning_rate": 7.913809487080564e-06,
+      "loss": 0.0868,
+      "step": 3833
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3870221264856809,
+      "learning_rate": 7.91270044228215e-06,
+      "loss": 0.1354,
+      "step": 3834
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5787301331289472,
+      "learning_rate": 7.91159118052884e-06,
+      "loss": 0.1025,
+      "step": 3835
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.34443228388239705,
+      "learning_rate": 7.910481701903257e-06,
+      "loss": 0.061,
+      "step": 3836
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4586440360893343,
+      "learning_rate": 7.909372006488041e-06,
+      "loss": 0.1127,
+      "step": 3837
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4419320629196965,
+      "learning_rate": 7.908262094365851e-06,
+      "loss": 0.088,
+      "step": 3838
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.47197517626942104,
+      "learning_rate": 7.907151965619358e-06,
+      "loss": 0.1297,
+      "step": 3839
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.40702805408889603,
+      "learning_rate": 7.906041620331251e-06,
+      "loss": 0.1143,
+      "step": 3840
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.22399706089817173,
+      "learning_rate": 7.904931058584235e-06,
+      "loss": 0.0755,
+      "step": 3841
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.1857511816799227,
+      "learning_rate": 7.903820280461032e-06,
+      "loss": 0.0518,
+      "step": 3842
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6232628861637834,
+      "learning_rate": 7.902709286044378e-06,
+      "loss": 0.1174,
+      "step": 3843
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3791585525964029,
+      "learning_rate": 7.901598075417027e-06,
+      "loss": 0.1125,
+      "step": 3844
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.21971738802441587,
+      "learning_rate": 7.900486648661748e-06,
+      "loss": 0.0606,
+      "step": 3845
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.46764976769635097,
+      "learning_rate": 7.899375005861327e-06,
+      "loss": 0.1001,
+      "step": 3846
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4755886785236143,
+      "learning_rate": 7.898263147098566e-06,
+      "loss": 0.1147,
+      "step": 3847
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4853486099080179,
+      "learning_rate": 7.897151072456281e-06,
+      "loss": 0.123,
+      "step": 3848
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2547251497808277,
+      "learning_rate": 7.896038782017308e-06,
+      "loss": 0.0617,
+      "step": 3849
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2255613555877384,
+      "learning_rate": 7.894926275864496e-06,
+      "loss": 0.0726,
+      "step": 3850
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4206119141303988,
+      "learning_rate": 7.89381355408071e-06,
+      "loss": 0.0997,
+      "step": 3851
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3801132782758231,
+      "learning_rate": 7.892700616748834e-06,
+      "loss": 0.1127,
+      "step": 3852
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.22865707795389592,
+      "learning_rate": 7.891587463951762e-06,
+      "loss": 0.051,
+      "step": 3853
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.48544415150394044,
+      "learning_rate": 7.890474095772413e-06,
+      "loss": 0.1265,
+      "step": 3854
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3376584033847918,
+      "learning_rate": 7.889360512293716e-06,
+      "loss": 0.0895,
+      "step": 3855
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4455884081721892,
+      "learning_rate": 7.888246713598615e-06,
+      "loss": 0.112,
+      "step": 3856
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3974832738480591,
+      "learning_rate": 7.887132699770074e-06,
+      "loss": 0.0776,
+      "step": 3857
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.25374421042763706,
+      "learning_rate": 7.88601847089107e-06,
+      "loss": 0.0919,
+      "step": 3858
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.362099186040887,
+      "learning_rate": 7.884904027044597e-06,
+      "loss": 0.1104,
+      "step": 3859
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.21619924652458278,
+      "learning_rate": 7.883789368313668e-06,
+      "loss": 0.0313,
+      "step": 3860
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.23705021872196716,
+      "learning_rate": 7.882674494781308e-06,
+      "loss": 0.0894,
+      "step": 3861
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.28662358042766967,
+      "learning_rate": 7.881559406530558e-06,
+      "loss": 0.0706,
+      "step": 3862
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.2872241798161116,
+      "learning_rate": 7.880444103644478e-06,
+      "loss": 0.0745,
+      "step": 3863
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.2556383958122709,
+      "learning_rate": 7.879328586206138e-06,
+      "loss": 0.0494,
+      "step": 3864
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.40214061625545827,
+      "learning_rate": 7.878212854298636e-06,
+      "loss": 0.1176,
+      "step": 3865
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.32447818837388304,
+      "learning_rate": 7.877096908005071e-06,
+      "loss": 0.0687,
+      "step": 3866
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4427484428195962,
+      "learning_rate": 7.87598074740857e-06,
+      "loss": 0.0773,
+      "step": 3867
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4785020316303284,
+      "learning_rate": 7.874864372592266e-06,
+      "loss": 0.1219,
+      "step": 3868
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.27507285393148845,
+      "learning_rate": 7.87374778363932e-06,
+      "loss": 0.0485,
+      "step": 3869
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3008880564571848,
+      "learning_rate": 7.872630980632896e-06,
+      "loss": 0.0923,
+      "step": 3870
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.282580828274891,
+      "learning_rate": 7.87151396365618e-06,
+      "loss": 0.0749,
+      "step": 3871
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4426910482050526,
+      "learning_rate": 7.87039673279238e-06,
+      "loss": 0.0902,
+      "step": 3872
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.34774976882377323,
+      "learning_rate": 7.86927928812471e-06,
+      "loss": 0.1083,
+      "step": 3873
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.32808868323170776,
+      "learning_rate": 7.868161629736401e-06,
+      "loss": 0.098,
+      "step": 3874
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.20919092817141202,
+      "learning_rate": 7.867043757710708e-06,
+      "loss": 0.0754,
+      "step": 3875
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.36780692192015213,
+      "learning_rate": 7.865925672130895e-06,
+      "loss": 0.1315,
+      "step": 3876
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.27328051781699075,
+      "learning_rate": 7.86480737308024e-06,
+      "loss": 0.0664,
+      "step": 3877
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7532479406464252,
+      "learning_rate": 7.863688860642044e-06,
+      "loss": 0.1199,
+      "step": 3878
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.267697721072923,
+      "learning_rate": 7.862570134899619e-06,
+      "loss": 0.0594,
+      "step": 3879
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.2661207783712932,
+      "learning_rate": 7.861451195936297e-06,
+      "loss": 0.0665,
+      "step": 3880
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.34507336322935234,
+      "learning_rate": 7.860332043835416e-06,
+      "loss": 0.074,
+      "step": 3881
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4842819332788676,
+      "learning_rate": 7.859212678680345e-06,
+      "loss": 0.1305,
+      "step": 3882
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.22775452056139886,
+      "learning_rate": 7.858093100554459e-06,
+      "loss": 0.0835,
+      "step": 3883
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.2649677826567675,
+      "learning_rate": 7.856973309541147e-06,
+      "loss": 0.0657,
+      "step": 3884
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.39299774013880595,
+      "learning_rate": 7.85585330572382e-06,
+      "loss": 0.0936,
+      "step": 3885
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.1688500630926694,
+      "learning_rate": 7.854733089185903e-06,
+      "loss": 0.0228,
+      "step": 3886
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3117916717171854,
+      "learning_rate": 7.853612660010834e-06,
+      "loss": 0.0553,
+      "step": 3887
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.337920200309975,
+      "learning_rate": 7.852492018282072e-06,
+      "loss": 0.0679,
+      "step": 3888
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.24995447289068923,
+      "learning_rate": 7.851371164083087e-06,
+      "loss": 0.0571,
+      "step": 3889
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3943027218458631,
+      "learning_rate": 7.850250097497366e-06,
+      "loss": 0.0815,
+      "step": 3890
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.29212582309694685,
+      "learning_rate": 7.849128818608416e-06,
+      "loss": 0.0743,
+      "step": 3891
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.315231818190116,
+      "learning_rate": 7.848007327499754e-06,
+      "loss": 0.065,
+      "step": 3892
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6298199122204137,
+      "learning_rate": 7.846885624254913e-06,
+      "loss": 0.1219,
+      "step": 3893
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.29929562240464264,
+      "learning_rate": 7.845763708957448e-06,
+      "loss": 0.0674,
+      "step": 3894
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4851510718287001,
+      "learning_rate": 7.844641581690925e-06,
+      "loss": 0.0949,
+      "step": 3895
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.24261751625604414,
+      "learning_rate": 7.843519242538925e-06,
+      "loss": 0.0578,
+      "step": 3896
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5517817474207253,
+      "learning_rate": 7.842396691585046e-06,
+      "loss": 0.1166,
+      "step": 3897
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.2546134293116837,
+      "learning_rate": 7.841273928912905e-06,
+      "loss": 0.0939,
+      "step": 3898
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.39441983782377865,
+      "learning_rate": 7.84015095460613e-06,
+      "loss": 0.0827,
+      "step": 3899
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3035935505016369,
+      "learning_rate": 7.839027768748364e-06,
+      "loss": 0.079,
+      "step": 3900
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.30435216409594557,
+      "learning_rate": 7.837904371423276e-06,
+      "loss": 0.0754,
+      "step": 3901
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.45913164376111093,
+      "learning_rate": 7.836780762714534e-06,
+      "loss": 0.1253,
+      "step": 3902
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3601705501532005,
+      "learning_rate": 7.835656942705839e-06,
+      "loss": 0.1068,
+      "step": 3903
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.40586107288741385,
+      "learning_rate": 7.834532911480893e-06,
+      "loss": 0.0982,
+      "step": 3904
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3106545064412594,
+      "learning_rate": 7.833408669123423e-06,
+      "loss": 0.1153,
+      "step": 3905
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.20534041734360922,
+      "learning_rate": 7.832284215717171e-06,
+      "loss": 0.0609,
+      "step": 3906
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.375585410253011,
+      "learning_rate": 7.831159551345892e-06,
+      "loss": 0.1162,
+      "step": 3907
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.18370658776983168,
+      "learning_rate": 7.830034676093356e-06,
+      "loss": 0.0697,
+      "step": 3908
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.20099520097408763,
+      "learning_rate": 7.828909590043351e-06,
+      "loss": 0.0625,
+      "step": 3909
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5327097194989433,
+      "learning_rate": 7.82778429327968e-06,
+      "loss": 0.1214,
+      "step": 3910
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.48473639875631025,
+      "learning_rate": 7.826658785886163e-06,
+      "loss": 0.1183,
+      "step": 3911
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3927484108510367,
+      "learning_rate": 7.825533067946632e-06,
+      "loss": 0.0989,
+      "step": 3912
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.40168002164475985,
+      "learning_rate": 7.824407139544938e-06,
+      "loss": 0.149,
+      "step": 3913
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.27247877096938683,
+      "learning_rate": 7.823281000764948e-06,
+      "loss": 0.0702,
+      "step": 3914
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.24980230421511473,
+      "learning_rate": 7.822154651690543e-06,
+      "loss": 0.0418,
+      "step": 3915
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.38864783944524817,
+      "learning_rate": 7.821028092405616e-06,
+      "loss": 0.0915,
+      "step": 3916
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5167848048494923,
+      "learning_rate": 7.819901322994085e-06,
+      "loss": 0.1205,
+      "step": 3917
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.27163177366930685,
+      "learning_rate": 7.818774343539876e-06,
+      "loss": 0.08,
+      "step": 3918
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3866988964807973,
+      "learning_rate": 7.817647154126933e-06,
+      "loss": 0.0611,
+      "step": 3919
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3371888403759303,
+      "learning_rate": 7.816519754839217e-06,
+      "loss": 0.0944,
+      "step": 3920
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.26309356733731043,
+      "learning_rate": 7.815392145760702e-06,
+      "loss": 0.089,
+      "step": 3921
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.359518063411512,
+      "learning_rate": 7.814264326975377e-06,
+      "loss": 0.0878,
+      "step": 3922
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3070311216569019,
+      "learning_rate": 7.813136298567254e-06,
+      "loss": 0.0717,
+      "step": 3923
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.271065586738851,
+      "learning_rate": 7.81200806062035e-06,
+      "loss": 0.0796,
+      "step": 3924
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.26696739368720934,
+      "learning_rate": 7.810879613218705e-06,
+      "loss": 0.0817,
+      "step": 3925
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3020440417645833,
+      "learning_rate": 7.80975095644637e-06,
+      "loss": 0.059,
+      "step": 3926
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.28228129717591405,
+      "learning_rate": 7.808622090387418e-06,
+      "loss": 0.0838,
+      "step": 3927
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.2760120273273923,
+      "learning_rate": 7.807493015125932e-06,
+      "loss": 0.0572,
+      "step": 3928
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.2516615265248732,
+      "learning_rate": 7.806363730746009e-06,
+      "loss": 0.0612,
+      "step": 3929
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3047945124657934,
+      "learning_rate": 7.805234237331767e-06,
+      "loss": 0.0896,
+      "step": 3930
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.21608690471088784,
+      "learning_rate": 7.804104534967339e-06,
+      "loss": 0.0401,
+      "step": 3931
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.32367240352112353,
+      "learning_rate": 7.80297462373687e-06,
+      "loss": 0.069,
+      "step": 3932
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.24022132954773023,
+      "learning_rate": 7.801844503724521e-06,
+      "loss": 0.091,
+      "step": 3933
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4523752191940846,
+      "learning_rate": 7.800714175014472e-06,
+      "loss": 0.0842,
+      "step": 3934
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4437494070946812,
+      "learning_rate": 7.799583637690917e-06,
+      "loss": 0.1166,
+      "step": 3935
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.25327393771307,
+      "learning_rate": 7.798452891838061e-06,
+      "loss": 0.0709,
+      "step": 3936
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.28885642118695637,
+      "learning_rate": 7.797321937540134e-06,
+      "loss": 0.0911,
+      "step": 3937
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4397357676924909,
+      "learning_rate": 7.796190774881374e-06,
+      "loss": 0.1429,
+      "step": 3938
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5784013673537458,
+      "learning_rate": 7.795059403946034e-06,
+      "loss": 0.1314,
+      "step": 3939
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3465212507978782,
+      "learning_rate": 7.793927824818387e-06,
+      "loss": 0.0992,
+      "step": 3940
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3967270184271305,
+      "learning_rate": 7.792796037582721e-06,
+      "loss": 0.0972,
+      "step": 3941
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.2423899022389975,
+      "learning_rate": 7.791664042323338e-06,
+      "loss": 0.0815,
+      "step": 3942
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3158363326405831,
+      "learning_rate": 7.790531839124554e-06,
+      "loss": 0.0933,
+      "step": 3943
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4994796120404779,
+      "learning_rate": 7.789399428070701e-06,
+      "loss": 0.1053,
+      "step": 3944
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.47922147707100066,
+      "learning_rate": 7.78826680924613e-06,
+      "loss": 0.1296,
+      "step": 3945
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4446338324211169,
+      "learning_rate": 7.787133982735204e-06,
+      "loss": 0.0934,
+      "step": 3946
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.642337923926641,
+      "learning_rate": 7.786000948622303e-06,
+      "loss": 0.13,
+      "step": 3947
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.34137154381810103,
+      "learning_rate": 7.784867706991822e-06,
+      "loss": 0.0972,
+      "step": 3948
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3816905833071871,
+      "learning_rate": 7.78373425792817e-06,
+      "loss": 0.1206,
+      "step": 3949
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.286984379845749,
+      "learning_rate": 7.782600601515776e-06,
+      "loss": 0.0627,
+      "step": 3950
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.34552144315568517,
+      "learning_rate": 7.781466737839076e-06,
+      "loss": 0.1077,
+      "step": 3951
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.2585494459406052,
+      "learning_rate": 7.780332666982533e-06,
+      "loss": 0.0883,
+      "step": 3952
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.2722975769808025,
+      "learning_rate": 7.779198389030616e-06,
+      "loss": 0.0647,
+      "step": 3953
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.17233674689960177,
+      "learning_rate": 7.778063904067812e-06,
+      "loss": 0.051,
+      "step": 3954
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.2831338892258267,
+      "learning_rate": 7.776929212178625e-06,
+      "loss": 0.049,
+      "step": 3955
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.18884196810759638,
+      "learning_rate": 7.775794313447574e-06,
+      "loss": 0.0677,
+      "step": 3956
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.32855299250160236,
+      "learning_rate": 7.774659207959192e-06,
+      "loss": 0.0806,
+      "step": 3957
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.24441013219395658,
+      "learning_rate": 7.773523895798029e-06,
+      "loss": 0.071,
+      "step": 3958
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3521650954837509,
+      "learning_rate": 7.77238837704865e-06,
+      "loss": 0.0881,
+      "step": 3959
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.40023080072597395,
+      "learning_rate": 7.771252651795634e-06,
+      "loss": 0.0766,
+      "step": 3960
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4300898088255061,
+      "learning_rate": 7.770116720123575e-06,
+      "loss": 0.1034,
+      "step": 3961
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.31574904581382807,
+      "learning_rate": 7.768980582117088e-06,
+      "loss": 0.0535,
+      "step": 3962
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.33147446430818694,
+      "learning_rate": 7.767844237860796e-06,
+      "loss": 0.0907,
+      "step": 3963
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.37455763022483446,
+      "learning_rate": 7.766707687439338e-06,
+      "loss": 0.1008,
+      "step": 3964
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.21217688007194535,
+      "learning_rate": 7.765570930937378e-06,
+      "loss": 0.0562,
+      "step": 3965
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.22957078114091542,
+      "learning_rate": 7.764433968439584e-06,
+      "loss": 0.0424,
+      "step": 3966
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.2494239311952329,
+      "learning_rate": 7.763296800030645e-06,
+      "loss": 0.0648,
+      "step": 3967
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.39853637975776846,
+      "learning_rate": 7.76215942579526e-06,
+      "loss": 0.1025,
+      "step": 3968
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.47141907447822345,
+      "learning_rate": 7.761021845818151e-06,
+      "loss": 0.1081,
+      "step": 3969
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.25839828207284204,
+      "learning_rate": 7.75988406018405e-06,
+      "loss": 0.0749,
+      "step": 3970
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.36266493309753617,
+      "learning_rate": 7.75874606897771e-06,
+      "loss": 0.1203,
+      "step": 3971
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6177078188743766,
+      "learning_rate": 7.75760787228389e-06,
+      "loss": 0.1198,
+      "step": 3972
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4094266014929528,
+      "learning_rate": 7.75646947018737e-06,
+      "loss": 0.0995,
+      "step": 3973
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3594921905797321,
+      "learning_rate": 7.755330862772948e-06,
+      "loss": 0.0701,
+      "step": 3974
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5153371565438242,
+      "learning_rate": 7.754192050125432e-06,
+      "loss": 0.0905,
+      "step": 3975
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4524479252696568,
+      "learning_rate": 7.753053032329649e-06,
+      "loss": 0.117,
+      "step": 3976
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5007515020466712,
+      "learning_rate": 7.751913809470436e-06,
+      "loss": 0.0886,
+      "step": 3977
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2674171332182288,
+      "learning_rate": 7.750774381632654e-06,
+      "loss": 0.0546,
+      "step": 3978
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.23899557220292109,
+      "learning_rate": 7.749634748901169e-06,
+      "loss": 0.0788,
+      "step": 3979
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.39561194783171605,
+      "learning_rate": 7.74849491136087e-06,
+      "loss": 0.0575,
+      "step": 3980
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.29513834962406893,
+      "learning_rate": 7.747354869096662e-06,
+      "loss": 0.0833,
+      "step": 3981
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.25178111276156356,
+      "learning_rate": 7.746214622193456e-06,
+      "loss": 0.0729,
+      "step": 3982
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.31989368785416455,
+      "learning_rate": 7.74507417073619e-06,
+      "loss": 0.0756,
+      "step": 3983
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.28770204019580486,
+      "learning_rate": 7.743933514809806e-06,
+      "loss": 0.0909,
+      "step": 3984
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4168901267760503,
+      "learning_rate": 7.742792654499266e-06,
+      "loss": 0.0874,
+      "step": 3985
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.36537482128960536,
+      "learning_rate": 7.741651589889557e-06,
+      "loss": 0.1088,
+      "step": 3986
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.23887990002328263,
+      "learning_rate": 7.740510321065664e-06,
+      "loss": 0.0638,
+      "step": 3987
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4304363520350159,
+      "learning_rate": 7.739368848112599e-06,
+      "loss": 0.0863,
+      "step": 3988
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2692463121263935,
+      "learning_rate": 7.738227171115382e-06,
+      "loss": 0.0624,
+      "step": 3989
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.34397976813151626,
+      "learning_rate": 7.737085290159054e-06,
+      "loss": 0.0945,
+      "step": 3990
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2808250520498091,
+      "learning_rate": 7.735943205328671e-06,
+      "loss": 0.0784,
+      "step": 3991
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2001538830970748,
+      "learning_rate": 7.734800916709297e-06,
+      "loss": 0.0653,
+      "step": 3992
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.239184288096797,
+      "learning_rate": 7.733658424386023e-06,
+      "loss": 0.0846,
+      "step": 3993
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.27395350249367867,
+      "learning_rate": 7.73251572844394e-06,
+      "loss": 0.0756,
+      "step": 3994
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.309972309252477,
+      "learning_rate": 7.731372828968173e-06,
+      "loss": 0.0526,
+      "step": 3995
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6218458113463283,
+      "learning_rate": 7.730229726043844e-06,
+      "loss": 0.0847,
+      "step": 3996
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4148747261021244,
+      "learning_rate": 7.729086419756098e-06,
+      "loss": 0.101,
+      "step": 3997
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.47661683101120744,
+      "learning_rate": 7.7279429101901e-06,
+      "loss": 0.1037,
+      "step": 3998
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.41667334874578343,
+      "learning_rate": 7.726799197431023e-06,
+      "loss": 0.1047,
+      "step": 3999
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.28002711284733733,
+      "learning_rate": 7.725655281564055e-06,
+      "loss": 0.0552,
+      "step": 4000
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.42819959084073117,
+      "learning_rate": 7.724511162674404e-06,
+      "loss": 0.1029,
+      "step": 4001
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.33068936577648816,
+      "learning_rate": 7.723366840847291e-06,
+      "loss": 0.0896,
+      "step": 4002
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6769773883105994,
+      "learning_rate": 7.722222316167952e-06,
+      "loss": 0.1748,
+      "step": 4003
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3447402535009358,
+      "learning_rate": 7.721077588721632e-06,
+      "loss": 0.0679,
+      "step": 4004
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.48399046976434695,
+      "learning_rate": 7.719932658593607e-06,
+      "loss": 0.0698,
+      "step": 4005
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4115589996149729,
+      "learning_rate": 7.71878752586915e-06,
+      "loss": 0.0737,
+      "step": 4006
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3244341082691695,
+      "learning_rate": 7.71764219063356e-06,
+      "loss": 0.1081,
+      "step": 4007
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4752411844244107,
+      "learning_rate": 7.71649665297215e-06,
+      "loss": 0.1112,
+      "step": 4008
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.42373652064661477,
+      "learning_rate": 7.715350912970243e-06,
+      "loss": 0.1088,
+      "step": 4009
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.35431428663881925,
+      "learning_rate": 7.714204970713182e-06,
+      "loss": 0.0579,
+      "step": 4010
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2930407518608745,
+      "learning_rate": 7.713058826286324e-06,
+      "loss": 0.079,
+      "step": 4011
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3590825257539356,
+      "learning_rate": 7.711912479775041e-06,
+      "loss": 0.1151,
+      "step": 4012
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4574084655550777,
+      "learning_rate": 7.710765931264716e-06,
+      "loss": 0.0808,
+      "step": 4013
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2720926805606191,
+      "learning_rate": 7.709619180840755e-06,
+      "loss": 0.0669,
+      "step": 4014
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.38781416518756745,
+      "learning_rate": 7.70847222858857e-06,
+      "loss": 0.1294,
+      "step": 4015
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.39553677643909313,
+      "learning_rate": 7.707325074593599e-06,
+      "loss": 0.1232,
+      "step": 4016
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2944292876842651,
+      "learning_rate": 7.706177718941284e-06,
+      "loss": 0.0838,
+      "step": 4017
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2756021041853114,
+      "learning_rate": 7.705030161717086e-06,
+      "loss": 0.0786,
+      "step": 4018
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2880406976479227,
+      "learning_rate": 7.703882403006484e-06,
+      "loss": 0.0683,
+      "step": 4019
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3760143685694372,
+      "learning_rate": 7.702734442894971e-06,
+      "loss": 0.0961,
+      "step": 4020
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3210768261952558,
+      "learning_rate": 7.701586281468052e-06,
+      "loss": 0.0756,
+      "step": 4021
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.20587020655063945,
+      "learning_rate": 7.700437918811247e-06,
+      "loss": 0.0603,
+      "step": 4022
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5094366062290338,
+      "learning_rate": 7.699289355010096e-06,
+      "loss": 0.1066,
+      "step": 4023
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4034650487557578,
+      "learning_rate": 7.69814059015015e-06,
+      "loss": 0.0982,
+      "step": 4024
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2741892725820665,
+      "learning_rate": 7.696991624316974e-06,
+      "loss": 0.0754,
+      "step": 4025
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3205332042832609,
+      "learning_rate": 7.69584245759615e-06,
+      "loss": 0.0964,
+      "step": 4026
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.43474262630995025,
+      "learning_rate": 7.694693090073275e-06,
+      "loss": 0.11,
+      "step": 4027
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.27107576963396923,
+      "learning_rate": 7.693543521833962e-06,
+      "loss": 0.0533,
+      "step": 4028
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2203427389352263,
+      "learning_rate": 7.692393752963837e-06,
+      "loss": 0.0464,
+      "step": 4029
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2533190497200509,
+      "learning_rate": 7.69124378354854e-06,
+      "loss": 0.0697,
+      "step": 4030
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2996798470197408,
+      "learning_rate": 7.690093613673727e-06,
+      "loss": 0.0858,
+      "step": 4031
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.21973213558292526,
+      "learning_rate": 7.688943243425073e-06,
+      "loss": 0.0489,
+      "step": 4032
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.48715801892400007,
+      "learning_rate": 7.687792672888261e-06,
+      "loss": 0.0912,
+      "step": 4033
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.30703317361014504,
+      "learning_rate": 7.686641902148994e-06,
+      "loss": 0.0747,
+      "step": 4034
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.23298684841048295,
+      "learning_rate": 7.685490931292987e-06,
+      "loss": 0.0632,
+      "step": 4035
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2711921061733807,
+      "learning_rate": 7.68433976040597e-06,
+      "loss": 0.0786,
+      "step": 4036
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.21030724459067965,
+      "learning_rate": 7.683188389573692e-06,
+      "loss": 0.0431,
+      "step": 4037
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.707414531849631,
+      "learning_rate": 7.682036818881913e-06,
+      "loss": 0.1177,
+      "step": 4038
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4889481316486415,
+      "learning_rate": 7.680885048416405e-06,
+      "loss": 0.1031,
+      "step": 4039
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3996951815634166,
+      "learning_rate": 7.679733078262963e-06,
+      "loss": 0.124,
+      "step": 4040
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.37750502644961265,
+      "learning_rate": 7.67858090850739e-06,
+      "loss": 0.0997,
+      "step": 4041
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5240621574382807,
+      "learning_rate": 7.677428539235508e-06,
+      "loss": 0.1012,
+      "step": 4042
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.43680227639711205,
+      "learning_rate": 7.67627597053315e-06,
+      "loss": 0.0938,
+      "step": 4043
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2689256479813027,
+      "learning_rate": 7.675123202486167e-06,
+      "loss": 0.0938,
+      "step": 4044
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.20647459481454553,
+      "learning_rate": 7.673970235180423e-06,
+      "loss": 0.0592,
+      "step": 4045
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4784859020930211,
+      "learning_rate": 7.6728170687018e-06,
+      "loss": 0.1002,
+      "step": 4046
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2845623831074091,
+      "learning_rate": 7.671663703136193e-06,
+      "loss": 0.0976,
+      "step": 4047
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4091619167054101,
+      "learning_rate": 7.67051013856951e-06,
+      "loss": 0.0957,
+      "step": 4048
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.43880039722637615,
+      "learning_rate": 7.669356375087673e-06,
+      "loss": 0.0811,
+      "step": 4049
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5165536880517452,
+      "learning_rate": 7.668202412776624e-06,
+      "loss": 0.064,
+      "step": 4050
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2512815741295981,
+      "learning_rate": 7.667048251722316e-06,
+      "loss": 0.0586,
+      "step": 4051
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3867109785146295,
+      "learning_rate": 7.665893892010717e-06,
+      "loss": 0.083,
+      "step": 4052
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4886579592787364,
+      "learning_rate": 7.66473933372781e-06,
+      "loss": 0.1243,
+      "step": 4053
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.28646513835626664,
+      "learning_rate": 7.663584576959596e-06,
+      "loss": 0.0654,
+      "step": 4054
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3553371167488568,
+      "learning_rate": 7.662429621792085e-06,
+      "loss": 0.093,
+      "step": 4055
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3522209563233432,
+      "learning_rate": 7.661274468311306e-06,
+      "loss": 0.1049,
+      "step": 4056
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.18011262515380627,
+      "learning_rate": 7.660119116603301e-06,
+      "loss": 0.0298,
+      "step": 4057
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4011966156190251,
+      "learning_rate": 7.65896356675413e-06,
+      "loss": 0.0537,
+      "step": 4058
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5989935728998027,
+      "learning_rate": 7.657807818849861e-06,
+      "loss": 0.1262,
+      "step": 4059
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3411606643183269,
+      "learning_rate": 7.656651872976584e-06,
+      "loss": 0.0974,
+      "step": 4060
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5844102571786082,
+      "learning_rate": 7.6554957292204e-06,
+      "loss": 0.1586,
+      "step": 4061
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.33484023894948817,
+      "learning_rate": 7.654339387667423e-06,
+      "loss": 0.0653,
+      "step": 4062
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4843659168103004,
+      "learning_rate": 7.653182848403786e-06,
+      "loss": 0.0754,
+      "step": 4063
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.39481290357573484,
+      "learning_rate": 7.652026111515635e-06,
+      "loss": 0.072,
+      "step": 4064
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2845060003189976,
+      "learning_rate": 7.65086917708913e-06,
+      "loss": 0.0725,
+      "step": 4065
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.18394871914271696,
+      "learning_rate": 7.649712045210446e-06,
+      "loss": 0.0773,
+      "step": 4066
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.217378176775968,
+      "learning_rate": 7.648554715965772e-06,
+      "loss": 0.0627,
+      "step": 4067
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3090135678269327,
+      "learning_rate": 7.647397189441315e-06,
+      "loss": 0.0966,
+      "step": 4068
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3604927719882226,
+      "learning_rate": 7.646239465723293e-06,
+      "loss": 0.0904,
+      "step": 4069
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3213725105667064,
+      "learning_rate": 7.64508154489794e-06,
+      "loss": 0.099,
+      "step": 4070
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.27765185240490897,
+      "learning_rate": 7.643923427051506e-06,
+      "loss": 0.0717,
+      "step": 4071
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.26264450806378553,
+      "learning_rate": 7.642765112270251e-06,
+      "loss": 0.0637,
+      "step": 4072
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6326742163905014,
+      "learning_rate": 7.641606600640458e-06,
+      "loss": 0.1498,
+      "step": 4073
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2166582316883081,
+      "learning_rate": 7.640447892248416e-06,
+      "loss": 0.064,
+      "step": 4074
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.25109836178583345,
+      "learning_rate": 7.63928898718043e-06,
+      "loss": 0.0699,
+      "step": 4075
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3297433233606274,
+      "learning_rate": 7.638129885522829e-06,
+      "loss": 0.0866,
+      "step": 4076
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3077088854014417,
+      "learning_rate": 7.636970587361946e-06,
+      "loss": 0.0953,
+      "step": 4077
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4627542230524302,
+      "learning_rate": 7.63581109278413e-06,
+      "loss": 0.098,
+      "step": 4078
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3384242436075803,
+      "learning_rate": 7.63465140187575e-06,
+      "loss": 0.0761,
+      "step": 4079
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.40829844522700826,
+      "learning_rate": 7.633491514723185e-06,
+      "loss": 0.0902,
+      "step": 4080
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6590651614844071,
+      "learning_rate": 7.632331431412831e-06,
+      "loss": 0.0725,
+      "step": 4081
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6599271711843734,
+      "learning_rate": 7.631171152031099e-06,
+      "loss": 0.13,
+      "step": 4082
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.422721413179361,
+      "learning_rate": 7.63001067666441e-06,
+      "loss": 0.1149,
+      "step": 4083
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.44969440773774494,
+      "learning_rate": 7.628850005399205e-06,
+      "loss": 0.1118,
+      "step": 4084
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5397707870719491,
+      "learning_rate": 7.627689138321939e-06,
+      "loss": 0.094,
+      "step": 4085
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2977252998941182,
+      "learning_rate": 7.626528075519076e-06,
+      "loss": 0.0817,
+      "step": 4086
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2182586032932128,
+      "learning_rate": 7.625366817077103e-06,
+      "loss": 0.0703,
+      "step": 4087
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.23855785350685535,
+      "learning_rate": 7.624205363082515e-06,
+      "loss": 0.0713,
+      "step": 4088
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2706152001778837,
+      "learning_rate": 7.623043713621825e-06,
+      "loss": 0.0859,
+      "step": 4089
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.35517180063459086,
+      "learning_rate": 7.621881868781559e-06,
+      "loss": 0.096,
+      "step": 4090
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3343674798532962,
+      "learning_rate": 7.620719828648257e-06,
+      "loss": 0.0858,
+      "step": 4091
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3385285508028679,
+      "learning_rate": 7.6195575933084755e-06,
+      "loss": 0.086,
+      "step": 4092
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3116993432037076,
+      "learning_rate": 7.618395162848785e-06,
+      "loss": 0.0954,
+      "step": 4093
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3687705844903468,
+      "learning_rate": 7.6172325373557685e-06,
+      "loss": 0.0784,
+      "step": 4094
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.31312325750676134,
+      "learning_rate": 7.616069716916028e-06,
+      "loss": 0.0592,
+      "step": 4095
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.31934650431237904,
+      "learning_rate": 7.6149067016161735e-06,
+      "loss": 0.089,
+      "step": 4096
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.28491119834756096,
+      "learning_rate": 7.613743491542835e-06,
+      "loss": 0.0616,
+      "step": 4097
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3054768956489881,
+      "learning_rate": 7.612580086782655e-06,
+      "loss": 0.1074,
+      "step": 4098
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.29584524204119894,
+      "learning_rate": 7.6114164874222915e-06,
+      "loss": 0.0613,
+      "step": 4099
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4403849222399699,
+      "learning_rate": 7.610252693548416e-06,
+      "loss": 0.1221,
+      "step": 4100
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.376056319007393,
+      "learning_rate": 7.609088705247711e-06,
+      "loss": 0.091,
+      "step": 4101
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3235599324792098,
+      "learning_rate": 7.607924522606884e-06,
+      "loss": 0.1014,
+      "step": 4102
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.44820619060878325,
+      "learning_rate": 7.606760145712644e-06,
+      "loss": 0.1395,
+      "step": 4103
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3740918933968217,
+      "learning_rate": 7.605595574651724e-06,
+      "loss": 0.0905,
+      "step": 4104
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3529507226498215,
+      "learning_rate": 7.604430809510866e-06,
+      "loss": 0.1305,
+      "step": 4105
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.20612356771082319,
+      "learning_rate": 7.603265850376831e-06,
+      "loss": 0.0576,
+      "step": 4106
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3253870556899657,
+      "learning_rate": 7.602100697336387e-06,
+      "loss": 0.0977,
+      "step": 4107
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.25916445341616645,
+      "learning_rate": 7.600935350476328e-06,
+      "loss": 0.0912,
+      "step": 4108
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.33828030855006297,
+      "learning_rate": 7.599769809883451e-06,
+      "loss": 0.0554,
+      "step": 4109
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.32881785971908845,
+      "learning_rate": 7.598604075644574e-06,
+      "loss": 0.0959,
+      "step": 4110
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3222898441562243,
+      "learning_rate": 7.597438147846527e-06,
+      "loss": 0.0684,
+      "step": 4111
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.1910580485649205,
+      "learning_rate": 7.596272026576157e-06,
+      "loss": 0.064,
+      "step": 4112
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.27142651348028524,
+      "learning_rate": 7.595105711920321e-06,
+      "loss": 0.0789,
+      "step": 4113
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2801995910893233,
+      "learning_rate": 7.593939203965894e-06,
+      "loss": 0.1186,
+      "step": 4114
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3548840279446117,
+      "learning_rate": 7.592772502799763e-06,
+      "loss": 0.0726,
+      "step": 4115
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.41259025342836914,
+      "learning_rate": 7.591605608508834e-06,
+      "loss": 0.1342,
+      "step": 4116
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.33476649294004124,
+      "learning_rate": 7.590438521180021e-06,
+      "loss": 0.0493,
+      "step": 4117
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3191818976659879,
+      "learning_rate": 7.589271240900258e-06,
+      "loss": 0.065,
+      "step": 4118
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3365665200953937,
+      "learning_rate": 7.5881037677564886e-06,
+      "loss": 0.0693,
+      "step": 4119
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3216148159309258,
+      "learning_rate": 7.586936101835672e-06,
+      "loss": 0.0837,
+      "step": 4120
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3098741407220086,
+      "learning_rate": 7.585768243224787e-06,
+      "loss": 0.1195,
+      "step": 4121
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2511957876148925,
+      "learning_rate": 7.584600192010818e-06,
+      "loss": 0.0707,
+      "step": 4122
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3403499917598926,
+      "learning_rate": 7.5834319482807715e-06,
+      "loss": 0.1108,
+      "step": 4123
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.45511733623025663,
+      "learning_rate": 7.582263512121664e-06,
+      "loss": 0.0957,
+      "step": 4124
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3459484540901208,
+      "learning_rate": 7.5810948836205265e-06,
+      "loss": 0.0973,
+      "step": 4125
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3043133381206861,
+      "learning_rate": 7.579926062864406e-06,
+      "loss": 0.0492,
+      "step": 4126
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2938463178085005,
+      "learning_rate": 7.578757049940364e-06,
+      "loss": 0.1048,
+      "step": 4127
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.27691768795903104,
+      "learning_rate": 7.577587844935475e-06,
+      "loss": 0.0684,
+      "step": 4128
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.21021584985902717,
+      "learning_rate": 7.576418447936827e-06,
+      "loss": 0.0363,
+      "step": 4129
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.30641278176073256,
+      "learning_rate": 7.575248859031523e-06,
+      "loss": 0.0807,
+      "step": 4130
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.36701046998657166,
+      "learning_rate": 7.574079078306684e-06,
+      "loss": 0.1126,
+      "step": 4131
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.43908391053022533,
+      "learning_rate": 7.57290910584944e-06,
+      "loss": 0.1269,
+      "step": 4132
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.44113010548333725,
+      "learning_rate": 7.571738941746935e-06,
+      "loss": 0.1378,
+      "step": 4133
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5012581841078975,
+      "learning_rate": 7.570568586086336e-06,
+      "loss": 0.1007,
+      "step": 4134
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.40304205665830417,
+      "learning_rate": 7.569398038954812e-06,
+      "loss": 0.0718,
+      "step": 4135
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2966017866581755,
+      "learning_rate": 7.568227300439554e-06,
+      "loss": 0.0842,
+      "step": 4136
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5568593402814396,
+      "learning_rate": 7.567056370627766e-06,
+      "loss": 0.1111,
+      "step": 4137
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.26139722336636106,
+      "learning_rate": 7.565885249606666e-06,
+      "loss": 0.0727,
+      "step": 4138
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7530181094940503,
+      "learning_rate": 7.5647139374634835e-06,
+      "loss": 0.0765,
+      "step": 4139
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3267124310184364,
+      "learning_rate": 7.563542434285467e-06,
+      "loss": 0.0776,
+      "step": 4140
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8516435835676295,
+      "learning_rate": 7.562370740159877e-06,
+      "loss": 0.0573,
+      "step": 4141
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4116861578214954,
+      "learning_rate": 7.561198855173988e-06,
+      "loss": 0.0894,
+      "step": 4142
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5749317862054534,
+      "learning_rate": 7.560026779415086e-06,
+      "loss": 0.0985,
+      "step": 4143
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.31547999819508743,
+      "learning_rate": 7.558854512970477e-06,
+      "loss": 0.1059,
+      "step": 4144
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.41689606429873405,
+      "learning_rate": 7.557682055927478e-06,
+      "loss": 0.1119,
+      "step": 4145
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4386684940280412,
+      "learning_rate": 7.55650940837342e-06,
+      "loss": 0.067,
+      "step": 4146
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.397457858997808,
+      "learning_rate": 7.555336570395648e-06,
+      "loss": 0.0872,
+      "step": 4147
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.37713801683202114,
+      "learning_rate": 7.554163542081523e-06,
+      "loss": 0.1082,
+      "step": 4148
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3098048010208587,
+      "learning_rate": 7.552990323518417e-06,
+      "loss": 0.061,
+      "step": 4149
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.37481142438343534,
+      "learning_rate": 7.551816914793721e-06,
+      "loss": 0.0884,
+      "step": 4150
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.23284921278655388,
+      "learning_rate": 7.550643315994834e-06,
+      "loss": 0.0543,
+      "step": 4151
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4292008093071825,
+      "learning_rate": 7.549469527209175e-06,
+      "loss": 0.1323,
+      "step": 4152
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3233260079947718,
+      "learning_rate": 7.548295548524175e-06,
+      "loss": 0.0974,
+      "step": 4153
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4350542027997086,
+      "learning_rate": 7.5471213800272755e-06,
+      "loss": 0.0448,
+      "step": 4154
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.38890532705287545,
+      "learning_rate": 7.545947021805939e-06,
+      "loss": 0.092,
+      "step": 4155
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.24481828590691174,
+      "learning_rate": 7.5447724739476365e-06,
+      "loss": 0.0522,
+      "step": 4156
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3660009746965549,
+      "learning_rate": 7.543597736539856e-06,
+      "loss": 0.0807,
+      "step": 4157
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4262702730060066,
+      "learning_rate": 7.542422809670099e-06,
+      "loss": 0.1323,
+      "step": 4158
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.24779812890943945,
+      "learning_rate": 7.54124769342588e-06,
+      "loss": 0.0762,
+      "step": 4159
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.34417899737674845,
+      "learning_rate": 7.540072387894729e-06,
+      "loss": 0.1008,
+      "step": 4160
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2562297954923462,
+      "learning_rate": 7.538896893164189e-06,
+      "loss": 0.0764,
+      "step": 4161
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.22245925678477532,
+      "learning_rate": 7.5377212093218175e-06,
+      "loss": 0.0767,
+      "step": 4162
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.34927215054126065,
+      "learning_rate": 7.53654533645519e-06,
+      "loss": 0.1014,
+      "step": 4163
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.33852001573023194,
+      "learning_rate": 7.535369274651887e-06,
+      "loss": 0.1089,
+      "step": 4164
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3606040794537733,
+      "learning_rate": 7.534193023999512e-06,
+      "loss": 0.0818,
+      "step": 4165
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2653663237601913,
+      "learning_rate": 7.533016584585676e-06,
+      "loss": 0.0723,
+      "step": 4166
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.23641807673459653,
+      "learning_rate": 7.531839956498011e-06,
+      "loss": 0.0932,
+      "step": 4167
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2910779870897051,
+      "learning_rate": 7.5306631398241546e-06,
+      "loss": 0.0717,
+      "step": 4168
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2876788304586991,
+      "learning_rate": 7.529486134651767e-06,
+      "loss": 0.0909,
+      "step": 4169
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4687252274879929,
+      "learning_rate": 7.528308941068517e-06,
+      "loss": 0.1216,
+      "step": 4170
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.40151991465859976,
+      "learning_rate": 7.5271315591620875e-06,
+      "loss": 0.087,
+      "step": 4171
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.32262277655879124,
+      "learning_rate": 7.525953989020179e-06,
+      "loss": 0.1034,
+      "step": 4172
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2137020105433216,
+      "learning_rate": 7.5247762307305014e-06,
+      "loss": 0.0558,
+      "step": 4173
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6679533827417082,
+      "learning_rate": 7.523598284380784e-06,
+      "loss": 0.152,
+      "step": 4174
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2506628711585966,
+      "learning_rate": 7.5224201500587636e-06,
+      "loss": 0.0871,
+      "step": 4175
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.28187695699823245,
+      "learning_rate": 7.521241827852197e-06,
+      "loss": 0.0985,
+      "step": 4176
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.36163269822089655,
+      "learning_rate": 7.5200633178488516e-06,
+      "loss": 0.07,
+      "step": 4177
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.40897357867534684,
+      "learning_rate": 7.518884620136511e-06,
+      "loss": 0.0713,
+      "step": 4178
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.26324978767307883,
+      "learning_rate": 7.5177057348029685e-06,
+      "loss": 0.0911,
+      "step": 4179
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.34255479554418916,
+      "learning_rate": 7.516526661936037e-06,
+      "loss": 0.0737,
+      "step": 4180
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3927609344767061,
+      "learning_rate": 7.515347401623538e-06,
+      "loss": 0.0673,
+      "step": 4181
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3154516670476979,
+      "learning_rate": 7.514167953953315e-06,
+      "loss": 0.0429,
+      "step": 4182
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.39799054981395265,
+      "learning_rate": 7.512988319013215e-06,
+      "loss": 0.0885,
+      "step": 4183
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2823373720810965,
+      "learning_rate": 7.511808496891105e-06,
+      "loss": 0.0797,
+      "step": 4184
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.25413186905328616,
+      "learning_rate": 7.510628487674866e-06,
+      "loss": 0.0943,
+      "step": 4185
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5179136387412128,
+      "learning_rate": 7.509448291452392e-06,
+      "loss": 0.1069,
+      "step": 4186
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.22786725529163349,
+      "learning_rate": 7.50826790831159e-06,
+      "loss": 0.0449,
+      "step": 4187
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.44288703961664744,
+      "learning_rate": 7.5070873383403845e-06,
+      "loss": 0.1184,
+      "step": 4188
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5032653617230661,
+      "learning_rate": 7.5059065816267074e-06,
+      "loss": 0.1053,
+      "step": 4189
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4173607717122207,
+      "learning_rate": 7.50472563825851e-06,
+      "loss": 0.0983,
+      "step": 4190
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6177052969813429,
+      "learning_rate": 7.503544508323757e-06,
+      "loss": 0.1335,
+      "step": 4191
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3419605783288976,
+      "learning_rate": 7.502363191910424e-06,
+      "loss": 0.085,
+      "step": 4192
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.38536667607078806,
+      "learning_rate": 7.501181689106504e-06,
+      "loss": 0.083,
+      "step": 4193
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.382908617202299,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 0.1015,
+      "step": 4194
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2793272672166173,
+      "learning_rate": 7.4988181246789335e-06,
+      "loss": 0.0706,
+      "step": 4195
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3085585836213908,
+      "learning_rate": 7.497636063231336e-06,
+      "loss": 0.0896,
+      "step": 4196
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4486469496712699,
+      "learning_rate": 7.496453815745255e-06,
+      "loss": 0.0776,
+      "step": 4197
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4723629341062975,
+      "learning_rate": 7.495271382308751e-06,
+      "loss": 0.1314,
+      "step": 4198
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3067639473574543,
+      "learning_rate": 7.494088763009899e-06,
+      "loss": 0.0961,
+      "step": 4199
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.30803491687027423,
+      "learning_rate": 7.492905957936785e-06,
+      "loss": 0.0957,
+      "step": 4200
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.30490982915499015,
+      "learning_rate": 7.491722967177515e-06,
+      "loss": 0.0973,
+      "step": 4201
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3602679208411932,
+      "learning_rate": 7.490539790820202e-06,
+      "loss": 0.0892,
+      "step": 4202
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3453827824664803,
+      "learning_rate": 7.489356428952975e-06,
+      "loss": 0.0817,
+      "step": 4203
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3200416659825833,
+      "learning_rate": 7.488172881663982e-06,
+      "loss": 0.0572,
+      "step": 4204
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4568010529951095,
+      "learning_rate": 7.486989149041377e-06,
+      "loss": 0.1035,
+      "step": 4205
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.49987374517798855,
+      "learning_rate": 7.485805231173334e-06,
+      "loss": 0.0895,
+      "step": 4206
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4525307472605828,
+      "learning_rate": 7.484621128148035e-06,
+      "loss": 0.0923,
+      "step": 4207
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.271804623528407,
+      "learning_rate": 7.483436840053682e-06,
+      "loss": 0.0587,
+      "step": 4208
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.29785231539836793,
+      "learning_rate": 7.482252366978484e-06,
+      "loss": 0.0713,
+      "step": 4209
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.1617458282059297,
+      "learning_rate": 7.481067709010671e-06,
+      "loss": 0.0488,
+      "step": 4210
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.43792823878787984,
+      "learning_rate": 7.479882866238482e-06,
+      "loss": 0.0983,
+      "step": 4211
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.21414884785832047,
+      "learning_rate": 7.47869783875017e-06,
+      "loss": 0.0666,
+      "step": 4212
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.27697379653093857,
+      "learning_rate": 7.477512626634002e-06,
+      "loss": 0.0701,
+      "step": 4213
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4521924906831223,
+      "learning_rate": 7.476327229978263e-06,
+      "loss": 0.0977,
+      "step": 4214
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3222599408085639,
+      "learning_rate": 7.4751416488712455e-06,
+      "loss": 0.1198,
+      "step": 4215
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2772283639148759,
+      "learning_rate": 7.473955883401261e-06,
+      "loss": 0.0602,
+      "step": 4216
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2596234663136592,
+      "learning_rate": 7.4727699336566275e-06,
+      "loss": 0.0611,
+      "step": 4217
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3653756763184814,
+      "learning_rate": 7.471583799725686e-06,
+      "loss": 0.0911,
+      "step": 4218
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3515642998130048,
+      "learning_rate": 7.470397481696784e-06,
+      "loss": 0.1062,
+      "step": 4219
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.28273921880289354,
+      "learning_rate": 7.469210979658287e-06,
+      "loss": 0.0743,
+      "step": 4220
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.23584351591631172,
+      "learning_rate": 7.468024293698573e-06,
+      "loss": 0.0712,
+      "step": 4221
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2438092906701349,
+      "learning_rate": 7.466837423906031e-06,
+      "loss": 0.0823,
+      "step": 4222
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2979226629421547,
+      "learning_rate": 7.465650370369069e-06,
+      "loss": 0.0757,
+      "step": 4223
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.36829834848447557,
+      "learning_rate": 7.464463133176101e-06,
+      "loss": 0.0895,
+      "step": 4224
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7047535146889492,
+      "learning_rate": 7.463275712415564e-06,
+      "loss": 0.134,
+      "step": 4225
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6487316211899284,
+      "learning_rate": 7.462088108175902e-06,
+      "loss": 0.1576,
+      "step": 4226
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3763779670163988,
+      "learning_rate": 7.460900320545575e-06,
+      "loss": 0.0957,
+      "step": 4227
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.317205523517601,
+      "learning_rate": 7.459712349613058e-06,
+      "loss": 0.0517,
+      "step": 4228
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3522019042803247,
+      "learning_rate": 7.458524195466834e-06,
+      "loss": 0.0896,
+      "step": 4229
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.290961807312452,
+      "learning_rate": 7.457335858195408e-06,
+      "loss": 0.0848,
+      "step": 4230
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.22823859281522973,
+      "learning_rate": 7.4561473378872905e-06,
+      "loss": 0.0655,
+      "step": 4231
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.15120987780472853,
+      "learning_rate": 7.454958634631011e-06,
+      "loss": 0.0255,
+      "step": 4232
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4257866707613998,
+      "learning_rate": 7.4537697485151126e-06,
+      "loss": 0.084,
+      "step": 4233
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4059194992600163,
+      "learning_rate": 7.452580679628149e-06,
+      "loss": 0.0864,
+      "step": 4234
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4420201394157988,
+      "learning_rate": 7.4513914280586885e-06,
+      "loss": 0.109,
+      "step": 4235
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.38939669023136964,
+      "learning_rate": 7.450201993895314e-06,
+      "loss": 0.0983,
+      "step": 4236
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.37014027955169737,
+      "learning_rate": 7.449012377226621e-06,
+      "loss": 0.0963,
+      "step": 4237
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.25991086053387835,
+      "learning_rate": 7.44782257814122e-06,
+      "loss": 0.0527,
+      "step": 4238
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.30986282707186885,
+      "learning_rate": 7.446632596727734e-06,
+      "loss": 0.1031,
+      "step": 4239
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4952863843892052,
+      "learning_rate": 7.4454424330748e-06,
+      "loss": 0.0638,
+      "step": 4240
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.24812752302684063,
+      "learning_rate": 7.444252087271067e-06,
+      "loss": 0.0633,
+      "step": 4241
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3790164482487376,
+      "learning_rate": 7.4430615594052e-06,
+      "loss": 0.0655,
+      "step": 4242
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2841771283441297,
+      "learning_rate": 7.441870849565875e-06,
+      "loss": 0.0684,
+      "step": 4243
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5213306147747089,
+      "learning_rate": 7.440679957841786e-06,
+      "loss": 0.1346,
+      "step": 4244
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3833954140037391,
+      "learning_rate": 7.439488884321635e-06,
+      "loss": 0.0853,
+      "step": 4245
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.28636090140439124,
+      "learning_rate": 7.438297629094141e-06,
+      "loss": 0.071,
+      "step": 4246
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.260601557035118,
+      "learning_rate": 7.437106192248036e-06,
+      "loss": 0.0804,
+      "step": 4247
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.18756272795851933,
+      "learning_rate": 7.435914573872064e-06,
+      "loss": 0.043,
+      "step": 4248
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.25927868537453286,
+      "learning_rate": 7.434722774054983e-06,
+      "loss": 0.0789,
+      "step": 4249
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4349795420956213,
+      "learning_rate": 7.433530792885568e-06,
+      "loss": 0.1236,
+      "step": 4250
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3836609119259851,
+      "learning_rate": 7.4323386304526024e-06,
+      "loss": 0.0997,
+      "step": 4251
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5399210676146076,
+      "learning_rate": 7.431146286844888e-06,
+      "loss": 0.1203,
+      "step": 4252
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.42750565960692893,
+      "learning_rate": 7.429953762151234e-06,
+      "loss": 0.0996,
+      "step": 4253
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.24300432508148984,
+      "learning_rate": 7.4287610564604675e-06,
+      "loss": 0.0654,
+      "step": 4254
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2445315404402922,
+      "learning_rate": 7.427568169861431e-06,
+      "loss": 0.0731,
+      "step": 4255
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3099650758809224,
+      "learning_rate": 7.426375102442974e-06,
+      "loss": 0.0811,
+      "step": 4256
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.27910099420005163,
+      "learning_rate": 7.4251818542939654e-06,
+      "loss": 0.0848,
+      "step": 4257
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4491837183840561,
+      "learning_rate": 7.423988425503284e-06,
+      "loss": 0.1483,
+      "step": 4258
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3619488778756981,
+      "learning_rate": 7.422794816159824e-06,
+      "loss": 0.0577,
+      "step": 4259
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2378925630263716,
+      "learning_rate": 7.421601026352494e-06,
+      "loss": 0.0605,
+      "step": 4260
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.38285524554311845,
+      "learning_rate": 7.42040705617021e-06,
+      "loss": 0.1405,
+      "step": 4261
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4694541219337153,
+      "learning_rate": 7.419212905701911e-06,
+      "loss": 0.0977,
+      "step": 4262
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.34160608264501197,
+      "learning_rate": 7.418018575036539e-06,
+      "loss": 0.102,
+      "step": 4263
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2919471505769069,
+      "learning_rate": 7.41682406426306e-06,
+      "loss": 0.0541,
+      "step": 4264
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3673163091884878,
+      "learning_rate": 7.415629373470446e-06,
+      "loss": 0.1073,
+      "step": 4265
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2537527465018049,
+      "learning_rate": 7.414434502747682e-06,
+      "loss": 0.0745,
+      "step": 4266
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5081120227541285,
+      "learning_rate": 7.413239452183772e-06,
+      "loss": 0.1233,
+      "step": 4267
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2717066177084876,
+      "learning_rate": 7.412044221867729e-06,
+      "loss": 0.06,
+      "step": 4268
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.32611557622984544,
+      "learning_rate": 7.410848811888582e-06,
+      "loss": 0.1074,
+      "step": 4269
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5820155301713758,
+      "learning_rate": 7.409653222335371e-06,
+      "loss": 0.1007,
+      "step": 4270
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.38054317632231843,
+      "learning_rate": 7.408457453297151e-06,
+      "loss": 0.0993,
+      "step": 4271
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3344910941148318,
+      "learning_rate": 7.407261504862989e-06,
+      "loss": 0.0951,
+      "step": 4272
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2990307399820234,
+      "learning_rate": 7.406065377121967e-06,
+      "loss": 0.0638,
+      "step": 4273
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3346970639531175,
+      "learning_rate": 7.404869070163181e-06,
+      "loss": 0.1017,
+      "step": 4274
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7723632660353381,
+      "learning_rate": 7.403672584075734e-06,
+      "loss": 0.1693,
+      "step": 4275
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.25662223130860173,
+      "learning_rate": 7.402475918948753e-06,
+      "loss": 0.0633,
+      "step": 4276
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.26781687018007666,
+      "learning_rate": 7.401279074871369e-06,
+      "loss": 0.0642,
+      "step": 4277
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2852689940981586,
+      "learning_rate": 7.400082051932732e-06,
+      "loss": 0.1095,
+      "step": 4278
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.33201343511126774,
+      "learning_rate": 7.398884850222001e-06,
+      "loss": 0.0961,
+      "step": 4279
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2800210007439544,
+      "learning_rate": 7.397687469828353e-06,
+      "loss": 0.0562,
+      "step": 4280
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.26111695759027387,
+      "learning_rate": 7.3964899108409736e-06,
+      "loss": 0.0699,
+      "step": 4281
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.33090038162953533,
+      "learning_rate": 7.395292173349067e-06,
+      "loss": 0.0861,
+      "step": 4282
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2377455497978928,
+      "learning_rate": 7.394094257441843e-06,
+      "loss": 0.0788,
+      "step": 4283
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.1911629932700685,
+      "learning_rate": 7.392896163208534e-06,
+      "loss": 0.0626,
+      "step": 4284
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.32592342801297475,
+      "learning_rate": 7.391697890738379e-06,
+      "loss": 0.0989,
+      "step": 4285
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.43014910282404156,
+      "learning_rate": 7.3904994401206305e-06,
+      "loss": 0.0727,
+      "step": 4286
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.32603578617581186,
+      "learning_rate": 7.3893008114445595e-06,
+      "loss": 0.0859,
+      "step": 4287
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.47791746023235065,
+      "learning_rate": 7.388102004799445e-06,
+      "loss": 0.102,
+      "step": 4288
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2215172017050562,
+      "learning_rate": 7.3869030202745826e-06,
+      "loss": 0.0843,
+      "step": 4289
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.205709306705733,
+      "learning_rate": 7.385703857959277e-06,
+      "loss": 0.0528,
+      "step": 4290
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.26121843163748204,
+      "learning_rate": 7.384504517942852e-06,
+      "loss": 0.0758,
+      "step": 4291
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.20342928259837592,
+      "learning_rate": 7.383305000314638e-06,
+      "loss": 0.0403,
+      "step": 4292
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4642664675965222,
+      "learning_rate": 7.382105305163985e-06,
+      "loss": 0.131,
+      "step": 4293
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.31492577233828084,
+      "learning_rate": 7.380905432580251e-06,
+      "loss": 0.1056,
+      "step": 4294
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.30987995677540575,
+      "learning_rate": 7.379705382652811e-06,
+      "loss": 0.0901,
+      "step": 4295
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.45601195233350356,
+      "learning_rate": 7.378505155471052e-06,
+      "loss": 0.0946,
+      "step": 4296
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2612468015373667,
+      "learning_rate": 7.377304751124372e-06,
+      "loss": 0.0659,
+      "step": 4297
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.37997070327040067,
+      "learning_rate": 7.376104169702186e-06,
+      "loss": 0.0845,
+      "step": 4298
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.340355623409111,
+      "learning_rate": 7.374903411293919e-06,
+      "loss": 0.1022,
+      "step": 4299
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.19625677991097032,
+      "learning_rate": 7.373702475989011e-06,
+      "loss": 0.0554,
+      "step": 4300
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.28547970499009795,
+      "learning_rate": 7.372501363876916e-06,
+      "loss": 0.0672,
+      "step": 4301
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.478412425951132,
+      "learning_rate": 7.371300075047097e-06,
+      "loss": 0.155,
+      "step": 4302
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.40425070243568867,
+      "learning_rate": 7.370098609589037e-06,
+      "loss": 0.1026,
+      "step": 4303
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.22174731517684848,
+      "learning_rate": 7.368896967592224e-06,
+      "loss": 0.0521,
+      "step": 4304
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3201767814085296,
+      "learning_rate": 7.367695149146165e-06,
+      "loss": 0.0784,
+      "step": 4305
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4657728150406824,
+      "learning_rate": 7.3664931543403786e-06,
+      "loss": 0.0697,
+      "step": 4306
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3392540445899379,
+      "learning_rate": 7.365290983264396e-06,
+      "loss": 0.09,
+      "step": 4307
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.466388277451937,
+      "learning_rate": 7.364088636007763e-06,
+      "loss": 0.1246,
+      "step": 4308
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3090290866025594,
+      "learning_rate": 7.362886112660036e-06,
+      "loss": 0.0554,
+      "step": 4309
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2535047477503763,
+      "learning_rate": 7.3616834133107874e-06,
+      "loss": 0.0997,
+      "step": 4310
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.35809566285902594,
+      "learning_rate": 7.3604805380495984e-06,
+      "loss": 0.0873,
+      "step": 4311
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.27652032160288487,
+      "learning_rate": 7.35927748696607e-06,
+      "loss": 0.0862,
+      "step": 4312
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3537699966785868,
+      "learning_rate": 7.358074260149809e-06,
+      "loss": 0.0737,
+      "step": 4313
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.35875827955352657,
+      "learning_rate": 7.356870857690441e-06,
+      "loss": 0.0882,
+      "step": 4314
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2705893899795068,
+      "learning_rate": 7.355667279677602e-06,
+      "loss": 0.1007,
+      "step": 4315
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2438026333145513,
+      "learning_rate": 7.354463526200941e-06,
+      "loss": 0.0842,
+      "step": 4316
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3616875225960766,
+      "learning_rate": 7.3532595973501205e-06,
+      "loss": 0.0677,
+      "step": 4317
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4095175612627275,
+      "learning_rate": 7.352055493214818e-06,
+      "loss": 0.0612,
+      "step": 4318
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2719831860414376,
+      "learning_rate": 7.3508512138847185e-06,
+      "loss": 0.0916,
+      "step": 4319
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.27334555564793167,
+      "learning_rate": 7.349646759449527e-06,
+      "loss": 0.0695,
+      "step": 4320
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.24949931427990701,
+      "learning_rate": 7.348442129998956e-06,
+      "loss": 0.0705,
+      "step": 4321
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6742848721421911,
+      "learning_rate": 7.3472373256227356e-06,
+      "loss": 0.1002,
+      "step": 4322
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.28512955865307865,
+      "learning_rate": 7.346032346410605e-06,
+      "loss": 0.0665,
+      "step": 4323
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2801798710924794,
+      "learning_rate": 7.34482719245232e-06,
+      "loss": 0.0844,
+      "step": 4324
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2343678444683059,
+      "learning_rate": 7.343621863837644e-06,
+      "loss": 0.0839,
+      "step": 4325
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4005533387618653,
+      "learning_rate": 7.34241636065636e-06,
+      "loss": 0.1362,
+      "step": 4326
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.26963481204494405,
+      "learning_rate": 7.341210682998261e-06,
+      "loss": 0.0931,
+      "step": 4327
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.42451083675964074,
+      "learning_rate": 7.340004830953149e-06,
+      "loss": 0.126,
+      "step": 4328
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3573707140213417,
+      "learning_rate": 7.338798804610847e-06,
+      "loss": 0.055,
+      "step": 4329
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.32194781214678536,
+      "learning_rate": 7.337592604061187e-06,
+      "loss": 0.0811,
+      "step": 4330
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.29655329621819715,
+      "learning_rate": 7.336386229394011e-06,
+      "loss": 0.0784,
+      "step": 4331
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.21850906753868482,
+      "learning_rate": 7.335179680699178e-06,
+      "loss": 0.0612,
+      "step": 4332
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.49927903593591255,
+      "learning_rate": 7.33397295806656e-06,
+      "loss": 0.1366,
+      "step": 4333
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3577708260538142,
+      "learning_rate": 7.3327660615860375e-06,
+      "loss": 0.0484,
+      "step": 4334
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4454095433442848,
+      "learning_rate": 7.331558991347512e-06,
+      "loss": 0.1086,
+      "step": 4335
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.23999068555625297,
+      "learning_rate": 7.330351747440888e-06,
+      "loss": 0.0809,
+      "step": 4336
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4103516917991739,
+      "learning_rate": 7.329144329956093e-06,
+      "loss": 0.1051,
+      "step": 4337
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4162759438458064,
+      "learning_rate": 7.327936738983058e-06,
+      "loss": 0.1078,
+      "step": 4338
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3275332949992294,
+      "learning_rate": 7.326728974611736e-06,
+      "loss": 0.0884,
+      "step": 4339
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.28251117058098785,
+      "learning_rate": 7.325521036932083e-06,
+      "loss": 0.0728,
+      "step": 4340
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.18909284090936726,
+      "learning_rate": 7.324312926034076e-06,
+      "loss": 0.0597,
+      "step": 4341
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3738375139878811,
+      "learning_rate": 7.323104642007704e-06,
+      "loss": 0.1018,
+      "step": 4342
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.19427880971931988,
+      "learning_rate": 7.3218961849429635e-06,
+      "loss": 0.0508,
+      "step": 4343
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.40532333884198357,
+      "learning_rate": 7.320687554929871e-06,
+      "loss": 0.1016,
+      "step": 4344
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3229727574160323,
+      "learning_rate": 7.319478752058449e-06,
+      "loss": 0.0726,
+      "step": 4345
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.38166721496828687,
+      "learning_rate": 7.318269776418738e-06,
+      "loss": 0.0945,
+      "step": 4346
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3453011602507455,
+      "learning_rate": 7.31706062810079e-06,
+      "loss": 0.0768,
+      "step": 4347
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.45076466141969534,
+      "learning_rate": 7.315851307194668e-06,
+      "loss": 0.0914,
+      "step": 4348
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6466746095245473,
+      "learning_rate": 7.314641813790449e-06,
+      "loss": 0.1331,
+      "step": 4349
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3392762434734487,
+      "learning_rate": 7.313432147978225e-06,
+      "loss": 0.0998,
+      "step": 4350
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.27683348733956964,
+      "learning_rate": 7.312222309848096e-06,
+      "loss": 0.0758,
+      "step": 4351
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3197989311735069,
+      "learning_rate": 7.311012299490182e-06,
+      "loss": 0.1033,
+      "step": 4352
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.49980893745225385,
+      "learning_rate": 7.309802116994608e-06,
+      "loss": 0.0809,
+      "step": 4353
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5765369405183879,
+      "learning_rate": 7.308591762451518e-06,
+      "loss": 0.139,
+      "step": 4354
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.30240568795392647,
+      "learning_rate": 7.307381235951063e-06,
+      "loss": 0.0495,
+      "step": 4355
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.27540500074467517,
+      "learning_rate": 7.306170537583413e-06,
+      "loss": 0.0934,
+      "step": 4356
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.335836855952277,
+      "learning_rate": 7.304959667438747e-06,
+      "loss": 0.0723,
+      "step": 4357
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.23020509653818272,
+      "learning_rate": 7.303748625607256e-06,
+      "loss": 0.0607,
+      "step": 4358
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.21433700953282805,
+      "learning_rate": 7.302537412179151e-06,
+      "loss": 0.066,
+      "step": 4359
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3642182633304163,
+      "learning_rate": 7.301326027244643e-06,
+      "loss": 0.1029,
+      "step": 4360
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2882530126919763,
+      "learning_rate": 7.300114470893968e-06,
+      "loss": 0.0857,
+      "step": 4361
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.20496859589516553,
+      "learning_rate": 7.2989027432173665e-06,
+      "loss": 0.0479,
+      "step": 4362
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.20590779088106023,
+      "learning_rate": 7.297690844305096e-06,
+      "loss": 0.0578,
+      "step": 4363
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.077340475566252,
+      "learning_rate": 7.296478774247428e-06,
+      "loss": 0.1395,
+      "step": 4364
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2673874528426199,
+      "learning_rate": 7.2952665331346426e-06,
+      "loss": 0.0687,
+      "step": 4365
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2576477308017653,
+      "learning_rate": 7.294054121057036e-06,
+      "loss": 0.0744,
+      "step": 4366
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.46834896153213434,
+      "learning_rate": 7.292841538104913e-06,
+      "loss": 0.13,
+      "step": 4367
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.17731293448854304,
+      "learning_rate": 7.291628784368595e-06,
+      "loss": 0.0616,
+      "step": 4368
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.32469769551803657,
+      "learning_rate": 7.2904158599384165e-06,
+      "loss": 0.114,
+      "step": 4369
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3390900253002126,
+      "learning_rate": 7.289202764904722e-06,
+      "loss": 0.0889,
+      "step": 4370
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.33692052488196417,
+      "learning_rate": 7.28798949935787e-06,
+      "loss": 0.08,
+      "step": 4371
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.35201374552178705,
+      "learning_rate": 7.286776063388231e-06,
+      "loss": 0.0831,
+      "step": 4372
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2824729277993885,
+      "learning_rate": 7.2855624570861905e-06,
+      "loss": 0.074,
+      "step": 4373
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3707127187786412,
+      "learning_rate": 7.284348680542144e-06,
+      "loss": 0.0881,
+      "step": 4374
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.26636636851544493,
+      "learning_rate": 7.283134733846499e-06,
+      "loss": 0.0862,
+      "step": 4375
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.31642570045095875,
+      "learning_rate": 7.28192061708968e-06,
+      "loss": 0.0648,
+      "step": 4376
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2592228677425772,
+      "learning_rate": 7.280706330362119e-06,
+      "loss": 0.0848,
+      "step": 4377
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3719727459868276,
+      "learning_rate": 7.279491873754265e-06,
+      "loss": 0.1168,
+      "step": 4378
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3843527606252072,
+      "learning_rate": 7.278277247356577e-06,
+      "loss": 0.0999,
+      "step": 4379
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2809575807692699,
+      "learning_rate": 7.277062451259529e-06,
+      "loss": 0.0909,
+      "step": 4380
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.28417691026021047,
+      "learning_rate": 7.275847485553602e-06,
+      "loss": 0.0695,
+      "step": 4381
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.40211499905870285,
+      "learning_rate": 7.274632350329299e-06,
+      "loss": 0.1088,
+      "step": 4382
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.1991233353544238,
+      "learning_rate": 7.273417045677127e-06,
+      "loss": 0.0838,
+      "step": 4383
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2734156955822557,
+      "learning_rate": 7.272201571687611e-06,
+      "loss": 0.0824,
+      "step": 4384
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3403644439128153,
+      "learning_rate": 7.270985928451283e-06,
+      "loss": 0.0688,
+      "step": 4385
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.34701787925141103,
+      "learning_rate": 7.269770116058695e-06,
+      "loss": 0.0793,
+      "step": 4386
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.23436470402478968,
+      "learning_rate": 7.268554134600407e-06,
+      "loss": 0.0519,
+      "step": 4387
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.35342647884045747,
+      "learning_rate": 7.267337984166991e-06,
+      "loss": 0.0695,
+      "step": 4388
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.22537326706406283,
+      "learning_rate": 7.266121664849033e-06,
+      "loss": 0.0695,
+      "step": 4389
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.39240502785855536,
+      "learning_rate": 7.264905176737134e-06,
+      "loss": 0.0934,
+      "step": 4390
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.43820444685264803,
+      "learning_rate": 7.263688519921901e-06,
+      "loss": 0.1176,
+      "step": 4391
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3230419627260889,
+      "learning_rate": 7.262471694493961e-06,
+      "loss": 0.107,
+      "step": 4392
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.23885547011906072,
+      "learning_rate": 7.261254700543951e-06,
+      "loss": 0.1052,
+      "step": 4393
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.22362323047075436,
+      "learning_rate": 7.2600375381625155e-06,
+      "loss": 0.0578,
+      "step": 4394
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4307300132716547,
+      "learning_rate": 7.25882020744032e-06,
+      "loss": 0.1062,
+      "step": 4395
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.37070712588669785,
+      "learning_rate": 7.257602708468036e-06,
+      "loss": 0.083,
+      "step": 4396
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.25962095105683713,
+      "learning_rate": 7.256385041336352e-06,
+      "loss": 0.0461,
+      "step": 4397
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.29170078736808197,
+      "learning_rate": 7.255167206135964e-06,
+      "loss": 0.0813,
+      "step": 4398
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5257765925315406,
+      "learning_rate": 7.2539492029575865e-06,
+      "loss": 0.1292,
+      "step": 4399
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3304763287260301,
+      "learning_rate": 7.252731031891943e-06,
+      "loss": 0.0971,
+      "step": 4400
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3190937303575025,
+      "learning_rate": 7.251512693029767e-06,
+      "loss": 0.0786,
+      "step": 4401
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.29075318897506036,
+      "learning_rate": 7.250294186461811e-06,
+      "loss": 0.0771,
+      "step": 4402
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.315261714070033,
+      "learning_rate": 7.249075512278835e-06,
+      "loss": 0.0906,
+      "step": 4403
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.25880370681209053,
+      "learning_rate": 7.247856670571613e-06,
+      "loss": 0.0672,
+      "step": 4404
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.24168920872624683,
+      "learning_rate": 7.246637661430931e-06,
+      "loss": 0.0814,
+      "step": 4405
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.24182136866013298,
+      "learning_rate": 7.245418484947588e-06,
+      "loss": 0.0587,
+      "step": 4406
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2524587110529705,
+      "learning_rate": 7.244199141212397e-06,
+      "loss": 0.0464,
+      "step": 4407
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.29652535553248915,
+      "learning_rate": 7.242979630316178e-06,
+      "loss": 0.0886,
+      "step": 4408
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3655872846260919,
+      "learning_rate": 7.2417599523497715e-06,
+      "loss": 0.0736,
+      "step": 4409
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5003253098559661,
+      "learning_rate": 7.240540107404026e-06,
+      "loss": 0.1424,
+      "step": 4410
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.24440258021406092,
+      "learning_rate": 7.239320095569798e-06,
+      "loss": 0.0711,
+      "step": 4411
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.29849592555062504,
+      "learning_rate": 7.238099916937967e-06,
+      "loss": 0.0678,
+      "step": 4412
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.1933788462165639,
+      "learning_rate": 7.236879571599415e-06,
+      "loss": 0.0603,
+      "step": 4413
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2793616599522201,
+      "learning_rate": 7.2356590596450425e-06,
+      "loss": 0.0849,
+      "step": 4414
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.44358257401033724,
+      "learning_rate": 7.234438381165759e-06,
+      "loss": 0.1118,
+      "step": 4415
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.32619148622601785,
+      "learning_rate": 7.233217536252489e-06,
+      "loss": 0.0557,
+      "step": 4416
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.38300263098741666,
+      "learning_rate": 7.231996524996167e-06,
+      "loss": 0.1245,
+      "step": 4417
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.32762409360761097,
+      "learning_rate": 7.230775347487742e-06,
+      "loss": 0.0933,
+      "step": 4418
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3291932808459943,
+      "learning_rate": 7.229554003818172e-06,
+      "loss": 0.0731,
+      "step": 4419
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4863714235244281,
+      "learning_rate": 7.228332494078434e-06,
+      "loss": 0.0798,
+      "step": 4420
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.24056915976906,
+      "learning_rate": 7.227110818359509e-06,
+      "loss": 0.053,
+      "step": 4421
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.36118812232274367,
+      "learning_rate": 7.225888976752398e-06,
+      "loss": 0.0513,
+      "step": 4422
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4606595015375231,
+      "learning_rate": 7.224666969348107e-06,
+      "loss": 0.1252,
+      "step": 4423
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.38433818350830407,
+      "learning_rate": 7.223444796237663e-06,
+      "loss": 0.085,
+      "step": 4424
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.31759313152307744,
+      "learning_rate": 7.2222224575120955e-06,
+      "loss": 0.0989,
+      "step": 4425
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4305259783160022,
+      "learning_rate": 7.220999953262455e-06,
+      "loss": 0.0931,
+      "step": 4426
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3379742170354169,
+      "learning_rate": 7.219777283579801e-06,
+      "loss": 0.0802,
+      "step": 4427
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.33606016540800654,
+      "learning_rate": 7.218554448555203e-06,
+      "loss": 0.0734,
+      "step": 4428
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.19773440207354814,
+      "learning_rate": 7.217331448279745e-06,
+      "loss": 0.0371,
+      "step": 4429
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.23918815767867072,
+      "learning_rate": 7.216108282844525e-06,
+      "loss": 0.0604,
+      "step": 4430
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4463802928039175,
+      "learning_rate": 7.2148849523406485e-06,
+      "loss": 0.0807,
+      "step": 4431
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.1840464037980593,
+      "learning_rate": 7.21366145685924e-06,
+      "loss": 0.0557,
+      "step": 4432
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.24062627832840017,
+      "learning_rate": 7.212437796491429e-06,
+      "loss": 0.0638,
+      "step": 4433
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.23466465731384345,
+      "learning_rate": 7.211213971328364e-06,
+      "loss": 0.0638,
+      "step": 4434
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2004013942249298,
+      "learning_rate": 7.2099899814612005e-06,
+      "loss": 0.0502,
+      "step": 4435
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3551902406961578,
+      "learning_rate": 7.208765826981109e-06,
+      "loss": 0.101,
+      "step": 4436
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2925379078798644,
+      "learning_rate": 7.2075415079792725e-06,
+      "loss": 0.0796,
+      "step": 4437
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.24888825117853458,
+      "learning_rate": 7.2063170245468824e-06,
+      "loss": 0.0873,
+      "step": 4438
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.628007136146665,
+      "learning_rate": 7.205092376775149e-06,
+      "loss": 0.1327,
+      "step": 4439
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3756592008196272,
+      "learning_rate": 7.203867564755289e-06,
+      "loss": 0.0739,
+      "step": 4440
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3217638584536869,
+      "learning_rate": 7.202642588578536e-06,
+      "loss": 0.1105,
+      "step": 4441
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.47164003380577274,
+      "learning_rate": 7.201417448336131e-06,
+      "loss": 0.0904,
+      "step": 4442
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3271624543576845,
+      "learning_rate": 7.200192144119329e-06,
+      "loss": 0.1075,
+      "step": 4443
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.38859696516958403,
+      "learning_rate": 7.198966676019401e-06,
+      "loss": 0.1192,
+      "step": 4444
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4714428443026246,
+      "learning_rate": 7.197741044127623e-06,
+      "loss": 0.0718,
+      "step": 4445
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.30058585516599096,
+      "learning_rate": 7.196515248535292e-06,
+      "loss": 0.0974,
+      "step": 4446
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.45191510979623795,
+      "learning_rate": 7.195289289333706e-06,
+      "loss": 0.0863,
+      "step": 4447
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.23339346480264778,
+      "learning_rate": 7.194063166614188e-06,
+      "loss": 0.0802,
+      "step": 4448
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.24143093250850814,
+      "learning_rate": 7.1928368804680636e-06,
+      "loss": 0.0328,
+      "step": 4449
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3800915660306654,
+      "learning_rate": 7.191610430986673e-06,
+      "loss": 0.1163,
+      "step": 4450
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3815436742845,
+      "learning_rate": 7.190383818261372e-06,
+      "loss": 0.1036,
+      "step": 4451
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5924616699354528,
+      "learning_rate": 7.189157042383523e-06,
+      "loss": 0.1707,
+      "step": 4452
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.40447090231590643,
+      "learning_rate": 7.187930103444505e-06,
+      "loss": 0.0697,
+      "step": 4453
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.31096012007847545,
+      "learning_rate": 7.186703001535707e-06,
+      "loss": 0.1008,
+      "step": 4454
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.2956273659078011,
+      "learning_rate": 7.185475736748532e-06,
+      "loss": 0.0791,
+      "step": 4455
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.2687782417486176,
+      "learning_rate": 7.184248309174392e-06,
+      "loss": 0.0347,
+      "step": 4456
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.39311432015158854,
+      "learning_rate": 7.183020718904714e-06,
+      "loss": 0.0656,
+      "step": 4457
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3604342669181342,
+      "learning_rate": 7.181792966030936e-06,
+      "loss": 0.1247,
+      "step": 4458
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.32957690369770604,
+      "learning_rate": 7.180565050644507e-06,
+      "loss": 0.0655,
+      "step": 4459
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3078306724532581,
+      "learning_rate": 7.17933697283689e-06,
+      "loss": 0.0777,
+      "step": 4460
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.41934440998114414,
+      "learning_rate": 7.178108732699563e-06,
+      "loss": 0.049,
+      "step": 4461
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.33856948453393737,
+      "learning_rate": 7.176880330324006e-06,
+      "loss": 0.089,
+      "step": 4462
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.1978172132743615,
+      "learning_rate": 7.175651765801721e-06,
+      "loss": 0.0492,
+      "step": 4463
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3435390649805543,
+      "learning_rate": 7.174423039224219e-06,
+      "loss": 0.095,
+      "step": 4464
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.18487806934137355,
+      "learning_rate": 7.173194150683022e-06,
+      "loss": 0.0384,
+      "step": 4465
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.45692108777722396,
+      "learning_rate": 7.171965100269664e-06,
+      "loss": 0.1263,
+      "step": 4466
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.18543258613937713,
+      "learning_rate": 7.170735888075694e-06,
+      "loss": 0.0393,
+      "step": 4467
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.29826467306355947,
+      "learning_rate": 7.16950651419267e-06,
+      "loss": 0.0503,
+      "step": 4468
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.390754866364476,
+      "learning_rate": 7.1682769787121605e-06,
+      "loss": 0.1058,
+      "step": 4469
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.2529586669336332,
+      "learning_rate": 7.167047281725752e-06,
+      "loss": 0.0469,
+      "step": 4470
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.27740849612526397,
+      "learning_rate": 7.165817423325037e-06,
+      "loss": 0.0678,
+      "step": 4471
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3238305300639644,
+      "learning_rate": 7.164587403601624e-06,
+      "loss": 0.0897,
+      "step": 4472
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.40256805226178577,
+      "learning_rate": 7.163357222647132e-06,
+      "loss": 0.1159,
+      "step": 4473
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.27320217497394783,
+      "learning_rate": 7.162126880553191e-06,
+      "loss": 0.0703,
+      "step": 4474
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3012757814252815,
+      "learning_rate": 7.160896377411446e-06,
+      "loss": 0.0778,
+      "step": 4475
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.33930309759192934,
+      "learning_rate": 7.159665713313549e-06,
+      "loss": 0.0905,
+      "step": 4476
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.43918828338493454,
+      "learning_rate": 7.158434888351169e-06,
+      "loss": 0.0934,
+      "step": 4477
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3143226321284182,
+      "learning_rate": 7.157203902615984e-06,
+      "loss": 0.1043,
+      "step": 4478
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.28527351755703373,
+      "learning_rate": 7.155972756199688e-06,
+      "loss": 0.1093,
+      "step": 4479
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.24373037037810458,
+      "learning_rate": 7.154741449193979e-06,
+      "loss": 0.0503,
+      "step": 4480
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5836879325180363,
+      "learning_rate": 7.1535099816905765e-06,
+      "loss": 0.1217,
+      "step": 4481
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3293130545818517,
+      "learning_rate": 7.152278353781205e-06,
+      "loss": 0.05,
+      "step": 4482
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.40905106472643826,
+      "learning_rate": 7.151046565557602e-06,
+      "loss": 0.0788,
+      "step": 4483
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.2786103898926736,
+      "learning_rate": 7.1498146171115234e-06,
+      "loss": 0.0898,
+      "step": 4484
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.23149565052398227,
+      "learning_rate": 7.148582508534726e-06,
+      "loss": 0.0727,
+      "step": 4485
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.22047645770055957,
+      "learning_rate": 7.1473502399189875e-06,
+      "loss": 0.0743,
+      "step": 4486
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5297343025317314,
+      "learning_rate": 7.146117811356094e-06,
+      "loss": 0.111,
+      "step": 4487
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.30991843567162825,
+      "learning_rate": 7.144885222937844e-06,
+      "loss": 0.0933,
+      "step": 4488
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.30575832293479766,
+      "learning_rate": 7.143652474756047e-06,
+      "loss": 0.0793,
+      "step": 4489
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.24623635944148872,
+      "learning_rate": 7.142419566902528e-06,
+      "loss": 0.0384,
+      "step": 4490
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3756769459497274,
+      "learning_rate": 7.141186499469117e-06,
+      "loss": 0.0846,
+      "step": 4491
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4207607937695023,
+      "learning_rate": 7.139953272547663e-06,
+      "loss": 0.1283,
+      "step": 4492
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.25843885632434227,
+      "learning_rate": 7.138719886230022e-06,
+      "loss": 0.0635,
+      "step": 4493
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.28195104279593974,
+      "learning_rate": 7.137486340608066e-06,
+      "loss": 0.094,
+      "step": 4494
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.28088401620762926,
+      "learning_rate": 7.1362526357736765e-06,
+      "loss": 0.1129,
+      "step": 4495
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.2892608047474482,
+      "learning_rate": 7.135018771818744e-06,
+      "loss": 0.0871,
+      "step": 4496
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5429554132612583,
+      "learning_rate": 7.133784748835177e-06,
+      "loss": 0.1375,
+      "step": 4497
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.31689206404312115,
+      "learning_rate": 7.132550566914892e-06,
+      "loss": 0.0714,
+      "step": 4498
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.36989331167756057,
+      "learning_rate": 7.131316226149818e-06,
+      "loss": 0.0826,
+      "step": 4499
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.32107091480601957,
+      "learning_rate": 7.130081726631895e-06,
+      "loss": 0.0713,
+      "step": 4500
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.29354001193379936,
+      "learning_rate": 7.128847068453076e-06,
+      "loss": 0.0849,
+      "step": 4501
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3026613125015102,
+      "learning_rate": 7.127612251705326e-06,
+      "loss": 0.0492,
+      "step": 4502
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3404872637488946,
+      "learning_rate": 7.126377276480622e-06,
+      "loss": 0.1008,
+      "step": 4503
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.43239200291067,
+      "learning_rate": 7.125142142870951e-06,
+      "loss": 0.1408,
+      "step": 4504
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.2140232836728838,
+      "learning_rate": 7.1239068509683126e-06,
+      "loss": 0.0347,
+      "step": 4505
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5432685958706636,
+      "learning_rate": 7.12267140086472e-06,
+      "loss": 0.075,
+      "step": 4506
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.40960067802505173,
+      "learning_rate": 7.121435792652196e-06,
+      "loss": 0.1218,
+      "step": 4507
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3903609894798665,
+      "learning_rate": 7.120200026422775e-06,
+      "loss": 0.0878,
+      "step": 4508
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5993004614364129,
+      "learning_rate": 7.1189641022685064e-06,
+      "loss": 0.1654,
+      "step": 4509
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.29129755236178356,
+      "learning_rate": 7.117728020281447e-06,
+      "loss": 0.064,
+      "step": 4510
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.48661698619386207,
+      "learning_rate": 7.116491780553668e-06,
+      "loss": 0.1218,
+      "step": 4511
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3990283108548014,
+      "learning_rate": 7.115255383177252e-06,
+      "loss": 0.1032,
+      "step": 4512
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.26757744786699955,
+      "learning_rate": 7.114018828244291e-06,
+      "loss": 0.0943,
+      "step": 4513
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5742791392232209,
+      "learning_rate": 7.112782115846894e-06,
+      "loss": 0.1582,
+      "step": 4514
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.286803319433409,
+      "learning_rate": 7.111545246077179e-06,
+      "loss": 0.0701,
+      "step": 4515
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.21214810768790418,
+      "learning_rate": 7.110308219027271e-06,
+      "loss": 0.0498,
+      "step": 4516
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3503200773857527,
+      "learning_rate": 7.109071034789314e-06,
+      "loss": 0.0654,
+      "step": 4517
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4210993995830312,
+      "learning_rate": 7.107833693455461e-06,
+      "loss": 0.1185,
+      "step": 4518
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.372882918440158,
+      "learning_rate": 7.106596195117873e-06,
+      "loss": 0.0766,
+      "step": 4519
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.25138807029501375,
+      "learning_rate": 7.105358539868731e-06,
+      "loss": 0.0768,
+      "step": 4520
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.33535327460005016,
+      "learning_rate": 7.104120727800219e-06,
+      "loss": 0.0756,
+      "step": 4521
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.23399303567254817,
+      "learning_rate": 7.102882759004539e-06,
+      "loss": 0.0493,
+      "step": 4522
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.516439559244914,
+      "learning_rate": 7.101644633573899e-06,
+      "loss": 0.125,
+      "step": 4523
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.32215312927783285,
+      "learning_rate": 7.1004063516005265e-06,
+      "loss": 0.0828,
+      "step": 4524
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.34557942017290244,
+      "learning_rate": 7.09916791317665e-06,
+      "loss": 0.105,
+      "step": 4525
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4373268497909671,
+      "learning_rate": 7.097929318394521e-06,
+      "loss": 0.1342,
+      "step": 4526
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3376293974335842,
+      "learning_rate": 7.096690567346394e-06,
+      "loss": 0.0969,
+      "step": 4527
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3187422636359759,
+      "learning_rate": 7.095451660124541e-06,
+      "loss": 0.0744,
+      "step": 4528
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.2693246089976058,
+      "learning_rate": 7.094212596821239e-06,
+      "loss": 0.069,
+      "step": 4529
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4123271220523029,
+      "learning_rate": 7.092973377528783e-06,
+      "loss": 0.1032,
+      "step": 4530
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.2670640501216721,
+      "learning_rate": 7.0917340023394795e-06,
+      "loss": 0.0602,
+      "step": 4531
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.19299859310112086,
+      "learning_rate": 7.090494471345641e-06,
+      "loss": 0.0679,
+      "step": 4532
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4056009924985113,
+      "learning_rate": 7.0892547846395975e-06,
+      "loss": 0.0912,
+      "step": 4533
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.24943599537177996,
+      "learning_rate": 7.088014942313686e-06,
+      "loss": 0.0966,
+      "step": 4534
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3231146495877769,
+      "learning_rate": 7.086774944460258e-06,
+      "loss": 0.0792,
+      "step": 4535
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.20979593245268607,
+      "learning_rate": 7.085534791171676e-06,
+      "loss": 0.0599,
+      "step": 4536
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.40319949068699457,
+      "learning_rate": 7.084294482540312e-06,
+      "loss": 0.0929,
+      "step": 4537
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.28956685189967385,
+      "learning_rate": 7.083054018658555e-06,
+      "loss": 0.0809,
+      "step": 4538
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3446692870375518,
+      "learning_rate": 7.081813399618799e-06,
+      "loss": 0.1038,
+      "step": 4539
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.38601453513696404,
+      "learning_rate": 7.080572625513453e-06,
+      "loss": 0.1035,
+      "step": 4540
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6281817949263598,
+      "learning_rate": 7.079331696434939e-06,
+      "loss": 0.149,
+      "step": 4541
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.20665617127371372,
+      "learning_rate": 7.078090612475686e-06,
+      "loss": 0.0395,
+      "step": 4542
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.2790853537886932,
+      "learning_rate": 7.076849373728141e-06,
+      "loss": 0.0701,
+      "step": 4543
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.302877243489361,
+      "learning_rate": 7.075607980284753e-06,
+      "loss": 0.063,
+      "step": 4544
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.35800164546769353,
+      "learning_rate": 7.074366432237993e-06,
+      "loss": 0.0896,
+      "step": 4545
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.2926497949919803,
+      "learning_rate": 7.073124729680336e-06,
+      "loss": 0.0911,
+      "step": 4546
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.2406824708776851,
+      "learning_rate": 7.071882872704273e-06,
+      "loss": 0.0684,
+      "step": 4547
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.25559412312241453,
+      "learning_rate": 7.070640861402305e-06,
+      "loss": 0.0688,
+      "step": 4548
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5451921562423949,
+      "learning_rate": 7.069398695866941e-06,
+      "loss": 0.1338,
+      "step": 4549
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.2991058186136316,
+      "learning_rate": 7.068156376190709e-06,
+      "loss": 0.0761,
+      "step": 4550
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.2907521569292752,
+      "learning_rate": 7.066913902466142e-06,
+      "loss": 0.1072,
+      "step": 4551
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.25959314747313417,
+      "learning_rate": 7.065671274785786e-06,
+      "loss": 0.0848,
+      "step": 4552
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4980979468877188,
+      "learning_rate": 7.0644284932422e-06,
+      "loss": 0.0899,
+      "step": 4553
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.2815966474336918,
+      "learning_rate": 7.063185557927954e-06,
+      "loss": 0.0599,
+      "step": 4554
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.29621962435486915,
+      "learning_rate": 7.061942468935629e-06,
+      "loss": 0.0804,
+      "step": 4555
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3709539384068646,
+      "learning_rate": 7.060699226357817e-06,
+      "loss": 0.1021,
+      "step": 4556
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.20716549058724973,
+      "learning_rate": 7.059455830287122e-06,
+      "loss": 0.0737,
+      "step": 4557
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.266064273887425,
+      "learning_rate": 7.05821228081616e-06,
+      "loss": 0.0626,
+      "step": 4558
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.2606200649832158,
+      "learning_rate": 7.056968578037556e-06,
+      "loss": 0.0519,
+      "step": 4559
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3190087466986366,
+      "learning_rate": 7.0557247220439514e-06,
+      "loss": 0.0846,
+      "step": 4560
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.2474997357838211,
+      "learning_rate": 7.0544807129279935e-06,
+      "loss": 0.0533,
+      "step": 4561
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.2813268579343367,
+      "learning_rate": 7.053236550782343e-06,
+      "loss": 0.056,
+      "step": 4562
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.2038261062017343,
+      "learning_rate": 7.051992235699674e-06,
+      "loss": 0.0522,
+      "step": 4563
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.45924830398045974,
+      "learning_rate": 7.050747767772669e-06,
+      "loss": 0.1049,
+      "step": 4564
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.2599416720240521,
+      "learning_rate": 7.0495031470940246e-06,
+      "loss": 0.0945,
+      "step": 4565
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.26395620351457916,
+      "learning_rate": 7.048258373756445e-06,
+      "loss": 0.0557,
+      "step": 4566
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.21162293152280376,
+      "learning_rate": 7.047013447852651e-06,
+      "loss": 0.0629,
+      "step": 4567
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.48992224361938397,
+      "learning_rate": 7.04576836947537e-06,
+      "loss": 0.1174,
+      "step": 4568
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.38844309636071467,
+      "learning_rate": 7.044523138717344e-06,
+      "loss": 0.0927,
+      "step": 4569
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2740228355056441,
+      "learning_rate": 7.043277755671325e-06,
+      "loss": 0.0674,
+      "step": 4570
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.24087935510242203,
+      "learning_rate": 7.042032220430074e-06,
+      "loss": 0.0657,
+      "step": 4571
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.32338038399839775,
+      "learning_rate": 7.040786533086369e-06,
+      "loss": 0.0821,
+      "step": 4572
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.31970494766109964,
+      "learning_rate": 7.039540693732994e-06,
+      "loss": 0.0747,
+      "step": 4573
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2981261456014911,
+      "learning_rate": 7.0382947024627465e-06,
+      "loss": 0.0943,
+      "step": 4574
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.31875198386949877,
+      "learning_rate": 7.037048559368436e-06,
+      "loss": 0.0818,
+      "step": 4575
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.469074319457,
+      "learning_rate": 7.035802264542881e-06,
+      "loss": 0.1001,
+      "step": 4576
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3937155124458543,
+      "learning_rate": 7.034555818078916e-06,
+      "loss": 0.0611,
+      "step": 4577
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.21983044139229677,
+      "learning_rate": 7.0333092200693805e-06,
+      "loss": 0.0573,
+      "step": 4578
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3320530314303248,
+      "learning_rate": 7.032062470607131e-06,
+      "loss": 0.1016,
+      "step": 4579
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.35522823385943625,
+      "learning_rate": 7.030815569785028e-06,
+      "loss": 0.0996,
+      "step": 4580
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2853870682707382,
+      "learning_rate": 7.029568517695952e-06,
+      "loss": 0.0828,
+      "step": 4581
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2533469884707315,
+      "learning_rate": 7.028321314432791e-06,
+      "loss": 0.0741,
+      "step": 4582
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3460787922153535,
+      "learning_rate": 7.027073960088442e-06,
+      "loss": 0.0978,
+      "step": 4583
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.23104150361551903,
+      "learning_rate": 7.025826454755815e-06,
+      "loss": 0.0694,
+      "step": 4584
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5683729296573715,
+      "learning_rate": 7.024578798527834e-06,
+      "loss": 0.1005,
+      "step": 4585
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2313644119344922,
+      "learning_rate": 7.023330991497429e-06,
+      "loss": 0.0736,
+      "step": 4586
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.24560630355133334,
+      "learning_rate": 7.0220830337575454e-06,
+      "loss": 0.0843,
+      "step": 4587
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.375194937879267,
+      "learning_rate": 7.020834925401138e-06,
+      "loss": 0.1012,
+      "step": 4588
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2802463715205821,
+      "learning_rate": 7.019586666521172e-06,
+      "loss": 0.0879,
+      "step": 4589
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3238147135815922,
+      "learning_rate": 7.018338257210627e-06,
+      "loss": 0.0814,
+      "step": 4590
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7947401677785902,
+      "learning_rate": 7.01708969756249e-06,
+      "loss": 0.1503,
+      "step": 4591
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.19029595064928856,
+      "learning_rate": 7.0158409876697646e-06,
+      "loss": 0.0717,
+      "step": 4592
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.24476472428894042,
+      "learning_rate": 7.0145921276254555e-06,
+      "loss": 0.0683,
+      "step": 4593
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.25057248777236724,
+      "learning_rate": 7.013343117522592e-06,
+      "loss": 0.0787,
+      "step": 4594
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5204783130474905,
+      "learning_rate": 7.012093957454202e-06,
+      "loss": 0.0944,
+      "step": 4595
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5037810067795928,
+      "learning_rate": 7.0108446475133356e-06,
+      "loss": 0.1523,
+      "step": 4596
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.24075357476484868,
+      "learning_rate": 7.009595187793043e-06,
+      "loss": 0.0663,
+      "step": 4597
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6943028078535718,
+      "learning_rate": 7.0083455783863955e-06,
+      "loss": 0.1712,
+      "step": 4598
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.37186329901988213,
+      "learning_rate": 7.00709581938647e-06,
+      "loss": 0.0701,
+      "step": 4599
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.15270887004180989,
+      "learning_rate": 7.005845910886355e-06,
+      "loss": 0.0571,
+      "step": 4600
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4636467211431442,
+      "learning_rate": 7.004595852979153e-06,
+      "loss": 0.1013,
+      "step": 4601
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.27646831570849695,
+      "learning_rate": 7.0033456457579715e-06,
+      "loss": 0.0667,
+      "step": 4602
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3197429005898988,
+      "learning_rate": 7.002095289315938e-06,
+      "loss": 0.0797,
+      "step": 4603
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4413457738987146,
+      "learning_rate": 7.000844783746185e-06,
+      "loss": 0.1036,
+      "step": 4604
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2531764616328079,
+      "learning_rate": 6.999594129141854e-06,
+      "loss": 0.0751,
+      "step": 4605
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.15963499278391408,
+      "learning_rate": 6.998343325596106e-06,
+      "loss": 0.0578,
+      "step": 4606
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6642626830466372,
+      "learning_rate": 6.997092373202106e-06,
+      "loss": 0.1174,
+      "step": 4607
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3708441577381597,
+      "learning_rate": 6.995841272053031e-06,
+      "loss": 0.0961,
+      "step": 4608
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.40781277400026017,
+      "learning_rate": 6.994590022242071e-06,
+      "loss": 0.0953,
+      "step": 4609
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4401928671007144,
+      "learning_rate": 6.993338623862428e-06,
+      "loss": 0.1091,
+      "step": 4610
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.22573047832048782,
+      "learning_rate": 6.992087077007312e-06,
+      "loss": 0.0809,
+      "step": 4611
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.37084836568178636,
+      "learning_rate": 6.990835381769946e-06,
+      "loss": 0.1166,
+      "step": 4612
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.27369582024088684,
+      "learning_rate": 6.989583538243563e-06,
+      "loss": 0.0911,
+      "step": 4613
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.21588924184149455,
+      "learning_rate": 6.988331546521408e-06,
+      "loss": 0.0676,
+      "step": 4614
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3194038077317392,
+      "learning_rate": 6.987079406696735e-06,
+      "loss": 0.075,
+      "step": 4615
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.18295596406516973,
+      "learning_rate": 6.9858271188628144e-06,
+      "loss": 0.039,
+      "step": 4616
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3164521029956316,
+      "learning_rate": 6.98457468311292e-06,
+      "loss": 0.106,
+      "step": 4617
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3496827451053549,
+      "learning_rate": 6.983322099540344e-06,
+      "loss": 0.1049,
+      "step": 4618
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.29804031091719513,
+      "learning_rate": 6.9820693682383825e-06,
+      "loss": 0.0629,
+      "step": 4619
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.31747830851939596,
+      "learning_rate": 6.9808164893003496e-06,
+      "loss": 0.0869,
+      "step": 4620
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4211134714581951,
+      "learning_rate": 6.979563462819564e-06,
+      "loss": 0.1098,
+      "step": 4621
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2849165694934356,
+      "learning_rate": 6.9783102888893615e-06,
+      "loss": 0.0915,
+      "step": 4622
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4936751434421673,
+      "learning_rate": 6.977056967603083e-06,
+      "loss": 0.0908,
+      "step": 4623
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.42582998041447295,
+      "learning_rate": 6.975803499054086e-06,
+      "loss": 0.1239,
+      "step": 4624
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.26669509073853215,
+      "learning_rate": 6.974549883335731e-06,
+      "loss": 0.0926,
+      "step": 4625
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.19023749707421994,
+      "learning_rate": 6.9732961205414016e-06,
+      "loss": 0.053,
+      "step": 4626
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.39579521014249525,
+      "learning_rate": 6.97204221076448e-06,
+      "loss": 0.1199,
+      "step": 4627
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.38001402498911424,
+      "learning_rate": 6.970788154098368e-06,
+      "loss": 0.0834,
+      "step": 4628
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.35363927317472754,
+      "learning_rate": 6.969533950636471e-06,
+      "loss": 0.1259,
+      "step": 4629
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.35084582737219505,
+      "learning_rate": 6.9682796004722145e-06,
+      "loss": 0.0949,
+      "step": 4630
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.30235315681779523,
+      "learning_rate": 6.967025103699026e-06,
+      "loss": 0.0793,
+      "step": 4631
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.49815176226630214,
+      "learning_rate": 6.96577046041035e-06,
+      "loss": 0.0801,
+      "step": 4632
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.36047701743173827,
+      "learning_rate": 6.964515670699638e-06,
+      "loss": 0.0667,
+      "step": 4633
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5693876994954974,
+      "learning_rate": 6.9632607346603545e-06,
+      "loss": 0.1272,
+      "step": 4634
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3551048192312459,
+      "learning_rate": 6.9620056523859745e-06,
+      "loss": 0.0987,
+      "step": 4635
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3176765401358284,
+      "learning_rate": 6.960750423969984e-06,
+      "loss": 0.0948,
+      "step": 4636
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.32885626929970263,
+      "learning_rate": 6.959495049505881e-06,
+      "loss": 0.0862,
+      "step": 4637
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.33441706535845866,
+      "learning_rate": 6.958239529087171e-06,
+      "loss": 0.0904,
+      "step": 4638
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3664674076428767,
+      "learning_rate": 6.956983862807371e-06,
+      "loss": 0.1395,
+      "step": 4639
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3643029000004587,
+      "learning_rate": 6.9557280507600165e-06,
+      "loss": 0.0944,
+      "step": 4640
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.39764525507965365,
+      "learning_rate": 6.954472093038641e-06,
+      "loss": 0.1085,
+      "step": 4641
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2462272943416633,
+      "learning_rate": 6.9532159897368e-06,
+      "loss": 0.0902,
+      "step": 4642
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6422862977200203,
+      "learning_rate": 6.951959740948053e-06,
+      "loss": 0.0977,
+      "step": 4643
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.35167146544361116,
+      "learning_rate": 6.950703346765974e-06,
+      "loss": 0.061,
+      "step": 4644
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.32838269354569666,
+      "learning_rate": 6.949446807284146e-06,
+      "loss": 0.0532,
+      "step": 4645
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.35395532582212236,
+      "learning_rate": 6.948190122596162e-06,
+      "loss": 0.1,
+      "step": 4646
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.34899245805070883,
+      "learning_rate": 6.946933292795631e-06,
+      "loss": 0.0911,
+      "step": 4647
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6800209985556319,
+      "learning_rate": 6.945676317976165e-06,
+      "loss": 0.0903,
+      "step": 4648
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.24444936195849173,
+      "learning_rate": 6.944419198231393e-06,
+      "loss": 0.0709,
+      "step": 4649
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3572113125000436,
+      "learning_rate": 6.943161933654952e-06,
+      "loss": 0.093,
+      "step": 4650
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4835815259553198,
+      "learning_rate": 6.941904524340489e-06,
+      "loss": 0.1242,
+      "step": 4651
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3221147294056924,
+      "learning_rate": 6.940646970381667e-06,
+      "loss": 0.1001,
+      "step": 4652
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.43774700711000897,
+      "learning_rate": 6.939389271872152e-06,
+      "loss": 0.1082,
+      "step": 4653
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.36203633198312873,
+      "learning_rate": 6.938131428905628e-06,
+      "loss": 0.0902,
+      "step": 4654
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2744331205096702,
+      "learning_rate": 6.936873441575782e-06,
+      "loss": 0.075,
+      "step": 4655
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.26726734788171613,
+      "learning_rate": 6.935615309976321e-06,
+      "loss": 0.0493,
+      "step": 4656
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.1789680585102601,
+      "learning_rate": 6.934357034200955e-06,
+      "loss": 0.0452,
+      "step": 4657
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6429217807588916,
+      "learning_rate": 6.93309861434341e-06,
+      "loss": 0.1236,
+      "step": 4658
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4419662043680832,
+      "learning_rate": 6.931840050497417e-06,
+      "loss": 0.1288,
+      "step": 4659
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3736932881348034,
+      "learning_rate": 6.9305813427567245e-06,
+      "loss": 0.1074,
+      "step": 4660
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.35688866018510507,
+      "learning_rate": 6.929322491215087e-06,
+      "loss": 0.0931,
+      "step": 4661
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2131496593756044,
+      "learning_rate": 6.9280634959662715e-06,
+      "loss": 0.0718,
+      "step": 4662
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.235904323384669,
+      "learning_rate": 6.9268043571040545e-06,
+      "loss": 0.0737,
+      "step": 4663
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3035957461968339,
+      "learning_rate": 6.925545074722225e-06,
+      "loss": 0.0871,
+      "step": 4664
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4489333302941714,
+      "learning_rate": 6.924285648914581e-06,
+      "loss": 0.0902,
+      "step": 4665
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.19537644175939278,
+      "learning_rate": 6.923026079774931e-06,
+      "loss": 0.054,
+      "step": 4666
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.32063151889596025,
+      "learning_rate": 6.921766367397097e-06,
+      "loss": 0.0573,
+      "step": 4667
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.34851712821721464,
+      "learning_rate": 6.920506511874909e-06,
+      "loss": 0.0728,
+      "step": 4668
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8600880905101923,
+      "learning_rate": 6.919246513302208e-06,
+      "loss": 0.1307,
+      "step": 4669
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2981160393683744,
+      "learning_rate": 6.917986371772847e-06,
+      "loss": 0.047,
+      "step": 4670
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.26524166829265733,
+      "learning_rate": 6.916726087380688e-06,
+      "loss": 0.0594,
+      "step": 4671
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.29683952650589285,
+      "learning_rate": 6.915465660219602e-06,
+      "loss": 0.0566,
+      "step": 4672
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4313234298898825,
+      "learning_rate": 6.914205090383477e-06,
+      "loss": 0.1119,
+      "step": 4673
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.23507208346845698,
+      "learning_rate": 6.912944377966205e-06,
+      "loss": 0.0736,
+      "step": 4674
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.24954692421620062,
+      "learning_rate": 6.911683523061693e-06,
+      "loss": 0.0807,
+      "step": 4675
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3524491627346,
+      "learning_rate": 6.910422525763855e-06,
+      "loss": 0.0712,
+      "step": 4676
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3509261483595118,
+      "learning_rate": 6.909161386166618e-06,
+      "loss": 0.06,
+      "step": 4677
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2331403126488029,
+      "learning_rate": 6.907900104363921e-06,
+      "loss": 0.0819,
+      "step": 4678
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3753832734274987,
+      "learning_rate": 6.906638680449709e-06,
+      "loss": 0.0831,
+      "step": 4679
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3123251492408427,
+      "learning_rate": 6.905377114517939e-06,
+      "loss": 0.076,
+      "step": 4680
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3943890690140778,
+      "learning_rate": 6.904115406662585e-06,
+      "loss": 0.1177,
+      "step": 4681
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2512982884076459,
+      "learning_rate": 6.902853556977623e-06,
+      "loss": 0.0431,
+      "step": 4682
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5464086302395986,
+      "learning_rate": 6.90159156555704e-06,
+      "loss": 0.1304,
+      "step": 4683
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.25476762630256355,
+      "learning_rate": 6.9003294324948424e-06,
+      "loss": 0.0611,
+      "step": 4684
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4199267752505593,
+      "learning_rate": 6.8990671578850375e-06,
+      "loss": 0.11,
+      "step": 4685
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.30638006343768687,
+      "learning_rate": 6.897804741821649e-06,
+      "loss": 0.0669,
+      "step": 4686
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.27595651815173466,
+      "learning_rate": 6.896542184398706e-06,
+      "loss": 0.0776,
+      "step": 4687
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.38570481456423733,
+      "learning_rate": 6.895279485710255e-06,
+      "loss": 0.0939,
+      "step": 4688
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8115734711332021,
+      "learning_rate": 6.8940166458503445e-06,
+      "loss": 0.0777,
+      "step": 4689
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.23435058375226153,
+      "learning_rate": 6.892753664913043e-06,
+      "loss": 0.0477,
+      "step": 4690
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2515452964404833,
+      "learning_rate": 6.891490542992421e-06,
+      "loss": 0.0651,
+      "step": 4691
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2996416988695253,
+      "learning_rate": 6.890227280182566e-06,
+      "loss": 0.0805,
+      "step": 4692
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.43598470353532426,
+      "learning_rate": 6.88896387657757e-06,
+      "loss": 0.1007,
+      "step": 4693
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5056086143166686,
+      "learning_rate": 6.887700332271542e-06,
+      "loss": 0.1029,
+      "step": 4694
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.47204727655141326,
+      "learning_rate": 6.886436647358595e-06,
+      "loss": 0.106,
+      "step": 4695
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.21778906506820928,
+      "learning_rate": 6.8851728219328596e-06,
+      "loss": 0.0563,
+      "step": 4696
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.23184222218221479,
+      "learning_rate": 6.8839088560884686e-06,
+      "loss": 0.0545,
+      "step": 4697
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.28575992476799483,
+      "learning_rate": 6.8826447499195716e-06,
+      "loss": 0.0779,
+      "step": 4698
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.23144626952205274,
+      "learning_rate": 6.881380503520326e-06,
+      "loss": 0.0483,
+      "step": 4699
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4076725202702417,
+      "learning_rate": 6.880116116984903e-06,
+      "loss": 0.1182,
+      "step": 4700
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2603814980769807,
+      "learning_rate": 6.878851590407476e-06,
+      "loss": 0.0636,
+      "step": 4701
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.39781841578384847,
+      "learning_rate": 6.877586923882239e-06,
+      "loss": 0.0959,
+      "step": 4702
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4672791986229205,
+      "learning_rate": 6.87632211750339e-06,
+      "loss": 0.118,
+      "step": 4703
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2774724526637543,
+      "learning_rate": 6.875057171365139e-06,
+      "loss": 0.0601,
+      "step": 4704
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3331048885944759,
+      "learning_rate": 6.873792085561708e-06,
+      "loss": 0.0908,
+      "step": 4705
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4808269838372655,
+      "learning_rate": 6.872526860187325e-06,
+      "loss": 0.0816,
+      "step": 4706
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2832218047384847,
+      "learning_rate": 6.871261495336234e-06,
+      "loss": 0.0793,
+      "step": 4707
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3558092756025122,
+      "learning_rate": 6.869995991102687e-06,
+      "loss": 0.0932,
+      "step": 4708
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.597405732365022,
+      "learning_rate": 6.868730347580943e-06,
+      "loss": 0.1281,
+      "step": 4709
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.35925914959064764,
+      "learning_rate": 6.8674645648652786e-06,
+      "loss": 0.1169,
+      "step": 4710
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.29376236223766683,
+      "learning_rate": 6.866198643049973e-06,
+      "loss": 0.0501,
+      "step": 4711
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.34245062628754164,
+      "learning_rate": 6.864932582229322e-06,
+      "loss": 0.0782,
+      "step": 4712
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3310320703503459,
+      "learning_rate": 6.863666382497628e-06,
+      "loss": 0.0968,
+      "step": 4713
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3239906387219654,
+      "learning_rate": 6.862400043949205e-06,
+      "loss": 0.0885,
+      "step": 4714
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2830266920446376,
+      "learning_rate": 6.861133566678379e-06,
+      "loss": 0.0783,
+      "step": 4715
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2570795726168445,
+      "learning_rate": 6.85986695077948e-06,
+      "loss": 0.0615,
+      "step": 4716
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3263729528481848,
+      "learning_rate": 6.8586001963468584e-06,
+      "loss": 0.1163,
+      "step": 4717
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.28581196559547745,
+      "learning_rate": 6.857333303474866e-06,
+      "loss": 0.0767,
+      "step": 4718
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2784226396919977,
+      "learning_rate": 6.85606627225787e-06,
+      "loss": 0.0893,
+      "step": 4719
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.37994558082634006,
+      "learning_rate": 6.854799102790245e-06,
+      "loss": 0.0695,
+      "step": 4720
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2534278134319433,
+      "learning_rate": 6.853531795166378e-06,
+      "loss": 0.0593,
+      "step": 4721
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3409146914360919,
+      "learning_rate": 6.852264349480666e-06,
+      "loss": 0.0768,
+      "step": 4722
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.29256550314214086,
+      "learning_rate": 6.850996765827513e-06,
+      "loss": 0.0809,
+      "step": 4723
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.16423458387963485,
+      "learning_rate": 6.849729044301341e-06,
+      "loss": 0.0554,
+      "step": 4724
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.19693369960416923,
+      "learning_rate": 6.848461184996572e-06,
+      "loss": 0.048,
+      "step": 4725
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.22390850530869377,
+      "learning_rate": 6.847193188007648e-06,
+      "loss": 0.0557,
+      "step": 4726
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.35208414274819,
+      "learning_rate": 6.845925053429012e-06,
+      "loss": 0.1081,
+      "step": 4727
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.34675431870085194,
+      "learning_rate": 6.844656781355127e-06,
+      "loss": 0.0722,
+      "step": 4728
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7770918014936488,
+      "learning_rate": 6.8433883718804576e-06,
+      "loss": 0.1634,
+      "step": 4729
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3428169366256438,
+      "learning_rate": 6.842119825099484e-06,
+      "loss": 0.0777,
+      "step": 4730
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4440970287684867,
+      "learning_rate": 6.8408511411066945e-06,
+      "loss": 0.131,
+      "step": 4731
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.39000181323981087,
+      "learning_rate": 6.83958231999659e-06,
+      "loss": 0.1129,
+      "step": 4732
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3463804017090998,
+      "learning_rate": 6.838313361863675e-06,
+      "loss": 0.0904,
+      "step": 4733
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4025518594208263,
+      "learning_rate": 6.837044266802475e-06,
+      "loss": 0.1163,
+      "step": 4734
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.31225765529396166,
+      "learning_rate": 6.835775034907515e-06,
+      "loss": 0.0423,
+      "step": 4735
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.24187770221824845,
+      "learning_rate": 6.8345056662733345e-06,
+      "loss": 0.0823,
+      "step": 4736
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.29409086590599975,
+      "learning_rate": 6.833236160994488e-06,
+      "loss": 0.0814,
+      "step": 4737
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4236879142949109,
+      "learning_rate": 6.831966519165531e-06,
+      "loss": 0.1076,
+      "step": 4738
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.32556564861773535,
+      "learning_rate": 6.830696740881036e-06,
+      "loss": 0.079,
+      "step": 4739
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2843365415641183,
+      "learning_rate": 6.8294268262355835e-06,
+      "loss": 0.0572,
+      "step": 4740
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3185509694895841,
+      "learning_rate": 6.828156775323763e-06,
+      "loss": 0.0999,
+      "step": 4741
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3337265724468897,
+      "learning_rate": 6.826886588240177e-06,
+      "loss": 0.1151,
+      "step": 4742
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.280392680161525,
+      "learning_rate": 6.825616265079435e-06,
+      "loss": 0.0806,
+      "step": 4743
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.26640941899131354,
+      "learning_rate": 6.824345805936159e-06,
+      "loss": 0.1016,
+      "step": 4744
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2886765840287694,
+      "learning_rate": 6.823075210904979e-06,
+      "loss": 0.0699,
+      "step": 4745
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3291778913934836,
+      "learning_rate": 6.821804480080538e-06,
+      "loss": 0.0838,
+      "step": 4746
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2792069447540215,
+      "learning_rate": 6.820533613557487e-06,
+      "loss": 0.074,
+      "step": 4747
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3678678595464938,
+      "learning_rate": 6.819262611430485e-06,
+      "loss": 0.0691,
+      "step": 4748
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.44540356711689905,
+      "learning_rate": 6.817991473794207e-06,
+      "loss": 0.0873,
+      "step": 4749
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.28630218956333225,
+      "learning_rate": 6.816720200743334e-06,
+      "loss": 0.0597,
+      "step": 4750
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.36136892266446796,
+      "learning_rate": 6.815448792372556e-06,
+      "loss": 0.0924,
+      "step": 4751
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.29983071113530757,
+      "learning_rate": 6.814177248776578e-06,
+      "loss": 0.0986,
+      "step": 4752
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6994318226042349,
+      "learning_rate": 6.812905570050108e-06,
+      "loss": 0.0643,
+      "step": 4753
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.313902942330313,
+      "learning_rate": 6.811633756287872e-06,
+      "loss": 0.0593,
+      "step": 4754
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4163270977390757,
+      "learning_rate": 6.810361807584599e-06,
+      "loss": 0.0756,
+      "step": 4755
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.287752436939843,
+      "learning_rate": 6.809089724035034e-06,
+      "loss": 0.0768,
+      "step": 4756
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.32523250985216284,
+      "learning_rate": 6.807817505733926e-06,
+      "loss": 0.087,
+      "step": 4757
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2858516588349136,
+      "learning_rate": 6.806545152776039e-06,
+      "loss": 0.071,
+      "step": 4758
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.24995676174002754,
+      "learning_rate": 6.8052726652561454e-06,
+      "loss": 0.0695,
+      "step": 4759
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6729513509788052,
+      "learning_rate": 6.804000043269028e-06,
+      "loss": 0.1223,
+      "step": 4760
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4392571685416266,
+      "learning_rate": 6.802727286909477e-06,
+      "loss": 0.1077,
+      "step": 4761
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.21284167256967465,
+      "learning_rate": 6.8014543962722976e-06,
+      "loss": 0.0697,
+      "step": 4762
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.38992822451407055,
+      "learning_rate": 6.8001813714523e-06,
+      "loss": 0.1006,
+      "step": 4763
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.15994176847069455,
+      "learning_rate": 6.798908212544306e-06,
+      "loss": 0.0295,
+      "step": 4764
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.20832386338669084,
+      "learning_rate": 6.797634919643149e-06,
+      "loss": 0.0717,
+      "step": 4765
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.24471129132564223,
+      "learning_rate": 6.796361492843673e-06,
+      "loss": 0.0699,
+      "step": 4766
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.35018391640569246,
+      "learning_rate": 6.795087932240726e-06,
+      "loss": 0.0686,
+      "step": 4767
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3019177341286526,
+      "learning_rate": 6.793814237929175e-06,
+      "loss": 0.0773,
+      "step": 4768
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.39358351764995964,
+      "learning_rate": 6.7925404100038875e-06,
+      "loss": 0.1116,
+      "step": 4769
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3347301918002869,
+      "learning_rate": 6.791266448559748e-06,
+      "loss": 0.101,
+      "step": 4770
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2614639320898981,
+      "learning_rate": 6.78999235369165e-06,
+      "loss": 0.061,
+      "step": 4771
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3479392756218102,
+      "learning_rate": 6.788718125494493e-06,
+      "loss": 0.0752,
+      "step": 4772
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.35653337709757765,
+      "learning_rate": 6.7874437640631916e-06,
+      "loss": 0.0966,
+      "step": 4773
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6706222577931181,
+      "learning_rate": 6.7861692694926654e-06,
+      "loss": 0.0924,
+      "step": 4774
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.28582327592269996,
+      "learning_rate": 6.784894641877845e-06,
+      "loss": 0.0827,
+      "step": 4775
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2989640083638522,
+      "learning_rate": 6.783619881313677e-06,
+      "loss": 0.0927,
+      "step": 4776
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.30277120949898,
+      "learning_rate": 6.782344987895106e-06,
+      "loss": 0.0742,
+      "step": 4777
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2969396982049214,
+      "learning_rate": 6.781069961717101e-06,
+      "loss": 0.0664,
+      "step": 4778
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2530215928439572,
+      "learning_rate": 6.779794802874627e-06,
+      "loss": 0.1033,
+      "step": 4779
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.30230400004630836,
+      "learning_rate": 6.77851951146267e-06,
+      "loss": 0.0486,
+      "step": 4780
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.23455831225433627,
+      "learning_rate": 6.777244087576218e-06,
+      "loss": 0.073,
+      "step": 4781
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3704920470617132,
+      "learning_rate": 6.775968531310273e-06,
+      "loss": 0.1362,
+      "step": 4782
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3942548124910074,
+      "learning_rate": 6.774692842759848e-06,
+      "loss": 0.1314,
+      "step": 4783
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4250598129507288,
+      "learning_rate": 6.773417022019959e-06,
+      "loss": 0.1248,
+      "step": 4784
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.431864465602126,
+      "learning_rate": 6.772141069185641e-06,
+      "loss": 0.1218,
+      "step": 4785
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2882545789030885,
+      "learning_rate": 6.770864984351933e-06,
+      "loss": 0.114,
+      "step": 4786
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.27376283045821603,
+      "learning_rate": 6.769588767613883e-06,
+      "loss": 0.0837,
+      "step": 4787
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3278122185874359,
+      "learning_rate": 6.768312419066555e-06,
+      "loss": 0.0889,
+      "step": 4788
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.40925665285111323,
+      "learning_rate": 6.767035938805015e-06,
+      "loss": 0.0938,
+      "step": 4789
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3091367832209835,
+      "learning_rate": 6.765759326924346e-06,
+      "loss": 0.0879,
+      "step": 4790
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.35413264990721044,
+      "learning_rate": 6.764482583519636e-06,
+      "loss": 0.0969,
+      "step": 4791
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2910013166318915,
+      "learning_rate": 6.763205708685984e-06,
+      "loss": 0.0878,
+      "step": 4792
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.25997263650834695,
+      "learning_rate": 6.761928702518499e-06,
+      "loss": 0.0806,
+      "step": 4793
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6882219264756096,
+      "learning_rate": 6.7606515651123e-06,
+      "loss": 0.1596,
+      "step": 4794
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4329014209363807,
+      "learning_rate": 6.759374296562516e-06,
+      "loss": 0.0897,
+      "step": 4795
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2444925578978857,
+      "learning_rate": 6.758096896964287e-06,
+      "loss": 0.0692,
+      "step": 4796
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.35319776297209743,
+      "learning_rate": 6.756819366412758e-06,
+      "loss": 0.115,
+      "step": 4797
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.32888963693556517,
+      "learning_rate": 6.755541705003088e-06,
+      "loss": 0.102,
+      "step": 4798
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.23233317529481087,
+      "learning_rate": 6.754263912830446e-06,
+      "loss": 0.0277,
+      "step": 4799
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3159533062090278,
+      "learning_rate": 6.752985989990007e-06,
+      "loss": 0.1159,
+      "step": 4800
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3801680857696001,
+      "learning_rate": 6.751707936576961e-06,
+      "loss": 0.0898,
+      "step": 4801
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3065403007931138,
+      "learning_rate": 6.7504297526865025e-06,
+      "loss": 0.0513,
+      "step": 4802
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3265341824875987,
+      "learning_rate": 6.749151438413839e-06,
+      "loss": 0.075,
+      "step": 4803
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6896437665252437,
+      "learning_rate": 6.747872993854188e-06,
+      "loss": 0.1378,
+      "step": 4804
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.35200937125392945,
+      "learning_rate": 6.746594419102773e-06,
+      "loss": 0.1084,
+      "step": 4805
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2811008958221108,
+      "learning_rate": 6.745315714254832e-06,
+      "loss": 0.1101,
+      "step": 4806
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.42188344891479634,
+      "learning_rate": 6.744036879405609e-06,
+      "loss": 0.126,
+      "step": 4807
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.29323847332343006,
+      "learning_rate": 6.742757914650358e-06,
+      "loss": 0.055,
+      "step": 4808
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.37034951264801164,
+      "learning_rate": 6.741478820084346e-06,
+      "loss": 0.0969,
+      "step": 4809
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3477401866129879,
+      "learning_rate": 6.740199595802849e-06,
+      "loss": 0.1133,
+      "step": 4810
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2169235414354096,
+      "learning_rate": 6.738920241901146e-06,
+      "loss": 0.0751,
+      "step": 4811
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.31316634728156545,
+      "learning_rate": 6.737640758474535e-06,
+      "loss": 0.1042,
+      "step": 4812
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4385252160361452,
+      "learning_rate": 6.736361145618317e-06,
+      "loss": 0.1106,
+      "step": 4813
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2078727550688456,
+      "learning_rate": 6.7350814034278055e-06,
+      "loss": 0.0545,
+      "step": 4814
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2883651508280866,
+      "learning_rate": 6.733801531998324e-06,
+      "loss": 0.0625,
+      "step": 4815
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3887164658302971,
+      "learning_rate": 6.7325215314252035e-06,
+      "loss": 0.114,
+      "step": 4816
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.19195405583857708,
+      "learning_rate": 6.7312414018037874e-06,
+      "loss": 0.0418,
+      "step": 4817
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.332848700826065,
+      "learning_rate": 6.729961143229427e-06,
+      "loss": 0.0648,
+      "step": 4818
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3977229443763912,
+      "learning_rate": 6.728680755797482e-06,
+      "loss": 0.1065,
+      "step": 4819
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.34984364624294406,
+      "learning_rate": 6.727400239603324e-06,
+      "loss": 0.0673,
+      "step": 4820
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.253708797316539,
+      "learning_rate": 6.726119594742333e-06,
+      "loss": 0.0745,
+      "step": 4821
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.39043095701220837,
+      "learning_rate": 6.7248388213099e-06,
+      "loss": 0.1121,
+      "step": 4822
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.22982872737022098,
+      "learning_rate": 6.723557919401423e-06,
+      "loss": 0.0486,
+      "step": 4823
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3204487931305174,
+      "learning_rate": 6.722276889112313e-06,
+      "loss": 0.0929,
+      "step": 4824
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.20821716975417381,
+      "learning_rate": 6.7209957305379854e-06,
+      "loss": 0.0571,
+      "step": 4825
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2794845191851916,
+      "learning_rate": 6.719714443773872e-06,
+      "loss": 0.0965,
+      "step": 4826
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.31605169667744376,
+      "learning_rate": 6.718433028915408e-06,
+      "loss": 0.072,
+      "step": 4827
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4475704121466054,
+      "learning_rate": 6.717151486058042e-06,
+      "loss": 0.1026,
+      "step": 4828
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.32852973687900044,
+      "learning_rate": 6.715869815297229e-06,
+      "loss": 0.0596,
+      "step": 4829
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.31312807089346767,
+      "learning_rate": 6.714588016728437e-06,
+      "loss": 0.0953,
+      "step": 4830
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3891474223614072,
+      "learning_rate": 6.713306090447142e-06,
+      "loss": 0.0978,
+      "step": 4831
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4289942857549009,
+      "learning_rate": 6.712024036548828e-06,
+      "loss": 0.094,
+      "step": 4832
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.36745781484813717,
+      "learning_rate": 6.7107418551289894e-06,
+      "loss": 0.0853,
+      "step": 4833
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3724514019305569,
+      "learning_rate": 6.709459546283134e-06,
+      "loss": 0.1004,
+      "step": 4834
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3683050927059635,
+      "learning_rate": 6.708177110106771e-06,
+      "loss": 0.0931,
+      "step": 4835
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.323883555395736,
+      "learning_rate": 6.706894546695427e-06,
+      "loss": 0.0881,
+      "step": 4836
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5019578078760347,
+      "learning_rate": 6.705611856144634e-06,
+      "loss": 0.1398,
+      "step": 4837
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6646548227223141,
+      "learning_rate": 6.704329038549933e-06,
+      "loss": 0.1351,
+      "step": 4838
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.20279640774519977,
+      "learning_rate": 6.703046094006878e-06,
+      "loss": 0.081,
+      "step": 4839
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.16055324276911118,
+      "learning_rate": 6.701763022611027e-06,
+      "loss": 0.0455,
+      "step": 4840
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.40491465301659757,
+      "learning_rate": 6.700479824457955e-06,
+      "loss": 0.1257,
+      "step": 4841
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2901370396281359,
+      "learning_rate": 6.699196499643239e-06,
+      "loss": 0.0826,
+      "step": 4842
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.35693484914402285,
+      "learning_rate": 6.697913048262469e-06,
+      "loss": 0.0934,
+      "step": 4843
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.43464986668649397,
+      "learning_rate": 6.696629470411244e-06,
+      "loss": 0.0594,
+      "step": 4844
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2695823756281699,
+      "learning_rate": 6.695345766185171e-06,
+      "loss": 0.0794,
+      "step": 4845
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2887731969791938,
+      "learning_rate": 6.694061935679871e-06,
+      "loss": 0.0772,
+      "step": 4846
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4428833451363169,
+      "learning_rate": 6.6927779789909685e-06,
+      "loss": 0.1118,
+      "step": 4847
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.314534917455563,
+      "learning_rate": 6.691493896214102e-06,
+      "loss": 0.0691,
+      "step": 4848
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.22681864268721977,
+      "learning_rate": 6.690209687444915e-06,
+      "loss": 0.0489,
+      "step": 4849
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4072737600006994,
+      "learning_rate": 6.688925352779065e-06,
+      "loss": 0.1084,
+      "step": 4850
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4008894558535504,
+      "learning_rate": 6.687640892312217e-06,
+      "loss": 0.1073,
+      "step": 4851
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3610936045362011,
+      "learning_rate": 6.686356306140043e-06,
+      "loss": 0.096,
+      "step": 4852
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5640241505132098,
+      "learning_rate": 6.685071594358228e-06,
+      "loss": 0.1044,
+      "step": 4853
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4610293347749703,
+      "learning_rate": 6.683786757062464e-06,
+      "loss": 0.1229,
+      "step": 4854
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2698101289616672,
+      "learning_rate": 6.682501794348453e-06,
+      "loss": 0.0804,
+      "step": 4855
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.42712417635640954,
+      "learning_rate": 6.681216706311908e-06,
+      "loss": 0.091,
+      "step": 4856
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.29674661186414414,
+      "learning_rate": 6.6799314930485485e-06,
+      "loss": 0.066,
+      "step": 4857
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.32561350124288796,
+      "learning_rate": 6.678646154654106e-06,
+      "loss": 0.0914,
+      "step": 4858
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.22562405702638078,
+      "learning_rate": 6.677360691224317e-06,
+      "loss": 0.0692,
+      "step": 4859
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5124364880383158,
+      "learning_rate": 6.676075102854934e-06,
+      "loss": 0.1002,
+      "step": 4860
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3116331878894401,
+      "learning_rate": 6.674789389641713e-06,
+      "loss": 0.0989,
+      "step": 4861
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3171396490608389,
+      "learning_rate": 6.673503551680423e-06,
+      "loss": 0.0901,
+      "step": 4862
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2031925588133999,
+      "learning_rate": 6.672217589066838e-06,
+      "loss": 0.0647,
+      "step": 4863
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2776175363039191,
+      "learning_rate": 6.670931501896747e-06,
+      "loss": 0.0649,
+      "step": 4864
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.16617942320294782,
+      "learning_rate": 6.669645290265944e-06,
+      "loss": 0.0359,
+      "step": 4865
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3253203596320841,
+      "learning_rate": 6.668358954270231e-06,
+      "loss": 0.0837,
+      "step": 4866
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.37950736762392717,
+      "learning_rate": 6.667072494005426e-06,
+      "loss": 0.102,
+      "step": 4867
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.30834604578980357,
+      "learning_rate": 6.6657859095673505e-06,
+      "loss": 0.0633,
+      "step": 4868
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.27392371282424016,
+      "learning_rate": 6.664499201051835e-06,
+      "loss": 0.0706,
+      "step": 4869
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.29401399993364213,
+      "learning_rate": 6.6632123685547235e-06,
+      "loss": 0.0627,
+      "step": 4870
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2792621981335468,
+      "learning_rate": 6.6619254121718655e-06,
+      "loss": 0.0607,
+      "step": 4871
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.28960157944553366,
+      "learning_rate": 6.660638331999123e-06,
+      "loss": 0.0813,
+      "step": 4872
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2604280537471189,
+      "learning_rate": 6.659351128132361e-06,
+      "loss": 0.0907,
+      "step": 4873
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.23430025588280498,
+      "learning_rate": 6.658063800667462e-06,
+      "loss": 0.0703,
+      "step": 4874
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.28789725622956963,
+      "learning_rate": 6.6567763497003125e-06,
+      "loss": 0.0696,
+      "step": 4875
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3475314181852648,
+      "learning_rate": 6.655488775326808e-06,
+      "loss": 0.1162,
+      "step": 4876
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3232197410406932,
+      "learning_rate": 6.654201077642857e-06,
+      "loss": 0.1097,
+      "step": 4877
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4052611086314303,
+      "learning_rate": 6.652913256744371e-06,
+      "loss": 0.1004,
+      "step": 4878
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5527571619649555,
+      "learning_rate": 6.651625312727278e-06,
+      "loss": 0.1314,
+      "step": 4879
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.28222193790591804,
+      "learning_rate": 6.650337245687511e-06,
+      "loss": 0.0683,
+      "step": 4880
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.22896453687477603,
+      "learning_rate": 6.6490490557210105e-06,
+      "loss": 0.0948,
+      "step": 4881
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.34335315066405003,
+      "learning_rate": 6.647760742923732e-06,
+      "loss": 0.1265,
+      "step": 4882
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.21088929704846662,
+      "learning_rate": 6.646472307391633e-06,
+      "loss": 0.0589,
+      "step": 4883
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3305695649921629,
+      "learning_rate": 6.645183749220685e-06,
+      "loss": 0.087,
+      "step": 4884
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3049404180221026,
+      "learning_rate": 6.643895068506868e-06,
+      "loss": 0.0835,
+      "step": 4885
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3026572288903209,
+      "learning_rate": 6.6426062653461696e-06,
+      "loss": 0.0999,
+      "step": 4886
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.48376280450561193,
+      "learning_rate": 6.641317339834589e-06,
+      "loss": 0.1293,
+      "step": 4887
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.30927416150747383,
+      "learning_rate": 6.640028292068129e-06,
+      "loss": 0.0769,
+      "step": 4888
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.421880856180528,
+      "learning_rate": 6.63873912214281e-06,
+      "loss": 0.1079,
+      "step": 4889
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.1830477950654268,
+      "learning_rate": 6.6374498301546546e-06,
+      "loss": 0.0537,
+      "step": 4890
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3088043635102654,
+      "learning_rate": 6.636160416199695e-06,
+      "loss": 0.0948,
+      "step": 4891
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2282217431261135,
+      "learning_rate": 6.6348708803739795e-06,
+      "loss": 0.0571,
+      "step": 4892
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8078319267244853,
+      "learning_rate": 6.633581222773554e-06,
+      "loss": 0.1469,
+      "step": 4893
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4184935824904322,
+      "learning_rate": 6.6322914434944855e-06,
+      "loss": 0.1187,
+      "step": 4894
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2708270831306791,
+      "learning_rate": 6.631001542632839e-06,
+      "loss": 0.0752,
+      "step": 4895
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4090851538482934,
+      "learning_rate": 6.6297115202846985e-06,
+      "loss": 0.0764,
+      "step": 4896
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3323025764577,
+      "learning_rate": 6.628421376546148e-06,
+      "loss": 0.0969,
+      "step": 4897
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3903908275663629,
+      "learning_rate": 6.627131111513289e-06,
+      "loss": 0.0918,
+      "step": 4898
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.384314944911904,
+      "learning_rate": 6.625840725282226e-06,
+      "loss": 0.0934,
+      "step": 4899
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.20659339269110363,
+      "learning_rate": 6.624550217949073e-06,
+      "loss": 0.071,
+      "step": 4900
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.362945188154971,
+      "learning_rate": 6.6232595896099565e-06,
+      "loss": 0.1113,
+      "step": 4901
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.27344255774067555,
+      "learning_rate": 6.6219688403610104e-06,
+      "loss": 0.0852,
+      "step": 4902
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2732616025996568,
+      "learning_rate": 6.6206779702983745e-06,
+      "loss": 0.0826,
+      "step": 4903
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.41497504940215346,
+      "learning_rate": 6.619386979518205e-06,
+      "loss": 0.0877,
+      "step": 4904
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.28920091835835127,
+      "learning_rate": 6.6180958681166585e-06,
+      "loss": 0.087,
+      "step": 4905
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.32556084547313535,
+      "learning_rate": 6.616804636189905e-06,
+      "loss": 0.0404,
+      "step": 4906
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.28324056672292114,
+      "learning_rate": 6.615513283834124e-06,
+      "loss": 0.0872,
+      "step": 4907
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.25415296747053184,
+      "learning_rate": 6.614221811145502e-06,
+      "loss": 0.0476,
+      "step": 4908
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2595607287247803,
+      "learning_rate": 6.612930218220237e-06,
+      "loss": 0.0862,
+      "step": 4909
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5484992492841477,
+      "learning_rate": 6.611638505154533e-06,
+      "loss": 0.0941,
+      "step": 4910
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.23320228301337387,
+      "learning_rate": 6.6103466720446055e-06,
+      "loss": 0.0625,
+      "step": 4911
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3998521546899515,
+      "learning_rate": 6.609054718986675e-06,
+      "loss": 0.1315,
+      "step": 4912
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.269050449869328,
+      "learning_rate": 6.6077626460769765e-06,
+      "loss": 0.0974,
+      "step": 4913
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.26866290209082827,
+      "learning_rate": 6.606470453411751e-06,
+      "loss": 0.0923,
+      "step": 4914
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3542843637335846,
+      "learning_rate": 6.605178141087246e-06,
+      "loss": 0.0739,
+      "step": 4915
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6122403980574995,
+      "learning_rate": 6.603885709199723e-06,
+      "loss": 0.1405,
+      "step": 4916
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5077689164686657,
+      "learning_rate": 6.60259315784545e-06,
+      "loss": 0.1054,
+      "step": 4917
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2046183785340652,
+      "learning_rate": 6.601300487120701e-06,
+      "loss": 0.0426,
+      "step": 4918
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5727733455061874,
+      "learning_rate": 6.600007697121765e-06,
+      "loss": 0.0974,
+      "step": 4919
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.23832815695287685,
+      "learning_rate": 6.598714787944934e-06,
+      "loss": 0.0869,
+      "step": 4920
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.43052502571506923,
+      "learning_rate": 6.597421759686513e-06,
+      "loss": 0.1008,
+      "step": 4921
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4695517326619301,
+      "learning_rate": 6.596128612442814e-06,
+      "loss": 0.0926,
+      "step": 4922
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.795711712940724,
+      "learning_rate": 6.594835346310158e-06,
+      "loss": 0.1327,
+      "step": 4923
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.21722893584109934,
+      "learning_rate": 6.593541961384874e-06,
+      "loss": 0.0531,
+      "step": 4924
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.18426186710451573,
+      "learning_rate": 6.5922484577633004e-06,
+      "loss": 0.0556,
+      "step": 4925
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.19348824410343896,
+      "learning_rate": 6.590954835541788e-06,
+      "loss": 0.065,
+      "step": 4926
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6738626166901616,
+      "learning_rate": 6.589661094816691e-06,
+      "loss": 0.1399,
+      "step": 4927
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3371391942009781,
+      "learning_rate": 6.5883672356843765e-06,
+      "loss": 0.1053,
+      "step": 4928
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2866511261628598,
+      "learning_rate": 6.587073258241215e-06,
+      "loss": 0.0638,
+      "step": 4929
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5065218711763665,
+      "learning_rate": 6.585779162583593e-06,
+      "loss": 0.0867,
+      "step": 4930
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2754065165449605,
+      "learning_rate": 6.584484948807901e-06,
+      "loss": 0.0445,
+      "step": 4931
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2068035734573744,
+      "learning_rate": 6.5831906170105406e-06,
+      "loss": 0.0442,
+      "step": 4932
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3332628508681216,
+      "learning_rate": 6.58189616728792e-06,
+      "loss": 0.0477,
+      "step": 4933
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.38297100287362024,
+      "learning_rate": 6.580601599736456e-06,
+      "loss": 0.0713,
+      "step": 4934
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.21283155505754833,
+      "learning_rate": 6.579306914452579e-06,
+      "loss": 0.0856,
+      "step": 4935
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4201297106188521,
+      "learning_rate": 6.578012111532722e-06,
+      "loss": 0.1071,
+      "step": 4936
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3810822275868338,
+      "learning_rate": 6.57671719107333e-06,
+      "loss": 0.0955,
+      "step": 4937
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2642408145797131,
+      "learning_rate": 6.5754221531708575e-06,
+      "loss": 0.0582,
+      "step": 4938
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3506623595730322,
+      "learning_rate": 6.574126997921765e-06,
+      "loss": 0.1052,
+      "step": 4939
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2785563172419917,
+      "learning_rate": 6.5728317254225235e-06,
+      "loss": 0.0881,
+      "step": 4940
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2702326913968486,
+      "learning_rate": 6.571536335769612e-06,
+      "loss": 0.054,
+      "step": 4941
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4163297981989955,
+      "learning_rate": 6.57024082905952e-06,
+      "loss": 0.0758,
+      "step": 4942
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.15162302990231197,
+      "learning_rate": 6.568945205388745e-06,
+      "loss": 0.0387,
+      "step": 4943
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3442304335066295,
+      "learning_rate": 6.567649464853789e-06,
+      "loss": 0.0801,
+      "step": 4944
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5387126750987744,
+      "learning_rate": 6.566353607551171e-06,
+      "loss": 0.0875,
+      "step": 4945
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.27601361889855264,
+      "learning_rate": 6.5650576335774105e-06,
+      "loss": 0.0793,
+      "step": 4946
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2503927548068538,
+      "learning_rate": 6.563761543029039e-06,
+      "loss": 0.0518,
+      "step": 4947
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.25470950084284744,
+      "learning_rate": 6.562465336002601e-06,
+      "loss": 0.0843,
+      "step": 4948
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3333534344818437,
+      "learning_rate": 6.561169012594641e-06,
+      "loss": 0.0941,
+      "step": 4949
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.39105577873215336,
+      "learning_rate": 6.559872572901719e-06,
+      "loss": 0.1027,
+      "step": 4950
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.47511744146933216,
+      "learning_rate": 6.558576017020401e-06,
+      "loss": 0.1022,
+      "step": 4951
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3351323068823734,
+      "learning_rate": 6.557279345047261e-06,
+      "loss": 0.0879,
+      "step": 4952
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4975771182781274,
+      "learning_rate": 6.555982557078884e-06,
+      "loss": 0.0836,
+      "step": 4953
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3179978187406493,
+      "learning_rate": 6.554685653211861e-06,
+      "loss": 0.0637,
+      "step": 4954
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2320710301638781,
+      "learning_rate": 6.553388633542795e-06,
+      "loss": 0.0694,
+      "step": 4955
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.400426818784265,
+      "learning_rate": 6.552091498168293e-06,
+      "loss": 0.1153,
+      "step": 4956
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.40578794588133793,
+      "learning_rate": 6.550794247184978e-06,
+      "loss": 0.1094,
+      "step": 4957
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3112202679021108,
+      "learning_rate": 6.54949688068947e-06,
+      "loss": 0.0905,
+      "step": 4958
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.39327923878477794,
+      "learning_rate": 6.548199398778409e-06,
+      "loss": 0.091,
+      "step": 4959
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.25205357815626855,
+      "learning_rate": 6.546901801548438e-06,
+      "loss": 0.075,
+      "step": 4960
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.44146306484474024,
+      "learning_rate": 6.545604089096209e-06,
+      "loss": 0.1304,
+      "step": 4961
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2623672458765783,
+      "learning_rate": 6.544306261518385e-06,
+      "loss": 0.0793,
+      "step": 4962
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3722855822940689,
+      "learning_rate": 6.5430083189116325e-06,
+      "loss": 0.0901,
+      "step": 4963
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.35561858737809965,
+      "learning_rate": 6.541710261372634e-06,
+      "loss": 0.0499,
+      "step": 4964
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.36620689849902505,
+      "learning_rate": 6.5404120889980715e-06,
+      "loss": 0.0997,
+      "step": 4965
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2778471532545692,
+      "learning_rate": 6.539113801884645e-06,
+      "loss": 0.0809,
+      "step": 4966
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.29347253779623605,
+      "learning_rate": 6.537815400129056e-06,
+      "loss": 0.0821,
+      "step": 4967
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3896457159713344,
+      "learning_rate": 6.536516883828017e-06,
+      "loss": 0.1091,
+      "step": 4968
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.214039369518944,
+      "learning_rate": 6.53521825307825e-06,
+      "loss": 0.0703,
+      "step": 4969
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.28368249130384643,
+      "learning_rate": 6.533919507976484e-06,
+      "loss": 0.0861,
+      "step": 4970
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.26842093576754844,
+      "learning_rate": 6.532620648619457e-06,
+      "loss": 0.0848,
+      "step": 4971
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2625932227753389,
+      "learning_rate": 6.531321675103917e-06,
+      "loss": 0.0433,
+      "step": 4972
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6763495141258672,
+      "learning_rate": 6.5300225875266165e-06,
+      "loss": 0.1401,
+      "step": 4973
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.24049173200348947,
+      "learning_rate": 6.528723385984322e-06,
+      "loss": 0.0758,
+      "step": 4974
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4541796484774277,
+      "learning_rate": 6.527424070573804e-06,
+      "loss": 0.0931,
+      "step": 4975
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3592578586107002,
+      "learning_rate": 6.5261246413918425e-06,
+      "loss": 0.0684,
+      "step": 4976
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.23034998026534714,
+      "learning_rate": 6.524825098535228e-06,
+      "loss": 0.0528,
+      "step": 4977
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3175670333424844,
+      "learning_rate": 6.523525442100758e-06,
+      "loss": 0.088,
+      "step": 4978
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.44402896122943847,
+      "learning_rate": 6.522225672185238e-06,
+      "loss": 0.1178,
+      "step": 4979
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.23563041358295203,
+      "learning_rate": 6.52092578888548e-06,
+      "loss": 0.05,
+      "step": 4980
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2544305556475782,
+      "learning_rate": 6.519625792298311e-06,
+      "loss": 0.086,
+      "step": 4981
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.30245869858540275,
+      "learning_rate": 6.518325682520561e-06,
+      "loss": 0.0565,
+      "step": 4982
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5057301390307193,
+      "learning_rate": 6.5170254596490675e-06,
+      "loss": 0.1195,
+      "step": 4983
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2504988483983231,
+      "learning_rate": 6.515725123780681e-06,
+      "loss": 0.0701,
+      "step": 4984
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.31464784055165174,
+      "learning_rate": 6.514424675012259e-06,
+      "loss": 0.0749,
+      "step": 4985
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3384801464424417,
+      "learning_rate": 6.513124113440664e-06,
+      "loss": 0.1057,
+      "step": 4986
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3354224291961883,
+      "learning_rate": 6.51182343916277e-06,
+      "loss": 0.0791,
+      "step": 4987
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4848671341745712,
+      "learning_rate": 6.5105226522754595e-06,
+      "loss": 0.134,
+      "step": 4988
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2591184890045394,
+      "learning_rate": 6.509221752875623e-06,
+      "loss": 0.0563,
+      "step": 4989
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4125943355897507,
+      "learning_rate": 6.507920741060158e-06,
+      "loss": 0.097,
+      "step": 4990
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.22542728972058484,
+      "learning_rate": 6.506619616925972e-06,
+      "loss": 0.0595,
+      "step": 4991
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.21044219942430617,
+      "learning_rate": 6.505318380569981e-06,
+      "loss": 0.0631,
+      "step": 4992
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3066235737812797,
+      "learning_rate": 6.504017032089106e-06,
+      "loss": 0.0669,
+      "step": 4993
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.34139031567789946,
+      "learning_rate": 6.502715571580282e-06,
+      "loss": 0.0777,
+      "step": 4994
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6212185286482573,
+      "learning_rate": 6.501413999140448e-06,
+      "loss": 0.1636,
+      "step": 4995
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3045151125150769,
+      "learning_rate": 6.500112314866553e-06,
+      "loss": 0.0769,
+      "step": 4996
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.23479195078238124,
+      "learning_rate": 6.498810518855555e-06,
+      "loss": 0.0658,
+      "step": 4997
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2935654694196785,
+      "learning_rate": 6.497508611204417e-06,
+      "loss": 0.0748,
+      "step": 4998
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3815338718524593,
+      "learning_rate": 6.496206592010114e-06,
+      "loss": 0.1042,
+      "step": 4999
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.31586932692974995,
+      "learning_rate": 6.49490446136963e-06,
+      "loss": 0.0986,
+      "step": 5000
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.26299610805225904,
+      "learning_rate": 6.493602219379953e-06,
+      "loss": 0.0696,
+      "step": 5001
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.21825462729415152,
+      "learning_rate": 6.49229986613808e-06,
+      "loss": 0.0501,
+      "step": 5002
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4494672288061924,
+      "learning_rate": 6.49099740174102e-06,
+      "loss": 0.1074,
+      "step": 5003
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.27087500929515734,
+      "learning_rate": 6.4896948262857885e-06,
+      "loss": 0.0669,
+      "step": 5004
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.24618049818754711,
+      "learning_rate": 6.488392139869407e-06,
+      "loss": 0.0919,
+      "step": 5005
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.28070133798760527,
+      "learning_rate": 6.4870893425889104e-06,
+      "loss": 0.0921,
+      "step": 5006
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3287066414001967,
+      "learning_rate": 6.485786434541335e-06,
+      "loss": 0.0805,
+      "step": 5007
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.26893340573755653,
+      "learning_rate": 6.484483415823731e-06,
+      "loss": 0.067,
+      "step": 5008
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.26221526526767114,
+      "learning_rate": 6.483180286533155e-06,
+      "loss": 0.0676,
+      "step": 5009
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4930738325877774,
+      "learning_rate": 6.4818770467666695e-06,
+      "loss": 0.0992,
+      "step": 5010
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.31196998865454895,
+      "learning_rate": 6.480573696621351e-06,
+      "loss": 0.0682,
+      "step": 5011
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3381343640589562,
+      "learning_rate": 6.479270236194277e-06,
+      "loss": 0.1295,
+      "step": 5012
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.28425764608852916,
+      "learning_rate": 6.47796666558254e-06,
+      "loss": 0.054,
+      "step": 5013
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3619156883705878,
+      "learning_rate": 6.476662984883234e-06,
+      "loss": 0.0941,
+      "step": 5014
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.33919449563345655,
+      "learning_rate": 6.475359194193469e-06,
+      "loss": 0.097,
+      "step": 5015
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.29035217697509946,
+      "learning_rate": 6.474055293610355e-06,
+      "loss": 0.0858,
+      "step": 5016
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3070955024663146,
+      "learning_rate": 6.472751283231016e-06,
+      "loss": 0.08,
+      "step": 5017
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2767586795416823,
+      "learning_rate": 6.4714471631525825e-06,
+      "loss": 0.0788,
+      "step": 5018
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.37509409073377437,
+      "learning_rate": 6.470142933472191e-06,
+      "loss": 0.083,
+      "step": 5019
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3364692149702103,
+      "learning_rate": 6.4688385942869915e-06,
+      "loss": 0.0933,
+      "step": 5020
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2771744733565043,
+      "learning_rate": 6.467534145694135e-06,
+      "loss": 0.0749,
+      "step": 5021
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5487536079881584,
+      "learning_rate": 6.4662295877907865e-06,
+      "loss": 0.1075,
+      "step": 5022
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3466256618887153,
+      "learning_rate": 6.464924920674119e-06,
+      "loss": 0.0991,
+      "step": 5023
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.25854092705346277,
+      "learning_rate": 6.463620144441308e-06,
+      "loss": 0.0864,
+      "step": 5024
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2954208713445148,
+      "learning_rate": 6.4623152591895434e-06,
+      "loss": 0.0731,
+      "step": 5025
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3121212028468487,
+      "learning_rate": 6.461010265016019e-06,
+      "loss": 0.0706,
+      "step": 5026
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3138915821341992,
+      "learning_rate": 6.4597051620179385e-06,
+      "loss": 0.0888,
+      "step": 5027
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2329737267059408,
+      "learning_rate": 6.458399950292515e-06,
+      "loss": 0.0479,
+      "step": 5028
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2588992022431299,
+      "learning_rate": 6.457094629936966e-06,
+      "loss": 0.0828,
+      "step": 5029
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.52556991309081,
+      "learning_rate": 6.455789201048523e-06,
+      "loss": 0.1128,
+      "step": 5030
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2768393560829019,
+      "learning_rate": 6.454483663724418e-06,
+      "loss": 0.0515,
+      "step": 5031
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6221289764629585,
+      "learning_rate": 6.453178018061899e-06,
+      "loss": 0.1123,
+      "step": 5032
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2005402435005419,
+      "learning_rate": 6.451872264158213e-06,
+      "loss": 0.0913,
+      "step": 5033
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.36300848899648425,
+      "learning_rate": 6.4505664021106255e-06,
+      "loss": 0.0944,
+      "step": 5034
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.30923208105798594,
+      "learning_rate": 6.4492604320164e-06,
+      "loss": 0.0933,
+      "step": 5035
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6123579343408642,
+      "learning_rate": 6.447954353972816e-06,
+      "loss": 0.1626,
+      "step": 5036
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.29662012201504695,
+      "learning_rate": 6.446648168077157e-06,
+      "loss": 0.0982,
+      "step": 5037
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.34551469150360464,
+      "learning_rate": 6.445341874426714e-06,
+      "loss": 0.098,
+      "step": 5038
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3596943507309376,
+      "learning_rate": 6.444035473118788e-06,
+      "loss": 0.095,
+      "step": 5039
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3355914464076175,
+      "learning_rate": 6.442728964250691e-06,
+      "loss": 0.0948,
+      "step": 5040
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4124424022259644,
+      "learning_rate": 6.441422347919733e-06,
+      "loss": 0.1159,
+      "step": 5041
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.27819871482231173,
+      "learning_rate": 6.4401156242232435e-06,
+      "loss": 0.069,
+      "step": 5042
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.32939357764806776,
+      "learning_rate": 6.438808793258552e-06,
+      "loss": 0.0807,
+      "step": 5043
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.42129641238343774,
+      "learning_rate": 6.437501855123e-06,
+      "loss": 0.0913,
+      "step": 5044
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2686422173490415,
+      "learning_rate": 6.436194809913936e-06,
+      "loss": 0.0819,
+      "step": 5045
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2581328510174416,
+      "learning_rate": 6.434887657728716e-06,
+      "loss": 0.0579,
+      "step": 5046
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4023706940830099,
+      "learning_rate": 6.433580398664705e-06,
+      "loss": 0.0999,
+      "step": 5047
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.27784003810290475,
+      "learning_rate": 6.432273032819273e-06,
+      "loss": 0.0681,
+      "step": 5048
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.21007825624320725,
+      "learning_rate": 6.430965560289803e-06,
+      "loss": 0.0691,
+      "step": 5049
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.30586526560767807,
+      "learning_rate": 6.429657981173683e-06,
+      "loss": 0.1043,
+      "step": 5050
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.307201982009912,
+      "learning_rate": 6.4283502955683065e-06,
+      "loss": 0.0707,
+      "step": 5051
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.31925414456324913,
+      "learning_rate": 6.427042503571081e-06,
+      "loss": 0.081,
+      "step": 5052
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.22836829621777202,
+      "learning_rate": 6.425734605279414e-06,
+      "loss": 0.0809,
+      "step": 5053
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.346242521474331,
+      "learning_rate": 6.4244266007907304e-06,
+      "loss": 0.0901,
+      "step": 5054
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.30662731802114246,
+      "learning_rate": 6.423118490202456e-06,
+      "loss": 0.0734,
+      "step": 5055
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.28587163179015795,
+      "learning_rate": 6.421810273612023e-06,
+      "loss": 0.0806,
+      "step": 5056
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.25576872866339184,
+      "learning_rate": 6.420501951116882e-06,
+      "loss": 0.0931,
+      "step": 5057
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.23019490471316545,
+      "learning_rate": 6.419193522814478e-06,
+      "loss": 0.0698,
+      "step": 5058
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2987212789799524,
+      "learning_rate": 6.417884988802275e-06,
+      "loss": 0.0646,
+      "step": 5059
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.1940393800406046,
+      "learning_rate": 6.4165763491777365e-06,
+      "loss": 0.0524,
+      "step": 5060
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4575970419058964,
+      "learning_rate": 6.415267604038342e-06,
+      "loss": 0.1352,
+      "step": 5061
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.22992657309965417,
+      "learning_rate": 6.413958753481569e-06,
+      "loss": 0.0842,
+      "step": 5062
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.49487432506669243,
+      "learning_rate": 6.4126497976049115e-06,
+      "loss": 0.1274,
+      "step": 5063
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.22582812423344603,
+      "learning_rate": 6.411340736505869e-06,
+      "loss": 0.0581,
+      "step": 5064
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.33076147771303804,
+      "learning_rate": 6.410031570281946e-06,
+      "loss": 0.0864,
+      "step": 5065
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3878302169174842,
+      "learning_rate": 6.408722299030658e-06,
+      "loss": 0.1248,
+      "step": 5066
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2896374240751767,
+      "learning_rate": 6.4074129228495265e-06,
+      "loss": 0.0693,
+      "step": 5067
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2799746839039772,
+      "learning_rate": 6.406103441836082e-06,
+      "loss": 0.0854,
+      "step": 5068
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.24553549859245666,
+      "learning_rate": 6.404793856087863e-06,
+      "loss": 0.074,
+      "step": 5069
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2873029701818784,
+      "learning_rate": 6.403484165702411e-06,
+      "loss": 0.0845,
+      "step": 5070
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.43435135945478764,
+      "learning_rate": 6.402174370777285e-06,
+      "loss": 0.1004,
+      "step": 5071
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.39628190239385086,
+      "learning_rate": 6.400864471410043e-06,
+      "loss": 0.0876,
+      "step": 5072
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.35314246297869617,
+      "learning_rate": 6.399554467698255e-06,
+      "loss": 0.064,
+      "step": 5073
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3910024156789632,
+      "learning_rate": 6.398244359739496e-06,
+      "loss": 0.098,
+      "step": 5074
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2535591210565408,
+      "learning_rate": 6.396934147631352e-06,
+      "loss": 0.0592,
+      "step": 5075
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3224967862438788,
+      "learning_rate": 6.395623831471416e-06,
+      "loss": 0.0664,
+      "step": 5076
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3907403649438304,
+      "learning_rate": 6.394313411357286e-06,
+      "loss": 0.0742,
+      "step": 5077
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.37685386758012246,
+      "learning_rate": 6.393002887386571e-06,
+      "loss": 0.0754,
+      "step": 5078
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.29114945029243683,
+      "learning_rate": 6.391692259656885e-06,
+      "loss": 0.0863,
+      "step": 5079
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4426190922804583,
+      "learning_rate": 6.390381528265853e-06,
+      "loss": 0.1026,
+      "step": 5080
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.33878778832001055,
+      "learning_rate": 6.389070693311106e-06,
+      "loss": 0.1073,
+      "step": 5081
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4866856531092809,
+      "learning_rate": 6.387759754890281e-06,
+      "loss": 0.1023,
+      "step": 5082
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.21902928550672768,
+      "learning_rate": 6.386448713101025e-06,
+      "loss": 0.0593,
+      "step": 5083
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.35981973601339395,
+      "learning_rate": 6.385137568040993e-06,
+      "loss": 0.0885,
+      "step": 5084
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.26959102786023414,
+      "learning_rate": 6.383826319807845e-06,
+      "loss": 0.078,
+      "step": 5085
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3996308244919742,
+      "learning_rate": 6.382514968499253e-06,
+      "loss": 0.1414,
+      "step": 5086
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3732410680366653,
+      "learning_rate": 6.3812035142128905e-06,
+      "loss": 0.0762,
+      "step": 5087
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.32304479548192916,
+      "learning_rate": 6.379891957046445e-06,
+      "loss": 0.0774,
+      "step": 5088
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.16943076262752108,
+      "learning_rate": 6.378580297097607e-06,
+      "loss": 0.0526,
+      "step": 5089
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2990947234421667,
+      "learning_rate": 6.3772685344640795e-06,
+      "loss": 0.0782,
+      "step": 5090
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3447900748068334,
+      "learning_rate": 6.375956669243568e-06,
+      "loss": 0.0893,
+      "step": 5091
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3455460735605349,
+      "learning_rate": 6.374644701533788e-06,
+      "loss": 0.0728,
+      "step": 5092
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.30185887116342336,
+      "learning_rate": 6.373332631432462e-06,
+      "loss": 0.0859,
+      "step": 5093
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4259665868930272,
+      "learning_rate": 6.372020459037321e-06,
+      "loss": 0.0833,
+      "step": 5094
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2919816365827332,
+      "learning_rate": 6.370708184446105e-06,
+      "loss": 0.0606,
+      "step": 5095
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.33598112225093646,
+      "learning_rate": 6.369395807756557e-06,
+      "loss": 0.0826,
+      "step": 5096
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.22720882870985729,
+      "learning_rate": 6.368083329066432e-06,
+      "loss": 0.0661,
+      "step": 5097
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9077888446790929,
+      "learning_rate": 6.366770748473493e-06,
+      "loss": 0.1163,
+      "step": 5098
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.48209921357093444,
+      "learning_rate": 6.365458066075505e-06,
+      "loss": 0.1113,
+      "step": 5099
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.27089999608786985,
+      "learning_rate": 6.364145281970247e-06,
+      "loss": 0.073,
+      "step": 5100
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2325847909295708,
+      "learning_rate": 6.3628323962555e-06,
+      "loss": 0.0691,
+      "step": 5101
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4536122346623705,
+      "learning_rate": 6.361519409029059e-06,
+      "loss": 0.1187,
+      "step": 5102
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.27630816563069316,
+      "learning_rate": 6.36020632038872e-06,
+      "loss": 0.0728,
+      "step": 5103
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.253075229197033,
+      "learning_rate": 6.3588931304322905e-06,
+      "loss": 0.0596,
+      "step": 5104
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.39107903850366826,
+      "learning_rate": 6.3575798392575846e-06,
+      "loss": 0.0904,
+      "step": 5105
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.37701026116821723,
+      "learning_rate": 6.3562664469624235e-06,
+      "loss": 0.0991,
+      "step": 5106
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.487250291271705,
+      "learning_rate": 6.354952953644636e-06,
+      "loss": 0.1233,
+      "step": 5107
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5460562821845559,
+      "learning_rate": 6.353639359402061e-06,
+      "loss": 0.1069,
+      "step": 5108
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.39750831422381533,
+      "learning_rate": 6.352325664332539e-06,
+      "loss": 0.1039,
+      "step": 5109
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.638983494765347,
+      "learning_rate": 6.351011868533925e-06,
+      "loss": 0.1118,
+      "step": 5110
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5905910254166707,
+      "learning_rate": 6.349697972104076e-06,
+      "loss": 0.0851,
+      "step": 5111
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.24767387021109064,
+      "learning_rate": 6.348383975140862e-06,
+      "loss": 0.073,
+      "step": 5112
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.294039794615177,
+      "learning_rate": 6.347069877742151e-06,
+      "loss": 0.0716,
+      "step": 5113
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2886262464525754,
+      "learning_rate": 6.34575568000583e-06,
+      "loss": 0.0808,
+      "step": 5114
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4128316445687048,
+      "learning_rate": 6.344441382029787e-06,
+      "loss": 0.1062,
+      "step": 5115
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.38302340055662404,
+      "learning_rate": 6.343126983911916e-06,
+      "loss": 0.0892,
+      "step": 5116
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3411276395220711,
+      "learning_rate": 6.341812485750124e-06,
+      "loss": 0.0882,
+      "step": 5117
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.26660827309731244,
+      "learning_rate": 6.3404978876423205e-06,
+      "loss": 0.0649,
+      "step": 5118
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4145834756708014,
+      "learning_rate": 6.339183189686426e-06,
+      "loss": 0.1093,
+      "step": 5119
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.25612640293289035,
+      "learning_rate": 6.337868391980367e-06,
+      "loss": 0.0902,
+      "step": 5120
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4237678002107076,
+      "learning_rate": 6.336553494622077e-06,
+      "loss": 0.0982,
+      "step": 5121
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4361783140165895,
+      "learning_rate": 6.335238497709495e-06,
+      "loss": 0.1223,
+      "step": 5122
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.27802344553059266,
+      "learning_rate": 6.333923401340572e-06,
+      "loss": 0.0786,
+      "step": 5123
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2993314997940703,
+      "learning_rate": 6.332608205613264e-06,
+      "loss": 0.1006,
+      "step": 5124
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.35561042437654855,
+      "learning_rate": 6.331292910625533e-06,
+      "loss": 0.0536,
+      "step": 5125
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3949167641028964,
+      "learning_rate": 6.3299775164753495e-06,
+      "loss": 0.1104,
+      "step": 5126
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.49954067696945664,
+      "learning_rate": 6.328662023260696e-06,
+      "loss": 0.0854,
+      "step": 5127
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2538857532554658,
+      "learning_rate": 6.327346431079553e-06,
+      "loss": 0.059,
+      "step": 5128
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.23857050563486423,
+      "learning_rate": 6.326030740029916e-06,
+      "loss": 0.1012,
+      "step": 5129
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.28855503017043477,
+      "learning_rate": 6.324714950209784e-06,
+      "loss": 0.0741,
+      "step": 5130
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3936977739714567,
+      "learning_rate": 6.323399061717165e-06,
+      "loss": 0.0918,
+      "step": 5131
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7527307133192241,
+      "learning_rate": 6.322083074650076e-06,
+      "loss": 0.137,
+      "step": 5132
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4874735487182669,
+      "learning_rate": 6.320766989106537e-06,
+      "loss": 0.1113,
+      "step": 5133
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2870276148346911,
+      "learning_rate": 6.319450805184579e-06,
+      "loss": 0.0674,
+      "step": 5134
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4325594125606785,
+      "learning_rate": 6.318134522982239e-06,
+      "loss": 0.1096,
+      "step": 5135
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.18184279655658253,
+      "learning_rate": 6.316818142597562e-06,
+      "loss": 0.0693,
+      "step": 5136
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.289487516660886,
+      "learning_rate": 6.315501664128596e-06,
+      "loss": 0.0764,
+      "step": 5137
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2621708938617668,
+      "learning_rate": 6.314185087673404e-06,
+      "loss": 0.0833,
+      "step": 5138
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.33076434156815726,
+      "learning_rate": 6.312868413330052e-06,
+      "loss": 0.1065,
+      "step": 5139
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4055905024743049,
+      "learning_rate": 6.311551641196612e-06,
+      "loss": 0.1007,
+      "step": 5140
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5533421547127563,
+      "learning_rate": 6.310234771371165e-06,
+      "loss": 0.1889,
+      "step": 5141
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.38411423123472727,
+      "learning_rate": 6.308917803951799e-06,
+      "loss": 0.1193,
+      "step": 5142
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3172037829610849,
+      "learning_rate": 6.30760073903661e-06,
+      "loss": 0.0848,
+      "step": 5143
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.29420732700167507,
+      "learning_rate": 6.306283576723702e-06,
+      "loss": 0.0852,
+      "step": 5144
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.33592250478861607,
+      "learning_rate": 6.304966317111183e-06,
+      "loss": 0.097,
+      "step": 5145
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7618676670534758,
+      "learning_rate": 6.30364896029717e-06,
+      "loss": 0.1775,
+      "step": 5146
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3046605235732828,
+      "learning_rate": 6.302331506379789e-06,
+      "loss": 0.0601,
+      "step": 5147
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3172669658320801,
+      "learning_rate": 6.301013955457169e-06,
+      "loss": 0.0835,
+      "step": 5148
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.21588165888739544,
+      "learning_rate": 6.2996963076274535e-06,
+      "loss": 0.053,
+      "step": 5149
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3379512095788092,
+      "learning_rate": 6.298378562988783e-06,
+      "loss": 0.1037,
+      "step": 5150
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.32430884701003304,
+      "learning_rate": 6.2970607216393154e-06,
+      "loss": 0.0842,
+      "step": 5151
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.34240021569397755,
+      "learning_rate": 6.2957427836772075e-06,
+      "loss": 0.0836,
+      "step": 5152
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3437104889896471,
+      "learning_rate": 6.294424749200629e-06,
+      "loss": 0.127,
+      "step": 5153
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3417632146071153,
+      "learning_rate": 6.293106618307757e-06,
+      "loss": 0.0921,
+      "step": 5154
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.18062797482194565,
+      "learning_rate": 6.291788391096769e-06,
+      "loss": 0.0638,
+      "step": 5155
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3082937560199461,
+      "learning_rate": 6.290470067665858e-06,
+      "loss": 0.0757,
+      "step": 5156
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5238024527913562,
+      "learning_rate": 6.289151648113217e-06,
+      "loss": 0.1118,
+      "step": 5157
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2651797135705574,
+      "learning_rate": 6.287833132537053e-06,
+      "loss": 0.0627,
+      "step": 5158
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5273501862063448,
+      "learning_rate": 6.286514521035574e-06,
+      "loss": 0.1159,
+      "step": 5159
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2637778275733897,
+      "learning_rate": 6.285195813706999e-06,
+      "loss": 0.0912,
+      "step": 5160
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.22590963850321447,
+      "learning_rate": 6.2838770106495535e-06,
+      "loss": 0.0678,
+      "step": 5161
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.29320572091763986,
+      "learning_rate": 6.282558111961469e-06,
+      "loss": 0.0925,
+      "step": 5162
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5279643431001139,
+      "learning_rate": 6.281239117740985e-06,
+      "loss": 0.1056,
+      "step": 5163
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6136431513238707,
+      "learning_rate": 6.279920028086347e-06,
+      "loss": 0.1705,
+      "step": 5164
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.19863772121577403,
+      "learning_rate": 6.278600843095809e-06,
+      "loss": 0.0699,
+      "step": 5165
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.25559180686480315,
+      "learning_rate": 6.2772815628676335e-06,
+      "loss": 0.0986,
+      "step": 5166
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.36503701428093804,
+      "learning_rate": 6.275962187500084e-06,
+      "loss": 0.1019,
+      "step": 5167
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.32045961123847094,
+      "learning_rate": 6.2746427170914405e-06,
+      "loss": 0.0555,
+      "step": 5168
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.26837908560880663,
+      "learning_rate": 6.27332315173998e-06,
+      "loss": 0.1003,
+      "step": 5169
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.26231090290125614,
+      "learning_rate": 6.272003491543995e-06,
+      "loss": 0.0681,
+      "step": 5170
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.34422761593659307,
+      "learning_rate": 6.270683736601779e-06,
+      "loss": 0.1146,
+      "step": 5171
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.37413142457805704,
+      "learning_rate": 6.269363887011637e-06,
+      "loss": 0.0924,
+      "step": 5172
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5052662802453262,
+      "learning_rate": 6.2680439428718784e-06,
+      "loss": 0.073,
+      "step": 5173
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.48362656473537935,
+      "learning_rate": 6.2667239042808185e-06,
+      "loss": 0.0894,
+      "step": 5174
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.22522435208419184,
+      "learning_rate": 6.265403771336784e-06,
+      "loss": 0.0557,
+      "step": 5175
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.30394007244896476,
+      "learning_rate": 6.264083544138104e-06,
+      "loss": 0.0644,
+      "step": 5176
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2664992122379471,
+      "learning_rate": 6.2627632227831194e-06,
+      "loss": 0.0848,
+      "step": 5177
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.19992770427447581,
+      "learning_rate": 6.261442807370173e-06,
+      "loss": 0.0518,
+      "step": 5178
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4142092988716273,
+      "learning_rate": 6.260122297997617e-06,
+      "loss": 0.0907,
+      "step": 5179
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.41650049081107376,
+      "learning_rate": 6.258801694763813e-06,
+      "loss": 0.113,
+      "step": 5180
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.41747702078138804,
+      "learning_rate": 6.257480997767125e-06,
+      "loss": 0.1082,
+      "step": 5181
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4009676149867442,
+      "learning_rate": 6.256160207105925e-06,
+      "loss": 0.0788,
+      "step": 5182
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.31045376154767057,
+      "learning_rate": 6.254839322878598e-06,
+      "loss": 0.0988,
+      "step": 5183
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3322045467570745,
+      "learning_rate": 6.253518345183527e-06,
+      "loss": 0.0678,
+      "step": 5184
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7139718357199312,
+      "learning_rate": 6.252197274119108e-06,
+      "loss": 0.0785,
+      "step": 5185
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3674088151430725,
+      "learning_rate": 6.250876109783741e-06,
+      "loss": 0.0972,
+      "step": 5186
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.1996350170747824,
+      "learning_rate": 6.249554852275835e-06,
+      "loss": 0.0587,
+      "step": 5187
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.21907129191713076,
+      "learning_rate": 6.248233501693803e-06,
+      "loss": 0.0592,
+      "step": 5188
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.34069128790062725,
+      "learning_rate": 6.24691205813607e-06,
+      "loss": 0.0913,
+      "step": 5189
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.26382432394386207,
+      "learning_rate": 6.245590521701063e-06,
+      "loss": 0.076,
+      "step": 5190
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3866148334623173,
+      "learning_rate": 6.244268892487217e-06,
+      "loss": 0.1073,
+      "step": 5191
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6201201587266143,
+      "learning_rate": 6.242947170592975e-06,
+      "loss": 0.1169,
+      "step": 5192
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.30369501699517293,
+      "learning_rate": 6.241625356116787e-06,
+      "loss": 0.0589,
+      "step": 5193
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.44604671621660175,
+      "learning_rate": 6.2403034491571105e-06,
+      "loss": 0.0878,
+      "step": 5194
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2179603961583201,
+      "learning_rate": 6.238981449812408e-06,
+      "loss": 0.0653,
+      "step": 5195
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2624839034369519,
+      "learning_rate": 6.237659358181149e-06,
+      "loss": 0.0784,
+      "step": 5196
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.34287357104560556,
+      "learning_rate": 6.23633717436181e-06,
+      "loss": 0.0779,
+      "step": 5197
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3008478494192351,
+      "learning_rate": 6.235014898452878e-06,
+      "loss": 0.077,
+      "step": 5198
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.24984795210907965,
+      "learning_rate": 6.23369253055284e-06,
+      "loss": 0.0597,
+      "step": 5199
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.35262018409534923,
+      "learning_rate": 6.2323700707601976e-06,
+      "loss": 0.1057,
+      "step": 5200
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.33599549759870734,
+      "learning_rate": 6.231047519173451e-06,
+      "loss": 0.0979,
+      "step": 5201
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.43934572836200547,
+      "learning_rate": 6.2297248758911155e-06,
+      "loss": 0.085,
+      "step": 5202
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2707930690238642,
+      "learning_rate": 6.228402141011706e-06,
+      "loss": 0.0363,
+      "step": 5203
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2857447055533513,
+      "learning_rate": 6.227079314633752e-06,
+      "loss": 0.0743,
+      "step": 5204
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.35087000536764146,
+      "learning_rate": 6.22575639685578e-06,
+      "loss": 0.0652,
+      "step": 5205
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.29241830136015806,
+      "learning_rate": 6.224433387776332e-06,
+      "loss": 0.0875,
+      "step": 5206
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4078855011623417,
+      "learning_rate": 6.223110287493952e-06,
+      "loss": 0.0724,
+      "step": 5207
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.32130080820204526,
+      "learning_rate": 6.221787096107192e-06,
+      "loss": 0.0836,
+      "step": 5208
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.32063810201623266,
+      "learning_rate": 6.220463813714614e-06,
+      "loss": 0.0632,
+      "step": 5209
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3177714705667108,
+      "learning_rate": 6.219140440414778e-06,
+      "loss": 0.0957,
+      "step": 5210
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.24373807562540997,
+      "learning_rate": 6.2178169763062614e-06,
+      "loss": 0.0536,
+      "step": 5211
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4192546314142608,
+      "learning_rate": 6.216493421487643e-06,
+      "loss": 0.1231,
+      "step": 5212
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.455772876853517,
+      "learning_rate": 6.215169776057507e-06,
+      "loss": 0.084,
+      "step": 5213
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.33006179647951395,
+      "learning_rate": 6.2138460401144475e-06,
+      "loss": 0.0295,
+      "step": 5214
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8422524354794191,
+      "learning_rate": 6.212522213757064e-06,
+      "loss": 0.1792,
+      "step": 5215
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.18670697645153825,
+      "learning_rate": 6.211198297083961e-06,
+      "loss": 0.038,
+      "step": 5216
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2764012360566822,
+      "learning_rate": 6.209874290193755e-06,
+      "loss": 0.093,
+      "step": 5217
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4644613658355343,
+      "learning_rate": 6.208550193185063e-06,
+      "loss": 0.1174,
+      "step": 5218
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.16372993976411004,
+      "learning_rate": 6.207226006156512e-06,
+      "loss": 0.0455,
+      "step": 5219
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.268957705241708,
+      "learning_rate": 6.205901729206736e-06,
+      "loss": 0.0723,
+      "step": 5220
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.1879080371373582,
+      "learning_rate": 6.204577362434375e-06,
+      "loss": 0.0395,
+      "step": 5221
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.18541568260756455,
+      "learning_rate": 6.203252905938074e-06,
+      "loss": 0.0589,
+      "step": 5222
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.21184262704789444,
+      "learning_rate": 6.201928359816487e-06,
+      "loss": 0.0651,
+      "step": 5223
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.24888988392936323,
+      "learning_rate": 6.200603724168275e-06,
+      "loss": 0.0859,
+      "step": 5224
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.20855703975492812,
+      "learning_rate": 6.199278999092102e-06,
+      "loss": 0.053,
+      "step": 5225
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.32001081232409684,
+      "learning_rate": 6.197954184686644e-06,
+      "loss": 0.0718,
+      "step": 5226
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.40238514890779303,
+      "learning_rate": 6.196629281050579e-06,
+      "loss": 0.0674,
+      "step": 5227
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.34154181015128565,
+      "learning_rate": 6.195304288282594e-06,
+      "loss": 0.0895,
+      "step": 5228
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.23415557798416675,
+      "learning_rate": 6.1939792064813846e-06,
+      "loss": 0.0898,
+      "step": 5229
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.374382807663893,
+      "learning_rate": 6.192654035745646e-06,
+      "loss": 0.1142,
+      "step": 5230
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.28467842116791237,
+      "learning_rate": 6.191328776174088e-06,
+      "loss": 0.0915,
+      "step": 5231
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3292240351475248,
+      "learning_rate": 6.190003427865422e-06,
+      "loss": 0.1224,
+      "step": 5232
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.25746160314004013,
+      "learning_rate": 6.188677990918369e-06,
+      "loss": 0.0737,
+      "step": 5233
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.18337006351164734,
+      "learning_rate": 6.187352465431655e-06,
+      "loss": 0.0237,
+      "step": 5234
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.16741479630572878,
+      "learning_rate": 6.186026851504012e-06,
+      "loss": 0.046,
+      "step": 5235
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.31670212416517907,
+      "learning_rate": 6.184701149234181e-06,
+      "loss": 0.1054,
+      "step": 5236
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3256155635478663,
+      "learning_rate": 6.183375358720907e-06,
+      "loss": 0.0973,
+      "step": 5237
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3372282515677099,
+      "learning_rate": 6.1820494800629415e-06,
+      "loss": 0.0573,
+      "step": 5238
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.14162739478644318,
+      "learning_rate": 6.180723513359045e-06,
+      "loss": 0.0258,
+      "step": 5239
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2614610261841953,
+      "learning_rate": 6.179397458707984e-06,
+      "loss": 0.0743,
+      "step": 5240
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.45798692754897236,
+      "learning_rate": 6.178071316208529e-06,
+      "loss": 0.1214,
+      "step": 5241
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2606433617573775,
+      "learning_rate": 6.17674508595946e-06,
+      "loss": 0.0687,
+      "step": 5242
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3858788305461409,
+      "learning_rate": 6.175418768059562e-06,
+      "loss": 0.0864,
+      "step": 5243
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.22879257836160302,
+      "learning_rate": 6.174092362607627e-06,
+      "loss": 0.0567,
+      "step": 5244
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.31052984314295756,
+      "learning_rate": 6.1727658697024505e-06,
+      "loss": 0.0564,
+      "step": 5245
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3993084983310892,
+      "learning_rate": 6.171439289442843e-06,
+      "loss": 0.0819,
+      "step": 5246
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.21731570647463253,
+      "learning_rate": 6.170112621927613e-06,
+      "loss": 0.0482,
+      "step": 5247
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4334208200535504,
+      "learning_rate": 6.168785867255578e-06,
+      "loss": 0.1036,
+      "step": 5248
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2757908625035082,
+      "learning_rate": 6.1674590255255625e-06,
+      "loss": 0.0694,
+      "step": 5249
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.29254399750667776,
+      "learning_rate": 6.166132096836399e-06,
+      "loss": 0.0488,
+      "step": 5250
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.36610907681238747,
+      "learning_rate": 6.164805081286922e-06,
+      "loss": 0.0788,
+      "step": 5251
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6990823988400623,
+      "learning_rate": 6.163477978975978e-06,
+      "loss": 0.0912,
+      "step": 5252
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3985405688150222,
+      "learning_rate": 6.162150790002417e-06,
+      "loss": 0.0715,
+      "step": 5253
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.32863635148769466,
+      "learning_rate": 6.160823514465093e-06,
+      "loss": 0.0805,
+      "step": 5254
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.1964920384892367,
+      "learning_rate": 6.1594961524628735e-06,
+      "loss": 0.0396,
+      "step": 5255
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3291273929905797,
+      "learning_rate": 6.158168704094625e-06,
+      "loss": 0.0524,
+      "step": 5256
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.39900875170454553,
+      "learning_rate": 6.156841169459225e-06,
+      "loss": 0.1108,
+      "step": 5257
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.34635101732129814,
+      "learning_rate": 6.155513548655556e-06,
+      "loss": 0.115,
+      "step": 5258
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5700847961066906,
+      "learning_rate": 6.154185841782505e-06,
+      "loss": 0.1235,
+      "step": 5259
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5093442194896075,
+      "learning_rate": 6.152858048938971e-06,
+      "loss": 0.0983,
+      "step": 5260
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.40187369354248026,
+      "learning_rate": 6.151530170223852e-06,
+      "loss": 0.0716,
+      "step": 5261
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.23064170997759925,
+      "learning_rate": 6.1502022057360575e-06,
+      "loss": 0.0573,
+      "step": 5262
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4203572895242865,
+      "learning_rate": 6.1488741555745036e-06,
+      "loss": 0.0931,
+      "step": 5263
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5294580157460433,
+      "learning_rate": 6.147546019838109e-06,
+      "loss": 0.1184,
+      "step": 5264
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.307256022814954,
+      "learning_rate": 6.146217798625803e-06,
+      "loss": 0.1003,
+      "step": 5265
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.26042322270279156,
+      "learning_rate": 6.144889492036516e-06,
+      "loss": 0.051,
+      "step": 5266
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.29410465015178416,
+      "learning_rate": 6.143561100169193e-06,
+      "loss": 0.0669,
+      "step": 5267
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5136536547199391,
+      "learning_rate": 6.142232623122775e-06,
+      "loss": 0.1134,
+      "step": 5268
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.20144584319530556,
+      "learning_rate": 6.140904060996218e-06,
+      "loss": 0.0462,
+      "step": 5269
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1016368380696153,
+      "learning_rate": 6.139575413888481e-06,
+      "loss": 0.136,
+      "step": 5270
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.33286888319463154,
+      "learning_rate": 6.138246681898528e-06,
+      "loss": 0.0805,
+      "step": 5271
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.30880960494374793,
+      "learning_rate": 6.1369178651253325e-06,
+      "loss": 0.0616,
+      "step": 5272
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.25163701847294884,
+      "learning_rate": 6.13558896366787e-06,
+      "loss": 0.0708,
+      "step": 5273
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5165211940623574,
+      "learning_rate": 6.134259977625128e-06,
+      "loss": 0.1065,
+      "step": 5274
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3129297548190298,
+      "learning_rate": 6.132930907096094e-06,
+      "loss": 0.0555,
+      "step": 5275
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.27842362431688267,
+      "learning_rate": 6.131601752179765e-06,
+      "loss": 0.0585,
+      "step": 5276
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2918712735621903,
+      "learning_rate": 6.130272512975147e-06,
+      "loss": 0.0741,
+      "step": 5277
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.330251987805454,
+      "learning_rate": 6.128943189581248e-06,
+      "loss": 0.1309,
+      "step": 5278
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.23201982937692855,
+      "learning_rate": 6.1276137820970825e-06,
+      "loss": 0.0666,
+      "step": 5279
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.25492387900104096,
+      "learning_rate": 6.126284290621674e-06,
+      "loss": 0.0933,
+      "step": 5280
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2718308543836977,
+      "learning_rate": 6.124954715254051e-06,
+      "loss": 0.1145,
+      "step": 5281
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.47840132991708734,
+      "learning_rate": 6.123625056093248e-06,
+      "loss": 0.0779,
+      "step": 5282
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2403083600987502,
+      "learning_rate": 6.122295313238304e-06,
+      "loss": 0.0731,
+      "step": 5283
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2322454846665232,
+      "learning_rate": 6.120965486788268e-06,
+      "loss": 0.0516,
+      "step": 5284
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.20172763897515666,
+      "learning_rate": 6.119635576842192e-06,
+      "loss": 0.0615,
+      "step": 5285
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.1665346362962438,
+      "learning_rate": 6.118305583499135e-06,
+      "loss": 0.0635,
+      "step": 5286
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2628097734887062,
+      "learning_rate": 6.116975506858166e-06,
+      "loss": 0.0616,
+      "step": 5287
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.30032685559774974,
+      "learning_rate": 6.115645347018352e-06,
+      "loss": 0.0531,
+      "step": 5288
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4209492939459478,
+      "learning_rate": 6.1143151040787755e-06,
+      "loss": 0.1033,
+      "step": 5289
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3261313461188077,
+      "learning_rate": 6.112984778138516e-06,
+      "loss": 0.0854,
+      "step": 5290
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.13389376858375338,
+      "learning_rate": 6.111654369296669e-06,
+      "loss": 0.0358,
+      "step": 5291
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3445051395951685,
+      "learning_rate": 6.110323877652328e-06,
+      "loss": 0.1095,
+      "step": 5292
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3742259378281943,
+      "learning_rate": 6.108993303304596e-06,
+      "loss": 0.1113,
+      "step": 5293
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3218806643812006,
+      "learning_rate": 6.107662646352584e-06,
+      "loss": 0.0993,
+      "step": 5294
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.25177948362326225,
+      "learning_rate": 6.106331906895403e-06,
+      "loss": 0.0786,
+      "step": 5295
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.316242319175783,
+      "learning_rate": 6.105001085032178e-06,
+      "loss": 0.083,
+      "step": 5296
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.20377348322281413,
+      "learning_rate": 6.103670180862036e-06,
+      "loss": 0.0405,
+      "step": 5297
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2710851318933867,
+      "learning_rate": 6.102339194484107e-06,
+      "loss": 0.0645,
+      "step": 5298
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3639422896700517,
+      "learning_rate": 6.1010081259975366e-06,
+      "loss": 0.1263,
+      "step": 5299
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.41266385451860965,
+      "learning_rate": 6.099676975501464e-06,
+      "loss": 0.1056,
+      "step": 5300
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5356794486774135,
+      "learning_rate": 6.0983457430950466e-06,
+      "loss": 0.1075,
+      "step": 5301
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.34534845977928014,
+      "learning_rate": 6.097014428877438e-06,
+      "loss": 0.1018,
+      "step": 5302
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3434918712563387,
+      "learning_rate": 6.095683032947805e-06,
+      "loss": 0.0964,
+      "step": 5303
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.45976494885796537,
+      "learning_rate": 6.0943515554053165e-06,
+      "loss": 0.1267,
+      "step": 5304
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.33969733743416625,
+      "learning_rate": 6.0930199963491496e-06,
+      "loss": 0.0755,
+      "step": 5305
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2712116293292752,
+      "learning_rate": 6.0916883558784865e-06,
+      "loss": 0.0781,
+      "step": 5306
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3796384927373055,
+      "learning_rate": 6.090356634092513e-06,
+      "loss": 0.074,
+      "step": 5307
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3556636637485662,
+      "learning_rate": 6.089024831090429e-06,
+      "loss": 0.0584,
+      "step": 5308
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.19135626383308632,
+      "learning_rate": 6.08769294697143e-06,
+      "loss": 0.0418,
+      "step": 5309
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.442098643807956,
+      "learning_rate": 6.0863609818347235e-06,
+      "loss": 0.1043,
+      "step": 5310
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.20451789830048572,
+      "learning_rate": 6.085028935779523e-06,
+      "loss": 0.0589,
+      "step": 5311
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4664615389241614,
+      "learning_rate": 6.083696808905046e-06,
+      "loss": 0.127,
+      "step": 5312
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.36286836154522245,
+      "learning_rate": 6.082364601310518e-06,
+      "loss": 0.0945,
+      "step": 5313
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.34145616084273633,
+      "learning_rate": 6.08103231309517e-06,
+      "loss": 0.0832,
+      "step": 5314
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.24564504275421514,
+      "learning_rate": 6.079699944358236e-06,
+      "loss": 0.0679,
+      "step": 5315
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2836701054721182,
+      "learning_rate": 6.078367495198963e-06,
+      "loss": 0.0812,
+      "step": 5316
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3013302779259131,
+      "learning_rate": 6.077034965716594e-06,
+      "loss": 0.123,
+      "step": 5317
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2398990555826029,
+      "learning_rate": 6.075702356010389e-06,
+      "loss": 0.0607,
+      "step": 5318
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3291590887858968,
+      "learning_rate": 6.0743696661796045e-06,
+      "loss": 0.0703,
+      "step": 5319
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.20390315647094814,
+      "learning_rate": 6.073036896323508e-06,
+      "loss": 0.0903,
+      "step": 5320
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.30712055957935314,
+      "learning_rate": 6.0717040465413755e-06,
+      "loss": 0.1032,
+      "step": 5321
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3077468754094887,
+      "learning_rate": 6.070371116932481e-06,
+      "loss": 0.0854,
+      "step": 5322
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.33880899280042637,
+      "learning_rate": 6.0690381075961106e-06,
+      "loss": 0.0722,
+      "step": 5323
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.327678207447578,
+      "learning_rate": 6.067705018631553e-06,
+      "loss": 0.0594,
+      "step": 5324
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.35960992797103836,
+      "learning_rate": 6.066371850138106e-06,
+      "loss": 0.105,
+      "step": 5325
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5613633055349678,
+      "learning_rate": 6.065038602215073e-06,
+      "loss": 0.128,
+      "step": 5326
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.45614752188697993,
+      "learning_rate": 6.06370527496176e-06,
+      "loss": 0.0599,
+      "step": 5327
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.344320021583728,
+      "learning_rate": 6.0623718684774816e-06,
+      "loss": 0.1022,
+      "step": 5328
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2617636852821932,
+      "learning_rate": 6.061038382861557e-06,
+      "loss": 0.0788,
+      "step": 5329
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.531924953798233,
+      "learning_rate": 6.0597048182133135e-06,
+      "loss": 0.1478,
+      "step": 5330
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2865681723873686,
+      "learning_rate": 6.058371174632081e-06,
+      "loss": 0.0645,
+      "step": 5331
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5234901913648033,
+      "learning_rate": 6.057037452217198e-06,
+      "loss": 0.0883,
+      "step": 5332
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.1575635283649197,
+      "learning_rate": 6.055703651068009e-06,
+      "loss": 0.0469,
+      "step": 5333
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3830883326555604,
+      "learning_rate": 6.054369771283861e-06,
+      "loss": 0.11,
+      "step": 5334
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2565821085185686,
+      "learning_rate": 6.053035812964111e-06,
+      "loss": 0.0677,
+      "step": 5335
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.28226342991490844,
+      "learning_rate": 6.051701776208119e-06,
+      "loss": 0.0742,
+      "step": 5336
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.284647734978893,
+      "learning_rate": 6.050367661115251e-06,
+      "loss": 0.0705,
+      "step": 5337
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.35946088997305786,
+      "learning_rate": 6.049033467784883e-06,
+      "loss": 0.0903,
+      "step": 5338
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.20383545277218626,
+      "learning_rate": 6.047699196316389e-06,
+      "loss": 0.0753,
+      "step": 5339
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4639088175511074,
+      "learning_rate": 6.046364846809157e-06,
+      "loss": 0.0562,
+      "step": 5340
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2276505730846899,
+      "learning_rate": 6.045030419362575e-06,
+      "loss": 0.0656,
+      "step": 5341
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.44853535073360584,
+      "learning_rate": 6.043695914076041e-06,
+      "loss": 0.1008,
+      "step": 5342
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2719509404814123,
+      "learning_rate": 6.042361331048955e-06,
+      "loss": 0.0903,
+      "step": 5343
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4498368931815228,
+      "learning_rate": 6.041026670380723e-06,
+      "loss": 0.0878,
+      "step": 5344
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.42161491126361156,
+      "learning_rate": 6.039691932170762e-06,
+      "loss": 0.0938,
+      "step": 5345
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.45854043176362996,
+      "learning_rate": 6.038357116518489e-06,
+      "loss": 0.1193,
+      "step": 5346
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.26355164424531924,
+      "learning_rate": 6.037022223523328e-06,
+      "loss": 0.0837,
+      "step": 5347
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.28987998376162316,
+      "learning_rate": 6.035687253284711e-06,
+      "loss": 0.0949,
+      "step": 5348
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5031540499759544,
+      "learning_rate": 6.034352205902074e-06,
+      "loss": 0.113,
+      "step": 5349
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3088277986263113,
+      "learning_rate": 6.03301708147486e-06,
+      "loss": 0.0788,
+      "step": 5350
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5342678592105653,
+      "learning_rate": 6.031681880102515e-06,
+      "loss": 0.1196,
+      "step": 5351
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.17505078876655264,
+      "learning_rate": 6.030346601884494e-06,
+      "loss": 0.0412,
+      "step": 5352
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.23158006036776607,
+      "learning_rate": 6.029011246920257e-06,
+      "loss": 0.0669,
+      "step": 5353
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3866824063089273,
+      "learning_rate": 6.027675815309265e-06,
+      "loss": 0.068,
+      "step": 5354
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.507954179326806,
+      "learning_rate": 6.0263403071509955e-06,
+      "loss": 0.1028,
+      "step": 5355
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.201664224253523,
+      "learning_rate": 6.025004722544918e-06,
+      "loss": 0.0621,
+      "step": 5356
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2728142522919522,
+      "learning_rate": 6.023669061590519e-06,
+      "loss": 0.0799,
+      "step": 5357
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4242176150960372,
+      "learning_rate": 6.022333324387284e-06,
+      "loss": 0.094,
+      "step": 5358
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.253479321393804,
+      "learning_rate": 6.020997511034709e-06,
+      "loss": 0.083,
+      "step": 5359
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.34327452023089416,
+      "learning_rate": 6.01966162163229e-06,
+      "loss": 0.1044,
+      "step": 5360
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.419842383178155,
+      "learning_rate": 6.018325656279535e-06,
+      "loss": 0.1473,
+      "step": 5361
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.21744188842082793,
+      "learning_rate": 6.016989615075952e-06,
+      "loss": 0.0761,
+      "step": 5362
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.38278722044476043,
+      "learning_rate": 6.015653498121058e-06,
+      "loss": 0.1036,
+      "step": 5363
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2568964799739123,
+      "learning_rate": 6.014317305514376e-06,
+      "loss": 0.0578,
+      "step": 5364
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3228270671877277,
+      "learning_rate": 6.01298103735543e-06,
+      "loss": 0.0428,
+      "step": 5365
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4603803037031919,
+      "learning_rate": 6.011644693743756e-06,
+      "loss": 0.1034,
+      "step": 5366
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.29767446653696644,
+      "learning_rate": 6.010308274778893e-06,
+      "loss": 0.0595,
+      "step": 5367
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.24804175699966854,
+      "learning_rate": 6.008971780560384e-06,
+      "loss": 0.0615,
+      "step": 5368
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9415976368171958,
+      "learning_rate": 6.0076352111877785e-06,
+      "loss": 0.1487,
+      "step": 5369
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4642217571132315,
+      "learning_rate": 6.0062985667606336e-06,
+      "loss": 0.1361,
+      "step": 5370
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.24236977539073618,
+      "learning_rate": 6.004961847378508e-06,
+      "loss": 0.0696,
+      "step": 5371
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.31704934013023744,
+      "learning_rate": 6.003625053140971e-06,
+      "loss": 0.0877,
+      "step": 5372
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.20176125855510404,
+      "learning_rate": 6.0022881841475926e-06,
+      "loss": 0.0282,
+      "step": 5373
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4372488176025192,
+      "learning_rate": 6.000951240497952e-06,
+      "loss": 0.1032,
+      "step": 5374
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.38813171225862125,
+      "learning_rate": 5.999614222291631e-06,
+      "loss": 0.0944,
+      "step": 5375
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.20913893591875704,
+      "learning_rate": 5.998277129628221e-06,
+      "loss": 0.0531,
+      "step": 5376
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.36742366619963834,
+      "learning_rate": 5.996939962607314e-06,
+      "loss": 0.0854,
+      "step": 5377
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3281824636655971,
+      "learning_rate": 5.995602721328511e-06,
+      "loss": 0.1037,
+      "step": 5378
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4207613209814052,
+      "learning_rate": 5.9942654058914184e-06,
+      "loss": 0.0579,
+      "step": 5379
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3772755919900469,
+      "learning_rate": 5.992928016395646e-06,
+      "loss": 0.0953,
+      "step": 5380
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3206313646534068,
+      "learning_rate": 5.9915905529408105e-06,
+      "loss": 0.1031,
+      "step": 5381
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.1735927131276033,
+      "learning_rate": 5.990253015626534e-06,
+      "loss": 0.0428,
+      "step": 5382
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.35059837178469994,
+      "learning_rate": 5.988915404552443e-06,
+      "loss": 0.1164,
+      "step": 5383
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.22522239498583646,
+      "learning_rate": 5.987577719818174e-06,
+      "loss": 0.0467,
+      "step": 5384
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.30030009583985073,
+      "learning_rate": 5.986239961523362e-06,
+      "loss": 0.0923,
+      "step": 5385
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2846834646501469,
+      "learning_rate": 5.984902129767654e-06,
+      "loss": 0.0869,
+      "step": 5386
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.23855077691604534,
+      "learning_rate": 5.983564224650696e-06,
+      "loss": 0.0715,
+      "step": 5387
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2828577573257001,
+      "learning_rate": 5.982226246272145e-06,
+      "loss": 0.0609,
+      "step": 5388
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.32854683418853087,
+      "learning_rate": 5.980888194731664e-06,
+      "loss": 0.0906,
+      "step": 5389
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.24760203766003816,
+      "learning_rate": 5.979550070128912e-06,
+      "loss": 0.0753,
+      "step": 5390
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.18559918192886488,
+      "learning_rate": 5.978211872563567e-06,
+      "loss": 0.0629,
+      "step": 5391
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4420638735545099,
+      "learning_rate": 5.9768736021353026e-06,
+      "loss": 0.1124,
+      "step": 5392
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.40955422044635365,
+      "learning_rate": 5.9755352589438035e-06,
+      "loss": 0.1171,
+      "step": 5393
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.36434886894430873,
+      "learning_rate": 5.974196843088752e-06,
+      "loss": 0.0861,
+      "step": 5394
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.35350580419269484,
+      "learning_rate": 5.972858354669846e-06,
+      "loss": 0.0844,
+      "step": 5395
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.314795612280872,
+      "learning_rate": 5.971519793786783e-06,
+      "loss": 0.1084,
+      "step": 5396
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.28266219250668395,
+      "learning_rate": 5.9701811605392655e-06,
+      "loss": 0.0904,
+      "step": 5397
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.37303693149487516,
+      "learning_rate": 5.968842455027004e-06,
+      "loss": 0.1007,
+      "step": 5398
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.292483999253837,
+      "learning_rate": 5.967503677349713e-06,
+      "loss": 0.0819,
+      "step": 5399
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.384219057211456,
+      "learning_rate": 5.96616482760711e-06,
+      "loss": 0.0672,
+      "step": 5400
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.271373697367901,
+      "learning_rate": 5.964825905898924e-06,
+      "loss": 0.0663,
+      "step": 5401
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.28328150845718114,
+      "learning_rate": 5.963486912324884e-06,
+      "loss": 0.0832,
+      "step": 5402
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.19124636918266488,
+      "learning_rate": 5.962147846984726e-06,
+      "loss": 0.0711,
+      "step": 5403
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2866461625574067,
+      "learning_rate": 5.960808709978191e-06,
+      "loss": 0.0808,
+      "step": 5404
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.24223384690506577,
+      "learning_rate": 5.959469501405025e-06,
+      "loss": 0.0645,
+      "step": 5405
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.17515017669779412,
+      "learning_rate": 5.958130221364984e-06,
+      "loss": 0.0454,
+      "step": 5406
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.44491585820785906,
+      "learning_rate": 5.956790869957821e-06,
+      "loss": 0.0672,
+      "step": 5407
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3001242145935283,
+      "learning_rate": 5.9554514472833e-06,
+      "loss": 0.0695,
+      "step": 5408
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.522825204285594,
+      "learning_rate": 5.95411195344119e-06,
+      "loss": 0.0952,
+      "step": 5409
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.29118676662374365,
+      "learning_rate": 5.952772388531263e-06,
+      "loss": 0.1118,
+      "step": 5410
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.34546595544419983,
+      "learning_rate": 5.951432752653299e-06,
+      "loss": 0.1014,
+      "step": 5411
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2479246892137602,
+      "learning_rate": 5.950093045907081e-06,
+      "loss": 0.1029,
+      "step": 5412
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2801591137502635,
+      "learning_rate": 5.948753268392397e-06,
+      "loss": 0.0699,
+      "step": 5413
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.25349991222741874,
+      "learning_rate": 5.9474134202090435e-06,
+      "loss": 0.0557,
+      "step": 5414
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4200883843164058,
+      "learning_rate": 5.946073501456819e-06,
+      "loss": 0.1118,
+      "step": 5415
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.32163358365090916,
+      "learning_rate": 5.94473351223553e-06,
+      "loss": 0.086,
+      "step": 5416
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.38542617248132943,
+      "learning_rate": 5.9433934526449835e-06,
+      "loss": 0.1317,
+      "step": 5417
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.15562172662486587,
+      "learning_rate": 5.942053322784999e-06,
+      "loss": 0.0415,
+      "step": 5418
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.34882786063223997,
+      "learning_rate": 5.940713122755394e-06,
+      "loss": 0.0768,
+      "step": 5419
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6119491340846164,
+      "learning_rate": 5.939372852655996e-06,
+      "loss": 0.1381,
+      "step": 5420
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5051897299308056,
+      "learning_rate": 5.9380325125866355e-06,
+      "loss": 0.1367,
+      "step": 5421
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.33105786807140036,
+      "learning_rate": 5.936692102647149e-06,
+      "loss": 0.0863,
+      "step": 5422
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.24647067650994745,
+      "learning_rate": 5.9353516229373775e-06,
+      "loss": 0.0679,
+      "step": 5423
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.26741282140702005,
+      "learning_rate": 5.934011073557169e-06,
+      "loss": 0.0577,
+      "step": 5424
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2973598602973295,
+      "learning_rate": 5.932670454606375e-06,
+      "loss": 0.0761,
+      "step": 5425
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.25193519769041045,
+      "learning_rate": 5.9313297661848514e-06,
+      "loss": 0.0483,
+      "step": 5426
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.20721343235490933,
+      "learning_rate": 5.9299890083924625e-06,
+      "loss": 0.0633,
+      "step": 5427
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.17814131641387448,
+      "learning_rate": 5.9286481813290745e-06,
+      "loss": 0.0608,
+      "step": 5428
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2612157632117362,
+      "learning_rate": 5.927307285094559e-06,
+      "loss": 0.0432,
+      "step": 5429
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.43869738915439743,
+      "learning_rate": 5.925966319788796e-06,
+      "loss": 0.1045,
+      "step": 5430
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.33942885639881754,
+      "learning_rate": 5.924625285511667e-06,
+      "loss": 0.092,
+      "step": 5431
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2767786740655042,
+      "learning_rate": 5.9232841823630605e-06,
+      "loss": 0.0879,
+      "step": 5432
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2264272886620654,
+      "learning_rate": 5.92194301044287e-06,
+      "loss": 0.052,
+      "step": 5433
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.24241049628349756,
+      "learning_rate": 5.9206017698509925e-06,
+      "loss": 0.0737,
+      "step": 5434
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2693447988420022,
+      "learning_rate": 5.919260460687333e-06,
+      "loss": 0.072,
+      "step": 5435
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3303712941790021,
+      "learning_rate": 5.917919083051798e-06,
+      "loss": 0.0873,
+      "step": 5436
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.34635469598794016,
+      "learning_rate": 5.916577637044304e-06,
+      "loss": 0.0756,
+      "step": 5437
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3020661412990145,
+      "learning_rate": 5.915236122764767e-06,
+      "loss": 0.1049,
+      "step": 5438
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3609340741305951,
+      "learning_rate": 5.913894540313113e-06,
+      "loss": 0.0979,
+      "step": 5439
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3188562655569674,
+      "learning_rate": 5.912552889789269e-06,
+      "loss": 0.0553,
+      "step": 5440
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2735123421782616,
+      "learning_rate": 5.911211171293169e-06,
+      "loss": 0.0649,
+      "step": 5441
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.23665961725644732,
+      "learning_rate": 5.909869384924756e-06,
+      "loss": 0.0521,
+      "step": 5442
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.45275528368388723,
+      "learning_rate": 5.908527530783967e-06,
+      "loss": 0.0726,
+      "step": 5443
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.39781500716977575,
+      "learning_rate": 5.907185608970758e-06,
+      "loss": 0.0866,
+      "step": 5444
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9431804063748954,
+      "learning_rate": 5.905843619585079e-06,
+      "loss": 0.1445,
+      "step": 5445
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.23786262645916229,
+      "learning_rate": 5.90450156272689e-06,
+      "loss": 0.0646,
+      "step": 5446
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.29672031518352865,
+      "learning_rate": 5.903159438496156e-06,
+      "loss": 0.08,
+      "step": 5447
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.27118386912964354,
+      "learning_rate": 5.901817246992845e-06,
+      "loss": 0.0862,
+      "step": 5448
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5252336068494479,
+      "learning_rate": 5.900474988316932e-06,
+      "loss": 0.1175,
+      "step": 5449
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.30824635135407047,
+      "learning_rate": 5.8991326625683965e-06,
+      "loss": 0.0831,
+      "step": 5450
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2939412791118779,
+      "learning_rate": 5.897790269847221e-06,
+      "loss": 0.1046,
+      "step": 5451
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4516006533442279,
+      "learning_rate": 5.896447810253399e-06,
+      "loss": 0.0836,
+      "step": 5452
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2632207720016078,
+      "learning_rate": 5.895105283886919e-06,
+      "loss": 0.0886,
+      "step": 5453
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3941103841087465,
+      "learning_rate": 5.893762690847786e-06,
+      "loss": 0.0782,
+      "step": 5454
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3606122750644074,
+      "learning_rate": 5.8924200312359995e-06,
+      "loss": 0.0976,
+      "step": 5455
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.18572721386996988,
+      "learning_rate": 5.891077305151571e-06,
+      "loss": 0.0506,
+      "step": 5456
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.27288165293169897,
+      "learning_rate": 5.889734512694512e-06,
+      "loss": 0.063,
+      "step": 5457
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.46128066092372755,
+      "learning_rate": 5.888391653964845e-06,
+      "loss": 0.0736,
+      "step": 5458
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4810538279899771,
+      "learning_rate": 5.887048729062593e-06,
+      "loss": 0.1153,
+      "step": 5459
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2792174271140684,
+      "learning_rate": 5.885705738087784e-06,
+      "loss": 0.1043,
+      "step": 5460
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.38596880016960144,
+      "learning_rate": 5.884362681140451e-06,
+      "loss": 0.0692,
+      "step": 5461
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.36190712670968567,
+      "learning_rate": 5.8830195583206336e-06,
+      "loss": 0.0876,
+      "step": 5462
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3695955082362342,
+      "learning_rate": 5.8816763697283764e-06,
+      "loss": 0.075,
+      "step": 5463
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2345099539619166,
+      "learning_rate": 5.880333115463727e-06,
+      "loss": 0.0604,
+      "step": 5464
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3954974425213387,
+      "learning_rate": 5.878989795626739e-06,
+      "loss": 0.0826,
+      "step": 5465
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.42798817688051183,
+      "learning_rate": 5.87764641031747e-06,
+      "loss": 0.0752,
+      "step": 5466
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.43157879208887184,
+      "learning_rate": 5.876302959635983e-06,
+      "loss": 0.0783,
+      "step": 5467
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.23685597293540014,
+      "learning_rate": 5.874959443682347e-06,
+      "loss": 0.0688,
+      "step": 5468
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.27552423101129225,
+      "learning_rate": 5.873615862556636e-06,
+      "loss": 0.0674,
+      "step": 5469
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3357522739684893,
+      "learning_rate": 5.872272216358925e-06,
+      "loss": 0.0631,
+      "step": 5470
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.27823073306115076,
+      "learning_rate": 5.8709285051892995e-06,
+      "loss": 0.0589,
+      "step": 5471
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.29433978868119975,
+      "learning_rate": 5.869584729147843e-06,
+      "loss": 0.0481,
+      "step": 5472
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3932134499635916,
+      "learning_rate": 5.8682408883346535e-06,
+      "loss": 0.097,
+      "step": 5473
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.32592081166041487,
+      "learning_rate": 5.866896982849821e-06,
+      "loss": 0.0976,
+      "step": 5474
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4376139618061136,
+      "learning_rate": 5.865553012793453e-06,
+      "loss": 0.0981,
+      "step": 5475
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2688216931576172,
+      "learning_rate": 5.864208978265656e-06,
+      "loss": 0.0774,
+      "step": 5476
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6363658614798618,
+      "learning_rate": 5.862864879366537e-06,
+      "loss": 0.1516,
+      "step": 5477
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2935507933750343,
+      "learning_rate": 5.861520716196218e-06,
+      "loss": 0.0774,
+      "step": 5478
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.27191425804936137,
+      "learning_rate": 5.860176488854815e-06,
+      "loss": 0.063,
+      "step": 5479
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3581244371332679,
+      "learning_rate": 5.8588321974424565e-06,
+      "loss": 0.0825,
+      "step": 5480
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.456551247047345,
+      "learning_rate": 5.857487842059274e-06,
+      "loss": 0.0488,
+      "step": 5481
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4851685836683691,
+      "learning_rate": 5.856143422805401e-06,
+      "loss": 0.1165,
+      "step": 5482
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4050814229820206,
+      "learning_rate": 5.854798939780979e-06,
+      "loss": 0.1029,
+      "step": 5483
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.30698153578493365,
+      "learning_rate": 5.853454393086152e-06,
+      "loss": 0.0889,
+      "step": 5484
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.25032996767820265,
+      "learning_rate": 5.85210978282107e-06,
+      "loss": 0.0831,
+      "step": 5485
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.42936133407711347,
+      "learning_rate": 5.850765109085888e-06,
+      "loss": 0.1057,
+      "step": 5486
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3808180925542593,
+      "learning_rate": 5.849420371980764e-06,
+      "loss": 0.1183,
+      "step": 5487
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3070775081588745,
+      "learning_rate": 5.848075571605863e-06,
+      "loss": 0.0858,
+      "step": 5488
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.31755065766205454,
+      "learning_rate": 5.846730708061353e-06,
+      "loss": 0.0594,
+      "step": 5489
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2731228189169844,
+      "learning_rate": 5.845385781447407e-06,
+      "loss": 0.0636,
+      "step": 5490
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5978250146115706,
+      "learning_rate": 5.844040791864203e-06,
+      "loss": 0.1253,
+      "step": 5491
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2935693534547744,
+      "learning_rate": 5.842695739411923e-06,
+      "loss": 0.0783,
+      "step": 5492
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.25102607364980223,
+      "learning_rate": 5.841350624190757e-06,
+      "loss": 0.054,
+      "step": 5493
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.21224210574241656,
+      "learning_rate": 5.8400054463008945e-06,
+      "loss": 0.0554,
+      "step": 5494
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.22159124127410115,
+      "learning_rate": 5.8386602058425336e-06,
+      "loss": 0.0735,
+      "step": 5495
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2708569485746134,
+      "learning_rate": 5.837314902915874e-06,
+      "loss": 0.0427,
+      "step": 5496
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3022036308764928,
+      "learning_rate": 5.835969537621124e-06,
+      "loss": 0.1,
+      "step": 5497
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.34493755313254176,
+      "learning_rate": 5.834624110058492e-06,
+      "loss": 0.097,
+      "step": 5498
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3102325199305184,
+      "learning_rate": 5.8332786203281936e-06,
+      "loss": 0.0965,
+      "step": 5499
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3379664814601461,
+      "learning_rate": 5.831933068530451e-06,
+      "loss": 0.0751,
+      "step": 5500
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.33880655909892915,
+      "learning_rate": 5.830587454765486e-06,
+      "loss": 0.0912,
+      "step": 5501
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.289699079667515,
+      "learning_rate": 5.82924177913353e-06,
+      "loss": 0.0713,
+      "step": 5502
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.28724861029267074,
+      "learning_rate": 5.827896041734814e-06,
+      "loss": 0.0871,
+      "step": 5503
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4289693560624725,
+      "learning_rate": 5.826550242669579e-06,
+      "loss": 0.0723,
+      "step": 5504
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.48773536208127166,
+      "learning_rate": 5.825204382038066e-06,
+      "loss": 0.1196,
+      "step": 5505
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2262072501009228,
+      "learning_rate": 5.823858459940524e-06,
+      "loss": 0.046,
+      "step": 5506
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5949291079928839,
+      "learning_rate": 5.822512476477205e-06,
+      "loss": 0.1239,
+      "step": 5507
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.36735769816055003,
+      "learning_rate": 5.821166431748365e-06,
+      "loss": 0.0783,
+      "step": 5508
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.304974037590092,
+      "learning_rate": 5.8198203258542654e-06,
+      "loss": 0.1056,
+      "step": 5509
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2879068494404196,
+      "learning_rate": 5.818474158895172e-06,
+      "loss": 0.0689,
+      "step": 5510
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3358417956913452,
+      "learning_rate": 5.8171279309713546e-06,
+      "loss": 0.1044,
+      "step": 5511
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.316950721634035,
+      "learning_rate": 5.81578164218309e-06,
+      "loss": 0.0913,
+      "step": 5512
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.30487195742167206,
+      "learning_rate": 5.814435292630658e-06,
+      "loss": 0.0874,
+      "step": 5513
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.301986113268227,
+      "learning_rate": 5.8130888824143384e-06,
+      "loss": 0.0808,
+      "step": 5514
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.25533843297048364,
+      "learning_rate": 5.811742411634424e-06,
+      "loss": 0.0735,
+      "step": 5515
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.33011822574590644,
+      "learning_rate": 5.810395880391203e-06,
+      "loss": 0.0663,
+      "step": 5516
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4078296517395853,
+      "learning_rate": 5.809049288784979e-06,
+      "loss": 0.1183,
+      "step": 5517
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3808957534783275,
+      "learning_rate": 5.807702636916049e-06,
+      "loss": 0.0823,
+      "step": 5518
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2430180721782435,
+      "learning_rate": 5.806355924884722e-06,
+      "loss": 0.0499,
+      "step": 5519
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.262472133568686,
+      "learning_rate": 5.805009152791309e-06,
+      "loss": 0.0864,
+      "step": 5520
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2514964345410544,
+      "learning_rate": 5.803662320736123e-06,
+      "loss": 0.0415,
+      "step": 5521
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.40876828981550006,
+      "learning_rate": 5.802315428819488e-06,
+      "loss": 0.1013,
+      "step": 5522
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.39789771047624384,
+      "learning_rate": 5.800968477141724e-06,
+      "loss": 0.0935,
+      "step": 5523
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.33977255097863057,
+      "learning_rate": 5.799621465803163e-06,
+      "loss": 0.1011,
+      "step": 5524
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2873917561215367,
+      "learning_rate": 5.7982743949041375e-06,
+      "loss": 0.0613,
+      "step": 5525
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.23837984934378806,
+      "learning_rate": 5.796927264544984e-06,
+      "loss": 0.062,
+      "step": 5526
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.19161710253473258,
+      "learning_rate": 5.795580074826045e-06,
+      "loss": 0.0238,
+      "step": 5527
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.21426544929669433,
+      "learning_rate": 5.794232825847668e-06,
+      "loss": 0.075,
+      "step": 5528
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2555229898399971,
+      "learning_rate": 5.792885517710204e-06,
+      "loss": 0.0554,
+      "step": 5529
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3436322031031592,
+      "learning_rate": 5.791538150514007e-06,
+      "loss": 0.0688,
+      "step": 5530
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.21500653929996094,
+      "learning_rate": 5.790190724359437e-06,
+      "loss": 0.0573,
+      "step": 5531
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3618648236786172,
+      "learning_rate": 5.788843239346859e-06,
+      "loss": 0.0645,
+      "step": 5532
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.23069586447728047,
+      "learning_rate": 5.787495695576642e-06,
+      "loss": 0.0435,
+      "step": 5533
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2910665008102202,
+      "learning_rate": 5.786148093149158e-06,
+      "loss": 0.0678,
+      "step": 5534
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.39998732650592017,
+      "learning_rate": 5.7848004321647834e-06,
+      "loss": 0.0874,
+      "step": 5535
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4703277204959219,
+      "learning_rate": 5.783452712723902e-06,
+      "loss": 0.0675,
+      "step": 5536
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5372705539226282,
+      "learning_rate": 5.782104934926897e-06,
+      "loss": 0.1022,
+      "step": 5537
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.47481471149161253,
+      "learning_rate": 5.78075709887416e-06,
+      "loss": 0.1082,
+      "step": 5538
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3450454044999949,
+      "learning_rate": 5.779409204666089e-06,
+      "loss": 0.077,
+      "step": 5539
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.35855623659021096,
+      "learning_rate": 5.778061252403077e-06,
+      "loss": 0.0936,
+      "step": 5540
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2721416815134429,
+      "learning_rate": 5.776713242185532e-06,
+      "loss": 0.0699,
+      "step": 5541
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2392716736204212,
+      "learning_rate": 5.77536517411386e-06,
+      "loss": 0.0607,
+      "step": 5542
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5508789040434487,
+      "learning_rate": 5.774017048288472e-06,
+      "loss": 0.1035,
+      "step": 5543
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2899073107527558,
+      "learning_rate": 5.772668864809786e-06,
+      "loss": 0.0909,
+      "step": 5544
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3312688244821426,
+      "learning_rate": 5.771320623778222e-06,
+      "loss": 0.1091,
+      "step": 5545
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.44022967666575574,
+      "learning_rate": 5.769972325294204e-06,
+      "loss": 0.0891,
+      "step": 5546
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.17073865233776703,
+      "learning_rate": 5.768623969458164e-06,
+      "loss": 0.0562,
+      "step": 5547
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.25083429444298777,
+      "learning_rate": 5.767275556370531e-06,
+      "loss": 0.1015,
+      "step": 5548
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.18966072650472876,
+      "learning_rate": 5.765927086131745e-06,
+      "loss": 0.0633,
+      "step": 5549
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.29976787787873366,
+      "learning_rate": 5.764578558842247e-06,
+      "loss": 0.0702,
+      "step": 5550
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6077257866343774,
+      "learning_rate": 5.763229974602487e-06,
+      "loss": 0.1474,
+      "step": 5551
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.47220680770563483,
+      "learning_rate": 5.7618813335129096e-06,
+      "loss": 0.1059,
+      "step": 5552
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2428644419018441,
+      "learning_rate": 5.760532635673973e-06,
+      "loss": 0.0685,
+      "step": 5553
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3041494908645644,
+      "learning_rate": 5.759183881186136e-06,
+      "loss": 0.0855,
+      "step": 5554
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.30260185328118305,
+      "learning_rate": 5.7578350701498585e-06,
+      "loss": 0.0972,
+      "step": 5555
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3272026213165436,
+      "learning_rate": 5.756486202665613e-06,
+      "loss": 0.0839,
+      "step": 5556
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.23151635629596667,
+      "learning_rate": 5.755137278833867e-06,
+      "loss": 0.0641,
+      "step": 5557
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.281132565847536,
+      "learning_rate": 5.753788298755097e-06,
+      "loss": 0.0776,
+      "step": 5558
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.329153405046996,
+      "learning_rate": 5.752439262529784e-06,
+      "loss": 0.0878,
+      "step": 5559
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.21577696935457444,
+      "learning_rate": 5.751090170258411e-06,
+      "loss": 0.0473,
+      "step": 5560
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.30112588880134533,
+      "learning_rate": 5.749741022041468e-06,
+      "loss": 0.0555,
+      "step": 5561
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.23695833935562333,
+      "learning_rate": 5.7483918179794455e-06,
+      "loss": 0.0535,
+      "step": 5562
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3717634800904838,
+      "learning_rate": 5.74704255817284e-06,
+      "loss": 0.0817,
+      "step": 5563
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.37744465779504965,
+      "learning_rate": 5.745693242722154e-06,
+      "loss": 0.0724,
+      "step": 5564
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.22368476125756517,
+      "learning_rate": 5.744343871727891e-06,
+      "loss": 0.0633,
+      "step": 5565
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5010375452613077,
+      "learning_rate": 5.742994445290559e-06,
+      "loss": 0.1203,
+      "step": 5566
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2204684839660284,
+      "learning_rate": 5.741644963510673e-06,
+      "loss": 0.0738,
+      "step": 5567
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3736840192163062,
+      "learning_rate": 5.740295426488751e-06,
+      "loss": 0.103,
+      "step": 5568
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2882556582912704,
+      "learning_rate": 5.738945834325311e-06,
+      "loss": 0.0826,
+      "step": 5569
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.34886051486822206,
+      "learning_rate": 5.737596187120883e-06,
+      "loss": 0.0871,
+      "step": 5570
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3522572809195416,
+      "learning_rate": 5.736246484975993e-06,
+      "loss": 0.0833,
+      "step": 5571
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3353737834337096,
+      "learning_rate": 5.734896727991175e-06,
+      "loss": 0.0953,
+      "step": 5572
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5509551293991473,
+      "learning_rate": 5.7335469162669684e-06,
+      "loss": 0.0995,
+      "step": 5573
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2464567524491927,
+      "learning_rate": 5.732197049903915e-06,
+      "loss": 0.0645,
+      "step": 5574
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2483681554925709,
+      "learning_rate": 5.730847129002559e-06,
+      "loss": 0.0619,
+      "step": 5575
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2610507995413378,
+      "learning_rate": 5.729497153663451e-06,
+      "loss": 0.0642,
+      "step": 5576
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3006625763711666,
+      "learning_rate": 5.728147123987147e-06,
+      "loss": 0.0703,
+      "step": 5577
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3182189379090218,
+      "learning_rate": 5.726797040074202e-06,
+      "loss": 0.0992,
+      "step": 5578
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3266703415070439,
+      "learning_rate": 5.72544690202518e-06,
+      "loss": 0.0549,
+      "step": 5579
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2233672082402826,
+      "learning_rate": 5.724096709940646e-06,
+      "loss": 0.0681,
+      "step": 5580
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.34636489326505143,
+      "learning_rate": 5.722746463921173e-06,
+      "loss": 0.0523,
+      "step": 5581
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4627520039721311,
+      "learning_rate": 5.721396164067331e-06,
+      "loss": 0.1276,
+      "step": 5582
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2597117947591045,
+      "learning_rate": 5.720045810479702e-06,
+      "loss": 0.0683,
+      "step": 5583
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.28018862022772884,
+      "learning_rate": 5.718695403258865e-06,
+      "loss": 0.0694,
+      "step": 5584
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.1873632986795777,
+      "learning_rate": 5.717344942505408e-06,
+      "loss": 0.0582,
+      "step": 5585
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.47741410793170386,
+      "learning_rate": 5.71599442831992e-06,
+      "loss": 0.1212,
+      "step": 5586
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.18485283410659728,
+      "learning_rate": 5.714643860802997e-06,
+      "loss": 0.0579,
+      "step": 5587
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.31753321833564613,
+      "learning_rate": 5.713293240055235e-06,
+      "loss": 0.1015,
+      "step": 5588
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4289710166430089,
+      "learning_rate": 5.711942566177239e-06,
+      "loss": 0.1073,
+      "step": 5589
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.20871209154545425,
+      "learning_rate": 5.710591839269613e-06,
+      "loss": 0.0683,
+      "step": 5590
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.32200975472776333,
+      "learning_rate": 5.709241059432966e-06,
+      "loss": 0.1044,
+      "step": 5591
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3069164887541522,
+      "learning_rate": 5.707890226767915e-06,
+      "loss": 0.1243,
+      "step": 5592
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.24478970836371816,
+      "learning_rate": 5.7065393413750745e-06,
+      "loss": 0.0785,
+      "step": 5593
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4204799714300244,
+      "learning_rate": 5.705188403355069e-06,
+      "loss": 0.0936,
+      "step": 5594
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.39321693575458266,
+      "learning_rate": 5.703837412808523e-06,
+      "loss": 0.0585,
+      "step": 5595
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3816088083027971,
+      "learning_rate": 5.702486369836066e-06,
+      "loss": 0.1096,
+      "step": 5596
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3420344999923336,
+      "learning_rate": 5.701135274538332e-06,
+      "loss": 0.078,
+      "step": 5597
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3588946040476363,
+      "learning_rate": 5.699784127015958e-06,
+      "loss": 0.0934,
+      "step": 5598
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.390002747615039,
+      "learning_rate": 5.698432927369586e-06,
+      "loss": 0.0968,
+      "step": 5599
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2861890574058236,
+      "learning_rate": 5.697081675699861e-06,
+      "loss": 0.0673,
+      "step": 5600
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2606084938917433,
+      "learning_rate": 5.695730372107431e-06,
+      "loss": 0.0636,
+      "step": 5601
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2116570568524385,
+      "learning_rate": 5.694379016692953e-06,
+      "loss": 0.0641,
+      "step": 5602
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.32236574141529023,
+      "learning_rate": 5.693027609557077e-06,
+      "loss": 0.0676,
+      "step": 5603
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.23067120768289576,
+      "learning_rate": 5.69167615080047e-06,
+      "loss": 0.0574,
+      "step": 5604
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3479213003383328,
+      "learning_rate": 5.6903246405237925e-06,
+      "loss": 0.0477,
+      "step": 5605
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2760653672522455,
+      "learning_rate": 5.688973078827714e-06,
+      "loss": 0.0679,
+      "step": 5606
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4573203780161493,
+      "learning_rate": 5.687621465812909e-06,
+      "loss": 0.1129,
+      "step": 5607
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5464482398141809,
+      "learning_rate": 5.68626980158005e-06,
+      "loss": 0.1591,
+      "step": 5608
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.27751221135060417,
+      "learning_rate": 5.6849180862298205e-06,
+      "loss": 0.0678,
+      "step": 5609
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6131120547521957,
+      "learning_rate": 5.683566319862899e-06,
+      "loss": 0.1489,
+      "step": 5610
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3840083354820757,
+      "learning_rate": 5.682214502579978e-06,
+      "loss": 0.0922,
+      "step": 5611
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.504500977166992,
+      "learning_rate": 5.680862634481746e-06,
+      "loss": 0.1234,
+      "step": 5612
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4544333815305592,
+      "learning_rate": 5.679510715668898e-06,
+      "loss": 0.0949,
+      "step": 5613
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5624046200684606,
+      "learning_rate": 5.678158746242134e-06,
+      "loss": 0.0745,
+      "step": 5614
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3124316296871797,
+      "learning_rate": 5.6768067263021575e-06,
+      "loss": 0.0962,
+      "step": 5615
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4099163638457547,
+      "learning_rate": 5.675454655949672e-06,
+      "loss": 0.074,
+      "step": 5616
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.34444542518798293,
+      "learning_rate": 5.674102535285388e-06,
+      "loss": 0.0832,
+      "step": 5617
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3876046215038411,
+      "learning_rate": 5.672750364410022e-06,
+      "loss": 0.1015,
+      "step": 5618
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.29850263005369165,
+      "learning_rate": 5.671398143424288e-06,
+      "loss": 0.1032,
+      "step": 5619
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.37656503515646045,
+      "learning_rate": 5.67004587242891e-06,
+      "loss": 0.1134,
+      "step": 5620
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7099567169699808,
+      "learning_rate": 5.668693551524614e-06,
+      "loss": 0.1365,
+      "step": 5621
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3362720418653712,
+      "learning_rate": 5.667341180812122e-06,
+      "loss": 0.0518,
+      "step": 5622
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7227858759375333,
+      "learning_rate": 5.665988760392175e-06,
+      "loss": 0.1635,
+      "step": 5623
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.34420597132915093,
+      "learning_rate": 5.664636290365505e-06,
+      "loss": 0.1058,
+      "step": 5624
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.26579031508845113,
+      "learning_rate": 5.663283770832852e-06,
+      "loss": 0.0574,
+      "step": 5625
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2803913870350494,
+      "learning_rate": 5.6619312018949606e-06,
+      "loss": 0.0628,
+      "step": 5626
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.21091289948698302,
+      "learning_rate": 5.660578583652576e-06,
+      "loss": 0.042,
+      "step": 5627
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3534792353523456,
+      "learning_rate": 5.659225916206451e-06,
+      "loss": 0.1189,
+      "step": 5628
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4752144203220441,
+      "learning_rate": 5.657873199657339e-06,
+      "loss": 0.0669,
+      "step": 5629
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2414156303105738,
+      "learning_rate": 5.656520434105999e-06,
+      "loss": 0.062,
+      "step": 5630
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.44630109384612976,
+      "learning_rate": 5.655167619653192e-06,
+      "loss": 0.108,
+      "step": 5631
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3440341420918822,
+      "learning_rate": 5.653814756399685e-06,
+      "loss": 0.0891,
+      "step": 5632
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.30230005052701264,
+      "learning_rate": 5.652461844446246e-06,
+      "loss": 0.1032,
+      "step": 5633
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.26432678382436975,
+      "learning_rate": 5.651108883893648e-06,
+      "loss": 0.0679,
+      "step": 5634
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3292432517057025,
+      "learning_rate": 5.649755874842666e-06,
+      "loss": 0.0681,
+      "step": 5635
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3226872765953513,
+      "learning_rate": 5.648402817394083e-06,
+      "loss": 0.0992,
+      "step": 5636
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6761321770399754,
+      "learning_rate": 5.64704971164868e-06,
+      "loss": 0.1163,
+      "step": 5637
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.36251828083933924,
+      "learning_rate": 5.6456965577072455e-06,
+      "loss": 0.1079,
+      "step": 5638
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.36111582223424094,
+      "learning_rate": 5.64434335567057e-06,
+      "loss": 0.0806,
+      "step": 5639
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3461952598619096,
+      "learning_rate": 5.642990105639447e-06,
+      "loss": 0.073,
+      "step": 5640
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.2749002554082094,
+      "learning_rate": 5.641636807714677e-06,
+      "loss": 0.0819,
+      "step": 5641
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.29281198780191914,
+      "learning_rate": 5.640283461997058e-06,
+      "loss": 0.0883,
+      "step": 5642
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3111774504830054,
+      "learning_rate": 5.6389300685873984e-06,
+      "loss": 0.1034,
+      "step": 5643
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.2151425102716583,
+      "learning_rate": 5.637576627586503e-06,
+      "loss": 0.067,
+      "step": 5644
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.2545000311954673,
+      "learning_rate": 5.636223139095188e-06,
+      "loss": 0.0803,
+      "step": 5645
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4608188534182907,
+      "learning_rate": 5.634869603214267e-06,
+      "loss": 0.0718,
+      "step": 5646
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.37046394131674043,
+      "learning_rate": 5.6335160200445595e-06,
+      "loss": 0.1085,
+      "step": 5647
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3740440366772873,
+      "learning_rate": 5.6321623896868885e-06,
+      "loss": 0.0583,
+      "step": 5648
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.27509748450576443,
+      "learning_rate": 5.630808712242081e-06,
+      "loss": 0.0765,
+      "step": 5649
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.482930793489925,
+      "learning_rate": 5.629454987810964e-06,
+      "loss": 0.1509,
+      "step": 5650
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.26073556383640606,
+      "learning_rate": 5.628101216494374e-06,
+      "loss": 0.0925,
+      "step": 5651
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.22663094642167075,
+      "learning_rate": 5.626747398393145e-06,
+      "loss": 0.065,
+      "step": 5652
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5500517239429134,
+      "learning_rate": 5.62539353360812e-06,
+      "loss": 0.1789,
+      "step": 5653
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.21734146046063216,
+      "learning_rate": 5.62403962224014e-06,
+      "loss": 0.0731,
+      "step": 5654
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4658442455755646,
+      "learning_rate": 5.622685664390056e-06,
+      "loss": 0.1222,
+      "step": 5655
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4624750731992381,
+      "learning_rate": 5.6213316601587144e-06,
+      "loss": 0.0882,
+      "step": 5656
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.2561310086827649,
+      "learning_rate": 5.619977609646971e-06,
+      "loss": 0.0722,
+      "step": 5657
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.44924519934666585,
+      "learning_rate": 5.618623512955686e-06,
+      "loss": 0.1068,
+      "step": 5658
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.23712888561973036,
+      "learning_rate": 5.617269370185715e-06,
+      "loss": 0.066,
+      "step": 5659
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.47920327494316406,
+      "learning_rate": 5.615915181437928e-06,
+      "loss": 0.1066,
+      "step": 5660
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.33331104594023203,
+      "learning_rate": 5.614560946813188e-06,
+      "loss": 0.0517,
+      "step": 5661
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.44260127535128885,
+      "learning_rate": 5.6132066664123705e-06,
+      "loss": 0.1171,
+      "step": 5662
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.34322407566172985,
+      "learning_rate": 5.611852340336348e-06,
+      "loss": 0.0986,
+      "step": 5663
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.36644487013452426,
+      "learning_rate": 5.610497968685997e-06,
+      "loss": 0.0796,
+      "step": 5664
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.25972183438657026,
+      "learning_rate": 5.609143551562203e-06,
+      "loss": 0.0649,
+      "step": 5665
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.39474931569518107,
+      "learning_rate": 5.607789089065847e-06,
+      "loss": 0.1068,
+      "step": 5666
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.31139319987199193,
+      "learning_rate": 5.606434581297821e-06,
+      "loss": 0.0961,
+      "step": 5667
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.37227392553888705,
+      "learning_rate": 5.605080028359013e-06,
+      "loss": 0.0799,
+      "step": 5668
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.37354813746092325,
+      "learning_rate": 5.6037254303503205e-06,
+      "loss": 0.094,
+      "step": 5669
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4497863241086802,
+      "learning_rate": 5.602370787372642e-06,
+      "loss": 0.1098,
+      "step": 5670
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.413406392049191,
+      "learning_rate": 5.601016099526876e-06,
+      "loss": 0.1301,
+      "step": 5671
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.22696516091607688,
+      "learning_rate": 5.599661366913931e-06,
+      "loss": 0.0702,
+      "step": 5672
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3063495252430633,
+      "learning_rate": 5.5983065896347135e-06,
+      "loss": 0.084,
+      "step": 5673
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.39186784208687114,
+      "learning_rate": 5.596951767790136e-06,
+      "loss": 0.0636,
+      "step": 5674
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.22955859999304504,
+      "learning_rate": 5.595596901481114e-06,
+      "loss": 0.0969,
+      "step": 5675
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3600930975295021,
+      "learning_rate": 5.594241990808565e-06,
+      "loss": 0.0837,
+      "step": 5676
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.23089071127195185,
+      "learning_rate": 5.59288703587341e-06,
+      "loss": 0.0568,
+      "step": 5677
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.423516362733909,
+      "learning_rate": 5.5915320367765745e-06,
+      "loss": 0.123,
+      "step": 5678
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3340064185986187,
+      "learning_rate": 5.590176993618989e-06,
+      "loss": 0.1019,
+      "step": 5679
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.1413243354166399,
+      "learning_rate": 5.58882190650158e-06,
+      "loss": 0.0393,
+      "step": 5680
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.2028496639971009,
+      "learning_rate": 5.587466775525287e-06,
+      "loss": 0.0654,
+      "step": 5681
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.18628100996373606,
+      "learning_rate": 5.586111600791047e-06,
+      "loss": 0.0629,
+      "step": 5682
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.42744836615686443,
+      "learning_rate": 5.584756382399801e-06,
+      "loss": 0.0956,
+      "step": 5683
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4940653348606991,
+      "learning_rate": 5.5834011204524905e-06,
+      "loss": 0.0959,
+      "step": 5684
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.30569883879524296,
+      "learning_rate": 5.582045815050069e-06,
+      "loss": 0.0884,
+      "step": 5685
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.40931278459331844,
+      "learning_rate": 5.580690466293482e-06,
+      "loss": 0.0999,
+      "step": 5686
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6138589350260053,
+      "learning_rate": 5.579335074283688e-06,
+      "loss": 0.1126,
+      "step": 5687
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7521137790545585,
+      "learning_rate": 5.5779796391216426e-06,
+      "loss": 0.1017,
+      "step": 5688
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.47098228306483425,
+      "learning_rate": 5.576624160908306e-06,
+      "loss": 0.107,
+      "step": 5689
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.427566993867442,
+      "learning_rate": 5.575268639744643e-06,
+      "loss": 0.0871,
+      "step": 5690
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.29042767680987536,
+      "learning_rate": 5.573913075731621e-06,
+      "loss": 0.0627,
+      "step": 5691
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4916901087083386,
+      "learning_rate": 5.572557468970209e-06,
+      "loss": 0.1025,
+      "step": 5692
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.32846475135583586,
+      "learning_rate": 5.571201819561381e-06,
+      "loss": 0.0726,
+      "step": 5693
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3223533666151447,
+      "learning_rate": 5.569846127606115e-06,
+      "loss": 0.0764,
+      "step": 5694
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3162263028394242,
+      "learning_rate": 5.568490393205389e-06,
+      "loss": 0.0625,
+      "step": 5695
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3597534976214187,
+      "learning_rate": 5.567134616460188e-06,
+      "loss": 0.0938,
+      "step": 5696
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3491695240286068,
+      "learning_rate": 5.565778797471496e-06,
+      "loss": 0.0806,
+      "step": 5697
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5666901458981042,
+      "learning_rate": 5.564422936340304e-06,
+      "loss": 0.0943,
+      "step": 5698
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.30673995229933726,
+      "learning_rate": 5.563067033167603e-06,
+      "loss": 0.064,
+      "step": 5699
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.32408433337671116,
+      "learning_rate": 5.561711088054389e-06,
+      "loss": 0.0666,
+      "step": 5700
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.2766375171785388,
+      "learning_rate": 5.560355101101662e-06,
+      "loss": 0.1018,
+      "step": 5701
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.551243941580642,
+      "learning_rate": 5.558999072410423e-06,
+      "loss": 0.0745,
+      "step": 5702
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3241903474123637,
+      "learning_rate": 5.557643002081674e-06,
+      "loss": 0.0637,
+      "step": 5703
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3284709302287849,
+      "learning_rate": 5.556286890216429e-06,
+      "loss": 0.0698,
+      "step": 5704
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3348684344022521,
+      "learning_rate": 5.554930736915694e-06,
+      "loss": 0.0985,
+      "step": 5705
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.2583363734599848,
+      "learning_rate": 5.5535745422804855e-06,
+      "loss": 0.0838,
+      "step": 5706
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5085799631078692,
+      "learning_rate": 5.55221830641182e-06,
+      "loss": 0.1432,
+      "step": 5707
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.42168498473145294,
+      "learning_rate": 5.5508620294107185e-06,
+      "loss": 0.1154,
+      "step": 5708
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4416564343069031,
+      "learning_rate": 5.549505711378204e-06,
+      "loss": 0.11,
+      "step": 5709
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.32444438881896137,
+      "learning_rate": 5.548149352415302e-06,
+      "loss": 0.059,
+      "step": 5710
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.22523026473208216,
+      "learning_rate": 5.546792952623044e-06,
+      "loss": 0.0638,
+      "step": 5711
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3886911800859997,
+      "learning_rate": 5.545436512102461e-06,
+      "loss": 0.0818,
+      "step": 5712
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.31775115692365624,
+      "learning_rate": 5.544080030954591e-06,
+      "loss": 0.0964,
+      "step": 5713
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.21575205617215942,
+      "learning_rate": 5.542723509280469e-06,
+      "loss": 0.0502,
+      "step": 5714
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.27910642023359394,
+      "learning_rate": 5.5413669471811385e-06,
+      "loss": 0.0569,
+      "step": 5715
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.45358836130889546,
+      "learning_rate": 5.5400103447576444e-06,
+      "loss": 0.1172,
+      "step": 5716
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.23814668839737924,
+      "learning_rate": 5.538653702111035e-06,
+      "loss": 0.0852,
+      "step": 5717
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.22656444488042088,
+      "learning_rate": 5.537297019342358e-06,
+      "loss": 0.0537,
+      "step": 5718
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.296385686962184,
+      "learning_rate": 5.535940296552671e-06,
+      "loss": 0.0927,
+      "step": 5719
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.2700843957690607,
+      "learning_rate": 5.534583533843028e-06,
+      "loss": 0.0609,
+      "step": 5720
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.24369352964737995,
+      "learning_rate": 5.533226731314489e-06,
+      "loss": 0.0722,
+      "step": 5721
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7172085429352948,
+      "learning_rate": 5.5318698890681166e-06,
+      "loss": 0.1509,
+      "step": 5722
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.37429696243513727,
+      "learning_rate": 5.530513007204977e-06,
+      "loss": 0.1158,
+      "step": 5723
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.20434959965219343,
+      "learning_rate": 5.529156085826138e-06,
+      "loss": 0.0518,
+      "step": 5724
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.23170871561108292,
+      "learning_rate": 5.527799125032671e-06,
+      "loss": 0.0442,
+      "step": 5725
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4605852887510632,
+      "learning_rate": 5.5264421249256495e-06,
+      "loss": 0.1091,
+      "step": 5726
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3543527114197729,
+      "learning_rate": 5.525085085606152e-06,
+      "loss": 0.0841,
+      "step": 5727
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4411884704545098,
+      "learning_rate": 5.523728007175259e-06,
+      "loss": 0.1037,
+      "step": 5728
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.27820584968170453,
+      "learning_rate": 5.5223708897340544e-06,
+      "loss": 0.0615,
+      "step": 5729
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.45218606064918976,
+      "learning_rate": 5.521013733383622e-06,
+      "loss": 0.1336,
+      "step": 5730
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.30549844456248904,
+      "learning_rate": 5.519656538225051e-06,
+      "loss": 0.0695,
+      "step": 5731
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4291096704631432,
+      "learning_rate": 5.5182993043594335e-06,
+      "loss": 0.0997,
+      "step": 5732
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.2571889427529331,
+      "learning_rate": 5.516942031887866e-06,
+      "loss": 0.0354,
+      "step": 5733
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.2757948996462117,
+      "learning_rate": 5.515584720911443e-06,
+      "loss": 0.0673,
+      "step": 5734
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5385883576334488,
+      "learning_rate": 5.5142273715312675e-06,
+      "loss": 0.1248,
+      "step": 5735
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3577852273370118,
+      "learning_rate": 5.512869983848441e-06,
+      "loss": 0.0629,
+      "step": 5736
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3682384777193351,
+      "learning_rate": 5.511512557964072e-06,
+      "loss": 0.0903,
+      "step": 5737
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.27309780141062473,
+      "learning_rate": 5.510155093979266e-06,
+      "loss": 0.0631,
+      "step": 5738
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.19783353713567856,
+      "learning_rate": 5.5087975919951374e-06,
+      "loss": 0.0391,
+      "step": 5739
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3012027119446076,
+      "learning_rate": 5.507440052112801e-06,
+      "loss": 0.0789,
+      "step": 5740
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.19164559440211185,
+      "learning_rate": 5.506082474433372e-06,
+      "loss": 0.0532,
+      "step": 5741
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5413036723531496,
+      "learning_rate": 5.504724859057974e-06,
+      "loss": 0.1287,
+      "step": 5742
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.23589077038115117,
+      "learning_rate": 5.503367206087727e-06,
+      "loss": 0.0592,
+      "step": 5743
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3436411318282866,
+      "learning_rate": 5.5020095156237575e-06,
+      "loss": 0.084,
+      "step": 5744
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.24285203077891754,
+      "learning_rate": 5.500651787767196e-06,
+      "loss": 0.0535,
+      "step": 5745
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.2931560683055988,
+      "learning_rate": 5.499294022619172e-06,
+      "loss": 0.0683,
+      "step": 5746
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.2899218719864031,
+      "learning_rate": 5.497936220280822e-06,
+      "loss": 0.0673,
+      "step": 5747
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5093583918373007,
+      "learning_rate": 5.496578380853281e-06,
+      "loss": 0.1355,
+      "step": 5748
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6146155527407061,
+      "learning_rate": 5.495220504437688e-06,
+      "loss": 0.1031,
+      "step": 5749
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3455043133951765,
+      "learning_rate": 5.4938625911351885e-06,
+      "loss": 0.0701,
+      "step": 5750
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.31276324853517523,
+      "learning_rate": 5.492504641046926e-06,
+      "loss": 0.1002,
+      "step": 5751
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5500843100576319,
+      "learning_rate": 5.491146654274049e-06,
+      "loss": 0.1221,
+      "step": 5752
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3300547141537957,
+      "learning_rate": 5.4897886309177085e-06,
+      "loss": 0.0668,
+      "step": 5753
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3974818080692812,
+      "learning_rate": 5.488430571079055e-06,
+      "loss": 0.0915,
+      "step": 5754
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.1615940111343022,
+      "learning_rate": 5.487072474859251e-06,
+      "loss": 0.0387,
+      "step": 5755
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.41050205838945275,
+      "learning_rate": 5.485714342359448e-06,
+      "loss": 0.0863,
+      "step": 5756
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.23596620228530818,
+      "learning_rate": 5.4843561736808134e-06,
+      "loss": 0.0896,
+      "step": 5757
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3934464826452503,
+      "learning_rate": 5.4829979689245095e-06,
+      "loss": 0.1321,
+      "step": 5758
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.38238645983439473,
+      "learning_rate": 5.481639728191703e-06,
+      "loss": 0.0837,
+      "step": 5759
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.39471242910149673,
+      "learning_rate": 5.480281451583564e-06,
+      "loss": 0.0862,
+      "step": 5760
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2176033924452278,
+      "learning_rate": 5.478923139201262e-06,
+      "loss": 0.05,
+      "step": 5761
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3581866885092856,
+      "learning_rate": 5.477564791145978e-06,
+      "loss": 0.0608,
+      "step": 5762
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.35832318353186404,
+      "learning_rate": 5.476206407518885e-06,
+      "loss": 0.0634,
+      "step": 5763
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3915307010977396,
+      "learning_rate": 5.474847988421165e-06,
+      "loss": 0.0916,
+      "step": 5764
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3376870603746107,
+      "learning_rate": 5.473489533954e-06,
+      "loss": 0.0711,
+      "step": 5765
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3554314616111118,
+      "learning_rate": 5.472131044218576e-06,
+      "loss": 0.0968,
+      "step": 5766
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4873010073319551,
+      "learning_rate": 5.470772519316082e-06,
+      "loss": 0.1127,
+      "step": 5767
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4002819925752563,
+      "learning_rate": 5.469413959347708e-06,
+      "loss": 0.105,
+      "step": 5768
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2927058346348805,
+      "learning_rate": 5.468055364414649e-06,
+      "loss": 0.0714,
+      "step": 5769
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.42967571320779985,
+      "learning_rate": 5.4666967346180985e-06,
+      "loss": 0.0947,
+      "step": 5770
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.35678173520469486,
+      "learning_rate": 5.465338070059257e-06,
+      "loss": 0.0978,
+      "step": 5771
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.386854159437913,
+      "learning_rate": 5.463979370839325e-06,
+      "loss": 0.0788,
+      "step": 5772
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3165962410990284,
+      "learning_rate": 5.462620637059507e-06,
+      "loss": 0.0668,
+      "step": 5773
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3409480840710618,
+      "learning_rate": 5.461261868821009e-06,
+      "loss": 0.0456,
+      "step": 5774
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.36521692370020237,
+      "learning_rate": 5.45990306622504e-06,
+      "loss": 0.0999,
+      "step": 5775
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.302673743631347,
+      "learning_rate": 5.458544229372811e-06,
+      "loss": 0.0755,
+      "step": 5776
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.34620960731028855,
+      "learning_rate": 5.457185358365537e-06,
+      "loss": 0.0961,
+      "step": 5777
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.22435024972206655,
+      "learning_rate": 5.455826453304433e-06,
+      "loss": 0.0497,
+      "step": 5778
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8149195174956303,
+      "learning_rate": 5.454467514290721e-06,
+      "loss": 0.1441,
+      "step": 5779
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.25484534732694647,
+      "learning_rate": 5.45310854142562e-06,
+      "loss": 0.0774,
+      "step": 5780
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7616130610160319,
+      "learning_rate": 5.4517495348103566e-06,
+      "loss": 0.1015,
+      "step": 5781
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.20326276038898253,
+      "learning_rate": 5.450390494546155e-06,
+      "loss": 0.0676,
+      "step": 5782
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.26004035244569695,
+      "learning_rate": 5.449031420734246e-06,
+      "loss": 0.0596,
+      "step": 5783
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3825553413830431,
+      "learning_rate": 5.44767231347586e-06,
+      "loss": 0.0639,
+      "step": 5784
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2925192137676939,
+      "learning_rate": 5.446313172872234e-06,
+      "loss": 0.0474,
+      "step": 5785
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5562649537342925,
+      "learning_rate": 5.444953999024602e-06,
+      "loss": 0.0984,
+      "step": 5786
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5448080211413895,
+      "learning_rate": 5.4435947920342045e-06,
+      "loss": 0.1255,
+      "step": 5787
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.40403460525905366,
+      "learning_rate": 5.4422355520022814e-06,
+      "loss": 0.1249,
+      "step": 5788
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3862065898774094,
+      "learning_rate": 5.440876279030081e-06,
+      "loss": 0.1225,
+      "step": 5789
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3385398640118941,
+      "learning_rate": 5.439516973218844e-06,
+      "loss": 0.0981,
+      "step": 5790
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.29746574279341037,
+      "learning_rate": 5.438157634669825e-06,
+      "loss": 0.066,
+      "step": 5791
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.26264185043030625,
+      "learning_rate": 5.436798263484271e-06,
+      "loss": 0.0776,
+      "step": 5792
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4388424294379245,
+      "learning_rate": 5.4354388597634385e-06,
+      "loss": 0.1062,
+      "step": 5793
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.35413387057772644,
+      "learning_rate": 5.434079423608584e-06,
+      "loss": 0.0922,
+      "step": 5794
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.482696367897711,
+      "learning_rate": 5.432719955120963e-06,
+      "loss": 0.1144,
+      "step": 5795
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4412554583422484,
+      "learning_rate": 5.431360454401842e-06,
+      "loss": 0.1303,
+      "step": 5796
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5570839878973987,
+      "learning_rate": 5.430000921552479e-06,
+      "loss": 0.1261,
+      "step": 5797
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3142435168390734,
+      "learning_rate": 5.428641356674144e-06,
+      "loss": 0.0502,
+      "step": 5798
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.23914828727147622,
+      "learning_rate": 5.427281759868101e-06,
+      "loss": 0.0817,
+      "step": 5799
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3124252898243474,
+      "learning_rate": 5.425922131235627e-06,
+      "loss": 0.1185,
+      "step": 5800
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.39326557894380637,
+      "learning_rate": 5.42456247087799e-06,
+      "loss": 0.0983,
+      "step": 5801
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.45163575489107505,
+      "learning_rate": 5.4232027788964655e-06,
+      "loss": 0.0904,
+      "step": 5802
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4718931757609372,
+      "learning_rate": 5.421843055392335e-06,
+      "loss": 0.1321,
+      "step": 5803
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.48401073611050005,
+      "learning_rate": 5.420483300466874e-06,
+      "loss": 0.1029,
+      "step": 5804
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.43973430989405965,
+      "learning_rate": 5.419123514221369e-06,
+      "loss": 0.0903,
+      "step": 5805
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2378611941770208,
+      "learning_rate": 5.417763696757103e-06,
+      "loss": 0.0659,
+      "step": 5806
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2681876612943705,
+      "learning_rate": 5.416403848175362e-06,
+      "loss": 0.0744,
+      "step": 5807
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3069016596884403,
+      "learning_rate": 5.415043968577439e-06,
+      "loss": 0.067,
+      "step": 5808
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2513450866195137,
+      "learning_rate": 5.413684058064621e-06,
+      "loss": 0.0473,
+      "step": 5809
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3838998526448848,
+      "learning_rate": 5.412324116738207e-06,
+      "loss": 0.0854,
+      "step": 5810
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2910001845774251,
+      "learning_rate": 5.410964144699487e-06,
+      "loss": 0.1049,
+      "step": 5811
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2809230421844356,
+      "learning_rate": 5.4096041420497656e-06,
+      "loss": 0.0934,
+      "step": 5812
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.32864647845632317,
+      "learning_rate": 5.408244108890342e-06,
+      "loss": 0.0869,
+      "step": 5813
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4249609302752765,
+      "learning_rate": 5.406884045322519e-06,
+      "loss": 0.1318,
+      "step": 5814
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.22202876283011805,
+      "learning_rate": 5.405523951447602e-06,
+      "loss": 0.0677,
+      "step": 5815
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3102489618890112,
+      "learning_rate": 5.404163827366898e-06,
+      "loss": 0.0925,
+      "step": 5816
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4326009188272736,
+      "learning_rate": 5.402803673181719e-06,
+      "loss": 0.1203,
+      "step": 5817
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.27032184169262746,
+      "learning_rate": 5.4014434889933755e-06,
+      "loss": 0.0852,
+      "step": 5818
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2881934693378529,
+      "learning_rate": 5.400083274903181e-06,
+      "loss": 0.0737,
+      "step": 5819
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.24999747585160442,
+      "learning_rate": 5.398723031012455e-06,
+      "loss": 0.1029,
+      "step": 5820
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3233752137246742,
+      "learning_rate": 5.397362757422515e-06,
+      "loss": 0.1,
+      "step": 5821
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2929702287674229,
+      "learning_rate": 5.396002454234681e-06,
+      "loss": 0.0827,
+      "step": 5822
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.28087136090119724,
+      "learning_rate": 5.394642121550279e-06,
+      "loss": 0.0962,
+      "step": 5823
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4008858294120123,
+      "learning_rate": 5.39328175947063e-06,
+      "loss": 0.094,
+      "step": 5824
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.41433994844952604,
+      "learning_rate": 5.391921368097067e-06,
+      "loss": 0.088,
+      "step": 5825
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.33171759257417444,
+      "learning_rate": 5.3905609475309164e-06,
+      "loss": 0.0916,
+      "step": 5826
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2630866211488383,
+      "learning_rate": 5.389200497873512e-06,
+      "loss": 0.0671,
+      "step": 5827
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2786433332379624,
+      "learning_rate": 5.387840019226186e-06,
+      "loss": 0.0721,
+      "step": 5828
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.24235499895738113,
+      "learning_rate": 5.386479511690276e-06,
+      "loss": 0.0567,
+      "step": 5829
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.337008430306418,
+      "learning_rate": 5.385118975367122e-06,
+      "loss": 0.1405,
+      "step": 5830
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3284653822466512,
+      "learning_rate": 5.383758410358061e-06,
+      "loss": 0.0583,
+      "step": 5831
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.338068549738393,
+      "learning_rate": 5.38239781676444e-06,
+      "loss": 0.0719,
+      "step": 5832
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2877297217286004,
+      "learning_rate": 5.381037194687602e-06,
+      "loss": 0.0725,
+      "step": 5833
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4754223947167591,
+      "learning_rate": 5.379676544228894e-06,
+      "loss": 0.1158,
+      "step": 5834
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.21380103263903913,
+      "learning_rate": 5.378315865489664e-06,
+      "loss": 0.0536,
+      "step": 5835
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3600735838001535,
+      "learning_rate": 5.376955158571264e-06,
+      "loss": 0.0719,
+      "step": 5836
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3495979556173892,
+      "learning_rate": 5.37559442357505e-06,
+      "loss": 0.0818,
+      "step": 5837
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4074629636760797,
+      "learning_rate": 5.374233660602374e-06,
+      "loss": 0.1171,
+      "step": 5838
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3170574969120073,
+      "learning_rate": 5.372872869754595e-06,
+      "loss": 0.0518,
+      "step": 5839
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2757957544302347,
+      "learning_rate": 5.371512051133072e-06,
+      "loss": 0.0399,
+      "step": 5840
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.26860795578830654,
+      "learning_rate": 5.370151204839168e-06,
+      "loss": 0.0818,
+      "step": 5841
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3532850669058584,
+      "learning_rate": 5.368790330974246e-06,
+      "loss": 0.0755,
+      "step": 5842
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3123370490525611,
+      "learning_rate": 5.367429429639672e-06,
+      "loss": 0.0746,
+      "step": 5843
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.25344512682869874,
+      "learning_rate": 5.366068500936814e-06,
+      "loss": 0.0657,
+      "step": 5844
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2138844382270035,
+      "learning_rate": 5.3647075449670405e-06,
+      "loss": 0.083,
+      "step": 5845
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.47741541450968,
+      "learning_rate": 5.363346561831725e-06,
+      "loss": 0.1028,
+      "step": 5846
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.31529084173161936,
+      "learning_rate": 5.361985551632241e-06,
+      "loss": 0.0825,
+      "step": 5847
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.35532618180966635,
+      "learning_rate": 5.360624514469963e-06,
+      "loss": 0.0937,
+      "step": 5848
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.18780081210512914,
+      "learning_rate": 5.359263450446273e-06,
+      "loss": 0.05,
+      "step": 5849
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3027682616308561,
+      "learning_rate": 5.3579023596625455e-06,
+      "loss": 0.0728,
+      "step": 5850
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4961364988587402,
+      "learning_rate": 5.356541242220168e-06,
+      "loss": 0.1199,
+      "step": 5851
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2533616266521466,
+      "learning_rate": 5.3551800982205195e-06,
+      "loss": 0.06,
+      "step": 5852
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.632176783013107,
+      "learning_rate": 5.35381892776499e-06,
+      "loss": 0.0866,
+      "step": 5853
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3864514137484007,
+      "learning_rate": 5.352457730954964e-06,
+      "loss": 0.0846,
+      "step": 5854
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.37369094144340437,
+      "learning_rate": 5.351096507891834e-06,
+      "loss": 0.0923,
+      "step": 5855
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9331689389350514,
+      "learning_rate": 5.34973525867699e-06,
+      "loss": 0.1184,
+      "step": 5856
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3196874489671762,
+      "learning_rate": 5.348373983411829e-06,
+      "loss": 0.1118,
+      "step": 5857
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2525625080176364,
+      "learning_rate": 5.3470126821977405e-06,
+      "loss": 0.0862,
+      "step": 5858
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.28349128317882527,
+      "learning_rate": 5.3456513551361285e-06,
+      "loss": 0.0459,
+      "step": 5859
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.17846759299212983,
+      "learning_rate": 5.344290002328389e-06,
+      "loss": 0.0485,
+      "step": 5860
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.27301391258555086,
+      "learning_rate": 5.3429286238759255e-06,
+      "loss": 0.0995,
+      "step": 5861
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.22453682704146127,
+      "learning_rate": 5.34156721988014e-06,
+      "loss": 0.0796,
+      "step": 5862
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.27045335024644945,
+      "learning_rate": 5.3402057904424385e-06,
+      "loss": 0.0814,
+      "step": 5863
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2613688011478815,
+      "learning_rate": 5.338844335664227e-06,
+      "loss": 0.0796,
+      "step": 5864
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2976030247791636,
+      "learning_rate": 5.337482855646915e-06,
+      "loss": 0.048,
+      "step": 5865
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.298052194853526,
+      "learning_rate": 5.336121350491915e-06,
+      "loss": 0.0926,
+      "step": 5866
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3143718227900045,
+      "learning_rate": 5.334759820300639e-06,
+      "loss": 0.0805,
+      "step": 5867
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.46996072528057004,
+      "learning_rate": 5.333398265174501e-06,
+      "loss": 0.1226,
+      "step": 5868
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2358291568869031,
+      "learning_rate": 5.332036685214918e-06,
+      "loss": 0.0614,
+      "step": 5869
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4607818536874872,
+      "learning_rate": 5.330675080523308e-06,
+      "loss": 0.0703,
+      "step": 5870
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5683169559396875,
+      "learning_rate": 5.329313451201092e-06,
+      "loss": 0.1208,
+      "step": 5871
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.30224134679462283,
+      "learning_rate": 5.327951797349692e-06,
+      "loss": 0.0547,
+      "step": 5872
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4742975099063904,
+      "learning_rate": 5.326590119070532e-06,
+      "loss": 0.1124,
+      "step": 5873
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.20598101488227194,
+      "learning_rate": 5.3252284164650355e-06,
+      "loss": 0.0586,
+      "step": 5874
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.25352174154897184,
+      "learning_rate": 5.323866689634633e-06,
+      "loss": 0.0621,
+      "step": 5875
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3153490273424826,
+      "learning_rate": 5.322504938680754e-06,
+      "loss": 0.0884,
+      "step": 5876
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2667981001589733,
+      "learning_rate": 5.321143163704827e-06,
+      "loss": 0.0694,
+      "step": 5877
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.27128628278836237,
+      "learning_rate": 5.319781364808287e-06,
+      "loss": 0.0764,
+      "step": 5878
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4193948040620459,
+      "learning_rate": 5.3184195420925686e-06,
+      "loss": 0.0983,
+      "step": 5879
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2864954945659026,
+      "learning_rate": 5.317057695659108e-06,
+      "loss": 0.0632,
+      "step": 5880
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.21143711338179555,
+      "learning_rate": 5.315695825609341e-06,
+      "loss": 0.0517,
+      "step": 5881
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.34212078559431086,
+      "learning_rate": 5.314333932044711e-06,
+      "loss": 0.0922,
+      "step": 5882
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5890274574199412,
+      "learning_rate": 5.312972015066659e-06,
+      "loss": 0.1088,
+      "step": 5883
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.22022221024465025,
+      "learning_rate": 5.3116100747766284e-06,
+      "loss": 0.0759,
+      "step": 5884
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.33730724392780426,
+      "learning_rate": 5.310248111276064e-06,
+      "loss": 0.0949,
+      "step": 5885
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.38127342083392296,
+      "learning_rate": 5.308886124666413e-06,
+      "loss": 0.124,
+      "step": 5886
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4351611167669001,
+      "learning_rate": 5.3075241150491244e-06,
+      "loss": 0.0863,
+      "step": 5887
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.37539160964482904,
+      "learning_rate": 5.306162082525646e-06,
+      "loss": 0.0868,
+      "step": 5888
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.19796214184097738,
+      "learning_rate": 5.3048000271974356e-06,
+      "loss": 0.0634,
+      "step": 5889
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.35484654187889153,
+      "learning_rate": 5.303437949165941e-06,
+      "loss": 0.099,
+      "step": 5890
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0184687645592114,
+      "learning_rate": 5.302075848532622e-06,
+      "loss": 0.1141,
+      "step": 5891
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2830622445863827,
+      "learning_rate": 5.300713725398933e-06,
+      "loss": 0.0465,
+      "step": 5892
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2509815816529308,
+      "learning_rate": 5.2993515798663345e-06,
+      "loss": 0.079,
+      "step": 5893
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3436773409881198,
+      "learning_rate": 5.297989412036285e-06,
+      "loss": 0.1019,
+      "step": 5894
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5377545672526517,
+      "learning_rate": 5.29662722201025e-06,
+      "loss": 0.0824,
+      "step": 5895
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3873789928457281,
+      "learning_rate": 5.295265009889691e-06,
+      "loss": 0.096,
+      "step": 5896
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4018760807378341,
+      "learning_rate": 5.293902775776074e-06,
+      "loss": 0.0691,
+      "step": 5897
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.32093257797482977,
+      "learning_rate": 5.292540519770864e-06,
+      "loss": 0.1028,
+      "step": 5898
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.281243435284015,
+      "learning_rate": 5.291178241975534e-06,
+      "loss": 0.0728,
+      "step": 5899
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5005730057655735,
+      "learning_rate": 5.289815942491553e-06,
+      "loss": 0.102,
+      "step": 5900
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.36259547387027613,
+      "learning_rate": 5.2884536214203906e-06,
+      "loss": 0.0862,
+      "step": 5901
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.40905649306165726,
+      "learning_rate": 5.287091278863524e-06,
+      "loss": 0.1181,
+      "step": 5902
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2893865074567721,
+      "learning_rate": 5.285728914922425e-06,
+      "loss": 0.0539,
+      "step": 5903
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3114824576784453,
+      "learning_rate": 5.284366529698572e-06,
+      "loss": 0.0905,
+      "step": 5904
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2703730092553051,
+      "learning_rate": 5.283004123293445e-06,
+      "loss": 0.0489,
+      "step": 5905
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3075813974631438,
+      "learning_rate": 5.281641695808519e-06,
+      "loss": 0.1066,
+      "step": 5906
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.258751587127309,
+      "learning_rate": 5.280279247345283e-06,
+      "loss": 0.0505,
+      "step": 5907
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2828849318854819,
+      "learning_rate": 5.278916778005212e-06,
+      "loss": 0.0622,
+      "step": 5908
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.38363668036834864,
+      "learning_rate": 5.277554287889798e-06,
+      "loss": 0.0935,
+      "step": 5909
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.24332167433953486,
+      "learning_rate": 5.276191777100521e-06,
+      "loss": 0.0523,
+      "step": 5910
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.22784704747186477,
+      "learning_rate": 5.274829245738872e-06,
+      "loss": 0.0576,
+      "step": 5911
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2959052956269438,
+      "learning_rate": 5.27346669390634e-06,
+      "loss": 0.0648,
+      "step": 5912
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.45115793924552994,
+      "learning_rate": 5.272104121704417e-06,
+      "loss": 0.0996,
+      "step": 5913
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3299800514050582,
+      "learning_rate": 5.270741529234593e-06,
+      "loss": 0.1047,
+      "step": 5914
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4349666765386404,
+      "learning_rate": 5.269378916598362e-06,
+      "loss": 0.0673,
+      "step": 5915
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.39744157951274733,
+      "learning_rate": 5.268016283897219e-06,
+      "loss": 0.0557,
+      "step": 5916
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5614272704360066,
+      "learning_rate": 5.266653631232665e-06,
+      "loss": 0.1439,
+      "step": 5917
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5728925466558913,
+      "learning_rate": 5.265290958706193e-06,
+      "loss": 0.0947,
+      "step": 5918
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2966067025469743,
+      "learning_rate": 5.263928266419306e-06,
+      "loss": 0.0672,
+      "step": 5919
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3315483235269579,
+      "learning_rate": 5.262565554473503e-06,
+      "loss": 0.0785,
+      "step": 5920
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.28046838501260424,
+      "learning_rate": 5.26120282297029e-06,
+      "loss": 0.0964,
+      "step": 5921
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3593400502979721,
+      "learning_rate": 5.259840072011168e-06,
+      "loss": 0.1196,
+      "step": 5922
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4858668951485186,
+      "learning_rate": 5.258477301697643e-06,
+      "loss": 0.0774,
+      "step": 5923
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2852576439764612,
+      "learning_rate": 5.2571145121312225e-06,
+      "loss": 0.1049,
+      "step": 5924
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2921896146119645,
+      "learning_rate": 5.255751703413418e-06,
+      "loss": 0.0572,
+      "step": 5925
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.21550888446582636,
+      "learning_rate": 5.254388875645734e-06,
+      "loss": 0.0567,
+      "step": 5926
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7359562388505169,
+      "learning_rate": 5.253026028929687e-06,
+      "loss": 0.1002,
+      "step": 5927
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.30707085315385535,
+      "learning_rate": 5.251663163366785e-06,
+      "loss": 0.0568,
+      "step": 5928
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3113336742342824,
+      "learning_rate": 5.250300279058546e-06,
+      "loss": 0.0841,
+      "step": 5929
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5603573509038955,
+      "learning_rate": 5.248937376106483e-06,
+      "loss": 0.0965,
+      "step": 5930
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.22265730954603144,
+      "learning_rate": 5.247574454612115e-06,
+      "loss": 0.0607,
+      "step": 5931
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.366032201458281,
+      "learning_rate": 5.24621151467696e-06,
+      "loss": 0.1,
+      "step": 5932
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.34885586152348336,
+      "learning_rate": 5.244848556402535e-06,
+      "loss": 0.0996,
+      "step": 5933
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2458269058914401,
+      "learning_rate": 5.243485579890365e-06,
+      "loss": 0.0786,
+      "step": 5934
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.43203385607718814,
+      "learning_rate": 5.24212258524197e-06,
+      "loss": 0.1243,
+      "step": 5935
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5942081425326236,
+      "learning_rate": 5.240759572558874e-06,
+      "loss": 0.1134,
+      "step": 5936
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.29506986376868666,
+      "learning_rate": 5.239396541942603e-06,
+      "loss": 0.1055,
+      "step": 5937
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3658487014076287,
+      "learning_rate": 5.2380334934946816e-06,
+      "loss": 0.0859,
+      "step": 5938
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3940361911956606,
+      "learning_rate": 5.236670427316641e-06,
+      "loss": 0.107,
+      "step": 5939
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.17776821079862817,
+      "learning_rate": 5.235307343510008e-06,
+      "loss": 0.0365,
+      "step": 5940
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.49507112224321226,
+      "learning_rate": 5.2339442421763135e-06,
+      "loss": 0.0937,
+      "step": 5941
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.38761918451307387,
+      "learning_rate": 5.232581123417087e-06,
+      "loss": 0.0886,
+      "step": 5942
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3604602083178836,
+      "learning_rate": 5.231217987333865e-06,
+      "loss": 0.0808,
+      "step": 5943
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.41391591755741614,
+      "learning_rate": 5.229854834028181e-06,
+      "loss": 0.0761,
+      "step": 5944
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.34931570151947455,
+      "learning_rate": 5.228491663601568e-06,
+      "loss": 0.0905,
+      "step": 5945
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.32140015880840156,
+      "learning_rate": 5.227128476155566e-06,
+      "loss": 0.0574,
+      "step": 5946
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.35284860343490987,
+      "learning_rate": 5.2257652717917115e-06,
+      "loss": 0.0396,
+      "step": 5947
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3187629835346542,
+      "learning_rate": 5.224402050611547e-06,
+      "loss": 0.0537,
+      "step": 5948
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5469521529973783,
+      "learning_rate": 5.223038812716607e-06,
+      "loss": 0.1037,
+      "step": 5949
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.21118980617296457,
+      "learning_rate": 5.221675558208438e-06,
+      "loss": 0.039,
+      "step": 5950
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.47383463843984874,
+      "learning_rate": 5.220312287188583e-06,
+      "loss": 0.1454,
+      "step": 5951
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2717259588284945,
+      "learning_rate": 5.218948999758584e-06,
+      "loss": 0.0571,
+      "step": 5952
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.29768858507025275,
+      "learning_rate": 5.2175856960199896e-06,
+      "loss": 0.0891,
+      "step": 5953
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3108257345894743,
+      "learning_rate": 5.216222376074343e-06,
+      "loss": 0.0727,
+      "step": 5954
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.26790559583117884,
+      "learning_rate": 5.214859040023197e-06,
+      "loss": 0.1058,
+      "step": 5955
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.1729405371315702,
+      "learning_rate": 5.213495687968096e-06,
+      "loss": 0.0454,
+      "step": 5956
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2882213341580032,
+      "learning_rate": 5.212132320010593e-06,
+      "loss": 0.0742,
+      "step": 5957
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.35019410501393217,
+      "learning_rate": 5.21076893625224e-06,
+      "loss": 0.1134,
+      "step": 5958
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2590088417304105,
+      "learning_rate": 5.2094055367945874e-06,
+      "loss": 0.0512,
+      "step": 5959
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3632555304199441,
+      "learning_rate": 5.208042121739191e-06,
+      "loss": 0.135,
+      "step": 5960
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.19896321479280807,
+      "learning_rate": 5.2066786911876055e-06,
+      "loss": 0.0427,
+      "step": 5961
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3718919503776485,
+      "learning_rate": 5.205315245241387e-06,
+      "loss": 0.113,
+      "step": 5962
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3040236896069676,
+      "learning_rate": 5.203951784002094e-06,
+      "loss": 0.1111,
+      "step": 5963
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.26393562291799183,
+      "learning_rate": 5.202588307571282e-06,
+      "loss": 0.0666,
+      "step": 5964
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.24472455738753562,
+      "learning_rate": 5.201224816050514e-06,
+      "loss": 0.0498,
+      "step": 5965
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.25578600245277316,
+      "learning_rate": 5.199861309541348e-06,
+      "loss": 0.0499,
+      "step": 5966
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.259787230681287,
+      "learning_rate": 5.1984977881453495e-06,
+      "loss": 0.0984,
+      "step": 5967
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.27769885696048274,
+      "learning_rate": 5.197134251964079e-06,
+      "loss": 0.0841,
+      "step": 5968
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4147406018659988,
+      "learning_rate": 5.1957707010991e-06,
+      "loss": 0.115,
+      "step": 5969
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3659666632133375,
+      "learning_rate": 5.19440713565198e-06,
+      "loss": 0.0926,
+      "step": 5970
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3471676841698387,
+      "learning_rate": 5.1930435557242816e-06,
+      "loss": 0.1109,
+      "step": 5971
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.33187993988037656,
+      "learning_rate": 5.191679961417577e-06,
+      "loss": 0.0767,
+      "step": 5972
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2284556792821559,
+      "learning_rate": 5.190316352833432e-06,
+      "loss": 0.0497,
+      "step": 5973
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.16747608400686667,
+      "learning_rate": 5.188952730073416e-06,
+      "loss": 0.0446,
+      "step": 5974
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.39378973268290185,
+      "learning_rate": 5.187589093239099e-06,
+      "loss": 0.0803,
+      "step": 5975
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3611311725092675,
+      "learning_rate": 5.186225442432055e-06,
+      "loss": 0.102,
+      "step": 5976
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5105128066014489,
+      "learning_rate": 5.184861777753855e-06,
+      "loss": 0.0953,
+      "step": 5977
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.32264093418209583,
+      "learning_rate": 5.183498099306072e-06,
+      "loss": 0.0931,
+      "step": 5978
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.19557256959602395,
+      "learning_rate": 5.182134407190282e-06,
+      "loss": 0.0822,
+      "step": 5979
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.33454980657832073,
+      "learning_rate": 5.180770701508061e-06,
+      "loss": 0.0817,
+      "step": 5980
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6800409858291414,
+      "learning_rate": 5.1794069823609835e-06,
+      "loss": 0.121,
+      "step": 5981
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2530820712205607,
+      "learning_rate": 5.17804324985063e-06,
+      "loss": 0.0717,
+      "step": 5982
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.31550944695375893,
+      "learning_rate": 5.176679504078578e-06,
+      "loss": 0.0601,
+      "step": 5983
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7737975011438912,
+      "learning_rate": 5.175315745146406e-06,
+      "loss": 0.1115,
+      "step": 5984
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3591080191243384,
+      "learning_rate": 5.173951973155698e-06,
+      "loss": 0.0851,
+      "step": 5985
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.1899691479200827,
+      "learning_rate": 5.172588188208033e-06,
+      "loss": 0.0655,
+      "step": 5986
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.27839218167742374,
+      "learning_rate": 5.1712243904049956e-06,
+      "loss": 0.1004,
+      "step": 5987
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.315812940174292,
+      "learning_rate": 5.169860579848166e-06,
+      "loss": 0.0483,
+      "step": 5988
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.31524438755974604,
+      "learning_rate": 5.168496756639132e-06,
+      "loss": 0.0958,
+      "step": 5989
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.39547916985896014,
+      "learning_rate": 5.167132920879478e-06,
+      "loss": 0.076,
+      "step": 5990
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5516666827136503,
+      "learning_rate": 5.16576907267079e-06,
+      "loss": 0.0912,
+      "step": 5991
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.858550522906612,
+      "learning_rate": 5.164405212114656e-06,
+      "loss": 0.1894,
+      "step": 5992
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2727058603363414,
+      "learning_rate": 5.163041339312664e-06,
+      "loss": 0.0551,
+      "step": 5993
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.31847098732429524,
+      "learning_rate": 5.1616774543664025e-06,
+      "loss": 0.1041,
+      "step": 5994
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.28331826469159443,
+      "learning_rate": 5.1603135573774645e-06,
+      "loss": 0.1042,
+      "step": 5995
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2662291660836843,
+      "learning_rate": 5.158949648447436e-06,
+      "loss": 0.0677,
+      "step": 5996
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.22852635336766552,
+      "learning_rate": 5.157585727677914e-06,
+      "loss": 0.0462,
+      "step": 5997
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3893035322829395,
+      "learning_rate": 5.1562217951704885e-06,
+      "loss": 0.0867,
+      "step": 5998
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.28045538888243887,
+      "learning_rate": 5.154857851026754e-06,
+      "loss": 0.0699,
+      "step": 5999
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4662070547928093,
+      "learning_rate": 5.153493895348303e-06,
+      "loss": 0.1064,
+      "step": 6000
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2500785431668303,
+      "learning_rate": 5.152129928236733e-06,
+      "loss": 0.0584,
+      "step": 6001
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5059242494636329,
+      "learning_rate": 5.150765949793641e-06,
+      "loss": 0.1115,
+      "step": 6002
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.381436787512659,
+      "learning_rate": 5.149401960120621e-06,
+      "loss": 0.0792,
+      "step": 6003
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.26742655601512416,
+      "learning_rate": 5.148037959319275e-06,
+      "loss": 0.1012,
+      "step": 6004
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.439925332102308,
+      "learning_rate": 5.146673947491198e-06,
+      "loss": 0.1111,
+      "step": 6005
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.29128553636157606,
+      "learning_rate": 5.1453099247379915e-06,
+      "loss": 0.0631,
+      "step": 6006
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.33777613782662774,
+      "learning_rate": 5.143945891161254e-06,
+      "loss": 0.0902,
+      "step": 6007
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.22804846787291416,
+      "learning_rate": 5.142581846862589e-06,
+      "loss": 0.0589,
+      "step": 6008
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2072539903007768,
+      "learning_rate": 5.141217791943597e-06,
+      "loss": 0.041,
+      "step": 6009
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2625603392078211,
+      "learning_rate": 5.139853726505881e-06,
+      "loss": 0.0679,
+      "step": 6010
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3142276861074264,
+      "learning_rate": 5.138489650651046e-06,
+      "loss": 0.062,
+      "step": 6011
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5182941206071672,
+      "learning_rate": 5.137125564480694e-06,
+      "loss": 0.1312,
+      "step": 6012
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3000767163669192,
+      "learning_rate": 5.135761468096431e-06,
+      "loss": 0.0677,
+      "step": 6013
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.20578289587005374,
+      "learning_rate": 5.134397361599864e-06,
+      "loss": 0.0628,
+      "step": 6014
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5389329009272015,
+      "learning_rate": 5.133033245092597e-06,
+      "loss": 0.1489,
+      "step": 6015
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.44336897162634975,
+      "learning_rate": 5.131669118676241e-06,
+      "loss": 0.1006,
+      "step": 6016
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.28585383584612256,
+      "learning_rate": 5.1303049824524e-06,
+      "loss": 0.0516,
+      "step": 6017
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.19821235155327224,
+      "learning_rate": 5.128940836522686e-06,
+      "loss": 0.034,
+      "step": 6018
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2646533646251551,
+      "learning_rate": 5.127576680988708e-06,
+      "loss": 0.0814,
+      "step": 6019
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.25190132046275804,
+      "learning_rate": 5.126212515952073e-06,
+      "loss": 0.077,
+      "step": 6020
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2830587747496971,
+      "learning_rate": 5.1248483415143976e-06,
+      "loss": 0.0696,
+      "step": 6021
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4068967408667984,
+      "learning_rate": 5.1234841577772884e-06,
+      "loss": 0.0707,
+      "step": 6022
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3742501212046979,
+      "learning_rate": 5.122119964842361e-06,
+      "loss": 0.0882,
+      "step": 6023
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2364339289600392,
+      "learning_rate": 5.120755762811227e-06,
+      "loss": 0.073,
+      "step": 6024
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.30354415010300856,
+      "learning_rate": 5.119391551785499e-06,
+      "loss": 0.0725,
+      "step": 6025
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6922659091844138,
+      "learning_rate": 5.118027331866795e-06,
+      "loss": 0.1682,
+      "step": 6026
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5031979493723242,
+      "learning_rate": 5.1166631031567275e-06,
+      "loss": 0.0987,
+      "step": 6027
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.24862483290448176,
+      "learning_rate": 5.11529886575691e-06,
+      "loss": 0.0381,
+      "step": 6028
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4987042803812137,
+      "learning_rate": 5.113934619768963e-06,
+      "loss": 0.0931,
+      "step": 6029
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.42348936868222986,
+      "learning_rate": 5.1125703652945014e-06,
+      "loss": 0.1061,
+      "step": 6030
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.40289906708737316,
+      "learning_rate": 5.111206102435144e-06,
+      "loss": 0.0932,
+      "step": 6031
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.341997785024806,
+      "learning_rate": 5.109841831292508e-06,
+      "loss": 0.1009,
+      "step": 6032
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2748591004662362,
+      "learning_rate": 5.108477551968213e-06,
+      "loss": 0.0775,
+      "step": 6033
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2860964350554465,
+      "learning_rate": 5.107113264563876e-06,
+      "loss": 0.0628,
+      "step": 6034
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6405188951230092,
+      "learning_rate": 5.105748969181121e-06,
+      "loss": 0.0952,
+      "step": 6035
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2633049428036427,
+      "learning_rate": 5.1043846659215654e-06,
+      "loss": 0.092,
+      "step": 6036
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.42460474650644475,
+      "learning_rate": 5.103020354886832e-06,
+      "loss": 0.1037,
+      "step": 6037
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.41450603349474446,
+      "learning_rate": 5.1016560361785435e-06,
+      "loss": 0.1129,
+      "step": 6038
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.31181237576527454,
+      "learning_rate": 5.10029170989832e-06,
+      "loss": 0.0958,
+      "step": 6039
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.19679489420377286,
+      "learning_rate": 5.0989273761477876e-06,
+      "loss": 0.0699,
+      "step": 6040
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.35651059392153267,
+      "learning_rate": 5.097563035028565e-06,
+      "loss": 0.0824,
+      "step": 6041
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.49997600730956254,
+      "learning_rate": 5.096198686642281e-06,
+      "loss": 0.1499,
+      "step": 6042
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4810115303788511,
+      "learning_rate": 5.094834331090559e-06,
+      "loss": 0.1039,
+      "step": 6043
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3926215205293296,
+      "learning_rate": 5.093469968475022e-06,
+      "loss": 0.1109,
+      "step": 6044
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.20481746960437508,
+      "learning_rate": 5.0921055988972985e-06,
+      "loss": 0.0502,
+      "step": 6045
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8602901296057163,
+      "learning_rate": 5.090741222459014e-06,
+      "loss": 0.1349,
+      "step": 6046
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.44396170496319914,
+      "learning_rate": 5.089376839261793e-06,
+      "loss": 0.0745,
+      "step": 6047
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5643161108627387,
+      "learning_rate": 5.0880124494072644e-06,
+      "loss": 0.1151,
+      "step": 6048
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.43854214195706365,
+      "learning_rate": 5.086648052997056e-06,
+      "loss": 0.1003,
+      "step": 6049
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.42559501223458013,
+      "learning_rate": 5.085283650132797e-06,
+      "loss": 0.1038,
+      "step": 6050
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4892743225990912,
+      "learning_rate": 5.0839192409161145e-06,
+      "loss": 0.1401,
+      "step": 6051
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2545673550847033,
+      "learning_rate": 5.082554825448638e-06,
+      "loss": 0.0821,
+      "step": 6052
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.42593398863016335,
+      "learning_rate": 5.081190403831997e-06,
+      "loss": 0.1137,
+      "step": 6053
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2274941087951388,
+      "learning_rate": 5.079825976167821e-06,
+      "loss": 0.0625,
+      "step": 6054
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.1998892508266908,
+      "learning_rate": 5.0784615425577435e-06,
+      "loss": 0.0687,
+      "step": 6055
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.32022767270678354,
+      "learning_rate": 5.077097103103391e-06,
+      "loss": 0.1004,
+      "step": 6056
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2857402003596764,
+      "learning_rate": 5.075732657906398e-06,
+      "loss": 0.0595,
+      "step": 6057
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.28914356400075425,
+      "learning_rate": 5.074368207068395e-06,
+      "loss": 0.0617,
+      "step": 6058
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.24259193360553943,
+      "learning_rate": 5.073003750691016e-06,
+      "loss": 0.0655,
+      "step": 6059
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.34322116862894014,
+      "learning_rate": 5.071639288875891e-06,
+      "loss": 0.0704,
+      "step": 6060
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4182289942633452,
+      "learning_rate": 5.070274821724656e-06,
+      "loss": 0.1355,
+      "step": 6061
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.48638309639879845,
+      "learning_rate": 5.06891034933894e-06,
+      "loss": 0.0988,
+      "step": 6062
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3461678141176499,
+      "learning_rate": 5.067545871820383e-06,
+      "loss": 0.0774,
+      "step": 6063
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.44029205232972474,
+      "learning_rate": 5.0661813892706135e-06,
+      "loss": 0.0978,
+      "step": 6064
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.26941632569487267,
+      "learning_rate": 5.064816901791271e-06,
+      "loss": 0.0739,
+      "step": 6065
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3282484488175969,
+      "learning_rate": 5.063452409483986e-06,
+      "loss": 0.0868,
+      "step": 6066
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.24144385176604088,
+      "learning_rate": 5.062087912450397e-06,
+      "loss": 0.0775,
+      "step": 6067
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3277952699966483,
+      "learning_rate": 5.060723410792137e-06,
+      "loss": 0.0792,
+      "step": 6068
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.18504452230728521,
+      "learning_rate": 5.059358904610846e-06,
+      "loss": 0.028,
+      "step": 6069
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.369359005896634,
+      "learning_rate": 5.057994394008156e-06,
+      "loss": 0.1059,
+      "step": 6070
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5076208991672917,
+      "learning_rate": 5.056629879085706e-06,
+      "loss": 0.0758,
+      "step": 6071
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.28294199569622525,
+      "learning_rate": 5.055265359945134e-06,
+      "loss": 0.0514,
+      "step": 6072
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3744689800713614,
+      "learning_rate": 5.053900836688075e-06,
+      "loss": 0.0647,
+      "step": 6073
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3728414059460615,
+      "learning_rate": 5.052536309416168e-06,
+      "loss": 0.0957,
+      "step": 6074
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.34654683058409563,
+      "learning_rate": 5.051171778231051e-06,
+      "loss": 0.1005,
+      "step": 6075
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.24576838475027235,
+      "learning_rate": 5.04980724323436e-06,
+      "loss": 0.0578,
+      "step": 6076
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4943386021879641,
+      "learning_rate": 5.048442704527738e-06,
+      "loss": 0.1092,
+      "step": 6077
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5020615791784437,
+      "learning_rate": 5.047078162212819e-06,
+      "loss": 0.1106,
+      "step": 6078
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5014360441255473,
+      "learning_rate": 5.0457136163912454e-06,
+      "loss": 0.1264,
+      "step": 6079
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.47226602959703157,
+      "learning_rate": 5.044349067164656e-06,
+      "loss": 0.1275,
+      "step": 6080
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.291855423467308,
+      "learning_rate": 5.042984514634688e-06,
+      "loss": 0.0905,
+      "step": 6081
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.26835728922279606,
+      "learning_rate": 5.0416199589029845e-06,
+      "loss": 0.0802,
+      "step": 6082
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2662254900305024,
+      "learning_rate": 5.040255400071182e-06,
+      "loss": 0.0915,
+      "step": 6083
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.21033866312393795,
+      "learning_rate": 5.038890838240925e-06,
+      "loss": 0.0735,
+      "step": 6084
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4582072771153348,
+      "learning_rate": 5.037526273513851e-06,
+      "loss": 0.1122,
+      "step": 6085
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.28944041072694016,
+      "learning_rate": 5.0361617059916025e-06,
+      "loss": 0.0792,
+      "step": 6086
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3898622098781108,
+      "learning_rate": 5.034797135775818e-06,
+      "loss": 0.104,
+      "step": 6087
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4119372114693691,
+      "learning_rate": 5.033432562968139e-06,
+      "loss": 0.099,
+      "step": 6088
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.28604862202724446,
+      "learning_rate": 5.03206798767021e-06,
+      "loss": 0.0937,
+      "step": 6089
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4169722220453741,
+      "learning_rate": 5.03070340998367e-06,
+      "loss": 0.1129,
+      "step": 6090
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2675338684397392,
+      "learning_rate": 5.029338830010162e-06,
+      "loss": 0.0661,
+      "step": 6091
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.32700418677726056,
+      "learning_rate": 5.027974247851326e-06,
+      "loss": 0.0642,
+      "step": 6092
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.329996492913501,
+      "learning_rate": 5.026609663608806e-06,
+      "loss": 0.091,
+      "step": 6093
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5688733007835383,
+      "learning_rate": 5.025245077384241e-06,
+      "loss": 0.1369,
+      "step": 6094
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3344755744175377,
+      "learning_rate": 5.023880489279278e-06,
+      "loss": 0.092,
+      "step": 6095
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.41448752927253607,
+      "learning_rate": 5.022515899395556e-06,
+      "loss": 0.094,
+      "step": 6096
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.44026909409439197,
+      "learning_rate": 5.021151307834719e-06,
+      "loss": 0.143,
+      "step": 6097
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.23724197925826088,
+      "learning_rate": 5.019786714698409e-06,
+      "loss": 0.058,
+      "step": 6098
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.44323632492808446,
+      "learning_rate": 5.01842212008827e-06,
+      "loss": 0.1428,
+      "step": 6099
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.29654520443007787,
+      "learning_rate": 5.017057524105943e-06,
+      "loss": 0.1426,
+      "step": 6100
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.23975135627080457,
+      "learning_rate": 5.015692926853073e-06,
+      "loss": 0.0751,
+      "step": 6101
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5190032153670512,
+      "learning_rate": 5.014328328431301e-06,
+      "loss": 0.1205,
+      "step": 6102
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2606633688499103,
+      "learning_rate": 5.012963728942274e-06,
+      "loss": 0.077,
+      "step": 6103
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2896102191932014,
+      "learning_rate": 5.011599128487631e-06,
+      "loss": 0.0758,
+      "step": 6104
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.42154835801245427,
+      "learning_rate": 5.010234527169018e-06,
+      "loss": 0.0911,
+      "step": 6105
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5778534701003817,
+      "learning_rate": 5.0088699250880775e-06,
+      "loss": 0.1206,
+      "step": 6106
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2923775936553259,
+      "learning_rate": 5.007505322346455e-06,
+      "loss": 0.0789,
+      "step": 6107
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6714251280893911,
+      "learning_rate": 5.006140719045791e-06,
+      "loss": 0.1082,
+      "step": 6108
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3114395221429919,
+      "learning_rate": 5.004776115287729e-06,
+      "loss": 0.0844,
+      "step": 6109
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3623220140843325,
+      "learning_rate": 5.0034115111739155e-06,
+      "loss": 0.0874,
+      "step": 6110
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3462248039122916,
+      "learning_rate": 5.0020469068059944e-06,
+      "loss": 0.0984,
+      "step": 6111
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.32836708934653575,
+      "learning_rate": 5.000682302285606e-06,
+      "loss": 0.1043,
+      "step": 6112
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4518292686496185,
+      "learning_rate": 4.999317697714396e-06,
+      "loss": 0.1423,
+      "step": 6113
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.18977448273876807,
+      "learning_rate": 4.997953093194008e-06,
+      "loss": 0.0325,
+      "step": 6114
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.19822460095554992,
+      "learning_rate": 4.9965884888260845e-06,
+      "loss": 0.0372,
+      "step": 6115
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.525613808028575,
+      "learning_rate": 4.9952238847122716e-06,
+      "loss": 0.1213,
+      "step": 6116
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.224031031105772,
+      "learning_rate": 4.993859280954212e-06,
+      "loss": 0.0542,
+      "step": 6117
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.2888428991375344,
+      "learning_rate": 4.992494677653546e-06,
+      "loss": 0.0748,
+      "step": 6118
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.33916694725982105,
+      "learning_rate": 4.991130074911923e-06,
+      "loss": 0.0893,
+      "step": 6119
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.36721548092249173,
+      "learning_rate": 4.9897654728309835e-06,
+      "loss": 0.0826,
+      "step": 6120
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.25161914575843103,
+      "learning_rate": 4.988400871512371e-06,
+      "loss": 0.068,
+      "step": 6121
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.21672267062809975,
+      "learning_rate": 4.987036271057728e-06,
+      "loss": 0.0863,
+      "step": 6122
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.2740922523337386,
+      "learning_rate": 4.985671671568699e-06,
+      "loss": 0.0881,
+      "step": 6123
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.24116254830966544,
+      "learning_rate": 4.984307073146929e-06,
+      "loss": 0.0657,
+      "step": 6124
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4553925657289683,
+      "learning_rate": 4.982942475894059e-06,
+      "loss": 0.0912,
+      "step": 6125
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.19328634084111265,
+      "learning_rate": 4.981577879911732e-06,
+      "loss": 0.0524,
+      "step": 6126
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5352265551660182,
+      "learning_rate": 4.980213285301593e-06,
+      "loss": 0.0963,
+      "step": 6127
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4117215521420952,
+      "learning_rate": 4.978848692165282e-06,
+      "loss": 0.064,
+      "step": 6128
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.27829909400673064,
+      "learning_rate": 4.977484100604446e-06,
+      "loss": 0.0763,
+      "step": 6129
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.613680964042532,
+      "learning_rate": 4.976119510720723e-06,
+      "loss": 0.1162,
+      "step": 6130
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.21436500883046947,
+      "learning_rate": 4.97475492261576e-06,
+      "loss": 0.0428,
+      "step": 6131
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.2534080720507505,
+      "learning_rate": 4.9733903363911964e-06,
+      "loss": 0.0665,
+      "step": 6132
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.24252724256587946,
+      "learning_rate": 4.972025752148676e-06,
+      "loss": 0.041,
+      "step": 6133
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3769287194881874,
+      "learning_rate": 4.970661169989839e-06,
+      "loss": 0.089,
+      "step": 6134
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.2688266686908356,
+      "learning_rate": 4.9692965900163305e-06,
+      "loss": 0.0469,
+      "step": 6135
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.2768978589757916,
+      "learning_rate": 4.9679320123297906e-06,
+      "loss": 0.0811,
+      "step": 6136
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3728127703149894,
+      "learning_rate": 4.966567437031862e-06,
+      "loss": 0.1083,
+      "step": 6137
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.33874900795796165,
+      "learning_rate": 4.965202864224185e-06,
+      "loss": 0.0951,
+      "step": 6138
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.36637561647606703,
+      "learning_rate": 4.963838294008399e-06,
+      "loss": 0.1024,
+      "step": 6139
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3797706313518771,
+      "learning_rate": 4.9624737264861504e-06,
+      "loss": 0.0917,
+      "step": 6140
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.26124896950166415,
+      "learning_rate": 4.961109161759077e-06,
+      "loss": 0.0719,
+      "step": 6141
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.36908080892762407,
+      "learning_rate": 4.959744599928818e-06,
+      "loss": 0.0612,
+      "step": 6142
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3029762992871014,
+      "learning_rate": 4.958380041097017e-06,
+      "loss": 0.0946,
+      "step": 6143
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4126229304999895,
+      "learning_rate": 4.957015485365314e-06,
+      "loss": 0.1055,
+      "step": 6144
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.29787301507691455,
+      "learning_rate": 4.955650932835345e-06,
+      "loss": 0.0657,
+      "step": 6145
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3451998528500356,
+      "learning_rate": 4.954286383608755e-06,
+      "loss": 0.0824,
+      "step": 6146
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.37345329058272825,
+      "learning_rate": 4.952921837787182e-06,
+      "loss": 0.0699,
+      "step": 6147
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.28721887579417194,
+      "learning_rate": 4.951557295472265e-06,
+      "loss": 0.0819,
+      "step": 6148
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.29254083426368205,
+      "learning_rate": 4.95019275676564e-06,
+      "loss": 0.0917,
+      "step": 6149
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.23525709287225097,
+      "learning_rate": 4.948828221768951e-06,
+      "loss": 0.0835,
+      "step": 6150
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.27557004524727574,
+      "learning_rate": 4.947463690583834e-06,
+      "loss": 0.0859,
+      "step": 6151
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3586414347561702,
+      "learning_rate": 4.946099163311925e-06,
+      "loss": 0.1031,
+      "step": 6152
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.2614876758449999,
+      "learning_rate": 4.944734640054867e-06,
+      "loss": 0.0827,
+      "step": 6153
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.31298286827836347,
+      "learning_rate": 4.943370120914295e-06,
+      "loss": 0.0757,
+      "step": 6154
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3697490641944804,
+      "learning_rate": 4.942005605991846e-06,
+      "loss": 0.0667,
+      "step": 6155
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7486924062975983,
+      "learning_rate": 4.940641095389155e-06,
+      "loss": 0.1032,
+      "step": 6156
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3381615215701721,
+      "learning_rate": 4.9392765892078635e-06,
+      "loss": 0.0507,
+      "step": 6157
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3717706200808369,
+      "learning_rate": 4.937912087549606e-06,
+      "loss": 0.0794,
+      "step": 6158
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.2468357848572955,
+      "learning_rate": 4.936547590516016e-06,
+      "loss": 0.0701,
+      "step": 6159
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3120429182848458,
+      "learning_rate": 4.935183098208731e-06,
+      "loss": 0.0693,
+      "step": 6160
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.320773136050333,
+      "learning_rate": 4.933818610729388e-06,
+      "loss": 0.08,
+      "step": 6161
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.1693172104819401,
+      "learning_rate": 4.932454128179619e-06,
+      "loss": 0.0488,
+      "step": 6162
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.34422545861133563,
+      "learning_rate": 4.931089650661061e-06,
+      "loss": 0.0597,
+      "step": 6163
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.1610373497546505,
+      "learning_rate": 4.929725178275347e-06,
+      "loss": 0.0206,
+      "step": 6164
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4164289325134955,
+      "learning_rate": 4.92836071112411e-06,
+      "loss": 0.1167,
+      "step": 6165
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.33136098218215754,
+      "learning_rate": 4.926996249308986e-06,
+      "loss": 0.0478,
+      "step": 6166
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7019121529586255,
+      "learning_rate": 4.925631792931607e-06,
+      "loss": 0.1094,
+      "step": 6167
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.448148126412502,
+      "learning_rate": 4.924267342093603e-06,
+      "loss": 0.0971,
+      "step": 6168
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3316166854938567,
+      "learning_rate": 4.92290289689661e-06,
+      "loss": 0.0757,
+      "step": 6169
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.38271703714217054,
+      "learning_rate": 4.921538457442259e-06,
+      "loss": 0.0924,
+      "step": 6170
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.25306676155551033,
+      "learning_rate": 4.9201740238321805e-06,
+      "loss": 0.0693,
+      "step": 6171
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.30773801579248233,
+      "learning_rate": 4.9188095961680055e-06,
+      "loss": 0.0783,
+      "step": 6172
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.48992162934954697,
+      "learning_rate": 4.917445174551363e-06,
+      "loss": 0.133,
+      "step": 6173
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.2861373728579443,
+      "learning_rate": 4.916080759083887e-06,
+      "loss": 0.084,
+      "step": 6174
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.25371818306946403,
+      "learning_rate": 4.9147163498672055e-06,
+      "loss": 0.0633,
+      "step": 6175
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.31941508410803,
+      "learning_rate": 4.913351947002943e-06,
+      "loss": 0.0778,
+      "step": 6176
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.306464808475965,
+      "learning_rate": 4.911987550592736e-06,
+      "loss": 0.0566,
+      "step": 6177
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5865977308019092,
+      "learning_rate": 4.910623160738209e-06,
+      "loss": 0.1236,
+      "step": 6178
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.26491957139116934,
+      "learning_rate": 4.909258777540987e-06,
+      "loss": 0.0711,
+      "step": 6179
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.38791620493039386,
+      "learning_rate": 4.907894401102702e-06,
+      "loss": 0.1087,
+      "step": 6180
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.19452559509793094,
+      "learning_rate": 4.906530031524979e-06,
+      "loss": 0.0525,
+      "step": 6181
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.21244441790593097,
+      "learning_rate": 4.905165668909444e-06,
+      "loss": 0.0412,
+      "step": 6182
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.2277006057321096,
+      "learning_rate": 4.903801313357719e-06,
+      "loss": 0.0527,
+      "step": 6183
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3114487910339468,
+      "learning_rate": 4.9024369649714355e-06,
+      "loss": 0.076,
+      "step": 6184
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.33686043061717225,
+      "learning_rate": 4.901072623852215e-06,
+      "loss": 0.0867,
+      "step": 6185
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.2269644153529275,
+      "learning_rate": 4.8997082901016805e-06,
+      "loss": 0.0645,
+      "step": 6186
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.449950932502013,
+      "learning_rate": 4.898343963821457e-06,
+      "loss": 0.105,
+      "step": 6187
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.22543888091890615,
+      "learning_rate": 4.896979645113169e-06,
+      "loss": 0.0794,
+      "step": 6188
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.48887855084364,
+      "learning_rate": 4.895615334078437e-06,
+      "loss": 0.0891,
+      "step": 6189
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4087896293110071,
+      "learning_rate": 4.8942510308188804e-06,
+      "loss": 0.1072,
+      "step": 6190
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5042544020336369,
+      "learning_rate": 4.8928867354361254e-06,
+      "loss": 0.1369,
+      "step": 6191
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.2910483740254676,
+      "learning_rate": 4.891522448031791e-06,
+      "loss": 0.0799,
+      "step": 6192
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.26523813919658606,
+      "learning_rate": 4.890158168707494e-06,
+      "loss": 0.077,
+      "step": 6193
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.30976789111563174,
+      "learning_rate": 4.8887938975648565e-06,
+      "loss": 0.1063,
+      "step": 6194
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3603648007666127,
+      "learning_rate": 4.887429634705499e-06,
+      "loss": 0.0818,
+      "step": 6195
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4682107554286099,
+      "learning_rate": 4.8860653802310374e-06,
+      "loss": 0.1163,
+      "step": 6196
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.460316617414806,
+      "learning_rate": 4.884701134243091e-06,
+      "loss": 0.1076,
+      "step": 6197
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.248189547260516,
+      "learning_rate": 4.883336896843275e-06,
+      "loss": 0.0759,
+      "step": 6198
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.36616886101656765,
+      "learning_rate": 4.881972668133207e-06,
+      "loss": 0.1078,
+      "step": 6199
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.33528772254418476,
+      "learning_rate": 4.880608448214501e-06,
+      "loss": 0.092,
+      "step": 6200
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3360783303094265,
+      "learning_rate": 4.879244237188775e-06,
+      "loss": 0.1176,
+      "step": 6201
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.2585707002194257,
+      "learning_rate": 4.87788003515764e-06,
+      "loss": 0.0775,
+      "step": 6202
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.20350412578856908,
+      "learning_rate": 4.8765158422227115e-06,
+      "loss": 0.0475,
+      "step": 6203
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3315548891993331,
+      "learning_rate": 4.875151658485604e-06,
+      "loss": 0.0882,
+      "step": 6204
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6881886227780356,
+      "learning_rate": 4.873787484047927e-06,
+      "loss": 0.1248,
+      "step": 6205
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.21521139280178045,
+      "learning_rate": 4.872423319011295e-06,
+      "loss": 0.0717,
+      "step": 6206
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3109168715839414,
+      "learning_rate": 4.871059163477315e-06,
+      "loss": 0.0669,
+      "step": 6207
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.34510866006985336,
+      "learning_rate": 4.869695017547601e-06,
+      "loss": 0.0739,
+      "step": 6208
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4109845004855507,
+      "learning_rate": 4.8683308813237616e-06,
+      "loss": 0.1078,
+      "step": 6209
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.28230029370167403,
+      "learning_rate": 4.866966754907403e-06,
+      "loss": 0.0813,
+      "step": 6210
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.30954225669820873,
+      "learning_rate": 4.865602638400138e-06,
+      "loss": 0.0875,
+      "step": 6211
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.2646630837706112,
+      "learning_rate": 4.864238531903571e-06,
+      "loss": 0.0656,
+      "step": 6212
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3463217752013865,
+      "learning_rate": 4.862874435519307e-06,
+      "loss": 0.0581,
+      "step": 6213
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.37180422015513553,
+      "learning_rate": 4.861510349348955e-06,
+      "loss": 0.0703,
+      "step": 6214
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.27825131319640045,
+      "learning_rate": 4.86014627349412e-06,
+      "loss": 0.0976,
+      "step": 6215
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.31088679586848195,
+      "learning_rate": 4.858782208056405e-06,
+      "loss": 0.1008,
+      "step": 6216
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.271446927815984,
+      "learning_rate": 4.857418153137412e-06,
+      "loss": 0.0799,
+      "step": 6217
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.21354901226878356,
+      "learning_rate": 4.856054108838747e-06,
+      "loss": 0.0594,
+      "step": 6218
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.2921263439087751,
+      "learning_rate": 4.854690075262011e-06,
+      "loss": 0.065,
+      "step": 6219
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.2259457387344419,
+      "learning_rate": 4.853326052508802e-06,
+      "loss": 0.0579,
+      "step": 6220
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3523429891646404,
+      "learning_rate": 4.851962040680726e-06,
+      "loss": 0.0535,
+      "step": 6221
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.35054118297676823,
+      "learning_rate": 4.85059803987938e-06,
+      "loss": 0.0665,
+      "step": 6222
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.26621599729415873,
+      "learning_rate": 4.8492340502063615e-06,
+      "loss": 0.083,
+      "step": 6223
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3795396936731353,
+      "learning_rate": 4.8478700717632675e-06,
+      "loss": 0.0743,
+      "step": 6224
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5025331654852412,
+      "learning_rate": 4.846506104651698e-06,
+      "loss": 0.0992,
+      "step": 6225
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.37887447002997976,
+      "learning_rate": 4.845142148973249e-06,
+      "loss": 0.0858,
+      "step": 6226
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.27834869580519567,
+      "learning_rate": 4.843778204829513e-06,
+      "loss": 0.0925,
+      "step": 6227
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.2670083933259194,
+      "learning_rate": 4.8424142723220865e-06,
+      "loss": 0.0805,
+      "step": 6228
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4695305940520185,
+      "learning_rate": 4.841050351552565e-06,
+      "loss": 0.1178,
+      "step": 6229
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3898764340310893,
+      "learning_rate": 4.839686442622537e-06,
+      "loss": 0.0847,
+      "step": 6230
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.36204248280406587,
+      "learning_rate": 4.838322545633598e-06,
+      "loss": 0.0983,
+      "step": 6231
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3426300874546533,
+      "learning_rate": 4.8369586606873376e-06,
+      "loss": 0.0819,
+      "step": 6232
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.18740978688371399,
+      "learning_rate": 4.835594787885346e-06,
+      "loss": 0.0517,
+      "step": 6233
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3350686009514446,
+      "learning_rate": 4.8342309273292115e-06,
+      "loss": 0.1111,
+      "step": 6234
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.26491393326760276,
+      "learning_rate": 4.832867079120524e-06,
+      "loss": 0.0782,
+      "step": 6235
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.31604590788242415,
+      "learning_rate": 4.831503243360869e-06,
+      "loss": 0.0769,
+      "step": 6236
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4068779973617073,
+      "learning_rate": 4.830139420151835e-06,
+      "loss": 0.1109,
+      "step": 6237
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.43931047000082313,
+      "learning_rate": 4.828775609595007e-06,
+      "loss": 0.1237,
+      "step": 6238
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.26055486211230416,
+      "learning_rate": 4.8274118117919685e-06,
+      "loss": 0.0819,
+      "step": 6239
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2920040158218073,
+      "learning_rate": 4.826048026844304e-06,
+      "loss": 0.0857,
+      "step": 6240
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2061827867170571,
+      "learning_rate": 4.824684254853594e-06,
+      "loss": 0.0734,
+      "step": 6241
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4860217462059491,
+      "learning_rate": 4.823320495921424e-06,
+      "loss": 0.1377,
+      "step": 6242
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2899670608034245,
+      "learning_rate": 4.821956750149372e-06,
+      "loss": 0.0522,
+      "step": 6243
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4591723499865208,
+      "learning_rate": 4.820593017639017e-06,
+      "loss": 0.0772,
+      "step": 6244
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.37750475993514787,
+      "learning_rate": 4.819229298491941e-06,
+      "loss": 0.0773,
+      "step": 6245
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2843229307029026,
+      "learning_rate": 4.81786559280972e-06,
+      "loss": 0.1086,
+      "step": 6246
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.38438792982637776,
+      "learning_rate": 4.816501900693931e-06,
+      "loss": 0.0752,
+      "step": 6247
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.28842356582610174,
+      "learning_rate": 4.815138222246147e-06,
+      "loss": 0.1065,
+      "step": 6248
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.20391680627537037,
+      "learning_rate": 4.813774557567947e-06,
+      "loss": 0.0676,
+      "step": 6249
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.26989407069220805,
+      "learning_rate": 4.812410906760903e-06,
+      "loss": 0.0749,
+      "step": 6250
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.38659777572540693,
+      "learning_rate": 4.811047269926585e-06,
+      "loss": 0.1225,
+      "step": 6251
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.30223685482080054,
+      "learning_rate": 4.809683647166569e-06,
+      "loss": 0.093,
+      "step": 6252
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.31324597709209756,
+      "learning_rate": 4.808320038582424e-06,
+      "loss": 0.0756,
+      "step": 6253
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3323420498607163,
+      "learning_rate": 4.8069564442757176e-06,
+      "loss": 0.0838,
+      "step": 6254
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3404440843400772,
+      "learning_rate": 4.805592864348022e-06,
+      "loss": 0.0996,
+      "step": 6255
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5698618083517882,
+      "learning_rate": 4.8042292989009016e-06,
+      "loss": 0.1477,
+      "step": 6256
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.17998600127109962,
+      "learning_rate": 4.8028657480359236e-06,
+      "loss": 0.0397,
+      "step": 6257
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.19905570148235932,
+      "learning_rate": 4.801502211854651e-06,
+      "loss": 0.0544,
+      "step": 6258
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.32578127007529,
+      "learning_rate": 4.800138690458653e-06,
+      "loss": 0.0826,
+      "step": 6259
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.22892401593421635,
+      "learning_rate": 4.798775183949488e-06,
+      "loss": 0.0682,
+      "step": 6260
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2784866089187351,
+      "learning_rate": 4.7974116924287194e-06,
+      "loss": 0.0731,
+      "step": 6261
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.24154303058920804,
+      "learning_rate": 4.796048215997908e-06,
+      "loss": 0.0472,
+      "step": 6262
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.1882640249736183,
+      "learning_rate": 4.794684754758614e-06,
+      "loss": 0.0706,
+      "step": 6263
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.32166669477369153,
+      "learning_rate": 4.793321308812396e-06,
+      "loss": 0.0504,
+      "step": 6264
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3799128565414007,
+      "learning_rate": 4.791957878260811e-06,
+      "loss": 0.0779,
+      "step": 6265
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4835770038609911,
+      "learning_rate": 4.790594463205413e-06,
+      "loss": 0.0698,
+      "step": 6266
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3239104881339727,
+      "learning_rate": 4.789231063747763e-06,
+      "loss": 0.1054,
+      "step": 6267
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4536566216534488,
+      "learning_rate": 4.787867679989408e-06,
+      "loss": 0.1383,
+      "step": 6268
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4604642955481199,
+      "learning_rate": 4.786504312031905e-06,
+      "loss": 0.1098,
+      "step": 6269
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3122401583874322,
+      "learning_rate": 4.7851409599768045e-06,
+      "loss": 0.0616,
+      "step": 6270
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5975977364413436,
+      "learning_rate": 4.783777623925657e-06,
+      "loss": 0.1334,
+      "step": 6271
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.25105025721283447,
+      "learning_rate": 4.782414303980012e-06,
+      "loss": 0.099,
+      "step": 6272
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.28175402175268005,
+      "learning_rate": 4.781051000241417e-06,
+      "loss": 0.0957,
+      "step": 6273
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.28155934290594453,
+      "learning_rate": 4.779687712811419e-06,
+      "loss": 0.0595,
+      "step": 6274
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.28438737266784225,
+      "learning_rate": 4.778324441791562e-06,
+      "loss": 0.0909,
+      "step": 6275
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2715099948536962,
+      "learning_rate": 4.7769611872833944e-06,
+      "loss": 0.0524,
+      "step": 6276
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.45232396982033773,
+      "learning_rate": 4.775597949388457e-06,
+      "loss": 0.1592,
+      "step": 6277
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2592280354765477,
+      "learning_rate": 4.774234728208288e-06,
+      "loss": 0.0853,
+      "step": 6278
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.30877447044210227,
+      "learning_rate": 4.772871523844435e-06,
+      "loss": 0.0825,
+      "step": 6279
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2618951953030846,
+      "learning_rate": 4.771508336398433e-06,
+      "loss": 0.0712,
+      "step": 6280
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.44030031974701817,
+      "learning_rate": 4.770145165971823e-06,
+      "loss": 0.1257,
+      "step": 6281
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3478775458193837,
+      "learning_rate": 4.768782012666136e-06,
+      "loss": 0.0918,
+      "step": 6282
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.26107654248503626,
+      "learning_rate": 4.767418876582914e-06,
+      "loss": 0.0586,
+      "step": 6283
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.1992128709048062,
+      "learning_rate": 4.766055757823689e-06,
+      "loss": 0.0675,
+      "step": 6284
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.41925562761070967,
+      "learning_rate": 4.764692656489992e-06,
+      "loss": 0.1136,
+      "step": 6285
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3246617472946094,
+      "learning_rate": 4.76332957268336e-06,
+      "loss": 0.0976,
+      "step": 6286
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.37278841245889494,
+      "learning_rate": 4.761966506505319e-06,
+      "loss": 0.0709,
+      "step": 6287
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.48758677019116825,
+      "learning_rate": 4.760603458057397e-06,
+      "loss": 0.1171,
+      "step": 6288
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5101353739093077,
+      "learning_rate": 4.759240427441127e-06,
+      "loss": 0.176,
+      "step": 6289
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4578674618512451,
+      "learning_rate": 4.757877414758032e-06,
+      "loss": 0.0966,
+      "step": 6290
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3082300225027582,
+      "learning_rate": 4.756514420109638e-06,
+      "loss": 0.0654,
+      "step": 6291
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.24800809474449645,
+      "learning_rate": 4.755151443597466e-06,
+      "loss": 0.0899,
+      "step": 6292
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4708500818865379,
+      "learning_rate": 4.753788485323042e-06,
+      "loss": 0.1311,
+      "step": 6293
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6769839347135105,
+      "learning_rate": 4.752425545387887e-06,
+      "loss": 0.137,
+      "step": 6294
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3155863847727109,
+      "learning_rate": 4.751062623893518e-06,
+      "loss": 0.0923,
+      "step": 6295
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.27708340327039194,
+      "learning_rate": 4.749699720941455e-06,
+      "loss": 0.079,
+      "step": 6296
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.34628954262283923,
+      "learning_rate": 4.7483368366332166e-06,
+      "loss": 0.0974,
+      "step": 6297
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3294642223885916,
+      "learning_rate": 4.746973971070316e-06,
+      "loss": 0.1051,
+      "step": 6298
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.23616371939708927,
+      "learning_rate": 4.745611124354267e-06,
+      "loss": 0.0821,
+      "step": 6299
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.14198819248534972,
+      "learning_rate": 4.744248296586584e-06,
+      "loss": 0.0204,
+      "step": 6300
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4264165092607951,
+      "learning_rate": 4.742885487868778e-06,
+      "loss": 0.1042,
+      "step": 6301
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.24703185208719472,
+      "learning_rate": 4.741522698302358e-06,
+      "loss": 0.0432,
+      "step": 6302
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3771141246815471,
+      "learning_rate": 4.740159927988835e-06,
+      "loss": 0.1092,
+      "step": 6303
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3329086267049419,
+      "learning_rate": 4.738797177029711e-06,
+      "loss": 0.0672,
+      "step": 6304
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.19626758612884662,
+      "learning_rate": 4.737434445526497e-06,
+      "loss": 0.0532,
+      "step": 6305
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.23309166058036826,
+      "learning_rate": 4.736071733580695e-06,
+      "loss": 0.0217,
+      "step": 6306
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.45785665305736367,
+      "learning_rate": 4.7347090412938085e-06,
+      "loss": 0.1495,
+      "step": 6307
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.30693582900821176,
+      "learning_rate": 4.733346368767338e-06,
+      "loss": 0.1128,
+      "step": 6308
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5212671926191326,
+      "learning_rate": 4.73198371610278e-06,
+      "loss": 0.0827,
+      "step": 6309
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9839800350129269,
+      "learning_rate": 4.73062108340164e-06,
+      "loss": 0.104,
+      "step": 6310
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2438334409648715,
+      "learning_rate": 4.7292584707654096e-06,
+      "loss": 0.0869,
+      "step": 6311
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.33223036155517466,
+      "learning_rate": 4.727895878295584e-06,
+      "loss": 0.0658,
+      "step": 6312
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3745328593180325,
+      "learning_rate": 4.7265333060936605e-06,
+      "loss": 0.1039,
+      "step": 6313
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2989441145540522,
+      "learning_rate": 4.72517075426113e-06,
+      "loss": 0.0781,
+      "step": 6314
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.44429380754758774,
+      "learning_rate": 4.723808222899481e-06,
+      "loss": 0.0742,
+      "step": 6315
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4552469348014972,
+      "learning_rate": 4.722445712110204e-06,
+      "loss": 0.1164,
+      "step": 6316
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3150480923628805,
+      "learning_rate": 4.721083221994789e-06,
+      "loss": 0.0898,
+      "step": 6317
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.35756208420739405,
+      "learning_rate": 4.71972075265472e-06,
+      "loss": 0.096,
+      "step": 6318
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2967834577307035,
+      "learning_rate": 4.71835830419148e-06,
+      "loss": 0.1013,
+      "step": 6319
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.23760317233505612,
+      "learning_rate": 4.716995876706557e-06,
+      "loss": 0.0404,
+      "step": 6320
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2278106666694121,
+      "learning_rate": 4.715633470301429e-06,
+      "loss": 0.0602,
+      "step": 6321
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.25198557628750506,
+      "learning_rate": 4.7142710850775755e-06,
+      "loss": 0.0582,
+      "step": 6322
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2677006939004904,
+      "learning_rate": 4.712908721136477e-06,
+      "loss": 0.0701,
+      "step": 6323
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.31163436756327373,
+      "learning_rate": 4.71154637857961e-06,
+      "loss": 0.0857,
+      "step": 6324
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3426430292277268,
+      "learning_rate": 4.71018405750845e-06,
+      "loss": 0.0649,
+      "step": 6325
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3352365569695946,
+      "learning_rate": 4.708821758024466e-06,
+      "loss": 0.0759,
+      "step": 6326
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6054451185818653,
+      "learning_rate": 4.707459480229137e-06,
+      "loss": 0.1483,
+      "step": 6327
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2783469794141662,
+      "learning_rate": 4.706097224223929e-06,
+      "loss": 0.0544,
+      "step": 6328
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3319860176359143,
+      "learning_rate": 4.704734990110311e-06,
+      "loss": 0.0687,
+      "step": 6329
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.26565763150784566,
+      "learning_rate": 4.703372777989752e-06,
+      "loss": 0.0734,
+      "step": 6330
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.35536486295116987,
+      "learning_rate": 4.702010587963716e-06,
+      "loss": 0.1124,
+      "step": 6331
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4190772687855606,
+      "learning_rate": 4.700648420133669e-06,
+      "loss": 0.1122,
+      "step": 6332
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4046026844649208,
+      "learning_rate": 4.699286274601069e-06,
+      "loss": 0.0903,
+      "step": 6333
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.38589767793669816,
+      "learning_rate": 4.6979241514673805e-06,
+      "loss": 0.1003,
+      "step": 6334
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2460209757292161,
+      "learning_rate": 4.69656205083406e-06,
+      "loss": 0.0788,
+      "step": 6335
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.24243466273324263,
+      "learning_rate": 4.695199972802566e-06,
+      "loss": 0.0619,
+      "step": 6336
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4275353329535424,
+      "learning_rate": 4.693837917474355e-06,
+      "loss": 0.1012,
+      "step": 6337
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3746458077203752,
+      "learning_rate": 4.692475884950877e-06,
+      "loss": 0.0989,
+      "step": 6338
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3015386618161945,
+      "learning_rate": 4.691113875333588e-06,
+      "loss": 0.0742,
+      "step": 6339
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.31256212176039233,
+      "learning_rate": 4.6897518887239366e-06,
+      "loss": 0.1225,
+      "step": 6340
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2833146786825738,
+      "learning_rate": 4.688389925223372e-06,
+      "loss": 0.1078,
+      "step": 6341
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.38156956934619807,
+      "learning_rate": 4.687027984933343e-06,
+      "loss": 0.0871,
+      "step": 6342
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3313935835880675,
+      "learning_rate": 4.685666067955289e-06,
+      "loss": 0.1036,
+      "step": 6343
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2203990771803701,
+      "learning_rate": 4.6843041743906595e-06,
+      "loss": 0.0545,
+      "step": 6344
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2193640648282117,
+      "learning_rate": 4.682942304340895e-06,
+      "loss": 0.0704,
+      "step": 6345
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.16418517763049162,
+      "learning_rate": 4.681580457907432e-06,
+      "loss": 0.049,
+      "step": 6346
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.45874778995129695,
+      "learning_rate": 4.680218635191714e-06,
+      "loss": 0.1298,
+      "step": 6347
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.39042596649995576,
+      "learning_rate": 4.678856836295174e-06,
+      "loss": 0.0886,
+      "step": 6348
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.86316159869915,
+      "learning_rate": 4.677495061319248e-06,
+      "loss": 0.1508,
+      "step": 6349
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3565196450777301,
+      "learning_rate": 4.676133310365368e-06,
+      "loss": 0.0775,
+      "step": 6350
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5590742391171413,
+      "learning_rate": 4.674771583534965e-06,
+      "loss": 0.0717,
+      "step": 6351
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4080409352837086,
+      "learning_rate": 4.673409880929471e-06,
+      "loss": 0.0939,
+      "step": 6352
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3170856063196134,
+      "learning_rate": 4.672048202650309e-06,
+      "loss": 0.0793,
+      "step": 6353
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6902625030187478,
+      "learning_rate": 4.6706865487989095e-06,
+      "loss": 0.1317,
+      "step": 6354
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.23264450947519252,
+      "learning_rate": 4.669324919476695e-06,
+      "loss": 0.0528,
+      "step": 6355
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.37434976642229184,
+      "learning_rate": 4.667963314785085e-06,
+      "loss": 0.0911,
+      "step": 6356
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4088662011726529,
+      "learning_rate": 4.666601734825501e-06,
+      "loss": 0.0772,
+      "step": 6357
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.2908346145386005,
+      "learning_rate": 4.665240179699363e-06,
+      "loss": 0.0734,
+      "step": 6358
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4560207248379656,
+      "learning_rate": 4.663878649508087e-06,
+      "loss": 0.098,
+      "step": 6359
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.21655318704650411,
+      "learning_rate": 4.662517144353085e-06,
+      "loss": 0.0411,
+      "step": 6360
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3130930820213886,
+      "learning_rate": 4.6611556643357745e-06,
+      "loss": 0.0842,
+      "step": 6361
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.26301732524459226,
+      "learning_rate": 4.659794209557564e-06,
+      "loss": 0.0531,
+      "step": 6362
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.23218131776371206,
+      "learning_rate": 4.658432780119861e-06,
+      "loss": 0.0541,
+      "step": 6363
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.246466215368339,
+      "learning_rate": 4.657071376124075e-06,
+      "loss": 0.0585,
+      "step": 6364
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.303471655266001,
+      "learning_rate": 4.655709997671612e-06,
+      "loss": 0.0887,
+      "step": 6365
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.2096262742961997,
+      "learning_rate": 4.654348644863873e-06,
+      "loss": 0.0599,
+      "step": 6366
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3843488090242564,
+      "learning_rate": 4.65298731780226e-06,
+      "loss": 0.094,
+      "step": 6367
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3498700668141037,
+      "learning_rate": 4.651626016588173e-06,
+      "loss": 0.0947,
+      "step": 6368
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3623777144333756,
+      "learning_rate": 4.650264741323011e-06,
+      "loss": 0.1137,
+      "step": 6369
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.2531786888615636,
+      "learning_rate": 4.648903492108167e-06,
+      "loss": 0.0717,
+      "step": 6370
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.2930192755767202,
+      "learning_rate": 4.6475422690450375e-06,
+      "loss": 0.0752,
+      "step": 6371
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.2705824940198104,
+      "learning_rate": 4.6461810722350114e-06,
+      "loss": 0.0749,
+      "step": 6372
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.38572616716722485,
+      "learning_rate": 4.644819901779482e-06,
+      "loss": 0.1204,
+      "step": 6373
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3515170840366077,
+      "learning_rate": 4.643458757779834e-06,
+      "loss": 0.1058,
+      "step": 6374
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.31565159039521745,
+      "learning_rate": 4.642097640337455e-06,
+      "loss": 0.0745,
+      "step": 6375
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.17036273316881242,
+      "learning_rate": 4.640736549553731e-06,
+      "loss": 0.0463,
+      "step": 6376
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.20259798271804128,
+      "learning_rate": 4.6393754855300375e-06,
+      "loss": 0.0614,
+      "step": 6377
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.18105187501842224,
+      "learning_rate": 4.638014448367762e-06,
+      "loss": 0.07,
+      "step": 6378
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.23670305452410104,
+      "learning_rate": 4.636653438168277e-06,
+      "loss": 0.0978,
+      "step": 6379
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.2606657516327883,
+      "learning_rate": 4.63529245503296e-06,
+      "loss": 0.092,
+      "step": 6380
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3863081708077683,
+      "learning_rate": 4.633931499063188e-06,
+      "loss": 0.1243,
+      "step": 6381
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4810535772326138,
+      "learning_rate": 4.63257057036033e-06,
+      "loss": 0.0941,
+      "step": 6382
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.30794550757641614,
+      "learning_rate": 4.631209669025756e-06,
+      "loss": 0.0851,
+      "step": 6383
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3387356568195617,
+      "learning_rate": 4.629848795160832e-06,
+      "loss": 0.0928,
+      "step": 6384
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.26616877638101577,
+      "learning_rate": 4.628487948866929e-06,
+      "loss": 0.0489,
+      "step": 6385
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3209706381627441,
+      "learning_rate": 4.6271271302454075e-06,
+      "loss": 0.1002,
+      "step": 6386
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.22102149492760403,
+      "learning_rate": 4.625766339397627e-06,
+      "loss": 0.0671,
+      "step": 6387
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3741838329478444,
+      "learning_rate": 4.624405576424951e-06,
+      "loss": 0.1044,
+      "step": 6388
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.2127267148578572,
+      "learning_rate": 4.6230448414287375e-06,
+      "loss": 0.0518,
+      "step": 6389
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.23212303153536942,
+      "learning_rate": 4.6216841345103395e-06,
+      "loss": 0.0848,
+      "step": 6390
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5229934445295567,
+      "learning_rate": 4.620323455771108e-06,
+      "loss": 0.1313,
+      "step": 6391
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.382715392410636,
+      "learning_rate": 4.6189628053124e-06,
+      "loss": 0.1339,
+      "step": 6392
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.28198871823808924,
+      "learning_rate": 4.617602183235562e-06,
+      "loss": 0.0637,
+      "step": 6393
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3632364371299663,
+      "learning_rate": 4.616241589641939e-06,
+      "loss": 0.0817,
+      "step": 6394
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.27242411559463203,
+      "learning_rate": 4.61488102463288e-06,
+      "loss": 0.0923,
+      "step": 6395
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.22680724022912901,
+      "learning_rate": 4.613520488309726e-06,
+      "loss": 0.069,
+      "step": 6396
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.22214152369120563,
+      "learning_rate": 4.612159980773816e-06,
+      "loss": 0.0662,
+      "step": 6397
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.2687140450924824,
+      "learning_rate": 4.61079950212649e-06,
+      "loss": 0.0683,
+      "step": 6398
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.30089743828084026,
+      "learning_rate": 4.609439052469086e-06,
+      "loss": 0.0678,
+      "step": 6399
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.25960469151427323,
+      "learning_rate": 4.6080786319029355e-06,
+      "loss": 0.0763,
+      "step": 6400
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5100085063236124,
+      "learning_rate": 4.606718240529371e-06,
+      "loss": 0.1151,
+      "step": 6401
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.26377890049904607,
+      "learning_rate": 4.6053578784497235e-06,
+      "loss": 0.0846,
+      "step": 6402
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.2964093010522445,
+      "learning_rate": 4.603997545765321e-06,
+      "loss": 0.0677,
+      "step": 6403
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.30774535201730274,
+      "learning_rate": 4.602637242577487e-06,
+      "loss": 0.0716,
+      "step": 6404
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3822215528398892,
+      "learning_rate": 4.601276968987546e-06,
+      "loss": 0.109,
+      "step": 6405
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9348034496512732,
+      "learning_rate": 4.5999167250968195e-06,
+      "loss": 0.2038,
+      "step": 6406
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.2978573866782475,
+      "learning_rate": 4.598556511006627e-06,
+      "loss": 0.0804,
+      "step": 6407
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.23790420202030357,
+      "learning_rate": 4.597196326818282e-06,
+      "loss": 0.0644,
+      "step": 6408
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.17247229028449812,
+      "learning_rate": 4.595836172633103e-06,
+      "loss": 0.0424,
+      "step": 6409
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5509906436531441,
+      "learning_rate": 4.5944760485523995e-06,
+      "loss": 0.1528,
+      "step": 6410
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.2645332496176322,
+      "learning_rate": 4.593115954677481e-06,
+      "loss": 0.0741,
+      "step": 6411
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.47295324139615186,
+      "learning_rate": 4.591755891109659e-06,
+      "loss": 0.1155,
+      "step": 6412
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.378128363343617,
+      "learning_rate": 4.590395857950235e-06,
+      "loss": 0.1001,
+      "step": 6413
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.20675316225415696,
+      "learning_rate": 4.589035855300512e-06,
+      "loss": 0.048,
+      "step": 6414
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.22011757692618353,
+      "learning_rate": 4.587675883261795e-06,
+      "loss": 0.0549,
+      "step": 6415
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3848278714917212,
+      "learning_rate": 4.58631594193538e-06,
+      "loss": 0.0849,
+      "step": 6416
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.31712092225589816,
+      "learning_rate": 4.584956031422564e-06,
+      "loss": 0.1066,
+      "step": 6417
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4765075750990313,
+      "learning_rate": 4.583596151824637e-06,
+      "loss": 0.1205,
+      "step": 6418
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4094440861677164,
+      "learning_rate": 4.5822363032428986e-06,
+      "loss": 0.0728,
+      "step": 6419
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.23138690444833263,
+      "learning_rate": 4.580876485778632e-06,
+      "loss": 0.0739,
+      "step": 6420
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.25982815308343693,
+      "learning_rate": 4.5795166995331256e-06,
+      "loss": 0.059,
+      "step": 6421
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5117836589684587,
+      "learning_rate": 4.578156944607667e-06,
+      "loss": 0.0877,
+      "step": 6422
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3046804946771742,
+      "learning_rate": 4.576797221103535e-06,
+      "loss": 0.0542,
+      "step": 6423
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.2647933826997276,
+      "learning_rate": 4.575437529122013e-06,
+      "loss": 0.0893,
+      "step": 6424
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.36714819555118783,
+      "learning_rate": 4.574077868764374e-06,
+      "loss": 0.1014,
+      "step": 6425
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.2769774862157114,
+      "learning_rate": 4.5727182401318995e-06,
+      "loss": 0.0747,
+      "step": 6426
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.24724298503471734,
+      "learning_rate": 4.5713586433258595e-06,
+      "loss": 0.05,
+      "step": 6427
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3376021179700868,
+      "learning_rate": 4.569999078447522e-06,
+      "loss": 0.1112,
+      "step": 6428
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.2700747560632514,
+      "learning_rate": 4.568639545598161e-06,
+      "loss": 0.0875,
+      "step": 6429
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3183043502078776,
+      "learning_rate": 4.567280044879038e-06,
+      "loss": 0.0838,
+      "step": 6430
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.34764728283430596,
+      "learning_rate": 4.565920576391418e-06,
+      "loss": 0.0768,
+      "step": 6431
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.46059120026869094,
+      "learning_rate": 4.564561140236562e-06,
+      "loss": 0.1072,
+      "step": 6432
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.39292585206242203,
+      "learning_rate": 4.5632017365157306e-06,
+      "loss": 0.0938,
+      "step": 6433
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3342316422773509,
+      "learning_rate": 4.561842365330177e-06,
+      "loss": 0.1136,
+      "step": 6434
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4116874585579343,
+      "learning_rate": 4.5604830267811565e-06,
+      "loss": 0.0948,
+      "step": 6435
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3689668614487559,
+      "learning_rate": 4.559123720969921e-06,
+      "loss": 0.1092,
+      "step": 6436
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5256124062907164,
+      "learning_rate": 4.557764447997719e-06,
+      "loss": 0.1104,
+      "step": 6437
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.18786379736368916,
+      "learning_rate": 4.556405207965796e-06,
+      "loss": 0.0479,
+      "step": 6438
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.31623415460191534,
+      "learning_rate": 4.5550460009753995e-06,
+      "loss": 0.0906,
+      "step": 6439
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.29487594013215457,
+      "learning_rate": 4.553686827127767e-06,
+      "loss": 0.0541,
+      "step": 6440
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.2784189143466072,
+      "learning_rate": 4.552327686524141e-06,
+      "loss": 0.1023,
+      "step": 6441
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.29301999214702185,
+      "learning_rate": 4.5509685792657555e-06,
+      "loss": 0.0945,
+      "step": 6442
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.21769059741990507,
+      "learning_rate": 4.549609505453847e-06,
+      "loss": 0.0589,
+      "step": 6443
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.41797524544995035,
+      "learning_rate": 4.548250465189647e-06,
+      "loss": 0.09,
+      "step": 6444
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.495765215316743,
+      "learning_rate": 4.54689145857438e-06,
+      "loss": 0.0771,
+      "step": 6445
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3022934581411002,
+      "learning_rate": 4.54553248570928e-06,
+      "loss": 0.1161,
+      "step": 6446
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.290344045502107,
+      "learning_rate": 4.544173546695568e-06,
+      "loss": 0.0806,
+      "step": 6447
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3535851970282823,
+      "learning_rate": 4.542814641634464e-06,
+      "loss": 0.1135,
+      "step": 6448
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4108216841272214,
+      "learning_rate": 4.54145577062719e-06,
+      "loss": 0.1175,
+      "step": 6449
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.2927670510392883,
+      "learning_rate": 4.540096933774962e-06,
+      "loss": 0.0761,
+      "step": 6450
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5513451454564373,
+      "learning_rate": 4.538738131178994e-06,
+      "loss": 0.0568,
+      "step": 6451
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.23572640752750904,
+      "learning_rate": 4.537379362940494e-06,
+      "loss": 0.0637,
+      "step": 6452
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3779108115056334,
+      "learning_rate": 4.536020629160676e-06,
+      "loss": 0.1138,
+      "step": 6453
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.2825554717489204,
+      "learning_rate": 4.534661929940745e-06,
+      "loss": 0.0779,
+      "step": 6454
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4506875740666896,
+      "learning_rate": 4.533303265381902e-06,
+      "loss": 0.0884,
+      "step": 6455
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8333110093918049,
+      "learning_rate": 4.531944635585352e-06,
+      "loss": 0.1072,
+      "step": 6456
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.34097750357601764,
+      "learning_rate": 4.530586040652293e-06,
+      "loss": 0.1069,
+      "step": 6457
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.32057741693119035,
+      "learning_rate": 4.52922748068392e-06,
+      "loss": 0.0808,
+      "step": 6458
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.44342064654459556,
+      "learning_rate": 4.527868955781424e-06,
+      "loss": 0.079,
+      "step": 6459
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.349842968722777,
+      "learning_rate": 4.526510466046002e-06,
+      "loss": 0.0758,
+      "step": 6460
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.2797579875720313,
+      "learning_rate": 4.525152011578838e-06,
+      "loss": 0.0749,
+      "step": 6461
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3928411408207082,
+      "learning_rate": 4.523793592481116e-06,
+      "loss": 0.1213,
+      "step": 6462
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.2895674158472946,
+      "learning_rate": 4.5224352088540234e-06,
+      "loss": 0.076,
+      "step": 6463
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.20909806831362063,
+      "learning_rate": 4.521076860798739e-06,
+      "loss": 0.0655,
+      "step": 6464
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.30705476267721243,
+      "learning_rate": 4.519718548416439e-06,
+      "loss": 0.0744,
+      "step": 6465
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.166139207763279,
+      "learning_rate": 4.518360271808298e-06,
+      "loss": 0.0259,
+      "step": 6466
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.32803136209456396,
+      "learning_rate": 4.517002031075492e-06,
+      "loss": 0.0846,
+      "step": 6467
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.26106294813053654,
+      "learning_rate": 4.515643826319188e-06,
+      "loss": 0.0516,
+      "step": 6468
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.48806364193452056,
+      "learning_rate": 4.514285657640553e-06,
+      "loss": 0.1128,
+      "step": 6469
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2715210618264937,
+      "learning_rate": 4.512927525140751e-06,
+      "loss": 0.0616,
+      "step": 6470
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.31206443572645565,
+      "learning_rate": 4.511569428920946e-06,
+      "loss": 0.0727,
+      "step": 6471
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3000187433964627,
+      "learning_rate": 4.510211369082293e-06,
+      "loss": 0.0805,
+      "step": 6472
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2565792721089214,
+      "learning_rate": 4.508853345725953e-06,
+      "loss": 0.0687,
+      "step": 6473
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3332108846145187,
+      "learning_rate": 4.507495358953075e-06,
+      "loss": 0.1211,
+      "step": 6474
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.38359092998067684,
+      "learning_rate": 4.506137408864813e-06,
+      "loss": 0.0807,
+      "step": 6475
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2772647668516554,
+      "learning_rate": 4.5047794955623125e-06,
+      "loss": 0.1117,
+      "step": 6476
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3322565166109134,
+      "learning_rate": 4.503421619146721e-06,
+      "loss": 0.0843,
+      "step": 6477
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6104521462711345,
+      "learning_rate": 4.502063779719181e-06,
+      "loss": 0.1052,
+      "step": 6478
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.22461520342183397,
+      "learning_rate": 4.5007059773808285e-06,
+      "loss": 0.0679,
+      "step": 6479
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2767457628287545,
+      "learning_rate": 4.4993482122328056e-06,
+      "loss": 0.0852,
+      "step": 6480
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.39831476439807795,
+      "learning_rate": 4.497990484376244e-06,
+      "loss": 0.1126,
+      "step": 6481
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.24739671888086207,
+      "learning_rate": 4.496632793912276e-06,
+      "loss": 0.0486,
+      "step": 6482
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.35337212918139704,
+      "learning_rate": 4.495275140942028e-06,
+      "loss": 0.1081,
+      "step": 6483
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.25181653419940325,
+      "learning_rate": 4.493917525566629e-06,
+      "loss": 0.0738,
+      "step": 6484
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.21264521167202877,
+      "learning_rate": 4.4925599478872014e-06,
+      "loss": 0.061,
+      "step": 6485
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4840611475951059,
+      "learning_rate": 4.491202408004863e-06,
+      "loss": 0.1189,
+      "step": 6486
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.27833809987555064,
+      "learning_rate": 4.489844906020735e-06,
+      "loss": 0.0423,
+      "step": 6487
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4184173655011638,
+      "learning_rate": 4.4884874420359306e-06,
+      "loss": 0.1164,
+      "step": 6488
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3060376038248094,
+      "learning_rate": 4.487130016151559e-06,
+      "loss": 0.0954,
+      "step": 6489
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.28903430923172074,
+      "learning_rate": 4.485772628468733e-06,
+      "loss": 0.0721,
+      "step": 6490
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.36461604381872326,
+      "learning_rate": 4.484415279088558e-06,
+      "loss": 0.0959,
+      "step": 6491
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.40045127240355227,
+      "learning_rate": 4.4830579681121365e-06,
+      "loss": 0.0793,
+      "step": 6492
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2640842342708515,
+      "learning_rate": 4.4817006956405664e-06,
+      "loss": 0.07,
+      "step": 6493
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2453090620777981,
+      "learning_rate": 4.4803434617749506e-06,
+      "loss": 0.0479,
+      "step": 6494
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2796746483975629,
+      "learning_rate": 4.4789862666163805e-06,
+      "loss": 0.0865,
+      "step": 6495
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.21948158103857257,
+      "learning_rate": 4.477629110265946e-06,
+      "loss": 0.0596,
+      "step": 6496
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3709021388517017,
+      "learning_rate": 4.4762719928247414e-06,
+      "loss": 0.1073,
+      "step": 6497
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2714682208895978,
+      "learning_rate": 4.474914914393849e-06,
+      "loss": 0.0522,
+      "step": 6498
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2225930076623636,
+      "learning_rate": 4.473557875074353e-06,
+      "loss": 0.0635,
+      "step": 6499
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.36936040799266867,
+      "learning_rate": 4.472200874967331e-06,
+      "loss": 0.0879,
+      "step": 6500
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.36763221877689395,
+      "learning_rate": 4.470843914173864e-06,
+      "loss": 0.0594,
+      "step": 6501
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.292314120452045,
+      "learning_rate": 4.4694869927950255e-06,
+      "loss": 0.0742,
+      "step": 6502
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3214006042271544,
+      "learning_rate": 4.468130110931885e-06,
+      "loss": 0.0809,
+      "step": 6503
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4908728792498285,
+      "learning_rate": 4.466773268685512e-06,
+      "loss": 0.0838,
+      "step": 6504
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0177633212010726,
+      "learning_rate": 4.465416466156974e-06,
+      "loss": 0.137,
+      "step": 6505
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.33014505165767494,
+      "learning_rate": 4.46405970344733e-06,
+      "loss": 0.1096,
+      "step": 6506
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3130259957586664,
+      "learning_rate": 4.462702980657643e-06,
+      "loss": 0.1128,
+      "step": 6507
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.190959551758352,
+      "learning_rate": 4.461346297888967e-06,
+      "loss": 0.0389,
+      "step": 6508
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.29811166127780836,
+      "learning_rate": 4.459989655242356e-06,
+      "loss": 0.0761,
+      "step": 6509
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.20021881324505128,
+      "learning_rate": 4.458633052818862e-06,
+      "loss": 0.0299,
+      "step": 6510
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.16719373848909286,
+      "learning_rate": 4.457276490719533e-06,
+      "loss": 0.0476,
+      "step": 6511
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3118397805654914,
+      "learning_rate": 4.455919969045412e-06,
+      "loss": 0.0463,
+      "step": 6512
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2966646351038485,
+      "learning_rate": 4.454563487897539e-06,
+      "loss": 0.0661,
+      "step": 6513
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4898708810700045,
+      "learning_rate": 4.453207047376957e-06,
+      "loss": 0.0874,
+      "step": 6514
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2911432838287848,
+      "learning_rate": 4.451850647584699e-06,
+      "loss": 0.0905,
+      "step": 6515
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2525551286673735,
+      "learning_rate": 4.450494288621799e-06,
+      "loss": 0.0625,
+      "step": 6516
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.41991257024161677,
+      "learning_rate": 4.449137970589282e-06,
+      "loss": 0.0839,
+      "step": 6517
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.28389265194281565,
+      "learning_rate": 4.447781693588182e-06,
+      "loss": 0.0686,
+      "step": 6518
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.20028449902639217,
+      "learning_rate": 4.446425457719517e-06,
+      "loss": 0.0549,
+      "step": 6519
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.29375230328365315,
+      "learning_rate": 4.445069263084307e-06,
+      "loss": 0.0689,
+      "step": 6520
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3407383594957745,
+      "learning_rate": 4.443713109783573e-06,
+      "loss": 0.1036,
+      "step": 6521
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3821946864149125,
+      "learning_rate": 4.442356997918327e-06,
+      "loss": 0.1086,
+      "step": 6522
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4386480130906768,
+      "learning_rate": 4.441000927589578e-06,
+      "loss": 0.1016,
+      "step": 6523
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2991156377120141,
+      "learning_rate": 4.439644898898339e-06,
+      "loss": 0.0735,
+      "step": 6524
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.20814358976718328,
+      "learning_rate": 4.438288911945613e-06,
+      "loss": 0.0292,
+      "step": 6525
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4082484089460323,
+      "learning_rate": 4.436932966832399e-06,
+      "loss": 0.0928,
+      "step": 6526
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2842921992312938,
+      "learning_rate": 4.435577063659697e-06,
+      "loss": 0.0743,
+      "step": 6527
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5379947620468265,
+      "learning_rate": 4.434221202528506e-06,
+      "loss": 0.0631,
+      "step": 6528
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.46187590798492384,
+      "learning_rate": 4.432865383539814e-06,
+      "loss": 0.1245,
+      "step": 6529
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.29544556373717507,
+      "learning_rate": 4.43150960679461e-06,
+      "loss": 0.0725,
+      "step": 6530
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4520433736703095,
+      "learning_rate": 4.4301538723938855e-06,
+      "loss": 0.075,
+      "step": 6531
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5203593556740789,
+      "learning_rate": 4.42879818043862e-06,
+      "loss": 0.0905,
+      "step": 6532
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5199286723332162,
+      "learning_rate": 4.427442531029794e-06,
+      "loss": 0.1251,
+      "step": 6533
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.25428779738881147,
+      "learning_rate": 4.42608692426838e-06,
+      "loss": 0.032,
+      "step": 6534
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2693632599447435,
+      "learning_rate": 4.424731360255359e-06,
+      "loss": 0.074,
+      "step": 6535
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5095644037596729,
+      "learning_rate": 4.423375839091696e-06,
+      "loss": 0.0908,
+      "step": 6536
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3008626265323932,
+      "learning_rate": 4.42202036087836e-06,
+      "loss": 0.0973,
+      "step": 6537
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.382540018211114,
+      "learning_rate": 4.420664925716313e-06,
+      "loss": 0.058,
+      "step": 6538
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.269789237696033,
+      "learning_rate": 4.419309533706519e-06,
+      "loss": 0.0973,
+      "step": 6539
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3667640734370712,
+      "learning_rate": 4.417954184949933e-06,
+      "loss": 0.0931,
+      "step": 6540
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4574104866691601,
+      "learning_rate": 4.41659887954751e-06,
+      "loss": 0.0904,
+      "step": 6541
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.40677574553479706,
+      "learning_rate": 4.4152436176002014e-06,
+      "loss": 0.1117,
+      "step": 6542
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4571608181984217,
+      "learning_rate": 4.413888399208954e-06,
+      "loss": 0.1309,
+      "step": 6543
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.32249475260900573,
+      "learning_rate": 4.412533224474714e-06,
+      "loss": 0.0716,
+      "step": 6544
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.38824721471470003,
+      "learning_rate": 4.411178093498421e-06,
+      "loss": 0.1047,
+      "step": 6545
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5325411771488757,
+      "learning_rate": 4.4098230063810136e-06,
+      "loss": 0.1098,
+      "step": 6546
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4318340872256885,
+      "learning_rate": 4.4084679632234254e-06,
+      "loss": 0.0751,
+      "step": 6547
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3202167235232858,
+      "learning_rate": 4.407112964126591e-06,
+      "loss": 0.0816,
+      "step": 6548
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.20027733975146453,
+      "learning_rate": 4.405758009191438e-06,
+      "loss": 0.0332,
+      "step": 6549
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.42160213388819356,
+      "learning_rate": 4.404403098518889e-06,
+      "loss": 0.079,
+      "step": 6550
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.16381146577137853,
+      "learning_rate": 4.403048232209865e-06,
+      "loss": 0.0299,
+      "step": 6551
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.27391853676011063,
+      "learning_rate": 4.401693410365288e-06,
+      "loss": 0.0571,
+      "step": 6552
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.29360694785679214,
+      "learning_rate": 4.400338633086071e-06,
+      "loss": 0.0875,
+      "step": 6553
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.30235228605413994,
+      "learning_rate": 4.398983900473124e-06,
+      "loss": 0.0709,
+      "step": 6554
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.34727645535127044,
+      "learning_rate": 4.39762921262736e-06,
+      "loss": 0.081,
+      "step": 6555
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2536306594792515,
+      "learning_rate": 4.396274569649681e-06,
+      "loss": 0.0913,
+      "step": 6556
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2844560563309367,
+      "learning_rate": 4.394919971640987e-06,
+      "loss": 0.0682,
+      "step": 6557
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.30456877248181224,
+      "learning_rate": 4.3935654187021804e-06,
+      "loss": 0.0829,
+      "step": 6558
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4832987049403699,
+      "learning_rate": 4.392210910934154e-06,
+      "loss": 0.1396,
+      "step": 6559
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.28805171728095846,
+      "learning_rate": 4.390856448437799e-06,
+      "loss": 0.0752,
+      "step": 6560
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.29295491920806943,
+      "learning_rate": 4.389502031314003e-06,
+      "loss": 0.0724,
+      "step": 6561
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2030398017972443,
+      "learning_rate": 4.388147659663654e-06,
+      "loss": 0.0501,
+      "step": 6562
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.29820618135694626,
+      "learning_rate": 4.386793333587632e-06,
+      "loss": 0.0847,
+      "step": 6563
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.48852553690562567,
+      "learning_rate": 4.385439053186812e-06,
+      "loss": 0.1028,
+      "step": 6564
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4002705697125619,
+      "learning_rate": 4.384084818562074e-06,
+      "loss": 0.0647,
+      "step": 6565
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.20411258935067553,
+      "learning_rate": 4.382730629814286e-06,
+      "loss": 0.057,
+      "step": 6566
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3826902086502373,
+      "learning_rate": 4.381376487044317e-06,
+      "loss": 0.0924,
+      "step": 6567
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3576968768263061,
+      "learning_rate": 4.380022390353029e-06,
+      "loss": 0.1067,
+      "step": 6568
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.20069061578793765,
+      "learning_rate": 4.378668339841286e-06,
+      "loss": 0.056,
+      "step": 6569
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2513619123941133,
+      "learning_rate": 4.377314335609946e-06,
+      "loss": 0.0814,
+      "step": 6570
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.24642943554301217,
+      "learning_rate": 4.37596037775986e-06,
+      "loss": 0.0406,
+      "step": 6571
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3003188900237948,
+      "learning_rate": 4.374606466391881e-06,
+      "loss": 0.0767,
+      "step": 6572
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2965540754892544,
+      "learning_rate": 4.373252601606856e-06,
+      "loss": 0.0563,
+      "step": 6573
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2807779406758953,
+      "learning_rate": 4.371898783505628e-06,
+      "loss": 0.0686,
+      "step": 6574
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.38687186378380395,
+      "learning_rate": 4.3705450121890376e-06,
+      "loss": 0.0715,
+      "step": 6575
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5157531460428615,
+      "learning_rate": 4.3691912877579216e-06,
+      "loss": 0.0981,
+      "step": 6576
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3182718582461447,
+      "learning_rate": 4.367837610313113e-06,
+      "loss": 0.0767,
+      "step": 6577
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3674405258515762,
+      "learning_rate": 4.366483979955441e-06,
+      "loss": 0.0834,
+      "step": 6578
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7990043171834716,
+      "learning_rate": 4.3651303967857345e-06,
+      "loss": 0.0654,
+      "step": 6579
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.28333553062304423,
+      "learning_rate": 4.363776860904813e-06,
+      "loss": 0.0756,
+      "step": 6580
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.33339389876216513,
+      "learning_rate": 4.3624233724134965e-06,
+      "loss": 0.0954,
+      "step": 6581
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.32800952998066174,
+      "learning_rate": 4.361069931412604e-06,
+      "loss": 0.0733,
+      "step": 6582
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.39939839377022013,
+      "learning_rate": 4.359716538002944e-06,
+      "loss": 0.1242,
+      "step": 6583
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.20681982256677475,
+      "learning_rate": 4.3583631922853255e-06,
+      "loss": 0.065,
+      "step": 6584
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5095063250278454,
+      "learning_rate": 4.357009894360554e-06,
+      "loss": 0.0709,
+      "step": 6585
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.34737637314789915,
+      "learning_rate": 4.355656644329431e-06,
+      "loss": 0.0879,
+      "step": 6586
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.26083315376932353,
+      "learning_rate": 4.354303442292757e-06,
+      "loss": 0.0743,
+      "step": 6587
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4209411337948804,
+      "learning_rate": 4.352950288351321e-06,
+      "loss": 0.1108,
+      "step": 6588
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.40871683232666245,
+      "learning_rate": 4.351597182605918e-06,
+      "loss": 0.1181,
+      "step": 6589
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.31110171970121203,
+      "learning_rate": 4.350244125157336e-06,
+      "loss": 0.0847,
+      "step": 6590
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.30093018702765834,
+      "learning_rate": 4.348891116106353e-06,
+      "loss": 0.0795,
+      "step": 6591
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8227554288219854,
+      "learning_rate": 4.347538155553755e-06,
+      "loss": 0.1214,
+      "step": 6592
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.18912049663213973,
+      "learning_rate": 4.346185243600317e-06,
+      "loss": 0.0517,
+      "step": 6593
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.42360315161198403,
+      "learning_rate": 4.3448323803468105e-06,
+      "loss": 0.0891,
+      "step": 6594
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.30794181128478937,
+      "learning_rate": 4.343479565894002e-06,
+      "loss": 0.0791,
+      "step": 6595
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3229937361479185,
+      "learning_rate": 4.342126800342662e-06,
+      "loss": 0.0748,
+      "step": 6596
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3481892958922051,
+      "learning_rate": 4.340774083793551e-06,
+      "loss": 0.0655,
+      "step": 6597
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.32792680359665793,
+      "learning_rate": 4.339421416347425e-06,
+      "loss": 0.0857,
+      "step": 6598
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.34960538612059905,
+      "learning_rate": 4.338068798105041e-06,
+      "loss": 0.0429,
+      "step": 6599
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.38350293307697125,
+      "learning_rate": 4.336716229167149e-06,
+      "loss": 0.0639,
+      "step": 6600
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.370438110846986,
+      "learning_rate": 4.335363709634497e-06,
+      "loss": 0.1146,
+      "step": 6601
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.21355527642936822,
+      "learning_rate": 4.334011239607825e-06,
+      "loss": 0.037,
+      "step": 6602
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.41047064644897835,
+      "learning_rate": 4.3326588191878785e-06,
+      "loss": 0.0974,
+      "step": 6603
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.28104315034643984,
+      "learning_rate": 4.33130644847539e-06,
+      "loss": 0.0943,
+      "step": 6604
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4538779493536291,
+      "learning_rate": 4.329954127571092e-06,
+      "loss": 0.1005,
+      "step": 6605
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.29273096865011583,
+      "learning_rate": 4.3286018565757125e-06,
+      "loss": 0.0532,
+      "step": 6606
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3334467847777329,
+      "learning_rate": 4.327249635589981e-06,
+      "loss": 0.0683,
+      "step": 6607
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.37214306541461123,
+      "learning_rate": 4.325897464714614e-06,
+      "loss": 0.0757,
+      "step": 6608
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3084425665272504,
+      "learning_rate": 4.324545344050331e-06,
+      "loss": 0.0714,
+      "step": 6609
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.35980017020480787,
+      "learning_rate": 4.323193273697844e-06,
+      "loss": 0.1024,
+      "step": 6610
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.26441614609329506,
+      "learning_rate": 4.321841253757867e-06,
+      "loss": 0.094,
+      "step": 6611
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.2459246304684299,
+      "learning_rate": 4.320489284331103e-06,
+      "loss": 0.0747,
+      "step": 6612
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3570080426497132,
+      "learning_rate": 4.319137365518255e-06,
+      "loss": 0.1123,
+      "step": 6613
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3403456851868235,
+      "learning_rate": 4.317785497420025e-06,
+      "loss": 0.0632,
+      "step": 6614
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.24852310359694293,
+      "learning_rate": 4.316433680137101e-06,
+      "loss": 0.0804,
+      "step": 6615
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.2807662697401163,
+      "learning_rate": 4.315081913770182e-06,
+      "loss": 0.0828,
+      "step": 6616
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3255983647435315,
+      "learning_rate": 4.313730198419951e-06,
+      "loss": 0.0919,
+      "step": 6617
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.24988162410054074,
+      "learning_rate": 4.312378534187093e-06,
+      "loss": 0.06,
+      "step": 6618
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.32418165812042693,
+      "learning_rate": 4.311026921172285e-06,
+      "loss": 0.1037,
+      "step": 6619
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.28493245433731745,
+      "learning_rate": 4.309675359476208e-06,
+      "loss": 0.0815,
+      "step": 6620
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.27234886762354943,
+      "learning_rate": 4.3083238491995325e-06,
+      "loss": 0.0608,
+      "step": 6621
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.38223154598747355,
+      "learning_rate": 4.3069723904429224e-06,
+      "loss": 0.0953,
+      "step": 6622
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3723687079003514,
+      "learning_rate": 4.30562098330705e-06,
+      "loss": 0.0884,
+      "step": 6623
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.24222599185044982,
+      "learning_rate": 4.30426962789257e-06,
+      "loss": 0.0503,
+      "step": 6624
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4196252028510256,
+      "learning_rate": 4.302918324300142e-06,
+      "loss": 0.0644,
+      "step": 6625
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.24809022254787272,
+      "learning_rate": 4.301567072630415e-06,
+      "loss": 0.0422,
+      "step": 6626
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.253785640473781,
+      "learning_rate": 4.300215872984044e-06,
+      "loss": 0.0769,
+      "step": 6627
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4202377731230367,
+      "learning_rate": 4.2988647254616705e-06,
+      "loss": 0.1111,
+      "step": 6628
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3063537708552872,
+      "learning_rate": 4.297513630163935e-06,
+      "loss": 0.0906,
+      "step": 6629
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.32862826804926837,
+      "learning_rate": 4.296162587191479e-06,
+      "loss": 0.0883,
+      "step": 6630
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3336391256637595,
+      "learning_rate": 4.2948115966449335e-06,
+      "loss": 0.0998,
+      "step": 6631
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.32074450671741295,
+      "learning_rate": 4.293460658624926e-06,
+      "loss": 0.071,
+      "step": 6632
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5546690604661323,
+      "learning_rate": 4.292109773232087e-06,
+      "loss": 0.1163,
+      "step": 6633
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.2813061406648457,
+      "learning_rate": 4.290758940567035e-06,
+      "loss": 0.0819,
+      "step": 6634
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.15806140486415668,
+      "learning_rate": 4.28940816073039e-06,
+      "loss": 0.046,
+      "step": 6635
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.45895587518711134,
+      "learning_rate": 4.288057433822762e-06,
+      "loss": 0.1075,
+      "step": 6636
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.34311815796934825,
+      "learning_rate": 4.286706759944766e-06,
+      "loss": 0.0487,
+      "step": 6637
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.37008190399311425,
+      "learning_rate": 4.285356139197005e-06,
+      "loss": 0.0765,
+      "step": 6638
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.45012839078055383,
+      "learning_rate": 4.284005571680081e-06,
+      "loss": 0.1008,
+      "step": 6639
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.31837896495886214,
+      "learning_rate": 4.282655057494593e-06,
+      "loss": 0.0956,
+      "step": 6640
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.21381520167328397,
+      "learning_rate": 4.281304596741137e-06,
+      "loss": 0.0647,
+      "step": 6641
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.46929036407119457,
+      "learning_rate": 4.279954189520301e-06,
+      "loss": 0.0969,
+      "step": 6642
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.22438810150653105,
+      "learning_rate": 4.27860383593267e-06,
+      "loss": 0.0532,
+      "step": 6643
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7692156015720939,
+      "learning_rate": 4.277253536078829e-06,
+      "loss": 0.0907,
+      "step": 6644
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3026380882756604,
+      "learning_rate": 4.275903290059355e-06,
+      "loss": 0.0568,
+      "step": 6645
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.36395642704955194,
+      "learning_rate": 4.274553097974821e-06,
+      "loss": 0.0628,
+      "step": 6646
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4038333685472117,
+      "learning_rate": 4.2732029599258e-06,
+      "loss": 0.1113,
+      "step": 6647
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.2712562708269014,
+      "learning_rate": 4.271852876012855e-06,
+      "loss": 0.0695,
+      "step": 6648
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3752884877672261,
+      "learning_rate": 4.270502846336549e-06,
+      "loss": 0.084,
+      "step": 6649
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3016811852232453,
+      "learning_rate": 4.2691528709974415e-06,
+      "loss": 0.0944,
+      "step": 6650
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.20237388424269012,
+      "learning_rate": 4.267802950096087e-06,
+      "loss": 0.0491,
+      "step": 6651
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.39437098179936275,
+      "learning_rate": 4.266453083733033e-06,
+      "loss": 0.1141,
+      "step": 6652
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5884862383200715,
+      "learning_rate": 4.265103272008825e-06,
+      "loss": 0.0884,
+      "step": 6653
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.1606884357867657,
+      "learning_rate": 4.263753515024009e-06,
+      "loss": 0.0494,
+      "step": 6654
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.25377134952149355,
+      "learning_rate": 4.262403812879119e-06,
+      "loss": 0.0854,
+      "step": 6655
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4773211473069833,
+      "learning_rate": 4.261054165674688e-06,
+      "loss": 0.0915,
+      "step": 6656
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3566499101468165,
+      "learning_rate": 4.259704573511251e-06,
+      "loss": 0.0943,
+      "step": 6657
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.21462589045795086,
+      "learning_rate": 4.258355036489328e-06,
+      "loss": 0.051,
+      "step": 6658
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.25194025330417336,
+      "learning_rate": 4.257005554709443e-06,
+      "loss": 0.0585,
+      "step": 6659
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4047910446001853,
+      "learning_rate": 4.255656128272111e-06,
+      "loss": 0.1007,
+      "step": 6660
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5169029362885067,
+      "learning_rate": 4.254306757277848e-06,
+      "loss": 0.1011,
+      "step": 6661
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.2710250186959179,
+      "learning_rate": 4.252957441827162e-06,
+      "loss": 0.073,
+      "step": 6662
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.486120137621442,
+      "learning_rate": 4.251608182020555e-06,
+      "loss": 0.1205,
+      "step": 6663
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.2434319241789466,
+      "learning_rate": 4.250258977958533e-06,
+      "loss": 0.0756,
+      "step": 6664
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3115339028155324,
+      "learning_rate": 4.24890982974159e-06,
+      "loss": 0.0972,
+      "step": 6665
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3410972245840506,
+      "learning_rate": 4.247560737470216e-06,
+      "loss": 0.0904,
+      "step": 6666
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.433896903377954,
+      "learning_rate": 4.2462117012449036e-06,
+      "loss": 0.0734,
+      "step": 6667
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.31950064877730494,
+      "learning_rate": 4.244862721166135e-06,
+      "loss": 0.0683,
+      "step": 6668
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.47586824202799577,
+      "learning_rate": 4.24351379733439e-06,
+      "loss": 0.0722,
+      "step": 6669
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4467431146350214,
+      "learning_rate": 4.2421649298501415e-06,
+      "loss": 0.0975,
+      "step": 6670
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3080481959613241,
+      "learning_rate": 4.240816118813866e-06,
+      "loss": 0.0958,
+      "step": 6671
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.18986314121026768,
+      "learning_rate": 4.239467364326029e-06,
+      "loss": 0.0331,
+      "step": 6672
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3868883046940507,
+      "learning_rate": 4.238118666487092e-06,
+      "loss": 0.0788,
+      "step": 6673
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.33729334016385415,
+      "learning_rate": 4.236770025397515e-06,
+      "loss": 0.1042,
+      "step": 6674
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.264508847821021,
+      "learning_rate": 4.235421441157753e-06,
+      "loss": 0.0636,
+      "step": 6675
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.307877421132644,
+      "learning_rate": 4.234072913868257e-06,
+      "loss": 0.0763,
+      "step": 6676
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3351767497350244,
+      "learning_rate": 4.232724443629471e-06,
+      "loss": 0.0844,
+      "step": 6677
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.22647589931448567,
+      "learning_rate": 4.231376030541838e-06,
+      "loss": 0.0556,
+      "step": 6678
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5359067427416808,
+      "learning_rate": 4.2300276747057965e-06,
+      "loss": 0.1352,
+      "step": 6679
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.2847611911006017,
+      "learning_rate": 4.228679376221779e-06,
+      "loss": 0.086,
+      "step": 6680
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4064726869747234,
+      "learning_rate": 4.227331135190215e-06,
+      "loss": 0.094,
+      "step": 6681
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.41168425302614675,
+      "learning_rate": 4.22598295171153e-06,
+      "loss": 0.112,
+      "step": 6682
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4657560318216727,
+      "learning_rate": 4.224634825886141e-06,
+      "loss": 0.1181,
+      "step": 6683
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.37438699301306105,
+      "learning_rate": 4.22328675781447e-06,
+      "loss": 0.0763,
+      "step": 6684
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.40047767201660317,
+      "learning_rate": 4.221938747596924e-06,
+      "loss": 0.0785,
+      "step": 6685
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3412003601161075,
+      "learning_rate": 4.220590795333914e-06,
+      "loss": 0.1131,
+      "step": 6686
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.46762750466251835,
+      "learning_rate": 4.21924290112584e-06,
+      "loss": 0.1169,
+      "step": 6687
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.28760476594750617,
+      "learning_rate": 4.217895065073105e-06,
+      "loss": 0.0737,
+      "step": 6688
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.31362829091012934,
+      "learning_rate": 4.216547287276102e-06,
+      "loss": 0.0852,
+      "step": 6689
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.25126486213224736,
+      "learning_rate": 4.215199567835217e-06,
+      "loss": 0.0623,
+      "step": 6690
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.2922482543097791,
+      "learning_rate": 4.213851906850844e-06,
+      "loss": 0.0935,
+      "step": 6691
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.30804714623059265,
+      "learning_rate": 4.21250430442336e-06,
+      "loss": 0.067,
+      "step": 6692
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.340713815975127,
+      "learning_rate": 4.211156760653143e-06,
+      "loss": 0.089,
+      "step": 6693
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.16815655547344227,
+      "learning_rate": 4.209809275640564e-06,
+      "loss": 0.0517,
+      "step": 6694
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.36198770625380794,
+      "learning_rate": 4.208461849485995e-06,
+      "loss": 0.0936,
+      "step": 6695
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.1606019699919142,
+      "learning_rate": 4.2071144822897985e-06,
+      "loss": 0.0334,
+      "step": 6696
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3762244402287604,
+      "learning_rate": 4.205767174152332e-06,
+      "loss": 0.0902,
+      "step": 6697
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.28364316221310926,
+      "learning_rate": 4.204419925173956e-06,
+      "loss": 0.069,
+      "step": 6698
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3767236506449875,
+      "learning_rate": 4.203072735455018e-06,
+      "loss": 0.0715,
+      "step": 6699
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.39123230195369724,
+      "learning_rate": 4.201725605095863e-06,
+      "loss": 0.1014,
+      "step": 6700
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3160990980692873,
+      "learning_rate": 4.200378534196838e-06,
+      "loss": 0.0739,
+      "step": 6701
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.2827541470515303,
+      "learning_rate": 4.199031522858277e-06,
+      "loss": 0.0702,
+      "step": 6702
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3362880977363988,
+      "learning_rate": 4.197684571180515e-06,
+      "loss": 0.079,
+      "step": 6703
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3799807391828085,
+      "learning_rate": 4.196337679263877e-06,
+      "loss": 0.1208,
+      "step": 6704
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.38750061256211654,
+      "learning_rate": 4.194990847208693e-06,
+      "loss": 0.0849,
+      "step": 6705
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.23492537434783853,
+      "learning_rate": 4.193644075115279e-06,
+      "loss": 0.0664,
+      "step": 6706
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3812605257978098,
+      "learning_rate": 4.192297363083952e-06,
+      "loss": 0.0792,
+      "step": 6707
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.2494947143137089,
+      "learning_rate": 4.1909507112150225e-06,
+      "loss": 0.1029,
+      "step": 6708
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3435038879352796,
+      "learning_rate": 4.189604119608798e-06,
+      "loss": 0.095,
+      "step": 6709
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.40538101682228045,
+      "learning_rate": 4.1882575883655795e-06,
+      "loss": 0.1293,
+      "step": 6710
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.2693400560496811,
+      "learning_rate": 4.186911117585663e-06,
+      "loss": 0.0777,
+      "step": 6711
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3433320318061915,
+      "learning_rate": 4.185564707369344e-06,
+      "loss": 0.0806,
+      "step": 6712
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.41652511664925856,
+      "learning_rate": 4.184218357816911e-06,
+      "loss": 0.076,
+      "step": 6713
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4008158984150123,
+      "learning_rate": 4.182872069028645e-06,
+      "loss": 0.0587,
+      "step": 6714
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.24111092261750192,
+      "learning_rate": 4.181525841104829e-06,
+      "loss": 0.0856,
+      "step": 6715
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.2913536126368673,
+      "learning_rate": 4.180179674145737e-06,
+      "loss": 0.0717,
+      "step": 6716
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3898928730357047,
+      "learning_rate": 4.178833568251636e-06,
+      "loss": 0.0939,
+      "step": 6717
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.23194975690205097,
+      "learning_rate": 4.177487523522796e-06,
+      "loss": 0.05,
+      "step": 6718
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.26129087501936504,
+      "learning_rate": 4.176141540059477e-06,
+      "loss": 0.0775,
+      "step": 6719
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3249522802961273,
+      "learning_rate": 4.174795617961936e-06,
+      "loss": 0.0799,
+      "step": 6720
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3765665308350886,
+      "learning_rate": 4.173449757330422e-06,
+      "loss": 0.0889,
+      "step": 6721
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3848623002172383,
+      "learning_rate": 4.172103958265187e-06,
+      "loss": 0.0824,
+      "step": 6722
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.2947809595564576,
+      "learning_rate": 4.170758220866473e-06,
+      "loss": 0.0825,
+      "step": 6723
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.32852547407410143,
+      "learning_rate": 4.169412545234515e-06,
+      "loss": 0.0997,
+      "step": 6724
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.2917985091144354,
+      "learning_rate": 4.1680669314695506e-06,
+      "loss": 0.1093,
+      "step": 6725
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3676117906077204,
+      "learning_rate": 4.166721379671807e-06,
+      "loss": 0.0759,
+      "step": 6726
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4158686518769835,
+      "learning_rate": 4.165375889941511e-06,
+      "loss": 0.1175,
+      "step": 6727
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.33874976097916226,
+      "learning_rate": 4.164030462378877e-06,
+      "loss": 0.0879,
+      "step": 6728
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6191105705592647,
+      "learning_rate": 4.162685097084127e-06,
+      "loss": 0.1136,
+      "step": 6729
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3394354024915789,
+      "learning_rate": 4.161339794157469e-06,
+      "loss": 0.0702,
+      "step": 6730
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.18324020252193918,
+      "learning_rate": 4.159994553699106e-06,
+      "loss": 0.0681,
+      "step": 6731
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3528293076985753,
+      "learning_rate": 4.158649375809245e-06,
+      "loss": 0.1096,
+      "step": 6732
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.301369646015412,
+      "learning_rate": 4.157304260588078e-06,
+      "loss": 0.0429,
+      "step": 6733
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.21058740119242356,
+      "learning_rate": 4.1559592081358005e-06,
+      "loss": 0.0472,
+      "step": 6734
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.34059933226435074,
+      "learning_rate": 4.154614218552595e-06,
+      "loss": 0.1106,
+      "step": 6735
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.26349059294240984,
+      "learning_rate": 4.153269291938649e-06,
+      "loss": 0.0595,
+      "step": 6736
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.14022659714520191,
+      "learning_rate": 4.15192442839414e-06,
+      "loss": 0.0416,
+      "step": 6737
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3047515663740909,
+      "learning_rate": 4.150579628019237e-06,
+      "loss": 0.0757,
+      "step": 6738
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4675222058245978,
+      "learning_rate": 4.149234890914113e-06,
+      "loss": 0.1265,
+      "step": 6739
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.23816280354749028,
+      "learning_rate": 4.147890217178932e-06,
+      "loss": 0.0598,
+      "step": 6740
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.42222867059924685,
+      "learning_rate": 4.146545606913849e-06,
+      "loss": 0.1091,
+      "step": 6741
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9252420334686703,
+      "learning_rate": 4.145201060219022e-06,
+      "loss": 0.1097,
+      "step": 6742
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.2079929274586764,
+      "learning_rate": 4.1438565771946005e-06,
+      "loss": 0.0536,
+      "step": 6743
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.30994152102768036,
+      "learning_rate": 4.142512157940728e-06,
+      "loss": 0.0889,
+      "step": 6744
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.44675145354536105,
+      "learning_rate": 4.141167802557544e-06,
+      "loss": 0.1126,
+      "step": 6745
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5829347882920988,
+      "learning_rate": 4.1398235111451865e-06,
+      "loss": 0.1497,
+      "step": 6746
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.244406470226285,
+      "learning_rate": 4.138479283803785e-06,
+      "loss": 0.0625,
+      "step": 6747
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.31077116515650294,
+      "learning_rate": 4.137135120633464e-06,
+      "loss": 0.1108,
+      "step": 6748
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3622631766485714,
+      "learning_rate": 4.135791021734347e-06,
+      "loss": 0.0871,
+      "step": 6749
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3539181512870654,
+      "learning_rate": 4.134446987206548e-06,
+      "loss": 0.1246,
+      "step": 6750
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.19785518944316133,
+      "learning_rate": 4.13310301715018e-06,
+      "loss": 0.0451,
+      "step": 6751
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3095474536570035,
+      "learning_rate": 4.131759111665349e-06,
+      "loss": 0.0822,
+      "step": 6752
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.23929064271995967,
+      "learning_rate": 4.130415270852158e-06,
+      "loss": 0.0703,
+      "step": 6753
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.33406849249711207,
+      "learning_rate": 4.129071494810704e-06,
+      "loss": 0.1347,
+      "step": 6754
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3155874898517172,
+      "learning_rate": 4.127727783641076e-06,
+      "loss": 0.0812,
+      "step": 6755
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.34216318485910996,
+      "learning_rate": 4.126384137443365e-06,
+      "loss": 0.0747,
+      "step": 6756
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3134103989651475,
+      "learning_rate": 4.125040556317654e-06,
+      "loss": 0.0736,
+      "step": 6757
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3198481571532385,
+      "learning_rate": 4.123697040364018e-06,
+      "loss": 0.0645,
+      "step": 6758
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.2766614292535357,
+      "learning_rate": 4.122353589682531e-06,
+      "loss": 0.0874,
+      "step": 6759
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4309210900063584,
+      "learning_rate": 4.121010204373264e-06,
+      "loss": 0.0607,
+      "step": 6760
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3778376034182019,
+      "learning_rate": 4.1196668845362755e-06,
+      "loss": 0.1185,
+      "step": 6761
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.33557607650423654,
+      "learning_rate": 4.118323630271624e-06,
+      "loss": 0.0928,
+      "step": 6762
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3017778456538718,
+      "learning_rate": 4.116980441679368e-06,
+      "loss": 0.0588,
+      "step": 6763
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.38555669907917045,
+      "learning_rate": 4.115637318859551e-06,
+      "loss": 0.0925,
+      "step": 6764
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3107414615073564,
+      "learning_rate": 4.114294261912217e-06,
+      "loss": 0.1089,
+      "step": 6765
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.27989614983308553,
+      "learning_rate": 4.112951270937409e-06,
+      "loss": 0.0667,
+      "step": 6766
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.30633754155861675,
+      "learning_rate": 4.111608346035156e-06,
+      "loss": 0.0769,
+      "step": 6767
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.21228421401237366,
+      "learning_rate": 4.11026548730549e-06,
+      "loss": 0.0471,
+      "step": 6768
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.29747965032265683,
+      "learning_rate": 4.108922694848431e-06,
+      "loss": 0.0767,
+      "step": 6769
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4148808617750333,
+      "learning_rate": 4.107579968764003e-06,
+      "loss": 0.1203,
+      "step": 6770
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.31096328151370295,
+      "learning_rate": 4.106237309152216e-06,
+      "loss": 0.1061,
+      "step": 6771
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.1869184244570876,
+      "learning_rate": 4.104894716113081e-06,
+      "loss": 0.0676,
+      "step": 6772
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.2126062417106381,
+      "learning_rate": 4.103552189746603e-06,
+      "loss": 0.0507,
+      "step": 6773
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.20945189582317772,
+      "learning_rate": 4.10220973015278e-06,
+      "loss": 0.0674,
+      "step": 6774
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.314407288815462,
+      "learning_rate": 4.100867337431605e-06,
+      "loss": 0.0843,
+      "step": 6775
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3708977701115243,
+      "learning_rate": 4.099525011683069e-06,
+      "loss": 0.0916,
+      "step": 6776
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3258221600453027,
+      "learning_rate": 4.098182753007157e-06,
+      "loss": 0.0873,
+      "step": 6777
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.29359793059488043,
+      "learning_rate": 4.096840561503847e-06,
+      "loss": 0.0851,
+      "step": 6778
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.32472828410485166,
+      "learning_rate": 4.095498437273112e-06,
+      "loss": 0.0778,
+      "step": 6779
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3430649384796402,
+      "learning_rate": 4.094156380414922e-06,
+      "loss": 0.0923,
+      "step": 6780
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.2590180899862874,
+      "learning_rate": 4.092814391029244e-06,
+      "loss": 0.059,
+      "step": 6781
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3245146174371091,
+      "learning_rate": 4.091472469216033e-06,
+      "loss": 0.0889,
+      "step": 6782
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.2985357030479436,
+      "learning_rate": 4.090130615075246e-06,
+      "loss": 0.0837,
+      "step": 6783
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3032639462786435,
+      "learning_rate": 4.0887888287068315e-06,
+      "loss": 0.0835,
+      "step": 6784
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.37380290910973657,
+      "learning_rate": 4.0874471102107325e-06,
+      "loss": 0.0718,
+      "step": 6785
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.47892256937214184,
+      "learning_rate": 4.086105459686888e-06,
+      "loss": 0.0839,
+      "step": 6786
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.2533923036998749,
+      "learning_rate": 4.084763877235234e-06,
+      "loss": 0.0693,
+      "step": 6787
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.30186878019287416,
+      "learning_rate": 4.083422362955698e-06,
+      "loss": 0.0862,
+      "step": 6788
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.21807618499576664,
+      "learning_rate": 4.082080916948203e-06,
+      "loss": 0.0372,
+      "step": 6789
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.32440083919954177,
+      "learning_rate": 4.080739539312669e-06,
+      "loss": 0.0979,
+      "step": 6790
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.2716222596817052,
+      "learning_rate": 4.07939823014901e-06,
+      "loss": 0.0565,
+      "step": 6791
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.18935570222734865,
+      "learning_rate": 4.0780569895571305e-06,
+      "loss": 0.0476,
+      "step": 6792
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.23267733586860823,
+      "learning_rate": 4.07671581763694e-06,
+      "loss": 0.0468,
+      "step": 6793
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.40777128513992217,
+      "learning_rate": 4.075374714488335e-06,
+      "loss": 0.0718,
+      "step": 6794
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3291297067225434,
+      "learning_rate": 4.0740336802112065e-06,
+      "loss": 0.105,
+      "step": 6795
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.40748911200926774,
+      "learning_rate": 4.072692714905441e-06,
+      "loss": 0.1269,
+      "step": 6796
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.34523880170730464,
+      "learning_rate": 4.071351818670928e-06,
+      "loss": 0.0732,
+      "step": 6797
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.36771245433661803,
+      "learning_rate": 4.07001099160754e-06,
+      "loss": 0.1244,
+      "step": 6798
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.36303065929658823,
+      "learning_rate": 4.0686702338151485e-06,
+      "loss": 0.0919,
+      "step": 6799
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.27990487172064493,
+      "learning_rate": 4.067329545393626e-06,
+      "loss": 0.0766,
+      "step": 6800
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.41379882302685794,
+      "learning_rate": 4.065988926442833e-06,
+      "loss": 0.1033,
+      "step": 6801
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.41654549464294943,
+      "learning_rate": 4.064648377062624e-06,
+      "loss": 0.0857,
+      "step": 6802
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.27639101990603476,
+      "learning_rate": 4.063307897352853e-06,
+      "loss": 0.0875,
+      "step": 6803
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4579069408372044,
+      "learning_rate": 4.061967487413366e-06,
+      "loss": 0.0954,
+      "step": 6804
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5178593222827905,
+      "learning_rate": 4.060627147344006e-06,
+      "loss": 0.1168,
+      "step": 6805
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5750232642417462,
+      "learning_rate": 4.059286877244606e-06,
+      "loss": 0.1279,
+      "step": 6806
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.35758193837051583,
+      "learning_rate": 4.057946677215002e-06,
+      "loss": 0.0891,
+      "step": 6807
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3680931555999769,
+      "learning_rate": 4.056606547355017e-06,
+      "loss": 0.085,
+      "step": 6808
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3928735116157584,
+      "learning_rate": 4.055266487764472e-06,
+      "loss": 0.106,
+      "step": 6809
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.33825824393667336,
+      "learning_rate": 4.0539264985431815e-06,
+      "loss": 0.081,
+      "step": 6810
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.2077629439405697,
+      "learning_rate": 4.052586579790958e-06,
+      "loss": 0.0764,
+      "step": 6811
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.2307810704491016,
+      "learning_rate": 4.051246731607605e-06,
+      "loss": 0.0627,
+      "step": 6812
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5049284676722581,
+      "learning_rate": 4.049906954092922e-06,
+      "loss": 0.1031,
+      "step": 6813
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.41505499428228176,
+      "learning_rate": 4.048567247346702e-06,
+      "loss": 0.0903,
+      "step": 6814
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.20259650983511962,
+      "learning_rate": 4.047227611468738e-06,
+      "loss": 0.0571,
+      "step": 6815
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.40376554744034776,
+      "learning_rate": 4.045888046558811e-06,
+      "loss": 0.0813,
+      "step": 6816
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3720988875762975,
+      "learning_rate": 4.044548552716702e-06,
+      "loss": 0.1188,
+      "step": 6817
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3338429090223802,
+      "learning_rate": 4.043209130042182e-06,
+      "loss": 0.0827,
+      "step": 6818
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.18049145025515426,
+      "learning_rate": 4.041869778635018e-06,
+      "loss": 0.0396,
+      "step": 6819
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.31032778295999447,
+      "learning_rate": 4.040530498594975e-06,
+      "loss": 0.0602,
+      "step": 6820
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.20363874233113247,
+      "learning_rate": 4.039191290021811e-06,
+      "loss": 0.0399,
+      "step": 6821
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.18534918619017265,
+      "learning_rate": 4.0378521530152765e-06,
+      "loss": 0.0664,
+      "step": 6822
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.343690806614877,
+      "learning_rate": 4.0365130876751166e-06,
+      "loss": 0.0931,
+      "step": 6823
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.27651327933324993,
+      "learning_rate": 4.035174094101077e-06,
+      "loss": 0.0732,
+      "step": 6824
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7756232879371485,
+      "learning_rate": 4.033835172392891e-06,
+      "loss": 0.1068,
+      "step": 6825
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3801855831507543,
+      "learning_rate": 4.0324963226502875e-06,
+      "loss": 0.0621,
+      "step": 6826
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3525152499037248,
+      "learning_rate": 4.031157544972997e-06,
+      "loss": 0.083,
+      "step": 6827
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4165430351194655,
+      "learning_rate": 4.029818839460735e-06,
+      "loss": 0.1311,
+      "step": 6828
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.20623182200542317,
+      "learning_rate": 4.0284802062132186e-06,
+      "loss": 0.0482,
+      "step": 6829
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.16966207710350995,
+      "learning_rate": 4.027141645330154e-06,
+      "loss": 0.0448,
+      "step": 6830
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.31182183818868703,
+      "learning_rate": 4.025803156911249e-06,
+      "loss": 0.1121,
+      "step": 6831
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.38283168639700876,
+      "learning_rate": 4.0244647410562e-06,
+      "loss": 0.081,
+      "step": 6832
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4177354344355331,
+      "learning_rate": 4.023126397864697e-06,
+      "loss": 0.1224,
+      "step": 6833
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.31735514000818316,
+      "learning_rate": 4.021788127436434e-06,
+      "loss": 0.1078,
+      "step": 6834
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3094289346702054,
+      "learning_rate": 4.020449929871089e-06,
+      "loss": 0.1018,
+      "step": 6835
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.20002493537778415,
+      "learning_rate": 4.01911180526834e-06,
+      "loss": 0.045,
+      "step": 6836
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2180440884792028,
+      "learning_rate": 4.017773753727855e-06,
+      "loss": 0.0441,
+      "step": 6837
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.30165629935327953,
+      "learning_rate": 4.0164357753493054e-06,
+      "loss": 0.09,
+      "step": 6838
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.29417436071160563,
+      "learning_rate": 4.015097870232349e-06,
+      "loss": 0.052,
+      "step": 6839
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.38880938549244015,
+      "learning_rate": 4.013760038476638e-06,
+      "loss": 0.1034,
+      "step": 6840
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.32806614950758095,
+      "learning_rate": 4.012422280181827e-06,
+      "loss": 0.1055,
+      "step": 6841
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.19801827242258774,
+      "learning_rate": 4.011084595447558e-06,
+      "loss": 0.0354,
+      "step": 6842
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2273540229962862,
+      "learning_rate": 4.009746984373469e-06,
+      "loss": 0.0661,
+      "step": 6843
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3107326713018466,
+      "learning_rate": 4.00840944705919e-06,
+      "loss": 0.0822,
+      "step": 6844
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.33049864915313754,
+      "learning_rate": 4.007071983604356e-06,
+      "loss": 0.0761,
+      "step": 6845
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.26568217506266945,
+      "learning_rate": 4.005734594108583e-06,
+      "loss": 0.0515,
+      "step": 6846
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.13978706738923913,
+      "learning_rate": 4.00439727867149e-06,
+      "loss": 0.0345,
+      "step": 6847
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2982093507094647,
+      "learning_rate": 4.003060037392687e-06,
+      "loss": 0.0934,
+      "step": 6848
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3503974564286525,
+      "learning_rate": 4.00172287037178e-06,
+      "loss": 0.1076,
+      "step": 6849
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7170834209883835,
+      "learning_rate": 4.00038577770837e-06,
+      "loss": 0.0755,
+      "step": 6850
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.39267547573702943,
+      "learning_rate": 3.99904875950205e-06,
+      "loss": 0.0721,
+      "step": 6851
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3600723060561483,
+      "learning_rate": 3.99771181585241e-06,
+      "loss": 0.107,
+      "step": 6852
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4613315424069212,
+      "learning_rate": 3.996374946859031e-06,
+      "loss": 0.0875,
+      "step": 6853
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7428663279329595,
+      "learning_rate": 3.995038152621493e-06,
+      "loss": 0.0806,
+      "step": 6854
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.32532917000953604,
+      "learning_rate": 3.993701433239368e-06,
+      "loss": 0.0993,
+      "step": 6855
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2067053801008601,
+      "learning_rate": 3.992364788812223e-06,
+      "loss": 0.0558,
+      "step": 6856
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2842408093994554,
+      "learning_rate": 3.991028219439616e-06,
+      "loss": 0.1108,
+      "step": 6857
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.26767623446334693,
+      "learning_rate": 3.989691725221107e-06,
+      "loss": 0.0594,
+      "step": 6858
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3248118158462626,
+      "learning_rate": 3.988355306256245e-06,
+      "loss": 0.0371,
+      "step": 6859
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3112313666799693,
+      "learning_rate": 3.987018962644572e-06,
+      "loss": 0.0791,
+      "step": 6860
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5887704456283226,
+      "learning_rate": 3.985682694485627e-06,
+      "loss": 0.1529,
+      "step": 6861
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.38791183324123424,
+      "learning_rate": 3.984346501878944e-06,
+      "loss": 0.0866,
+      "step": 6862
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.302491237976391,
+      "learning_rate": 3.9830103849240505e-06,
+      "loss": 0.0855,
+      "step": 6863
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3255217508390222,
+      "learning_rate": 3.981674343720466e-06,
+      "loss": 0.0729,
+      "step": 6864
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.26594685480403674,
+      "learning_rate": 3.9803383783677105e-06,
+      "loss": 0.0743,
+      "step": 6865
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3306186169374885,
+      "learning_rate": 3.979002488965293e-06,
+      "loss": 0.0865,
+      "step": 6866
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.30203550558194897,
+      "learning_rate": 3.977666675612716e-06,
+      "loss": 0.0913,
+      "step": 6867
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2507093305463405,
+      "learning_rate": 3.9763309384094826e-06,
+      "loss": 0.0916,
+      "step": 6868
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.30186650052077385,
+      "learning_rate": 3.9749952774550835e-06,
+      "loss": 0.0917,
+      "step": 6869
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3288616087363759,
+      "learning_rate": 3.973659692849008e-06,
+      "loss": 0.0878,
+      "step": 6870
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0474023553411038,
+      "learning_rate": 3.972324184690735e-06,
+      "loss": 0.12,
+      "step": 6871
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2773767454393442,
+      "learning_rate": 3.9709887530797455e-06,
+      "loss": 0.079,
+      "step": 6872
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4087339597638231,
+      "learning_rate": 3.969653398115507e-06,
+      "loss": 0.0907,
+      "step": 6873
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7185180661234648,
+      "learning_rate": 3.9683181198974856e-06,
+      "loss": 0.1537,
+      "step": 6874
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3052978535512997,
+      "learning_rate": 3.966982918525141e-06,
+      "loss": 0.0871,
+      "step": 6875
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.23011818774772066,
+      "learning_rate": 3.965647794097927e-06,
+      "loss": 0.0656,
+      "step": 6876
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2906644857966385,
+      "learning_rate": 3.964312746715291e-06,
+      "loss": 0.09,
+      "step": 6877
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4580226570243548,
+      "learning_rate": 3.962977776476673e-06,
+      "loss": 0.126,
+      "step": 6878
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2179894609241442,
+      "learning_rate": 3.961642883481514e-06,
+      "loss": 0.0793,
+      "step": 6879
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3149798752114038,
+      "learning_rate": 3.960308067829241e-06,
+      "loss": 0.1031,
+      "step": 6880
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.36145864546505274,
+      "learning_rate": 3.958973329619278e-06,
+      "loss": 0.1012,
+      "step": 6881
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4653254560014289,
+      "learning_rate": 3.957638668951047e-06,
+      "loss": 0.1233,
+      "step": 6882
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.22409894563567676,
+      "learning_rate": 3.956304085923961e-06,
+      "loss": 0.0763,
+      "step": 6883
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3170448879218194,
+      "learning_rate": 3.954969580637426e-06,
+      "loss": 0.0684,
+      "step": 6884
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.36381311831357716,
+      "learning_rate": 3.953635153190845e-06,
+      "loss": 0.0919,
+      "step": 6885
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.35663766732705743,
+      "learning_rate": 3.952300803683612e-06,
+      "loss": 0.108,
+      "step": 6886
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.31699664296900637,
+      "learning_rate": 3.950966532215119e-06,
+      "loss": 0.0667,
+      "step": 6887
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.37314035144374846,
+      "learning_rate": 3.949632338884749e-06,
+      "loss": 0.1008,
+      "step": 6888
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2807584043156559,
+      "learning_rate": 3.9482982237918834e-06,
+      "loss": 0.0795,
+      "step": 6889
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.27005308049762156,
+      "learning_rate": 3.946964187035892e-06,
+      "loss": 0.0772,
+      "step": 6890
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.26411207342231685,
+      "learning_rate": 3.94563022871614e-06,
+      "loss": 0.0614,
+      "step": 6891
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3958383665053706,
+      "learning_rate": 3.944296348931993e-06,
+      "loss": 0.0768,
+      "step": 6892
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.21767363431582126,
+      "learning_rate": 3.942962547782804e-06,
+      "loss": 0.0579,
+      "step": 6893
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4735568060561059,
+      "learning_rate": 3.941628825367922e-06,
+      "loss": 0.1178,
+      "step": 6894
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.28398262507960714,
+      "learning_rate": 3.940295181786688e-06,
+      "loss": 0.0731,
+      "step": 6895
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2456623769835113,
+      "learning_rate": 3.938961617138445e-06,
+      "loss": 0.0947,
+      "step": 6896
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.34473921628690324,
+      "learning_rate": 3.937628131522521e-06,
+      "loss": 0.1109,
+      "step": 6897
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4360184885539232,
+      "learning_rate": 3.936294725038241e-06,
+      "loss": 0.1177,
+      "step": 6898
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2872847294398273,
+      "learning_rate": 3.9349613977849286e-06,
+      "loss": 0.0875,
+      "step": 6899
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.27738662823783844,
+      "learning_rate": 3.933628149861895e-06,
+      "loss": 0.0754,
+      "step": 6900
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7029633715831742,
+      "learning_rate": 3.932294981368447e-06,
+      "loss": 0.1495,
+      "step": 6901
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.19966650316120085,
+      "learning_rate": 3.930961892403892e-06,
+      "loss": 0.0475,
+      "step": 6902
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2816966978197229,
+      "learning_rate": 3.929628883067521e-06,
+      "loss": 0.0873,
+      "step": 6903
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.22519645617762957,
+      "learning_rate": 3.928295953458627e-06,
+      "loss": 0.06,
+      "step": 6904
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.45407714829663914,
+      "learning_rate": 3.926963103676492e-06,
+      "loss": 0.0948,
+      "step": 6905
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3753609797589767,
+      "learning_rate": 3.925630333820396e-06,
+      "loss": 0.0876,
+      "step": 6906
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.42273827838680217,
+      "learning_rate": 3.924297643989614e-06,
+      "loss": 0.1221,
+      "step": 6907
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.17436072127293975,
+      "learning_rate": 3.922965034283406e-06,
+      "loss": 0.0574,
+      "step": 6908
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.20532330459128384,
+      "learning_rate": 3.92163250480104e-06,
+      "loss": 0.044,
+      "step": 6909
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.26158196627028846,
+      "learning_rate": 3.920300055641765e-06,
+      "loss": 0.0828,
+      "step": 6910
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.33341292629031166,
+      "learning_rate": 3.918967686904834e-06,
+      "loss": 0.091,
+      "step": 6911
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.21122779945521636,
+      "learning_rate": 3.9176353986894826e-06,
+      "loss": 0.0619,
+      "step": 6912
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3879874587047194,
+      "learning_rate": 3.916303191094955e-06,
+      "loss": 0.1329,
+      "step": 6913
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.28900429982951475,
+      "learning_rate": 3.91497106422048e-06,
+      "loss": 0.0623,
+      "step": 6914
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.25485906322825685,
+      "learning_rate": 3.913639018165278e-06,
+      "loss": 0.0477,
+      "step": 6915
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3336056533499356,
+      "learning_rate": 3.912307053028571e-06,
+      "loss": 0.0957,
+      "step": 6916
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2855754068711921,
+      "learning_rate": 3.910975168909573e-06,
+      "loss": 0.0374,
+      "step": 6917
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.13585458714134174,
+      "learning_rate": 3.909643365907486e-06,
+      "loss": 0.0353,
+      "step": 6918
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4751721332480166,
+      "learning_rate": 3.908311644121515e-06,
+      "loss": 0.1246,
+      "step": 6919
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.26515906135054784,
+      "learning_rate": 3.906980003650852e-06,
+      "loss": 0.0777,
+      "step": 6920
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7268104034896427,
+      "learning_rate": 3.905648444594684e-06,
+      "loss": 0.0925,
+      "step": 6921
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.22550233843603237,
+      "learning_rate": 3.904316967052196e-06,
+      "loss": 0.0643,
+      "step": 6922
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.357974656274648,
+      "learning_rate": 3.902985571122563e-06,
+      "loss": 0.062,
+      "step": 6923
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2488885939456447,
+      "learning_rate": 3.901654256904957e-06,
+      "loss": 0.0562,
+      "step": 6924
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.24560733442498642,
+      "learning_rate": 3.900323024498536e-06,
+      "loss": 0.0711,
+      "step": 6925
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.316387671919573,
+      "learning_rate": 3.898991874002466e-06,
+      "loss": 0.0686,
+      "step": 6926
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3963199896160202,
+      "learning_rate": 3.897660805515894e-06,
+      "loss": 0.0977,
+      "step": 6927
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6910810913344372,
+      "learning_rate": 3.896329819137967e-06,
+      "loss": 0.1385,
+      "step": 6928
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.22774201678755376,
+      "learning_rate": 3.8949989149678224e-06,
+      "loss": 0.0448,
+      "step": 6929
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.40492149371090796,
+      "learning_rate": 3.893668093104598e-06,
+      "loss": 0.1083,
+      "step": 6930
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3799007974898237,
+      "learning_rate": 3.892337353647419e-06,
+      "loss": 0.0832,
+      "step": 6931
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.25474780141915054,
+      "learning_rate": 3.891006696695404e-06,
+      "loss": 0.0833,
+      "step": 6932
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.23560666806712227,
+      "learning_rate": 3.889676122347673e-06,
+      "loss": 0.0539,
+      "step": 6933
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.36829838265912307,
+      "learning_rate": 3.888345630703333e-06,
+      "loss": 0.0933,
+      "step": 6934
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.35923305151957297,
+      "learning_rate": 3.887015221861483e-06,
+      "loss": 0.0951,
+      "step": 6935
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3583063995201625,
+      "learning_rate": 3.885684895921226e-06,
+      "loss": 0.1175,
+      "step": 6936
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2571510882881295,
+      "learning_rate": 3.884354652981649e-06,
+      "loss": 0.0491,
+      "step": 6937
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5138359927628624,
+      "learning_rate": 3.883024493141836e-06,
+      "loss": 0.1114,
+      "step": 6938
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.35977749156900724,
+      "learning_rate": 3.881694416500865e-06,
+      "loss": 0.1144,
+      "step": 6939
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3227372429212325,
+      "learning_rate": 3.880364423157809e-06,
+      "loss": 0.0494,
+      "step": 6940
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3252328144018323,
+      "learning_rate": 3.879034513211734e-06,
+      "loss": 0.0733,
+      "step": 6941
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5130741662260205,
+      "learning_rate": 3.8777046867616965e-06,
+      "loss": 0.1128,
+      "step": 6942
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3472291509855456,
+      "learning_rate": 3.876374943906753e-06,
+      "loss": 0.0897,
+      "step": 6943
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3482848141859018,
+      "learning_rate": 3.87504528474595e-06,
+      "loss": 0.0718,
+      "step": 6944
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.21451089036494903,
+      "learning_rate": 3.873715709378327e-06,
+      "loss": 0.0721,
+      "step": 6945
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2860063289050506,
+      "learning_rate": 3.872386217902918e-06,
+      "loss": 0.0601,
+      "step": 6946
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2197182276614109,
+      "learning_rate": 3.871056810418754e-06,
+      "loss": 0.0573,
+      "step": 6947
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.21099458045111344,
+      "learning_rate": 3.869727487024854e-06,
+      "loss": 0.0528,
+      "step": 6948
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.27294417207788796,
+      "learning_rate": 3.8683982478202356e-06,
+      "loss": 0.0785,
+      "step": 6949
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2626738163970261,
+      "learning_rate": 3.867069092903908e-06,
+      "loss": 0.0627,
+      "step": 6950
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2153282411907624,
+      "learning_rate": 3.8657400223748746e-06,
+      "loss": 0.0641,
+      "step": 6951
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2517636050579485,
+      "learning_rate": 3.8644110363321305e-06,
+      "loss": 0.0681,
+      "step": 6952
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.26457393409906205,
+      "learning_rate": 3.863082134874669e-06,
+      "loss": 0.0715,
+      "step": 6953
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3022005503802715,
+      "learning_rate": 3.861753318101473e-06,
+      "loss": 0.1287,
+      "step": 6954
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2863978156496446,
+      "learning_rate": 3.86042458611152e-06,
+      "loss": 0.0975,
+      "step": 6955
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.325104885195721,
+      "learning_rate": 3.859095939003782e-06,
+      "loss": 0.085,
+      "step": 6956
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.31063211095875765,
+      "learning_rate": 3.857767376877226e-06,
+      "loss": 0.0492,
+      "step": 6957
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3137533087387582,
+      "learning_rate": 3.85643889983081e-06,
+      "loss": 0.0799,
+      "step": 6958
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.25382124918736076,
+      "learning_rate": 3.855110507963484e-06,
+      "loss": 0.0537,
+      "step": 6959
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3075078301860931,
+      "learning_rate": 3.853782201374198e-06,
+      "loss": 0.0706,
+      "step": 6960
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.19160549305484859,
+      "learning_rate": 3.852453980161893e-06,
+      "loss": 0.0641,
+      "step": 6961
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4451107614333401,
+      "learning_rate": 3.851125844425499e-06,
+      "loss": 0.0676,
+      "step": 6962
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2155923247640567,
+      "learning_rate": 3.849797794263943e-06,
+      "loss": 0.0568,
+      "step": 6963
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3975958788434452,
+      "learning_rate": 3.84846982977615e-06,
+      "loss": 0.1189,
+      "step": 6964
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5710769646276664,
+      "learning_rate": 3.847141951061032e-06,
+      "loss": 0.1152,
+      "step": 6965
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4692006955601725,
+      "learning_rate": 3.845814158217495e-06,
+      "loss": 0.0904,
+      "step": 6966
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.43762383087876533,
+      "learning_rate": 3.844486451344445e-06,
+      "loss": 0.0853,
+      "step": 6967
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3706605262386526,
+      "learning_rate": 3.843158830540777e-06,
+      "loss": 0.0963,
+      "step": 6968
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4985092999289467,
+      "learning_rate": 3.841831295905377e-06,
+      "loss": 0.1026,
+      "step": 6969
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.17626772584460504,
+      "learning_rate": 3.840503847537127e-06,
+      "loss": 0.0348,
+      "step": 6970
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.33472016390239234,
+      "learning_rate": 3.839176485534908e-06,
+      "loss": 0.0805,
+      "step": 6971
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4074588277535488,
+      "learning_rate": 3.837849209997586e-06,
+      "loss": 0.1013,
+      "step": 6972
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3219608630924946,
+      "learning_rate": 3.8365220210240225e-06,
+      "loss": 0.0837,
+      "step": 6973
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.24337186896112342,
+      "learning_rate": 3.8351949187130795e-06,
+      "loss": 0.0821,
+      "step": 6974
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6445673547171886,
+      "learning_rate": 3.8338679031636035e-06,
+      "loss": 0.0894,
+      "step": 6975
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.39257176043469416,
+      "learning_rate": 3.832540974474438e-06,
+      "loss": 0.1037,
+      "step": 6976
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.33674127779434965,
+      "learning_rate": 3.831214132744423e-06,
+      "loss": 0.0524,
+      "step": 6977
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2261301318414963,
+      "learning_rate": 3.829887378072389e-06,
+      "loss": 0.0588,
+      "step": 6978
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2155514377233801,
+      "learning_rate": 3.828560710557159e-06,
+      "loss": 0.0874,
+      "step": 6979
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3549295515673941,
+      "learning_rate": 3.8272341302975494e-06,
+      "loss": 0.0947,
+      "step": 6980
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4557513749274151,
+      "learning_rate": 3.825907637392375e-06,
+      "loss": 0.1036,
+      "step": 6981
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2870464472742229,
+      "learning_rate": 3.82458123194044e-06,
+      "loss": 0.043,
+      "step": 6982
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4886983894195904,
+      "learning_rate": 3.8232549140405416e-06,
+      "loss": 0.1041,
+      "step": 6983
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6183753344049833,
+      "learning_rate": 3.8219286837914716e-06,
+      "loss": 0.119,
+      "step": 6984
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.31722070462520496,
+      "learning_rate": 3.820602541292017e-06,
+      "loss": 0.0972,
+      "step": 6985
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2994385694468029,
+      "learning_rate": 3.819276486640956e-06,
+      "loss": 0.0953,
+      "step": 6986
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5208278610938926,
+      "learning_rate": 3.817950519937059e-06,
+      "loss": 0.0938,
+      "step": 6987
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.39204515448357136,
+      "learning_rate": 3.816624641279096e-06,
+      "loss": 0.0947,
+      "step": 6988
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2846704213515172,
+      "learning_rate": 3.81529885076582e-06,
+      "loss": 0.0706,
+      "step": 6989
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.32309530962331634,
+      "learning_rate": 3.813973148495988e-06,
+      "loss": 0.0953,
+      "step": 6990
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4249505702345002,
+      "learning_rate": 3.8126475345683455e-06,
+      "loss": 0.0768,
+      "step": 6991
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3563129260604548,
+      "learning_rate": 3.8113220090816323e-06,
+      "loss": 0.0864,
+      "step": 6992
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.28453006800038677,
+      "learning_rate": 3.8099965721345777e-06,
+      "loss": 0.0733,
+      "step": 6993
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.21862895412323283,
+      "learning_rate": 3.8086712238259134e-06,
+      "loss": 0.0456,
+      "step": 6994
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3060886413087305,
+      "learning_rate": 3.8073459642543558e-06,
+      "loss": 0.07,
+      "step": 6995
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.28349790840860656,
+      "learning_rate": 3.806020793518619e-06,
+      "loss": 0.0835,
+      "step": 6996
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4748548267099864,
+      "learning_rate": 3.8046957117174064e-06,
+      "loss": 0.1155,
+      "step": 6997
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.30080223676320494,
+      "learning_rate": 3.803370718949423e-06,
+      "loss": 0.0721,
+      "step": 6998
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.20757776942535627,
+      "learning_rate": 3.8020458153133586e-06,
+      "loss": 0.0395,
+      "step": 6999
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.29913997673728926,
+      "learning_rate": 3.800721000907899e-06,
+      "loss": 0.0624,
+      "step": 7000
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.25747983312633915,
+      "learning_rate": 3.799396275831727e-06,
+      "loss": 0.1024,
+      "step": 7001
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.24140307001140046,
+      "learning_rate": 3.798071640183515e-06,
+      "loss": 0.0554,
+      "step": 7002
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.26058474617581845,
+      "learning_rate": 3.796747094061929e-06,
+      "loss": 0.0755,
+      "step": 7003
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2793340283786342,
+      "learning_rate": 3.795422637565626e-06,
+      "loss": 0.0667,
+      "step": 7004
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.25598930970923833,
+      "learning_rate": 3.794098270793265e-06,
+      "loss": 0.0477,
+      "step": 7005
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2547974546742003,
+      "learning_rate": 3.7927739938434893e-06,
+      "loss": 0.0699,
+      "step": 7006
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3968049901002055,
+      "learning_rate": 3.7914498068149375e-06,
+      "loss": 0.1156,
+      "step": 7007
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3243338367429409,
+      "learning_rate": 3.790125709806246e-06,
+      "loss": 0.1036,
+      "step": 7008
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4550230170378503,
+      "learning_rate": 3.78880170291604e-06,
+      "loss": 0.1179,
+      "step": 7009
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2832282661186303,
+      "learning_rate": 3.787477786242937e-06,
+      "loss": 0.0788,
+      "step": 7010
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8166012434149398,
+      "learning_rate": 3.7861539598855533e-06,
+      "loss": 0.0881,
+      "step": 7011
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4151586992920546,
+      "learning_rate": 3.784830223942495e-06,
+      "loss": 0.0984,
+      "step": 7012
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.49025791287037995,
+      "learning_rate": 3.78350657851236e-06,
+      "loss": 0.0746,
+      "step": 7013
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3023400108625917,
+      "learning_rate": 3.782183023693739e-06,
+      "loss": 0.0964,
+      "step": 7014
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.37215096417704824,
+      "learning_rate": 3.780859559585223e-06,
+      "loss": 0.0783,
+      "step": 7015
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.35959119478526436,
+      "learning_rate": 3.7795361862853897e-06,
+      "loss": 0.12,
+      "step": 7016
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.17353135646058698,
+      "learning_rate": 3.7782129038928096e-06,
+      "loss": 0.039,
+      "step": 7017
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3013421968143348,
+      "learning_rate": 3.7768897125060494e-06,
+      "loss": 0.1069,
+      "step": 7018
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3226532495918191,
+      "learning_rate": 3.77556661222367e-06,
+      "loss": 0.0762,
+      "step": 7019
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3215069317216213,
+      "learning_rate": 3.774243603144222e-06,
+      "loss": 0.1008,
+      "step": 7020
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4862797368380538,
+      "learning_rate": 3.7729206853662497e-06,
+      "loss": 0.0659,
+      "step": 7021
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3655197820433912,
+      "learning_rate": 3.7715978589882944e-06,
+      "loss": 0.086,
+      "step": 7022
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5036309981146027,
+      "learning_rate": 3.7702751241088858e-06,
+      "loss": 0.0991,
+      "step": 7023
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3497530324309186,
+      "learning_rate": 3.7689524808265494e-06,
+      "loss": 0.1037,
+      "step": 7024
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2883804973343093,
+      "learning_rate": 3.7676299292398045e-06,
+      "loss": 0.0821,
+      "step": 7025
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5536771259550994,
+      "learning_rate": 3.766307469447161e-06,
+      "loss": 0.1333,
+      "step": 7026
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.30288012043242024,
+      "learning_rate": 3.7649851015471227e-06,
+      "loss": 0.0683,
+      "step": 7027
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.320755536701085,
+      "learning_rate": 3.763662825638191e-06,
+      "loss": 0.0768,
+      "step": 7028
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.250016506061455,
+      "learning_rate": 3.7623406418188533e-06,
+      "loss": 0.0544,
+      "step": 7029
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.23024196427022733,
+      "learning_rate": 3.761018550187595e-06,
+      "loss": 0.0897,
+      "step": 7030
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.37991801233326405,
+      "learning_rate": 3.7596965508428907e-06,
+      "loss": 0.0825,
+      "step": 7031
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5622701616598457,
+      "learning_rate": 3.7583746438832143e-06,
+      "loss": 0.0687,
+      "step": 7032
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4121675383586414,
+      "learning_rate": 3.7570528294070276e-06,
+      "loss": 0.1009,
+      "step": 7033
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.19817967193002944,
+      "learning_rate": 3.7557311075127846e-06,
+      "loss": 0.0498,
+      "step": 7034
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.33962798292450097,
+      "learning_rate": 3.7544094782989394e-06,
+      "loss": 0.0842,
+      "step": 7035
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.1967848655131811,
+      "learning_rate": 3.753087941863932e-06,
+      "loss": 0.0488,
+      "step": 7036
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.32763062678757426,
+      "learning_rate": 3.751766498306199e-06,
+      "loss": 0.1026,
+      "step": 7037
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.24105474449013065,
+      "learning_rate": 3.7504451477241667e-06,
+      "loss": 0.0736,
+      "step": 7038
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2016926848858177,
+      "learning_rate": 3.749123890216261e-06,
+      "loss": 0.0543,
+      "step": 7039
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.36091709707264324,
+      "learning_rate": 3.747802725880894e-06,
+      "loss": 0.1136,
+      "step": 7040
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.14312194905759096,
+      "learning_rate": 3.746481654816473e-06,
+      "loss": 0.0273,
+      "step": 7041
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4841918383180562,
+      "learning_rate": 3.7451606771214032e-06,
+      "loss": 0.0562,
+      "step": 7042
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6499918442721546,
+      "learning_rate": 3.7438397928940754e-06,
+      "loss": 0.1426,
+      "step": 7043
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3049548740958857,
+      "learning_rate": 3.7425190022328763e-06,
+      "loss": 0.0972,
+      "step": 7044
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.20533678197582406,
+      "learning_rate": 3.7411983052361887e-06,
+      "loss": 0.0606,
+      "step": 7045
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.27686810093525666,
+      "learning_rate": 3.7398777020023846e-06,
+      "loss": 0.0443,
+      "step": 7046
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.19279331839169905,
+      "learning_rate": 3.73855719262983e-06,
+      "loss": 0.0302,
+      "step": 7047
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.49075573565532965,
+      "learning_rate": 3.737236777216882e-06,
+      "loss": 0.1017,
+      "step": 7048
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5493096636993589,
+      "learning_rate": 3.7359164558618964e-06,
+      "loss": 0.1101,
+      "step": 7049
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3773128878578621,
+      "learning_rate": 3.734596228663218e-06,
+      "loss": 0.0894,
+      "step": 7050
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5139926133037153,
+      "learning_rate": 3.7332760957191828e-06,
+      "loss": 0.0371,
+      "step": 7051
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4056280877174042,
+      "learning_rate": 3.7319560571281232e-06,
+      "loss": 0.0779,
+      "step": 7052
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2956182499980726,
+      "learning_rate": 3.730636112988364e-06,
+      "loss": 0.0695,
+      "step": 7053
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.29453639301224877,
+      "learning_rate": 3.7293162633982218e-06,
+      "loss": 0.081,
+      "step": 7054
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4277574571612674,
+      "learning_rate": 3.7279965084560054e-06,
+      "loss": 0.0763,
+      "step": 7055
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5507344968431241,
+      "learning_rate": 3.7266768482600206e-06,
+      "loss": 0.1561,
+      "step": 7056
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.29962664683835755,
+      "learning_rate": 3.7253572829085604e-06,
+      "loss": 0.0989,
+      "step": 7057
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.40749641174583323,
+      "learning_rate": 3.7240378124999157e-06,
+      "loss": 0.0905,
+      "step": 7058
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.35887863029084166,
+      "learning_rate": 3.7227184371323678e-06,
+      "loss": 0.0859,
+      "step": 7059
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.22793340982692317,
+      "learning_rate": 3.721399156904192e-06,
+      "loss": 0.0865,
+      "step": 7060
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.44455990637492027,
+      "learning_rate": 3.720079971913654e-06,
+      "loss": 0.123,
+      "step": 7061
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.19123378058978224,
+      "learning_rate": 3.7187608822590164e-06,
+      "loss": 0.0594,
+      "step": 7062
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4022764974218865,
+      "learning_rate": 3.7174418880385333e-06,
+      "loss": 0.0918,
+      "step": 7063
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.22296929274764535,
+      "learning_rate": 3.716122989350449e-06,
+      "loss": 0.0364,
+      "step": 7064
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4068852381796377,
+      "learning_rate": 3.714804186293002e-06,
+      "loss": 0.0867,
+      "step": 7065
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.20005847141779912,
+      "learning_rate": 3.7134854789644276e-06,
+      "loss": 0.0592,
+      "step": 7066
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2120790917461066,
+      "learning_rate": 3.7121668674629497e-06,
+      "loss": 0.03,
+      "step": 7067
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5852540904333556,
+      "learning_rate": 3.710848351886783e-06,
+      "loss": 0.1213,
+      "step": 7068
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.38404801998372473,
+      "learning_rate": 3.709529932334144e-06,
+      "loss": 0.1191,
+      "step": 7069
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.32121427795862706,
+      "learning_rate": 3.708211608903233e-06,
+      "loss": 0.0801,
+      "step": 7070
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.31588263187885507,
+      "learning_rate": 3.7068933816922458e-06,
+      "loss": 0.061,
+      "step": 7071
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.21635739226761505,
+      "learning_rate": 3.7055752507993703e-06,
+      "loss": 0.0645,
+      "step": 7072
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3935744420770631,
+      "learning_rate": 3.7042572163227937e-06,
+      "loss": 0.1009,
+      "step": 7073
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3416488109587413,
+      "learning_rate": 3.702939278360687e-06,
+      "loss": 0.0668,
+      "step": 7074
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.28693921237537107,
+      "learning_rate": 3.7016214370112173e-06,
+      "loss": 0.0516,
+      "step": 7075
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.23247866998050992,
+      "learning_rate": 3.700303692372549e-06,
+      "loss": 0.0757,
+      "step": 7076
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3260319601780935,
+      "learning_rate": 3.6989860445428323e-06,
+      "loss": 0.0991,
+      "step": 7077
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.23245566701996806,
+      "learning_rate": 3.6976684936202116e-06,
+      "loss": 0.046,
+      "step": 7078
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.33818073060730136,
+      "learning_rate": 3.6963510397028314e-06,
+      "loss": 0.0852,
+      "step": 7079
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.23990985349862584,
+      "learning_rate": 3.695033682888819e-06,
+      "loss": 0.0488,
+      "step": 7080
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.19898496092289647,
+      "learning_rate": 3.6937164232763e-06,
+      "loss": 0.0655,
+      "step": 7081
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.334187534008223,
+      "learning_rate": 3.6923992609633897e-06,
+      "loss": 0.0913,
+      "step": 7082
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.34328319424172987,
+      "learning_rate": 3.691082196048202e-06,
+      "loss": 0.0676,
+      "step": 7083
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.47413862018557135,
+      "learning_rate": 3.689765228628837e-06,
+      "loss": 0.1294,
+      "step": 7084
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3251873131766871,
+      "learning_rate": 3.68844835880339e-06,
+      "loss": 0.0795,
+      "step": 7085
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3130937656927123,
+      "learning_rate": 3.6871315866699487e-06,
+      "loss": 0.0886,
+      "step": 7086
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.22048519587811288,
+      "learning_rate": 3.6858149123265962e-06,
+      "loss": 0.0592,
+      "step": 7087
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2401010322221783,
+      "learning_rate": 3.6844983358714047e-06,
+      "loss": 0.0629,
+      "step": 7088
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.29585843772610343,
+      "learning_rate": 3.6831818574024405e-06,
+      "loss": 0.0791,
+      "step": 7089
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.44718787829652257,
+      "learning_rate": 3.6818654770177626e-06,
+      "loss": 0.0872,
+      "step": 7090
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.24341815348838447,
+      "learning_rate": 3.6805491948154214e-06,
+      "loss": 0.0748,
+      "step": 7091
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3054959231265354,
+      "learning_rate": 3.679233010893463e-06,
+      "loss": 0.0886,
+      "step": 7092
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.277906863215993,
+      "learning_rate": 3.677916925349925e-06,
+      "loss": 0.0851,
+      "step": 7093
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2865706191090071,
+      "learning_rate": 3.676600938282836e-06,
+      "loss": 0.0603,
+      "step": 7094
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.34437711121338116,
+      "learning_rate": 3.6752850497902174e-06,
+      "loss": 0.0672,
+      "step": 7095
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4460454863222348,
+      "learning_rate": 3.673969259970086e-06,
+      "loss": 0.0942,
+      "step": 7096
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5874193689607091,
+      "learning_rate": 3.672653568920449e-06,
+      "loss": 0.1072,
+      "step": 7097
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3286084095064521,
+      "learning_rate": 3.671337976739307e-06,
+      "loss": 0.0953,
+      "step": 7098
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.602485039120254,
+      "learning_rate": 3.67002248352465e-06,
+      "loss": 0.1131,
+      "step": 7099
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3137692093687027,
+      "learning_rate": 3.6687070893744685e-06,
+      "loss": 0.1204,
+      "step": 7100
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.39656881011082407,
+      "learning_rate": 3.6673917943867386e-06,
+      "loss": 0.0951,
+      "step": 7101
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.26946064492873245,
+      "learning_rate": 3.6660765986594282e-06,
+      "loss": 0.0761,
+      "step": 7102
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.27844623347034936,
+      "learning_rate": 3.664761502290506e-06,
+      "loss": 0.094,
+      "step": 7103
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.47889147465179843,
+      "learning_rate": 3.6634465053779257e-06,
+      "loss": 0.0938,
+      "step": 7104
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.27852307460161446,
+      "learning_rate": 3.662131608019635e-06,
+      "loss": 0.0901,
+      "step": 7105
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.42453675918345757,
+      "learning_rate": 3.6608168103135737e-06,
+      "loss": 0.1019,
+      "step": 7106
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.21032317166867096,
+      "learning_rate": 3.6595021123576803e-06,
+      "loss": 0.0685,
+      "step": 7107
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.32767050823324617,
+      "learning_rate": 3.658187514249878e-06,
+      "loss": 0.0906,
+      "step": 7108
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.296831715977161,
+      "learning_rate": 3.656873016088085e-06,
+      "loss": 0.0788,
+      "step": 7109
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.32933418030133305,
+      "learning_rate": 3.655558617970215e-06,
+      "loss": 0.0853,
+      "step": 7110
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.271136525418101,
+      "learning_rate": 3.654244319994172e-06,
+      "loss": 0.0702,
+      "step": 7111
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2462001482687036,
+      "learning_rate": 3.6529301222578517e-06,
+      "loss": 0.0437,
+      "step": 7112
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3594241603997126,
+      "learning_rate": 3.6516160248591404e-06,
+      "loss": 0.0631,
+      "step": 7113
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.46842087785984116,
+      "learning_rate": 3.6503020278959244e-06,
+      "loss": 0.0755,
+      "step": 7114
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.45674193330505825,
+      "learning_rate": 3.648988131466077e-06,
+      "loss": 0.1104,
+      "step": 7115
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.344206651257694,
+      "learning_rate": 3.647674335667461e-06,
+      "loss": 0.1088,
+      "step": 7116
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3357409102647396,
+      "learning_rate": 3.6463606405979403e-06,
+      "loss": 0.1002,
+      "step": 7117
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2690026049259225,
+      "learning_rate": 3.645047046355365e-06,
+      "loss": 0.069,
+      "step": 7118
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.30496112825872285,
+      "learning_rate": 3.6437335530375778e-06,
+      "loss": 0.1042,
+      "step": 7119
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.20285205233021208,
+      "learning_rate": 3.6424201607424163e-06,
+      "loss": 0.0448,
+      "step": 7120
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.23580418383731042,
+      "learning_rate": 3.641106869567711e-06,
+      "loss": 0.0625,
+      "step": 7121
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2947611114259498,
+      "learning_rate": 3.639793679611282e-06,
+      "loss": 0.0822,
+      "step": 7122
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.29643890819337687,
+      "learning_rate": 3.638480590970942e-06,
+      "loss": 0.0825,
+      "step": 7123
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3693671498326471,
+      "learning_rate": 3.637167603744501e-06,
+      "loss": 0.0886,
+      "step": 7124
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.44237181526538577,
+      "learning_rate": 3.635854718029754e-06,
+      "loss": 0.0906,
+      "step": 7125
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3420100536857426,
+      "learning_rate": 3.6345419339244946e-06,
+      "loss": 0.0714,
+      "step": 7126
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3103941523748551,
+      "learning_rate": 3.6332292515265078e-06,
+      "loss": 0.1064,
+      "step": 7127
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4541477247644959,
+      "learning_rate": 3.6319166709335684e-06,
+      "loss": 0.0994,
+      "step": 7128
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3192004727331564,
+      "learning_rate": 3.6306041922434436e-06,
+      "loss": 0.0851,
+      "step": 7129
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.851652154574873,
+      "learning_rate": 3.629291815553896e-06,
+      "loss": 0.126,
+      "step": 7130
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3196016860985171,
+      "learning_rate": 3.62797954096268e-06,
+      "loss": 0.083,
+      "step": 7131
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3360810262656064,
+      "learning_rate": 3.62666736856754e-06,
+      "loss": 0.0769,
+      "step": 7132
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.32071694812099405,
+      "learning_rate": 3.625355298466213e-06,
+      "loss": 0.0977,
+      "step": 7133
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2502376544929433,
+      "learning_rate": 3.6240433307564337e-06,
+      "loss": 0.0789,
+      "step": 7134
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4073455920283119,
+      "learning_rate": 3.6227314655359226e-06,
+      "loss": 0.0812,
+      "step": 7135
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2160006907522294,
+      "learning_rate": 3.6214197029023923e-06,
+      "loss": 0.0559,
+      "step": 7136
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.27085113965736396,
+      "learning_rate": 3.6201080429535563e-06,
+      "loss": 0.105,
+      "step": 7137
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.41180008480737884,
+      "learning_rate": 3.6187964857871107e-06,
+      "loss": 0.1044,
+      "step": 7138
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.36068851672146485,
+      "learning_rate": 3.61748503150075e-06,
+      "loss": 0.0792,
+      "step": 7139
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.20841773566430538,
+      "learning_rate": 3.6161736801921554e-06,
+      "loss": 0.0702,
+      "step": 7140
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.27459567741339497,
+      "learning_rate": 3.6148624319590083e-06,
+      "loss": 0.0618,
+      "step": 7141
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3977736107589453,
+      "learning_rate": 3.6135512868989765e-06,
+      "loss": 0.0973,
+      "step": 7142
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.420698169307046,
+      "learning_rate": 3.6122402451097192e-06,
+      "loss": 0.0862,
+      "step": 7143
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3406006623975837,
+      "learning_rate": 3.610929306688895e-06,
+      "loss": 0.0885,
+      "step": 7144
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.28569623738410815,
+      "learning_rate": 3.6096184717341483e-06,
+      "loss": 0.0759,
+      "step": 7145
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2618131131678179,
+      "learning_rate": 3.6083077403431165e-06,
+      "loss": 0.0681,
+      "step": 7146
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3710443493857251,
+      "learning_rate": 3.6069971126134306e-06,
+      "loss": 0.0626,
+      "step": 7147
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.41453698327334637,
+      "learning_rate": 3.605686588642716e-06,
+      "loss": 0.124,
+      "step": 7148
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.16193254323420794,
+      "learning_rate": 3.604376168528587e-06,
+      "loss": 0.0345,
+      "step": 7149
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2696876898211043,
+      "learning_rate": 3.603065852368648e-06,
+      "loss": 0.0773,
+      "step": 7150
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.24805561066795712,
+      "learning_rate": 3.6017556402605054e-06,
+      "loss": 0.0617,
+      "step": 7151
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.25096618503627316,
+      "learning_rate": 3.6004455323017473e-06,
+      "loss": 0.048,
+      "step": 7152
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2558511966307502,
+      "learning_rate": 3.599135528589958e-06,
+      "loss": 0.0824,
+      "step": 7153
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2836724332371867,
+      "learning_rate": 3.5978256292227152e-06,
+      "loss": 0.06,
+      "step": 7154
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4096538406473501,
+      "learning_rate": 3.596515834297589e-06,
+      "loss": 0.0924,
+      "step": 7155
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5103851424960107,
+      "learning_rate": 3.59520614391214e-06,
+      "loss": 0.0988,
+      "step": 7156
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3657133228520561,
+      "learning_rate": 3.5938965581639195e-06,
+      "loss": 0.0661,
+      "step": 7157
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3912204006862977,
+      "learning_rate": 3.592587077150475e-06,
+      "loss": 0.1171,
+      "step": 7158
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2884624306791856,
+      "learning_rate": 3.5912777009693435e-06,
+      "loss": 0.0773,
+      "step": 7159
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5088823110812557,
+      "learning_rate": 3.589968429718055e-06,
+      "loss": 0.1177,
+      "step": 7160
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3899696117750577,
+      "learning_rate": 3.588659263494132e-06,
+      "loss": 0.0824,
+      "step": 7161
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.21575436392551625,
+      "learning_rate": 3.5873502023950898e-06,
+      "loss": 0.0577,
+      "step": 7162
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.38425591725953806,
+      "learning_rate": 3.5860412465184324e-06,
+      "loss": 0.0944,
+      "step": 7163
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.32378828785144875,
+      "learning_rate": 3.58473239596166e-06,
+      "loss": 0.0919,
+      "step": 7164
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.26621206431989625,
+      "learning_rate": 3.5834236508222643e-06,
+      "loss": 0.0888,
+      "step": 7165
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4117143748668806,
+      "learning_rate": 3.582115011197727e-06,
+      "loss": 0.1014,
+      "step": 7166
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.17836120445202577,
+      "learning_rate": 3.5808064771855216e-06,
+      "loss": 0.051,
+      "step": 7167
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.18995216725015457,
+      "learning_rate": 3.5794980488831194e-06,
+      "loss": 0.0615,
+      "step": 7168
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7707197645463754,
+      "learning_rate": 3.578189726387977e-06,
+      "loss": 0.0827,
+      "step": 7169
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2828497960713177,
+      "learning_rate": 3.5768815097975452e-06,
+      "loss": 0.0933,
+      "step": 7170
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.35325781465784833,
+      "learning_rate": 3.57557339920927e-06,
+      "loss": 0.0535,
+      "step": 7171
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.25975126129292264,
+      "learning_rate": 3.5742653947205865e-06,
+      "loss": 0.063,
+      "step": 7172
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2939556017225196,
+      "learning_rate": 3.572957496428922e-06,
+      "loss": 0.073,
+      "step": 7173
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3376509869795295,
+      "learning_rate": 3.5716497044316943e-06,
+      "loss": 0.0924,
+      "step": 7174
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2495971077391261,
+      "learning_rate": 3.5703420188263193e-06,
+      "loss": 0.0531,
+      "step": 7175
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7372077295233995,
+      "learning_rate": 3.5690344397101983e-06,
+      "loss": 0.1145,
+      "step": 7176
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.306580763620131,
+      "learning_rate": 3.567726967180727e-06,
+      "loss": 0.0717,
+      "step": 7177
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.34639346707495006,
+      "learning_rate": 3.566419601335297e-06,
+      "loss": 0.069,
+      "step": 7178
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2794510385541492,
+      "learning_rate": 3.5651123422712865e-06,
+      "loss": 0.0677,
+      "step": 7179
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.27498673063775847,
+      "learning_rate": 3.563805190086067e-06,
+      "loss": 0.086,
+      "step": 7180
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.24367884563086192,
+      "learning_rate": 3.5624981448770012e-06,
+      "loss": 0.072,
+      "step": 7181
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.26630958905056645,
+      "learning_rate": 3.56119120674145e-06,
+      "loss": 0.0846,
+      "step": 7182
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.310386756000949,
+      "learning_rate": 3.5598843757767595e-06,
+      "loss": 0.0578,
+      "step": 7183
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.41210458038505293,
+      "learning_rate": 3.5585776520802673e-06,
+      "loss": 0.1029,
+      "step": 7184
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3503584387298873,
+      "learning_rate": 3.557271035749311e-06,
+      "loss": 0.0519,
+      "step": 7185
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.24825980923834484,
+      "learning_rate": 3.5559645268812126e-06,
+      "loss": 0.0833,
+      "step": 7186
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.38934650203515087,
+      "learning_rate": 3.5546581255732872e-06,
+      "loss": 0.0953,
+      "step": 7187
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2910646303609189,
+      "learning_rate": 3.5533518319228445e-06,
+      "loss": 0.0938,
+      "step": 7188
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2859802459032237,
+      "learning_rate": 3.5520456460271858e-06,
+      "loss": 0.0708,
+      "step": 7189
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.24187029811429853,
+      "learning_rate": 3.550739567983602e-06,
+      "loss": 0.0553,
+      "step": 7190
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.27932892510458834,
+      "learning_rate": 3.5494335978893766e-06,
+      "loss": 0.058,
+      "step": 7191
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3030468633946572,
+      "learning_rate": 3.548127735841789e-06,
+      "loss": 0.0542,
+      "step": 7192
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.510280397826606,
+      "learning_rate": 3.5468219819381035e-06,
+      "loss": 0.0868,
+      "step": 7193
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2569317369229568,
+      "learning_rate": 3.545516336275582e-06,
+      "loss": 0.0614,
+      "step": 7194
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1593744124521028,
+      "learning_rate": 3.5442107989514783e-06,
+      "loss": 0.183,
+      "step": 7195
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6709256031053239,
+      "learning_rate": 3.5429053700630343e-06,
+      "loss": 0.0887,
+      "step": 7196
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2781811730022812,
+      "learning_rate": 3.5416000497074866e-06,
+      "loss": 0.0879,
+      "step": 7197
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.616379470887631,
+      "learning_rate": 3.5402948379820623e-06,
+      "loss": 0.0972,
+      "step": 7198
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3660574288429818,
+      "learning_rate": 3.538989734983983e-06,
+      "loss": 0.0858,
+      "step": 7199
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.26300874629795823,
+      "learning_rate": 3.537684740810459e-06,
+      "loss": 0.0662,
+      "step": 7200
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.23966330082773055,
+      "learning_rate": 3.5363798555586924e-06,
+      "loss": 0.0569,
+      "step": 7201
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3735382078055617,
+      "learning_rate": 3.5350750793258825e-06,
+      "loss": 0.0989,
+      "step": 7202
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.36859260403936434,
+      "learning_rate": 3.5337704122092144e-06,
+      "loss": 0.1041,
+      "step": 7203
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.24370026362096808,
+      "learning_rate": 3.532465854305867e-06,
+      "loss": 0.0866,
+      "step": 7204
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.38199543199599034,
+      "learning_rate": 3.53116140571301e-06,
+      "loss": 0.1313,
+      "step": 7205
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.38612073418466125,
+      "learning_rate": 3.5298570665278106e-06,
+      "loss": 0.0952,
+      "step": 7206
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4394731385307924,
+      "learning_rate": 3.528552836847421e-06,
+      "loss": 0.106,
+      "step": 7207
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3529553057936811,
+      "learning_rate": 3.527248716768985e-06,
+      "loss": 0.0782,
+      "step": 7208
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4342915743036655,
+      "learning_rate": 3.5259447063896468e-06,
+      "loss": 0.1164,
+      "step": 7209
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3296869251887252,
+      "learning_rate": 3.524640805806534e-06,
+      "loss": 0.05,
+      "step": 7210
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.19725458332355078,
+      "learning_rate": 3.5233370151167664e-06,
+      "loss": 0.0493,
+      "step": 7211
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3578378313893545,
+      "learning_rate": 3.522033334417462e-06,
+      "loss": 0.0802,
+      "step": 7212
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3629770845891983,
+      "learning_rate": 3.520729763805725e-06,
+      "loss": 0.1133,
+      "step": 7213
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.26450723990271996,
+      "learning_rate": 3.519426303378652e-06,
+      "loss": 0.0684,
+      "step": 7214
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.36609275708141564,
+      "learning_rate": 3.5181229532333304e-06,
+      "loss": 0.0967,
+      "step": 7215
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.21398705665598083,
+      "learning_rate": 3.516819713466847e-06,
+      "loss": 0.0496,
+      "step": 7216
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6343095212699651,
+      "learning_rate": 3.515516584176271e-06,
+      "loss": 0.1404,
+      "step": 7217
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3184382668078663,
+      "learning_rate": 3.5142135654586655e-06,
+      "loss": 0.0748,
+      "step": 7218
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.41082450263607484,
+      "learning_rate": 3.5129106574110916e-06,
+      "loss": 0.0866,
+      "step": 7219
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.20336614906203018,
+      "learning_rate": 3.5116078601305944e-06,
+      "loss": 0.07,
+      "step": 7220
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5156052422853074,
+      "learning_rate": 3.510305173714214e-06,
+      "loss": 0.1082,
+      "step": 7221
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3787695549469417,
+      "learning_rate": 3.5090025982589804e-06,
+      "loss": 0.0957,
+      "step": 7222
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3821850536403129,
+      "learning_rate": 3.507700133861922e-06,
+      "loss": 0.1093,
+      "step": 7223
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3570362528167716,
+      "learning_rate": 3.50639778062005e-06,
+      "loss": 0.1035,
+      "step": 7224
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0692125173788571,
+      "learning_rate": 3.505095538630371e-06,
+      "loss": 0.1579,
+      "step": 7225
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.33028632913736616,
+      "learning_rate": 3.503793407989886e-06,
+      "loss": 0.0543,
+      "step": 7226
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2511899919211044,
+      "learning_rate": 3.5024913887955838e-06,
+      "loss": 0.0647,
+      "step": 7227
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.498909533634594,
+      "learning_rate": 3.5011894811444457e-06,
+      "loss": 0.094,
+      "step": 7228
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.30543285911720114,
+      "learning_rate": 3.4998876851334473e-06,
+      "loss": 0.0398,
+      "step": 7229
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.27309554345108344,
+      "learning_rate": 3.498586000859553e-06,
+      "loss": 0.0597,
+      "step": 7230
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4788962021688593,
+      "learning_rate": 3.4972844284197194e-06,
+      "loss": 0.1755,
+      "step": 7231
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.33577069238052243,
+      "learning_rate": 3.495982967910895e-06,
+      "loss": 0.0862,
+      "step": 7232
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.45041548434189355,
+      "learning_rate": 3.4946816194300214e-06,
+      "loss": 0.0927,
+      "step": 7233
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.23446000809126838,
+      "learning_rate": 3.49338038307403e-06,
+      "loss": 0.043,
+      "step": 7234
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5058923573069056,
+      "learning_rate": 3.4920792589398423e-06,
+      "loss": 0.0563,
+      "step": 7235
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.34508190081391704,
+      "learning_rate": 3.4907782471243784e-06,
+      "loss": 0.0756,
+      "step": 7236
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2168039682729426,
+      "learning_rate": 3.489477347724542e-06,
+      "loss": 0.0693,
+      "step": 7237
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8416626268546971,
+      "learning_rate": 3.488176560837232e-06,
+      "loss": 0.1531,
+      "step": 7238
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.35826822003421627,
+      "learning_rate": 3.4868758865593374e-06,
+      "loss": 0.07,
+      "step": 7239
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.27368161822874465,
+      "learning_rate": 3.485575324987743e-06,
+      "loss": 0.0795,
+      "step": 7240
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2799987437948479,
+      "learning_rate": 3.4842748762193203e-06,
+      "loss": 0.0799,
+      "step": 7241
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.42396320867636134,
+      "learning_rate": 3.482974540350933e-06,
+      "loss": 0.0908,
+      "step": 7242
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3893674761808773,
+      "learning_rate": 3.481674317479441e-06,
+      "loss": 0.104,
+      "step": 7243
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.30654440707673486,
+      "learning_rate": 3.4803742077016913e-06,
+      "loss": 0.0814,
+      "step": 7244
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.32677813121857013,
+      "learning_rate": 3.4790742111145206e-06,
+      "loss": 0.0684,
+      "step": 7245
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.25611821595628015,
+      "learning_rate": 3.4777743278147645e-06,
+      "loss": 0.0664,
+      "step": 7246
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3176130736017971,
+      "learning_rate": 3.476474557899244e-06,
+      "loss": 0.0963,
+      "step": 7247
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2528567963636341,
+      "learning_rate": 3.475174901464774e-06,
+      "loss": 0.0832,
+      "step": 7248
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5049130924403095,
+      "learning_rate": 3.473875358608158e-06,
+      "loss": 0.0704,
+      "step": 7249
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2666556149932916,
+      "learning_rate": 3.472575929426197e-06,
+      "loss": 0.0729,
+      "step": 7250
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6513495967114249,
+      "learning_rate": 3.47127661401568e-06,
+      "loss": 0.1104,
+      "step": 7251
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5802113438065355,
+      "learning_rate": 3.469977412473383e-06,
+      "loss": 0.102,
+      "step": 7252
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.24800736729906492,
+      "learning_rate": 3.4686783248960843e-06,
+      "loss": 0.0675,
+      "step": 7253
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2271687024284112,
+      "learning_rate": 3.467379351380544e-06,
+      "loss": 0.0666,
+      "step": 7254
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.21486998138012595,
+      "learning_rate": 3.4660804920235184e-06,
+      "loss": 0.0903,
+      "step": 7255
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2933283723673711,
+      "learning_rate": 3.464781746921751e-06,
+      "loss": 0.0674,
+      "step": 7256
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5886895327579774,
+      "learning_rate": 3.4634831161719847e-06,
+      "loss": 0.0994,
+      "step": 7257
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.339142436612873,
+      "learning_rate": 3.462184599870947e-06,
+      "loss": 0.0444,
+      "step": 7258
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2788659062621638,
+      "learning_rate": 3.460886198115357e-06,
+      "loss": 0.0746,
+      "step": 7259
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3195635724142746,
+      "learning_rate": 3.4595879110019306e-06,
+      "loss": 0.0966,
+      "step": 7260
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.33568624617670695,
+      "learning_rate": 3.4582897386273683e-06,
+      "loss": 0.0605,
+      "step": 7261
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2729839246012114,
+      "learning_rate": 3.456991681088368e-06,
+      "loss": 0.0808,
+      "step": 7262
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.32401307201830587,
+      "learning_rate": 3.4556937384816167e-06,
+      "loss": 0.0607,
+      "step": 7263
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.19311700523388095,
+      "learning_rate": 3.4543959109037927e-06,
+      "loss": 0.0573,
+      "step": 7264
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1199505208179195,
+      "learning_rate": 3.453098198451563e-06,
+      "loss": 0.0815,
+      "step": 7265
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.34774625334082765,
+      "learning_rate": 3.4518006012215915e-06,
+      "loss": 0.0553,
+      "step": 7266
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.44682756864854534,
+      "learning_rate": 3.450503119310531e-06,
+      "loss": 0.0796,
+      "step": 7267
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.28661969624421535,
+      "learning_rate": 3.4492057528150245e-06,
+      "loss": 0.1129,
+      "step": 7268
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.27184027153051976,
+      "learning_rate": 3.447908501831706e-06,
+      "loss": 0.0645,
+      "step": 7269
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3619101616059619,
+      "learning_rate": 3.446611366457206e-06,
+      "loss": 0.0748,
+      "step": 7270
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2830447566215545,
+      "learning_rate": 3.44531434678814e-06,
+      "loss": 0.1141,
+      "step": 7271
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.33487652330292617,
+      "learning_rate": 3.4440174429211186e-06,
+      "loss": 0.0944,
+      "step": 7272
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3159947315615128,
+      "learning_rate": 3.4427206549527404e-06,
+      "loss": 0.0724,
+      "step": 7273
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3927528037145045,
+      "learning_rate": 3.4414239829796014e-06,
+      "loss": 0.1437,
+      "step": 7274
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5889045046118133,
+      "learning_rate": 3.4401274270982833e-06,
+      "loss": 0.1242,
+      "step": 7275
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.23341212144436196,
+      "learning_rate": 3.4388309874053593e-06,
+      "loss": 0.0468,
+      "step": 7276
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.27156731859335653,
+      "learning_rate": 3.4375346639974005e-06,
+      "loss": 0.1025,
+      "step": 7277
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.20075248458132203,
+      "learning_rate": 3.436238456970962e-06,
+      "loss": 0.0471,
+      "step": 7278
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3265523324186541,
+      "learning_rate": 3.43494236642259e-06,
+      "loss": 0.1012,
+      "step": 7279
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.1911025447309183,
+      "learning_rate": 3.43364639244883e-06,
+      "loss": 0.0369,
+      "step": 7280
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.34143365340398263,
+      "learning_rate": 3.4323505351462116e-06,
+      "loss": 0.0783,
+      "step": 7281
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.35100147408668453,
+      "learning_rate": 3.4310547946112573e-06,
+      "loss": 0.0902,
+      "step": 7282
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2238889101175209,
+      "learning_rate": 3.42975917094048e-06,
+      "loss": 0.0701,
+      "step": 7283
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.37107890748805655,
+      "learning_rate": 3.4284636642303885e-06,
+      "loss": 0.0644,
+      "step": 7284
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7028317820048504,
+      "learning_rate": 3.4271682745774786e-06,
+      "loss": 0.143,
+      "step": 7285
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4752189097069653,
+      "learning_rate": 3.425873002078236e-06,
+      "loss": 0.1018,
+      "step": 7286
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3003558705325369,
+      "learning_rate": 3.424577846829144e-06,
+      "loss": 0.0879,
+      "step": 7287
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.31777370239077735,
+      "learning_rate": 3.423282808926671e-06,
+      "loss": 0.0887,
+      "step": 7288
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.17369588604691094,
+      "learning_rate": 3.4219878884672807e-06,
+      "loss": 0.0247,
+      "step": 7289
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.25601900319042425,
+      "learning_rate": 3.4206930855474217e-06,
+      "loss": 0.0648,
+      "step": 7290
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5064267884878119,
+      "learning_rate": 3.419398400263545e-06,
+      "loss": 0.0789,
+      "step": 7291
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2507583373150979,
+      "learning_rate": 3.4181038327120822e-06,
+      "loss": 0.0812,
+      "step": 7292
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3658612007364356,
+      "learning_rate": 3.416809382989461e-06,
+      "loss": 0.0954,
+      "step": 7293
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2817766497967344,
+      "learning_rate": 3.4155150511921005e-06,
+      "loss": 0.0841,
+      "step": 7294
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.28812257852455986,
+      "learning_rate": 3.4142208374164077e-06,
+      "loss": 0.0714,
+      "step": 7295
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4013677716460676,
+      "learning_rate": 3.412926741758785e-06,
+      "loss": 0.1409,
+      "step": 7296
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.31206205628267586,
+      "learning_rate": 3.4116327643156256e-06,
+      "loss": 0.0467,
+      "step": 7297
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2916657268965481,
+      "learning_rate": 3.4103389051833104e-06,
+      "loss": 0.0486,
+      "step": 7298
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.1984090740492212,
+      "learning_rate": 3.409045164458213e-06,
+      "loss": 0.0674,
+      "step": 7299
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.472677919117807,
+      "learning_rate": 3.407751542236699e-06,
+      "loss": 0.123,
+      "step": 7300
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.24128546558172273,
+      "learning_rate": 3.406458038615128e-06,
+      "loss": 0.0506,
+      "step": 7301
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2964864172269203,
+      "learning_rate": 3.405164653689845e-06,
+      "loss": 0.0929,
+      "step": 7302
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.27177509258566596,
+      "learning_rate": 3.4038713875571865e-06,
+      "loss": 0.0682,
+      "step": 7303
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.22988649633746,
+      "learning_rate": 3.402578240313488e-06,
+      "loss": 0.0477,
+      "step": 7304
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.39519627710965466,
+      "learning_rate": 3.401285212055067e-06,
+      "loss": 0.0949,
+      "step": 7305
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.4447553832897663,
+      "learning_rate": 3.399992302878237e-06,
+      "loss": 0.13,
+      "step": 7306
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.35348263136679137,
+      "learning_rate": 3.3986995128792995e-06,
+      "loss": 0.0641,
+      "step": 7307
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.26923819017135264,
+      "learning_rate": 3.397406842154552e-06,
+      "loss": 0.0962,
+      "step": 7308
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2661925759353993,
+      "learning_rate": 3.3961142908002788e-06,
+      "loss": 0.0667,
+      "step": 7309
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2794693465144013,
+      "learning_rate": 3.3948218589127546e-06,
+      "loss": 0.1075,
+      "step": 7310
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3559780866132902,
+      "learning_rate": 3.3935295465882513e-06,
+      "loss": 0.0776,
+      "step": 7311
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3357754746711033,
+      "learning_rate": 3.392237353923026e-06,
+      "loss": 0.1276,
+      "step": 7312
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3534045442914703,
+      "learning_rate": 3.3909452810133258e-06,
+      "loss": 0.0973,
+      "step": 7313
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3439437243947871,
+      "learning_rate": 3.3896533279553965e-06,
+      "loss": 0.1033,
+      "step": 7314
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.18324978723755,
+      "learning_rate": 3.3883614948454692e-06,
+      "loss": 0.0426,
+      "step": 7315
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2048284909644386,
+      "learning_rate": 3.3870697817797647e-06,
+      "loss": 0.0576,
+      "step": 7316
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.35030533290719207,
+      "learning_rate": 3.385778188854498e-06,
+      "loss": 0.0855,
+      "step": 7317
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3072156923647518,
+      "learning_rate": 3.3844867161658775e-06,
+      "loss": 0.0696,
+      "step": 7318
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.36013910883288075,
+      "learning_rate": 3.3831953638100967e-06,
+      "loss": 0.0571,
+      "step": 7319
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.22589359986472995,
+      "learning_rate": 3.3819041318833423e-06,
+      "loss": 0.0537,
+      "step": 7320
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3297858010687959,
+      "learning_rate": 3.3806130204817967e-06,
+      "loss": 0.1055,
+      "step": 7321
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.497880815879118,
+      "learning_rate": 3.379322029701626e-06,
+      "loss": 0.1246,
+      "step": 7322
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.19213067700810027,
+      "learning_rate": 3.3780311596389925e-06,
+      "loss": 0.0465,
+      "step": 7323
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.20977806692211526,
+      "learning_rate": 3.3767404103900443e-06,
+      "loss": 0.043,
+      "step": 7324
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.25975951229543537,
+      "learning_rate": 3.3754497820509286e-06,
+      "loss": 0.0627,
+      "step": 7325
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3013359840330051,
+      "learning_rate": 3.374159274717777e-06,
+      "loss": 0.0679,
+      "step": 7326
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3724425507276075,
+      "learning_rate": 3.372868888486713e-06,
+      "loss": 0.1111,
+      "step": 7327
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.26164632783404873,
+      "learning_rate": 3.3715786234538535e-06,
+      "loss": 0.0787,
+      "step": 7328
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2705317428346432,
+      "learning_rate": 3.3702884797153036e-06,
+      "loss": 0.0824,
+      "step": 7329
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.30002090191429237,
+      "learning_rate": 3.3689984573671626e-06,
+      "loss": 0.0675,
+      "step": 7330
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3401056574839956,
+      "learning_rate": 3.367708556505517e-06,
+      "loss": 0.1174,
+      "step": 7331
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5492510053705704,
+      "learning_rate": 3.3664187772264465e-06,
+      "loss": 0.1041,
+      "step": 7332
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.24289093831202652,
+      "learning_rate": 3.3651291196260226e-06,
+      "loss": 0.0709,
+      "step": 7333
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.31834569301477755,
+      "learning_rate": 3.3638395838003047e-06,
+      "loss": 0.0926,
+      "step": 7334
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.30027725163024976,
+      "learning_rate": 3.362550169845348e-06,
+      "loss": 0.0688,
+      "step": 7335
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.35920159354771275,
+      "learning_rate": 3.361260877857192e-06,
+      "loss": 0.0993,
+      "step": 7336
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.36193896607786297,
+      "learning_rate": 3.3599717079318705e-06,
+      "loss": 0.0927,
+      "step": 7337
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.4561312958747651,
+      "learning_rate": 3.358682660165413e-06,
+      "loss": 0.0454,
+      "step": 7338
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.4716281307587733,
+      "learning_rate": 3.357393734653832e-06,
+      "loss": 0.0793,
+      "step": 7339
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.24014773629796418,
+      "learning_rate": 3.356104931493134e-06,
+      "loss": 0.0688,
+      "step": 7340
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.17549648648519187,
+      "learning_rate": 3.354816250779316e-06,
+      "loss": 0.0182,
+      "step": 7341
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.42810442730459763,
+      "learning_rate": 3.353527692608368e-06,
+      "loss": 0.0938,
+      "step": 7342
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5372456079457855,
+      "learning_rate": 3.352239257076271e-06,
+      "loss": 0.1099,
+      "step": 7343
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.4387094212202471,
+      "learning_rate": 3.350950944278989e-06,
+      "loss": 0.124,
+      "step": 7344
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.27863917736818483,
+      "learning_rate": 3.3496627543124904e-06,
+      "loss": 0.0732,
+      "step": 7345
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.42412365415940234,
+      "learning_rate": 3.3483746872727228e-06,
+      "loss": 0.1084,
+      "step": 7346
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.41643913594161863,
+      "learning_rate": 3.3470867432556307e-06,
+      "loss": 0.0874,
+      "step": 7347
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.28728368779798186,
+      "learning_rate": 3.345798922357145e-06,
+      "loss": 0.0773,
+      "step": 7348
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3309504340312473,
+      "learning_rate": 3.344511224673194e-06,
+      "loss": 0.0894,
+      "step": 7349
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5964580472130235,
+      "learning_rate": 3.34322365029969e-06,
+      "loss": 0.1727,
+      "step": 7350
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2896800268124212,
+      "learning_rate": 3.341936199332538e-06,
+      "loss": 0.0741,
+      "step": 7351
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2854547072253174,
+      "learning_rate": 3.34064887186764e-06,
+      "loss": 0.0778,
+      "step": 7352
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.34577572636504195,
+      "learning_rate": 3.33936166800088e-06,
+      "loss": 0.1097,
+      "step": 7353
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.4139979838833184,
+      "learning_rate": 3.3380745878281345e-06,
+      "loss": 0.108,
+      "step": 7354
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.29551235409294335,
+      "learning_rate": 3.3367876314452773e-06,
+      "loss": 0.1135,
+      "step": 7355
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.25174756766885115,
+      "learning_rate": 3.3355007989481668e-06,
+      "loss": 0.0565,
+      "step": 7356
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.31572549084114787,
+      "learning_rate": 3.3342140904326524e-06,
+      "loss": 0.1054,
+      "step": 7357
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.21828230464923454,
+      "learning_rate": 3.3329275059945747e-06,
+      "loss": 0.0706,
+      "step": 7358
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3904951675509458,
+      "learning_rate": 3.33164104572977e-06,
+      "loss": 0.0924,
+      "step": 7359
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2975274293301175,
+      "learning_rate": 3.3303547097340593e-06,
+      "loss": 0.0671,
+      "step": 7360
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3702635298505712,
+      "learning_rate": 3.3290684981032545e-06,
+      "loss": 0.0706,
+      "step": 7361
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3884181094518574,
+      "learning_rate": 3.3277824109331636e-06,
+      "loss": 0.1417,
+      "step": 7362
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.30903463681615256,
+      "learning_rate": 3.3264964483195784e-06,
+      "loss": 0.1039,
+      "step": 7363
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.32391025757051095,
+      "learning_rate": 3.3252106103582883e-06,
+      "loss": 0.0811,
+      "step": 7364
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3960958468995765,
+      "learning_rate": 3.3239248971450667e-06,
+      "loss": 0.1204,
+      "step": 7365
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3405914240571599,
+      "learning_rate": 3.3226393087756836e-06,
+      "loss": 0.0785,
+      "step": 7366
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.33048864176932524,
+      "learning_rate": 3.321353845345896e-06,
+      "loss": 0.0892,
+      "step": 7367
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.4854807937129951,
+      "learning_rate": 3.3200685069514523e-06,
+      "loss": 0.0883,
+      "step": 7368
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.216819190456203,
+      "learning_rate": 3.3187832936880937e-06,
+      "loss": 0.08,
+      "step": 7369
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.30879965060969233,
+      "learning_rate": 3.3174982056515485e-06,
+      "loss": 0.0863,
+      "step": 7370
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2531302495911648,
+      "learning_rate": 3.3162132429375372e-06,
+      "loss": 0.0799,
+      "step": 7371
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.31671883728501554,
+      "learning_rate": 3.3149284056417734e-06,
+      "loss": 0.1009,
+      "step": 7372
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.4015282545285348,
+      "learning_rate": 3.313643693859959e-06,
+      "loss": 0.091,
+      "step": 7373
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.34851302314913324,
+      "learning_rate": 3.312359107687786e-06,
+      "loss": 0.1097,
+      "step": 7374
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2692607578424626,
+      "learning_rate": 3.3110746472209354e-06,
+      "loss": 0.067,
+      "step": 7375
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.16629148272299638,
+      "learning_rate": 3.3097903125550857e-06,
+      "loss": 0.0449,
+      "step": 7376
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.24846075841033108,
+      "learning_rate": 3.3085061037859e-06,
+      "loss": 0.0761,
+      "step": 7377
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2516934027597921,
+      "learning_rate": 3.3072220210090315e-06,
+      "loss": 0.0901,
+      "step": 7378
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.38975211102828633,
+      "learning_rate": 3.30593806432013e-06,
+      "loss": 0.0575,
+      "step": 7379
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2711966316855158,
+      "learning_rate": 3.3046542338148302e-06,
+      "loss": 0.0772,
+      "step": 7380
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.23389505861708823,
+      "learning_rate": 3.303370529588759e-06,
+      "loss": 0.0425,
+      "step": 7381
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.19994936475620134,
+      "learning_rate": 3.3020869517375327e-06,
+      "loss": 0.072,
+      "step": 7382
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2328239407172048,
+      "learning_rate": 3.300803500356763e-06,
+      "loss": 0.0571,
+      "step": 7383
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3860696656748534,
+      "learning_rate": 3.299520175542047e-06,
+      "loss": 0.0909,
+      "step": 7384
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3961233647663579,
+      "learning_rate": 3.298236977388972e-06,
+      "loss": 0.0461,
+      "step": 7385
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3327187898042797,
+      "learning_rate": 3.296953905993123e-06,
+      "loss": 0.0931,
+      "step": 7386
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2623153457039759,
+      "learning_rate": 3.295670961450068e-06,
+      "loss": 0.062,
+      "step": 7387
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2500242220306109,
+      "learning_rate": 3.2943881438553665e-06,
+      "loss": 0.0553,
+      "step": 7388
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.27799651995084546,
+      "learning_rate": 3.2931054533045736e-06,
+      "loss": 0.0861,
+      "step": 7389
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.24129785802628034,
+      "learning_rate": 3.2918228898932304e-06,
+      "loss": 0.0563,
+      "step": 7390
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3216812873857199,
+      "learning_rate": 3.2905404537168687e-06,
+      "loss": 0.0781,
+      "step": 7391
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3016006457518386,
+      "learning_rate": 3.2892581448710105e-06,
+      "loss": 0.1051,
+      "step": 7392
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.36514021903641075,
+      "learning_rate": 3.2879759634511736e-06,
+      "loss": 0.0881,
+      "step": 7393
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.579465200900013,
+      "learning_rate": 3.2866939095528606e-06,
+      "loss": 0.1168,
+      "step": 7394
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.25839963791599485,
+      "learning_rate": 3.2854119832715637e-06,
+      "loss": 0.0975,
+      "step": 7395
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.27439272810427784,
+      "learning_rate": 3.284130184702773e-06,
+      "loss": 0.0777,
+      "step": 7396
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.4287215952810872,
+      "learning_rate": 3.2828485139419598e-06,
+      "loss": 0.0906,
+      "step": 7397
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.44182111832395715,
+      "learning_rate": 3.281566971084594e-06,
+      "loss": 0.1299,
+      "step": 7398
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3311099603997046,
+      "learning_rate": 3.2802855562261293e-06,
+      "loss": 0.0607,
+      "step": 7399
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2744813227370427,
+      "learning_rate": 3.2790042694620154e-06,
+      "loss": 0.077,
+      "step": 7400
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.21197500147752987,
+      "learning_rate": 3.277723110887688e-06,
+      "loss": 0.0547,
+      "step": 7401
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.27168772065251484,
+      "learning_rate": 3.2764420805985777e-06,
+      "loss": 0.078,
+      "step": 7402
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2149601811319796,
+      "learning_rate": 3.275161178690101e-06,
+      "loss": 0.0747,
+      "step": 7403
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2741569599668653,
+      "learning_rate": 3.2738804052576683e-06,
+      "loss": 0.0828,
+      "step": 7404
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.30684780417902063,
+      "learning_rate": 3.2725997603966763e-06,
+      "loss": 0.082,
+      "step": 7405
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5059134743669398,
+      "learning_rate": 3.2713192442025195e-06,
+      "loss": 0.1437,
+      "step": 7406
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.4577032491610798,
+      "learning_rate": 3.2700388567705755e-06,
+      "loss": 0.1013,
+      "step": 7407
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.27872862959896944,
+      "learning_rate": 3.2687585981962142e-06,
+      "loss": 0.0756,
+      "step": 7408
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.35960748796772535,
+      "learning_rate": 3.267478468574797e-06,
+      "loss": 0.1251,
+      "step": 7409
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2883856213497421,
+      "learning_rate": 3.2661984680016777e-06,
+      "loss": 0.0582,
+      "step": 7410
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.28983736253553816,
+      "learning_rate": 3.264918596572196e-06,
+      "loss": 0.0764,
+      "step": 7411
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.424724935430722,
+      "learning_rate": 3.2636388543816846e-06,
+      "loss": 0.1145,
+      "step": 7412
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.17338907510544171,
+      "learning_rate": 3.2623592415254668e-06,
+      "loss": 0.0516,
+      "step": 7413
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5885301216754125,
+      "learning_rate": 3.261079758098855e-06,
+      "loss": 0.1491,
+      "step": 7414
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.4242738373687518,
+      "learning_rate": 3.2598004041971544e-06,
+      "loss": 0.0921,
+      "step": 7415
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.24088113766842717,
+      "learning_rate": 3.2585211799156537e-06,
+      "loss": 0.0702,
+      "step": 7416
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3045936350204951,
+      "learning_rate": 3.257242085349643e-06,
+      "loss": 0.0748,
+      "step": 7417
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.42868643259688044,
+      "learning_rate": 3.255963120594394e-06,
+      "loss": 0.1094,
+      "step": 7418
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.28358696502776637,
+      "learning_rate": 3.2546842857451688e-06,
+      "loss": 0.088,
+      "step": 7419
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.39101465549547637,
+      "learning_rate": 3.253405580897228e-06,
+      "loss": 0.1042,
+      "step": 7420
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.37056056531663695,
+      "learning_rate": 3.2521270061458143e-06,
+      "loss": 0.107,
+      "step": 7421
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.45593406653836094,
+      "learning_rate": 3.2508485615861605e-06,
+      "loss": 0.0832,
+      "step": 7422
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4008710362891544,
+      "learning_rate": 3.2495702473134983e-06,
+      "loss": 0.0933,
+      "step": 7423
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.584780999806043,
+      "learning_rate": 3.248292063423041e-06,
+      "loss": 0.079,
+      "step": 7424
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2017669908599626,
+      "learning_rate": 3.2470140100099944e-06,
+      "loss": 0.0575,
+      "step": 7425
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6526943454967085,
+      "learning_rate": 3.2457360871695555e-06,
+      "loss": 0.111,
+      "step": 7426
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5545066672299871,
+      "learning_rate": 3.2444582949969126e-06,
+      "loss": 0.1066,
+      "step": 7427
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5836604015333878,
+      "learning_rate": 3.2431806335872438e-06,
+      "loss": 0.1675,
+      "step": 7428
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2635620055119552,
+      "learning_rate": 3.241903103035715e-06,
+      "loss": 0.0563,
+      "step": 7429
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5406997658665079,
+      "learning_rate": 3.240625703437485e-06,
+      "loss": 0.1324,
+      "step": 7430
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.25023229935125674,
+      "learning_rate": 3.239348434887701e-06,
+      "loss": 0.0337,
+      "step": 7431
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.23068573734113001,
+      "learning_rate": 3.238071297481503e-06,
+      "loss": 0.0552,
+      "step": 7432
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.16806121367917953,
+      "learning_rate": 3.2367942913140178e-06,
+      "loss": 0.0508,
+      "step": 7433
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.29090898578568836,
+      "learning_rate": 3.235517416480366e-06,
+      "loss": 0.1198,
+      "step": 7434
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7812114589650989,
+      "learning_rate": 3.234240673075655e-06,
+      "loss": 0.0778,
+      "step": 7435
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.32679717562924704,
+      "learning_rate": 3.2329640611949855e-06,
+      "loss": 0.0852,
+      "step": 7436
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.29535154160652316,
+      "learning_rate": 3.231687580933447e-06,
+      "loss": 0.0843,
+      "step": 7437
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.24215785554315591,
+      "learning_rate": 3.230411232386118e-06,
+      "loss": 0.05,
+      "step": 7438
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2537081196884538,
+      "learning_rate": 3.2291350156480685e-06,
+      "loss": 0.0578,
+      "step": 7439
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.527436623416943,
+      "learning_rate": 3.22785893081436e-06,
+      "loss": 0.1109,
+      "step": 7440
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2469326007158126,
+      "learning_rate": 3.226582977980042e-06,
+      "loss": 0.0754,
+      "step": 7441
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.24640223940945197,
+      "learning_rate": 3.225307157240155e-06,
+      "loss": 0.0602,
+      "step": 7442
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.25192200679490573,
+      "learning_rate": 3.224031468689727e-06,
+      "loss": 0.0652,
+      "step": 7443
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2965783969795636,
+      "learning_rate": 3.222755912423783e-06,
+      "loss": 0.0724,
+      "step": 7444
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3104032394575369,
+      "learning_rate": 3.221480488537332e-06,
+      "loss": 0.0968,
+      "step": 7445
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4107924142341617,
+      "learning_rate": 3.220205197125373e-06,
+      "loss": 0.1247,
+      "step": 7446
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3294485010096923,
+      "learning_rate": 3.2189300382829002e-06,
+      "loss": 0.0796,
+      "step": 7447
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.1961269796591617,
+      "learning_rate": 3.2176550121048944e-06,
+      "loss": 0.049,
+      "step": 7448
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3046730998588272,
+      "learning_rate": 3.2163801186863266e-06,
+      "loss": 0.0939,
+      "step": 7449
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.23586704429609778,
+      "learning_rate": 3.2151053581221545e-06,
+      "loss": 0.0597,
+      "step": 7450
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.26475644537801235,
+      "learning_rate": 3.2138307305073367e-06,
+      "loss": 0.0535,
+      "step": 7451
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3310516696459729,
+      "learning_rate": 3.2125562359368105e-06,
+      "loss": 0.1155,
+      "step": 7452
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3811301332478338,
+      "learning_rate": 3.2112818745055065e-06,
+      "loss": 0.1128,
+      "step": 7453
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4508032513921881,
+      "learning_rate": 3.2100076463083507e-06,
+      "loss": 0.1148,
+      "step": 7454
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3237442424519417,
+      "learning_rate": 3.2087335514402527e-06,
+      "loss": 0.1137,
+      "step": 7455
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3145742399360536,
+      "learning_rate": 3.2074595899961146e-06,
+      "loss": 0.0796,
+      "step": 7456
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.32677911581995633,
+      "learning_rate": 3.206185762070827e-06,
+      "loss": 0.0794,
+      "step": 7457
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4609880382334431,
+      "learning_rate": 3.2049120677592755e-06,
+      "loss": 0.1271,
+      "step": 7458
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.32738224998198867,
+      "learning_rate": 3.2036385071563302e-06,
+      "loss": 0.0999,
+      "step": 7459
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3748991117333819,
+      "learning_rate": 3.202365080356851e-06,
+      "loss": 0.116,
+      "step": 7460
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.49097756483682037,
+      "learning_rate": 3.2010917874556954e-06,
+      "loss": 0.1191,
+      "step": 7461
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5516323334164326,
+      "learning_rate": 3.199818628547702e-06,
+      "loss": 0.1301,
+      "step": 7462
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.1945615103618961,
+      "learning_rate": 3.198545603727704e-06,
+      "loss": 0.0617,
+      "step": 7463
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3783222525498935,
+      "learning_rate": 3.197272713090524e-06,
+      "loss": 0.116,
+      "step": 7464
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.24916970996288107,
+      "learning_rate": 3.195999956730973e-06,
+      "loss": 0.0861,
+      "step": 7465
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.20255172540595803,
+      "learning_rate": 3.194727334743856e-06,
+      "loss": 0.0415,
+      "step": 7466
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2838204883826325,
+      "learning_rate": 3.1934548472239617e-06,
+      "loss": 0.0674,
+      "step": 7467
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.24761518232478513,
+      "learning_rate": 3.1921824942660763e-06,
+      "loss": 0.0619,
+      "step": 7468
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.40029139447753037,
+      "learning_rate": 3.190910275964968e-06,
+      "loss": 0.1098,
+      "step": 7469
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.31293603460587804,
+      "learning_rate": 3.189638192415402e-06,
+      "loss": 0.0838,
+      "step": 7470
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.41762864022561214,
+      "learning_rate": 3.1883662437121297e-06,
+      "loss": 0.1005,
+      "step": 7471
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5067572040653646,
+      "learning_rate": 3.1870944299498928e-06,
+      "loss": 0.0625,
+      "step": 7472
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3461648499545234,
+      "learning_rate": 3.1858227512234254e-06,
+      "loss": 0.1192,
+      "step": 7473
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.29577072173602353,
+      "learning_rate": 3.1845512076274444e-06,
+      "loss": 0.0819,
+      "step": 7474
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6096168953491207,
+      "learning_rate": 3.1832797992566677e-06,
+      "loss": 0.1136,
+      "step": 7475
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.32072682282314624,
+      "learning_rate": 3.1820085262057943e-06,
+      "loss": 0.0771,
+      "step": 7476
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.26252219685391276,
+      "learning_rate": 3.180737388569515e-06,
+      "loss": 0.08,
+      "step": 7477
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.287183794688386,
+      "learning_rate": 3.1794663864425146e-06,
+      "loss": 0.0977,
+      "step": 7478
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2594711886747748,
+      "learning_rate": 3.1781955199194634e-06,
+      "loss": 0.0855,
+      "step": 7479
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.35828038605174267,
+      "learning_rate": 3.1769247890950206e-06,
+      "loss": 0.1002,
+      "step": 7480
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3046537943329612,
+      "learning_rate": 3.1756541940638417e-06,
+      "loss": 0.0888,
+      "step": 7481
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.33707491248391264,
+      "learning_rate": 3.1743837349205665e-06,
+      "loss": 0.0878,
+      "step": 7482
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.33828426149122465,
+      "learning_rate": 3.173113411759825e-06,
+      "loss": 0.068,
+      "step": 7483
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.32238878051012504,
+      "learning_rate": 3.1718432246762376e-06,
+      "loss": 0.0775,
+      "step": 7484
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.23557545569126495,
+      "learning_rate": 3.1705731737644186e-06,
+      "loss": 0.0579,
+      "step": 7485
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3840864540453718,
+      "learning_rate": 3.169303259118966e-06,
+      "loss": 0.0944,
+      "step": 7486
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2021329499404736,
+      "learning_rate": 3.1680334808344696e-06,
+      "loss": 0.0676,
+      "step": 7487
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3248235212173923,
+      "learning_rate": 3.166763839005514e-06,
+      "loss": 0.1202,
+      "step": 7488
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.35598711735489263,
+      "learning_rate": 3.1654943337266663e-06,
+      "loss": 0.0569,
+      "step": 7489
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.1922245355040462,
+      "learning_rate": 3.1642249650924883e-06,
+      "loss": 0.0768,
+      "step": 7490
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2651243935893133,
+      "learning_rate": 3.162955733197527e-06,
+      "loss": 0.0969,
+      "step": 7491
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.35811733768048903,
+      "learning_rate": 3.1616866381363255e-06,
+      "loss": 0.0717,
+      "step": 7492
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.47824437231040295,
+      "learning_rate": 3.1604176800034127e-06,
+      "loss": 0.0795,
+      "step": 7493
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.26514132231403953,
+      "learning_rate": 3.159148858893305e-06,
+      "loss": 0.0693,
+      "step": 7494
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.41285330699260103,
+      "learning_rate": 3.157880174900517e-06,
+      "loss": 0.12,
+      "step": 7495
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.26052959659735436,
+      "learning_rate": 3.1566116281195437e-06,
+      "loss": 0.0854,
+      "step": 7496
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2112470210598013,
+      "learning_rate": 3.1553432186448745e-06,
+      "loss": 0.0451,
+      "step": 7497
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2880792287848207,
+      "learning_rate": 3.154074946570989e-06,
+      "loss": 0.0807,
+      "step": 7498
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.41890540761238787,
+      "learning_rate": 3.1528068119923537e-06,
+      "loss": 0.0918,
+      "step": 7499
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5898363425109961,
+      "learning_rate": 3.151538815003429e-06,
+      "loss": 0.1364,
+      "step": 7500
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3266462844920886,
+      "learning_rate": 3.1502709556986606e-06,
+      "loss": 0.1139,
+      "step": 7501
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.40465090162948564,
+      "learning_rate": 3.1490032341724876e-06,
+      "loss": 0.1251,
+      "step": 7502
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.31940297971941356,
+      "learning_rate": 3.1477356505193357e-06,
+      "loss": 0.0801,
+      "step": 7503
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3053068160535285,
+      "learning_rate": 3.1464682048336224e-06,
+      "loss": 0.068,
+      "step": 7504
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.22115661040473558,
+      "learning_rate": 3.145200897209756e-06,
+      "loss": 0.0667,
+      "step": 7505
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3058206710117369,
+      "learning_rate": 3.1439337277421323e-06,
+      "loss": 0.092,
+      "step": 7506
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2740688334106272,
+      "learning_rate": 3.142666696525136e-06,
+      "loss": 0.0929,
+      "step": 7507
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3548648423804408,
+      "learning_rate": 3.1413998036531424e-06,
+      "loss": 0.1054,
+      "step": 7508
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.643187805530458,
+      "learning_rate": 3.1401330492205205e-06,
+      "loss": 0.1205,
+      "step": 7509
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.24681355199470112,
+      "learning_rate": 3.1388664333216246e-06,
+      "loss": 0.0876,
+      "step": 7510
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.44139657157666573,
+      "learning_rate": 3.1375999560507954e-06,
+      "loss": 0.1115,
+      "step": 7511
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2134277138639315,
+      "learning_rate": 3.136333617502373e-06,
+      "loss": 0.0639,
+      "step": 7512
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5950430369478585,
+      "learning_rate": 3.13506741777068e-06,
+      "loss": 0.1327,
+      "step": 7513
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.275102249907796,
+      "learning_rate": 3.133801356950027e-06,
+      "loss": 0.0705,
+      "step": 7514
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4307600565521753,
+      "learning_rate": 3.1325354351347227e-06,
+      "loss": 0.1,
+      "step": 7515
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.34454543120684683,
+      "learning_rate": 3.1312696524190577e-06,
+      "loss": 0.1178,
+      "step": 7516
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.40124859522257766,
+      "learning_rate": 3.130004008897316e-06,
+      "loss": 0.1009,
+      "step": 7517
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2732894227225401,
+      "learning_rate": 3.1287385046637665e-06,
+      "loss": 0.0436,
+      "step": 7518
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3304042071898356,
+      "learning_rate": 3.127473139812676e-06,
+      "loss": 0.0814,
+      "step": 7519
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6031544168331392,
+      "learning_rate": 3.1262079144382946e-06,
+      "loss": 0.0954,
+      "step": 7520
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2863460074938233,
+      "learning_rate": 3.1249428286348616e-06,
+      "loss": 0.0873,
+      "step": 7521
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5194621824346859,
+      "learning_rate": 3.1236778824966106e-06,
+      "loss": 0.1495,
+      "step": 7522
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.23213992496197666,
+      "learning_rate": 3.122413076117763e-06,
+      "loss": 0.0628,
+      "step": 7523
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4066952204403874,
+      "learning_rate": 3.121148409592526e-06,
+      "loss": 0.0956,
+      "step": 7524
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2863313488331882,
+      "learning_rate": 3.119883883015099e-06,
+      "loss": 0.104,
+      "step": 7525
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3574330701026415,
+      "learning_rate": 3.1186194964796744e-06,
+      "loss": 0.1006,
+      "step": 7526
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.1910168147289262,
+      "learning_rate": 3.1173552500804305e-06,
+      "loss": 0.0482,
+      "step": 7527
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.351227076390276,
+      "learning_rate": 3.1160911439115327e-06,
+      "loss": 0.1313,
+      "step": 7528
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3142577786051454,
+      "learning_rate": 3.1148271780671425e-06,
+      "loss": 0.0852,
+      "step": 7529
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2908722023233454,
+      "learning_rate": 3.113563352641406e-06,
+      "loss": 0.1013,
+      "step": 7530
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.24316715242066264,
+      "learning_rate": 3.1122996677284593e-06,
+      "loss": 0.0644,
+      "step": 7531
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.339780043367246,
+      "learning_rate": 3.1110361234224316e-06,
+      "loss": 0.0865,
+      "step": 7532
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3948221383374114,
+      "learning_rate": 3.1097727198174356e-06,
+      "loss": 0.0683,
+      "step": 7533
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.542824965001126,
+      "learning_rate": 3.1085094570075804e-06,
+      "loss": 0.1002,
+      "step": 7534
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3465377031481567,
+      "learning_rate": 3.1072463350869587e-06,
+      "loss": 0.1254,
+      "step": 7535
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2684367183488828,
+      "learning_rate": 3.1059833541496563e-06,
+      "loss": 0.0606,
+      "step": 7536
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2927199227029024,
+      "learning_rate": 3.1047205142897473e-06,
+      "loss": 0.0822,
+      "step": 7537
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.26780500429948084,
+      "learning_rate": 3.1034578156012947e-06,
+      "loss": 0.0862,
+      "step": 7538
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3409494701252806,
+      "learning_rate": 3.102195258178353e-06,
+      "loss": 0.0832,
+      "step": 7539
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.19368975188938772,
+      "learning_rate": 3.100932842114964e-06,
+      "loss": 0.0467,
+      "step": 7540
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2769635623097641,
+      "learning_rate": 3.0996705675051597e-06,
+      "loss": 0.0777,
+      "step": 7541
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.19825581210940144,
+      "learning_rate": 3.09840843444296e-06,
+      "loss": 0.0339,
+      "step": 7542
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3835708473794745,
+      "learning_rate": 3.09714644302238e-06,
+      "loss": 0.0824,
+      "step": 7543
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.41280720711689006,
+      "learning_rate": 3.095884593337417e-06,
+      "loss": 0.0857,
+      "step": 7544
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6029844371439689,
+      "learning_rate": 3.0946228854820603e-06,
+      "loss": 0.0462,
+      "step": 7545
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.43417514878235447,
+      "learning_rate": 3.093361319550293e-06,
+      "loss": 0.1178,
+      "step": 7546
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.22678908796391573,
+      "learning_rate": 3.0920998956360817e-06,
+      "loss": 0.0494,
+      "step": 7547
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.28458731029134526,
+      "learning_rate": 3.0908386138333812e-06,
+      "loss": 0.0858,
+      "step": 7548
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5136574362698686,
+      "learning_rate": 3.0895774742361463e-06,
+      "loss": 0.1197,
+      "step": 7549
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.16634165774189022,
+      "learning_rate": 3.088316476938309e-06,
+      "loss": 0.0539,
+      "step": 7550
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4630251740916094,
+      "learning_rate": 3.0870556220337964e-06,
+      "loss": 0.1193,
+      "step": 7551
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.30637543965006275,
+      "learning_rate": 3.085794909616524e-06,
+      "loss": 0.083,
+      "step": 7552
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2890721748017524,
+      "learning_rate": 3.0845343397803994e-06,
+      "loss": 0.0497,
+      "step": 7553
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3632288029514629,
+      "learning_rate": 3.0832739126193158e-06,
+      "loss": 0.09,
+      "step": 7554
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3533243182288359,
+      "learning_rate": 3.082013628227154e-06,
+      "loss": 0.119,
+      "step": 7555
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3205222088986445,
+      "learning_rate": 3.0807534866977928e-06,
+      "loss": 0.1015,
+      "step": 7556
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3894941259944348,
+      "learning_rate": 3.079493488125093e-06,
+      "loss": 0.078,
+      "step": 7557
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4771072371476969,
+      "learning_rate": 3.078233632602905e-06,
+      "loss": 0.095,
+      "step": 7558
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.27833711163922675,
+      "learning_rate": 3.07697392022507e-06,
+      "loss": 0.09,
+      "step": 7559
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2276439154504608,
+      "learning_rate": 3.0757143510854217e-06,
+      "loss": 0.0546,
+      "step": 7560
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3594107325058686,
+      "learning_rate": 3.074454925277778e-06,
+      "loss": 0.0815,
+      "step": 7561
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.1897825091893244,
+      "learning_rate": 3.0731956428959463e-06,
+      "loss": 0.0413,
+      "step": 7562
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5288974970070162,
+      "learning_rate": 3.07193650403373e-06,
+      "loss": 0.1671,
+      "step": 7563
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2618825749210449,
+      "learning_rate": 3.0706775087849143e-06,
+      "loss": 0.0688,
+      "step": 7564
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.28055677976640175,
+      "learning_rate": 3.0694186572432768e-06,
+      "loss": 0.0883,
+      "step": 7565
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3503467993303243,
+      "learning_rate": 3.0681599495025845e-06,
+      "loss": 0.1308,
+      "step": 7566
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3505450712437244,
+      "learning_rate": 3.066901385656592e-06,
+      "loss": 0.102,
+      "step": 7567
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.22559138454775035,
+      "learning_rate": 3.0656429657990462e-06,
+      "loss": 0.0502,
+      "step": 7568
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.30361790805215616,
+      "learning_rate": 3.06438469002368e-06,
+      "loss": 0.0387,
+      "step": 7569
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3317921809128949,
+      "learning_rate": 3.0631265584242194e-06,
+      "loss": 0.0889,
+      "step": 7570
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3118201982288268,
+      "learning_rate": 3.061868571094374e-06,
+      "loss": 0.091,
+      "step": 7571
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.22449881726134763,
+      "learning_rate": 3.0606107281278485e-06,
+      "loss": 0.0547,
+      "step": 7572
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.35839939438308904,
+      "learning_rate": 3.0593530296183345e-06,
+      "loss": 0.072,
+      "step": 7573
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.27299680614028915,
+      "learning_rate": 3.0580954756595115e-06,
+      "loss": 0.0577,
+      "step": 7574
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.28344973987279354,
+      "learning_rate": 3.056838066345051e-06,
+      "loss": 0.0587,
+      "step": 7575
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3718812552683147,
+      "learning_rate": 3.0555808017686084e-06,
+      "loss": 0.0664,
+      "step": 7576
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.26011765558266553,
+      "learning_rate": 3.0543236820238366e-06,
+      "loss": 0.0708,
+      "step": 7577
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.20279512283766082,
+      "learning_rate": 3.0530667072043718e-06,
+      "loss": 0.0486,
+      "step": 7578
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2158328986102784,
+      "learning_rate": 3.051809877403838e-06,
+      "loss": 0.0471,
+      "step": 7579
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.35759888191261263,
+      "learning_rate": 3.0505531927158556e-06,
+      "loss": 0.0552,
+      "step": 7580
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.27633923251662657,
+      "learning_rate": 3.049296653234028e-06,
+      "loss": 0.091,
+      "step": 7581
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.32840874081319354,
+      "learning_rate": 3.0480402590519493e-06,
+      "loss": 0.0788,
+      "step": 7582
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.35431497364871034,
+      "learning_rate": 3.0467840102632007e-06,
+      "loss": 0.0727,
+      "step": 7583
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.46701377723717535,
+      "learning_rate": 3.0455279069613596e-06,
+      "loss": 0.0829,
+      "step": 7584
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4477093426231771,
+      "learning_rate": 3.0442719492399865e-06,
+      "loss": 0.1326,
+      "step": 7585
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.355387574543283,
+      "learning_rate": 3.0430161371926283e-06,
+      "loss": 0.0794,
+      "step": 7586
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.27048287788175784,
+      "learning_rate": 3.041760470912831e-06,
+      "loss": 0.0605,
+      "step": 7587
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.40011901009465345,
+      "learning_rate": 3.0405049504941212e-06,
+      "loss": 0.0951,
+      "step": 7588
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3158251845240956,
+      "learning_rate": 3.039249576030016e-06,
+      "loss": 0.0822,
+      "step": 7589
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2837336618924544,
+      "learning_rate": 3.037994347614026e-06,
+      "loss": 0.0855,
+      "step": 7590
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2123223136818581,
+      "learning_rate": 3.0367392653396476e-06,
+      "loss": 0.0815,
+      "step": 7591
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.22147569414361187,
+      "learning_rate": 3.0354843293003644e-06,
+      "loss": 0.0576,
+      "step": 7592
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3002331992333214,
+      "learning_rate": 3.0342295395896516e-06,
+      "loss": 0.1006,
+      "step": 7593
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3940278974763917,
+      "learning_rate": 3.0329748963009754e-06,
+      "loss": 0.0908,
+      "step": 7594
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.40997655731125643,
+      "learning_rate": 3.0317203995277876e-06,
+      "loss": 0.1172,
+      "step": 7595
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2685186800301309,
+      "learning_rate": 3.0304660493635283e-06,
+      "loss": 0.0735,
+      "step": 7596
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3287420030588943,
+      "learning_rate": 3.0292118459016338e-06,
+      "loss": 0.0766,
+      "step": 7597
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6935983027051918,
+      "learning_rate": 3.0279577892355215e-06,
+      "loss": 0.1155,
+      "step": 7598
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.33427866364929465,
+      "learning_rate": 3.0267038794586005e-06,
+      "loss": 0.0955,
+      "step": 7599
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2982176294491736,
+      "learning_rate": 3.02545011666427e-06,
+      "loss": 0.0687,
+      "step": 7600
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2778211255949358,
+      "learning_rate": 3.024196500945916e-06,
+      "loss": 0.0696,
+      "step": 7601
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.24645067918227956,
+      "learning_rate": 3.0229430323969185e-06,
+      "loss": 0.071,
+      "step": 7602
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5074886491320254,
+      "learning_rate": 3.0216897111106397e-06,
+      "loss": 0.1213,
+      "step": 7603
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3699543368319706,
+      "learning_rate": 3.020436537180437e-06,
+      "loss": 0.0928,
+      "step": 7604
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3630029844294792,
+      "learning_rate": 3.0191835106996513e-06,
+      "loss": 0.097,
+      "step": 7605
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2644561583063568,
+      "learning_rate": 3.0179306317616175e-06,
+      "loss": 0.0756,
+      "step": 7606
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.35252190859469246,
+      "learning_rate": 3.0166779004596574e-06,
+      "loss": 0.0807,
+      "step": 7607
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.15544356822506278,
+      "learning_rate": 3.015425316887081e-06,
+      "loss": 0.0346,
+      "step": 7608
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.30891978930094066,
+      "learning_rate": 3.0141728811371885e-06,
+      "loss": 0.0985,
+      "step": 7609
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3512663288688792,
+      "learning_rate": 3.012920593303265e-06,
+      "loss": 0.1315,
+      "step": 7610
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.28492316852100275,
+      "learning_rate": 3.011668453478594e-06,
+      "loss": 0.0598,
+      "step": 7611
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.43692417010839485,
+      "learning_rate": 3.0104164617564393e-06,
+      "loss": 0.1343,
+      "step": 7612
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3972450705812554,
+      "learning_rate": 3.009164618230055e-06,
+      "loss": 0.0851,
+      "step": 7613
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.31786627481822394,
+      "learning_rate": 3.007912922992689e-06,
+      "loss": 0.1005,
+      "step": 7614
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.36484990001751216,
+      "learning_rate": 3.0066613761375734e-06,
+      "loss": 0.0757,
+      "step": 7615
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.29865652232610834,
+      "learning_rate": 3.0054099777579303e-06,
+      "loss": 0.1089,
+      "step": 7616
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2014425658868871,
+      "learning_rate": 3.0041587279469696e-06,
+      "loss": 0.0594,
+      "step": 7617
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2176238146550361,
+      "learning_rate": 3.002907626797896e-06,
+      "loss": 0.0568,
+      "step": 7618
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.30891447627532476,
+      "learning_rate": 3.0016566744038955e-06,
+      "loss": 0.0745,
+      "step": 7619
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2923663072973704,
+      "learning_rate": 3.0004058708581456e-06,
+      "loss": 0.0553,
+      "step": 7620
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.347270635943339,
+      "learning_rate": 2.9991552162538174e-06,
+      "loss": 0.0829,
+      "step": 7621
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.19497550949400194,
+      "learning_rate": 2.9979047106840634e-06,
+      "loss": 0.0539,
+      "step": 7622
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3107590316173691,
+      "learning_rate": 2.996654354242028e-06,
+      "loss": 0.1101,
+      "step": 7623
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.18428650511238057,
+      "learning_rate": 2.995404147020849e-06,
+      "loss": 0.0657,
+      "step": 7624
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5619812205864332,
+      "learning_rate": 2.9941540891136467e-06,
+      "loss": 0.1093,
+      "step": 7625
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3511298786457933,
+      "learning_rate": 2.9929041806135323e-06,
+      "loss": 0.0561,
+      "step": 7626
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.33180366620939994,
+      "learning_rate": 2.991654421613605e-06,
+      "loss": 0.0757,
+      "step": 7627
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3045304684837932,
+      "learning_rate": 2.990404812206958e-06,
+      "loss": 0.0954,
+      "step": 7628
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.20954155704102642,
+      "learning_rate": 2.989155352486667e-06,
+      "loss": 0.0792,
+      "step": 7629
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2537791676624729,
+      "learning_rate": 2.9879060425457974e-06,
+      "loss": 0.0234,
+      "step": 7630
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.36437972679109165,
+      "learning_rate": 2.9866568824774096e-06,
+      "loss": 0.0976,
+      "step": 7631
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.37627610736542844,
+      "learning_rate": 2.985407872374545e-06,
+      "loss": 0.102,
+      "step": 7632
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.36931605273029683,
+      "learning_rate": 2.984159012330239e-06,
+      "loss": 0.0912,
+      "step": 7633
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.34110256966372815,
+      "learning_rate": 2.9829103024375105e-06,
+      "loss": 0.0951,
+      "step": 7634
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3265722812173713,
+      "learning_rate": 2.981661742789374e-06,
+      "loss": 0.0993,
+      "step": 7635
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3624395643476571,
+      "learning_rate": 2.9804133334788297e-06,
+      "loss": 0.0852,
+      "step": 7636
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3264678749388332,
+      "learning_rate": 2.979165074598864e-06,
+      "loss": 0.0822,
+      "step": 7637
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4241695418593112,
+      "learning_rate": 2.9779169662424566e-06,
+      "loss": 0.0755,
+      "step": 7638
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2479543260289617,
+      "learning_rate": 2.9766690085025722e-06,
+      "loss": 0.0892,
+      "step": 7639
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.25915875201505956,
+      "learning_rate": 2.975421201472167e-06,
+      "loss": 0.0552,
+      "step": 7640
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.34036949312700226,
+      "learning_rate": 2.9741735452441855e-06,
+      "loss": 0.0846,
+      "step": 7641
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.21488368851748546,
+      "learning_rate": 2.97292603991156e-06,
+      "loss": 0.0394,
+      "step": 7642
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.20697780050330308,
+      "learning_rate": 2.9716786855672113e-06,
+      "loss": 0.0462,
+      "step": 7643
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3678616622894788,
+      "learning_rate": 2.970431482304048e-06,
+      "loss": 0.0878,
+      "step": 7644
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.41651679679682846,
+      "learning_rate": 2.969184430214973e-06,
+      "loss": 0.0953,
+      "step": 7645
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.36251147165396075,
+      "learning_rate": 2.9679375293928726e-06,
+      "loss": 0.0816,
+      "step": 7646
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2741761732758418,
+      "learning_rate": 2.96669077993062e-06,
+      "loss": 0.0678,
+      "step": 7647
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.240763451112145,
+      "learning_rate": 2.965444181921085e-06,
+      "loss": 0.0789,
+      "step": 7648
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2565469244560333,
+      "learning_rate": 2.9641977354571194e-06,
+      "loss": 0.0746,
+      "step": 7649
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.43655989860299493,
+      "learning_rate": 2.9629514406315663e-06,
+      "loss": 0.0883,
+      "step": 7650
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3119912018197633,
+      "learning_rate": 2.961705297537254e-06,
+      "loss": 0.0911,
+      "step": 7651
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.22531002021468668,
+      "learning_rate": 2.960459306267008e-06,
+      "loss": 0.0788,
+      "step": 7652
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.49296910053286613,
+      "learning_rate": 2.9592134669136334e-06,
+      "loss": 0.0983,
+      "step": 7653
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.21650063558728297,
+      "learning_rate": 2.957967779569926e-06,
+      "loss": 0.0568,
+      "step": 7654
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3485043769503001,
+      "learning_rate": 2.956722244328677e-06,
+      "loss": 0.094,
+      "step": 7655
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.19558482986653544,
+      "learning_rate": 2.955476861282658e-06,
+      "loss": 0.0473,
+      "step": 7656
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.30923844446709947,
+      "learning_rate": 2.95423163052463e-06,
+      "loss": 0.0614,
+      "step": 7657
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6232460291241838,
+      "learning_rate": 2.95298655214735e-06,
+      "loss": 0.0951,
+      "step": 7658
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.27995577607582195,
+      "learning_rate": 2.951741626243556e-06,
+      "loss": 0.0453,
+      "step": 7659
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.29337837941816786,
+      "learning_rate": 2.9504968529059784e-06,
+      "loss": 0.1032,
+      "step": 7660
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2296433823809354,
+      "learning_rate": 2.949252232227332e-06,
+      "loss": 0.0523,
+      "step": 7661
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2768467010035087,
+      "learning_rate": 2.948007764300328e-06,
+      "loss": 0.0605,
+      "step": 7662
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4289900829373465,
+      "learning_rate": 2.946763449217659e-06,
+      "loss": 0.1005,
+      "step": 7663
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.25394514404082935,
+      "learning_rate": 2.9455192870720073e-06,
+      "loss": 0.0567,
+      "step": 7664
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4441613527634566,
+      "learning_rate": 2.9442752779560502e-06,
+      "loss": 0.117,
+      "step": 7665
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.20181939604021895,
+      "learning_rate": 2.9430314219624446e-06,
+      "loss": 0.0666,
+      "step": 7666
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6510775557177738,
+      "learning_rate": 2.941787719183842e-06,
+      "loss": 0.1088,
+      "step": 7667
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.25728350642052045,
+      "learning_rate": 2.9405441697128795e-06,
+      "loss": 0.0434,
+      "step": 7668
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.31808476105024885,
+      "learning_rate": 2.939300773642184e-06,
+      "loss": 0.1049,
+      "step": 7669
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2932412813441007,
+      "learning_rate": 2.938057531064372e-06,
+      "loss": 0.0632,
+      "step": 7670
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2579541918096553,
+      "learning_rate": 2.936814442072047e-06,
+      "loss": 0.0512,
+      "step": 7671
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3456252510687903,
+      "learning_rate": 2.9355715067578015e-06,
+      "loss": 0.1066,
+      "step": 7672
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.27739207304873187,
+      "learning_rate": 2.934328725214215e-06,
+      "loss": 0.0719,
+      "step": 7673
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.505938677170983,
+      "learning_rate": 2.9330860975338592e-06,
+      "loss": 0.1642,
+      "step": 7674
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.22631945575667362,
+      "learning_rate": 2.9318436238092917e-06,
+      "loss": 0.074,
+      "step": 7675
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3580393207810524,
+      "learning_rate": 2.93060130413306e-06,
+      "loss": 0.0832,
+      "step": 7676
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2521019471724546,
+      "learning_rate": 2.929359138597698e-06,
+      "loss": 0.0735,
+      "step": 7677
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.28287885008670893,
+      "learning_rate": 2.9281171272957275e-06,
+      "loss": 0.0799,
+      "step": 7678
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5348398919167312,
+      "learning_rate": 2.9268752703196645e-06,
+      "loss": 0.1071,
+      "step": 7679
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.22248502795144798,
+      "learning_rate": 2.925633567762009e-06,
+      "loss": 0.0659,
+      "step": 7680
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4584132035139908,
+      "learning_rate": 2.924392019715247e-06,
+      "loss": 0.1124,
+      "step": 7681
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4183298450255073,
+      "learning_rate": 2.9231506262718612e-06,
+      "loss": 0.0842,
+      "step": 7682
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.26046234945738206,
+      "learning_rate": 2.9219093875243144e-06,
+      "loss": 0.0571,
+      "step": 7683
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4564274945246979,
+      "learning_rate": 2.920668303565063e-06,
+      "loss": 0.1033,
+      "step": 7684
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4381342566983733,
+      "learning_rate": 2.919427374486547e-06,
+      "loss": 0.1129,
+      "step": 7685
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3263740523874261,
+      "learning_rate": 2.9181866003812025e-06,
+      "loss": 0.0732,
+      "step": 7686
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.17927566082124533,
+      "learning_rate": 2.9169459813414477e-06,
+      "loss": 0.0603,
+      "step": 7687
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.24534618688691814,
+      "learning_rate": 2.915705517459688e-06,
+      "loss": 0.0631,
+      "step": 7688
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4702917318921455,
+      "learning_rate": 2.9144652088283264e-06,
+      "loss": 0.08,
+      "step": 7689
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6530921816652279,
+      "learning_rate": 2.9132250555397445e-06,
+      "loss": 0.1248,
+      "step": 7690
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.36946832985721206,
+      "learning_rate": 2.9119850576863175e-06,
+      "loss": 0.1047,
+      "step": 7691
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.237938050649642,
+      "learning_rate": 2.910745215360404e-06,
+      "loss": 0.0743,
+      "step": 7692
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5283002663052525,
+      "learning_rate": 2.9095055286543605e-06,
+      "loss": 0.1435,
+      "step": 7693
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3016769280838413,
+      "learning_rate": 2.9082659976605226e-06,
+      "loss": 0.0627,
+      "step": 7694
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.1457954373829542,
+      "learning_rate": 2.907026622471217e-06,
+      "loss": 0.0418,
+      "step": 7695
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3808649438057077,
+      "learning_rate": 2.9057874031787624e-06,
+      "loss": 0.0584,
+      "step": 7696
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.37938589666766376,
+      "learning_rate": 2.9045483398754628e-06,
+      "loss": 0.0953,
+      "step": 7697
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5125293923537707,
+      "learning_rate": 2.9033094326536063e-06,
+      "loss": 0.128,
+      "step": 7698
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.23634379723225254,
+      "learning_rate": 2.90207068160548e-06,
+      "loss": 0.0616,
+      "step": 7699
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4405512483281603,
+      "learning_rate": 2.9008320868233507e-06,
+      "loss": 0.067,
+      "step": 7700
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3237926244621433,
+      "learning_rate": 2.899593648399477e-06,
+      "loss": 0.0784,
+      "step": 7701
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.30430600796467955,
+      "learning_rate": 2.898355366426101e-06,
+      "loss": 0.0644,
+      "step": 7702
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4212062858735066,
+      "learning_rate": 2.897117240995463e-06,
+      "loss": 0.1017,
+      "step": 7703
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3278006754646287,
+      "learning_rate": 2.8958792721997823e-06,
+      "loss": 0.0791,
+      "step": 7704
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.26884786771859026,
+      "learning_rate": 2.8946414601312696e-06,
+      "loss": 0.0878,
+      "step": 7705
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6112089498822476,
+      "learning_rate": 2.893403804882127e-06,
+      "loss": 0.1099,
+      "step": 7706
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.35748296203113894,
+      "learning_rate": 2.8921663065445414e-06,
+      "loss": 0.0929,
+      "step": 7707
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2454367672306242,
+      "learning_rate": 2.8909289652106877e-06,
+      "loss": 0.049,
+      "step": 7708
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.29977887854425456,
+      "learning_rate": 2.889691780972729e-06,
+      "loss": 0.0704,
+      "step": 7709
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.20627984595493482,
+      "learning_rate": 2.8884547539228225e-06,
+      "loss": 0.0832,
+      "step": 7710
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.26327292206158076,
+      "learning_rate": 2.887217884153106e-06,
+      "loss": 0.0655,
+      "step": 7711
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.29190156002195655,
+      "learning_rate": 2.885981171755708e-06,
+      "loss": 0.0618,
+      "step": 7712
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6801916682029193,
+      "learning_rate": 2.8847446168227487e-06,
+      "loss": 0.1327,
+      "step": 7713
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3070634528299016,
+      "learning_rate": 2.8835082194463324e-06,
+      "loss": 0.0821,
+      "step": 7714
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.24887581119208113,
+      "learning_rate": 2.882271979718553e-06,
+      "loss": 0.0839,
+      "step": 7715
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2973307206098664,
+      "learning_rate": 2.8810358977314935e-06,
+      "loss": 0.0979,
+      "step": 7716
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.25893381919528624,
+      "learning_rate": 2.879799973577225e-06,
+      "loss": 0.0707,
+      "step": 7717
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.38457707097676397,
+      "learning_rate": 2.8785642073478054e-06,
+      "loss": 0.0962,
+      "step": 7718
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3778350828875611,
+      "learning_rate": 2.877328599135282e-06,
+      "loss": 0.1255,
+      "step": 7719
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.22353347419610237,
+      "learning_rate": 2.876093149031688e-06,
+      "loss": 0.0746,
+      "step": 7720
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.23773070229008716,
+      "learning_rate": 2.8748578571290507e-06,
+      "loss": 0.077,
+      "step": 7721
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.18234505482289376,
+      "learning_rate": 2.87362272351938e-06,
+      "loss": 0.0498,
+      "step": 7722
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3580999734792023,
+      "learning_rate": 2.8723877482946762e-06,
+      "loss": 0.1087,
+      "step": 7723
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3466034664032849,
+      "learning_rate": 2.871152931546925e-06,
+      "loss": 0.0784,
+      "step": 7724
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.27517761546731734,
+      "learning_rate": 2.869918273368107e-06,
+      "loss": 0.0503,
+      "step": 7725
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.33251455436398075,
+      "learning_rate": 2.868683773850185e-06,
+      "loss": 0.0788,
+      "step": 7726
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2634912591045007,
+      "learning_rate": 2.867449433085111e-06,
+      "loss": 0.0631,
+      "step": 7727
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3223917887124083,
+      "learning_rate": 2.866215251164824e-06,
+      "loss": 0.0736,
+      "step": 7728
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.390254519818775,
+      "learning_rate": 2.864981228181257e-06,
+      "loss": 0.0781,
+      "step": 7729
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.17689229295112066,
+      "learning_rate": 2.863747364226327e-06,
+      "loss": 0.0496,
+      "step": 7730
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3741799012458619,
+      "learning_rate": 2.8625136593919368e-06,
+      "loss": 0.1193,
+      "step": 7731
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.38117468634003,
+      "learning_rate": 2.8612801137699786e-06,
+      "loss": 0.0845,
+      "step": 7732
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.22853354874061838,
+      "learning_rate": 2.8600467274523392e-06,
+      "loss": 0.0614,
+      "step": 7733
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.17525877578123059,
+      "learning_rate": 2.858813500530886e-06,
+      "loss": 0.0533,
+      "step": 7734
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.327697450578931,
+      "learning_rate": 2.857580433097476e-06,
+      "loss": 0.0637,
+      "step": 7735
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.25955956707958583,
+      "learning_rate": 2.8563475252439544e-06,
+      "loss": 0.079,
+      "step": 7736
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.36288134821774254,
+      "learning_rate": 2.855114777062158e-06,
+      "loss": 0.0748,
+      "step": 7737
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.30444640532718537,
+      "learning_rate": 2.853882188643908e-06,
+      "loss": 0.0766,
+      "step": 7738
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3056367521384169,
+      "learning_rate": 2.8526497600810133e-06,
+      "loss": 0.0747,
+      "step": 7739
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2728911026014942,
+      "learning_rate": 2.851417491465276e-06,
+      "loss": 0.0404,
+      "step": 7740
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2262891081268102,
+      "learning_rate": 2.8501853828884795e-06,
+      "loss": 0.0705,
+      "step": 7741
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2633499713807462,
+      "learning_rate": 2.8489534344424e-06,
+      "loss": 0.0813,
+      "step": 7742
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3990478145803632,
+      "learning_rate": 2.847721646218796e-06,
+      "loss": 0.0926,
+      "step": 7743
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.28539402583828705,
+      "learning_rate": 2.8464900183094256e-06,
+      "loss": 0.0748,
+      "step": 7744
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3222072327034806,
+      "learning_rate": 2.8452585508060225e-06,
+      "loss": 0.1009,
+      "step": 7745
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.29571964393787387,
+      "learning_rate": 2.8440272438003136e-06,
+      "loss": 0.0684,
+      "step": 7746
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.29543316741360764,
+      "learning_rate": 2.842796097384016e-06,
+      "loss": 0.0945,
+      "step": 7747
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.189283231814599,
+      "learning_rate": 2.841565111648833e-06,
+      "loss": 0.1147,
+      "step": 7748
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3870682024027058,
+      "learning_rate": 2.8403342866864514e-06,
+      "loss": 0.1044,
+      "step": 7749
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2427672205820367,
+      "learning_rate": 2.839103622588556e-06,
+      "loss": 0.0792,
+      "step": 7750
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.37481290294364916,
+      "learning_rate": 2.8378731194468103e-06,
+      "loss": 0.0812,
+      "step": 7751
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3615292004364502,
+      "learning_rate": 2.83664277735287e-06,
+      "loss": 0.1025,
+      "step": 7752
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3066090494529635,
+      "learning_rate": 2.835412596398376e-06,
+      "loss": 0.095,
+      "step": 7753
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.40884492792958027,
+      "learning_rate": 2.8341825766749635e-06,
+      "loss": 0.0731,
+      "step": 7754
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3268826758852914,
+      "learning_rate": 2.83295271827425e-06,
+      "loss": 0.0558,
+      "step": 7755
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3354642759847708,
+      "learning_rate": 2.83172302128784e-06,
+      "loss": 0.0969,
+      "step": 7756
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.40071439121645547,
+      "learning_rate": 2.830493485807332e-06,
+      "loss": 0.0729,
+      "step": 7757
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.29087465458835,
+      "learning_rate": 2.8292641119243074e-06,
+      "loss": 0.0641,
+      "step": 7758
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.22827017699942953,
+      "learning_rate": 2.8280348997303376e-06,
+      "loss": 0.0623,
+      "step": 7759
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.508477524648532,
+      "learning_rate": 2.8268058493169793e-06,
+      "loss": 0.0763,
+      "step": 7760
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.465987828406582,
+      "learning_rate": 2.8255769607757826e-06,
+      "loss": 0.1108,
+      "step": 7761
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.21670620629936793,
+      "learning_rate": 2.8243482341982807e-06,
+      "loss": 0.0503,
+      "step": 7762
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.49318093587915685,
+      "learning_rate": 2.8231196696759956e-06,
+      "loss": 0.1072,
+      "step": 7763
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.267304769734604,
+      "learning_rate": 2.8218912673004394e-06,
+      "loss": 0.0723,
+      "step": 7764
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3441234836373086,
+      "learning_rate": 2.8206630271631108e-06,
+      "loss": 0.0616,
+      "step": 7765
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2819235176435247,
+      "learning_rate": 2.8194349493554935e-06,
+      "loss": 0.0548,
+      "step": 7766
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.26834172068321543,
+      "learning_rate": 2.818207033969066e-06,
+      "loss": 0.0559,
+      "step": 7767
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.327223231205228,
+      "learning_rate": 2.816979281095288e-06,
+      "loss": 0.0577,
+      "step": 7768
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4358538600455522,
+      "learning_rate": 2.81575169082561e-06,
+      "loss": 0.0863,
+      "step": 7769
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.22055821894129377,
+      "learning_rate": 2.814524263251469e-06,
+      "loss": 0.0477,
+      "step": 7770
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2574180088296187,
+      "learning_rate": 2.813296998464293e-06,
+      "loss": 0.0862,
+      "step": 7771
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3828656067902435,
+      "learning_rate": 2.8120698965554967e-06,
+      "loss": 0.1086,
+      "step": 7772
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2144904971797863,
+      "learning_rate": 2.8108429576164775e-06,
+      "loss": 0.0493,
+      "step": 7773
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5926798224449158,
+      "learning_rate": 2.809616181738629e-06,
+      "loss": 0.0884,
+      "step": 7774
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.39327353425927253,
+      "learning_rate": 2.808389569013328e-06,
+      "loss": 0.121,
+      "step": 7775
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.33730792100709706,
+      "learning_rate": 2.807163119531938e-06,
+      "loss": 0.1116,
+      "step": 7776
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2543890407827106,
+      "learning_rate": 2.805936833385812e-06,
+      "loss": 0.0796,
+      "step": 7777
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.17976443099663517,
+      "learning_rate": 2.804710710666294e-06,
+      "loss": 0.0337,
+      "step": 7778
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.23361126209148086,
+      "learning_rate": 2.80348475146471e-06,
+      "loss": 0.0402,
+      "step": 7779
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.29027813613608794,
+      "learning_rate": 2.8022589558723763e-06,
+      "loss": 0.0675,
+      "step": 7780
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3977364594731456,
+      "learning_rate": 2.8010333239805997e-06,
+      "loss": 0.0785,
+      "step": 7781
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.32304287409554566,
+      "learning_rate": 2.7998078558806712e-06,
+      "loss": 0.0812,
+      "step": 7782
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4061670729655189,
+      "learning_rate": 2.7985825516638685e-06,
+      "loss": 0.1093,
+      "step": 7783
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.23842240679660404,
+      "learning_rate": 2.797357411421464e-06,
+      "loss": 0.0225,
+      "step": 7784
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2305189449582252,
+      "learning_rate": 2.7961324352447106e-06,
+      "loss": 0.0631,
+      "step": 7785
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.45914660814205205,
+      "learning_rate": 2.7949076232248516e-06,
+      "loss": 0.0885,
+      "step": 7786
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.17680099406509622,
+      "learning_rate": 2.7936829754531192e-06,
+      "loss": 0.0495,
+      "step": 7787
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2053245544012522,
+      "learning_rate": 2.7924584920207287e-06,
+      "loss": 0.0355,
+      "step": 7788
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3549145613488633,
+      "learning_rate": 2.791234173018892e-06,
+      "loss": 0.0618,
+      "step": 7789
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.31950699897421087,
+      "learning_rate": 2.790010018538801e-06,
+      "loss": 0.0824,
+      "step": 7790
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3568178312543278,
+      "learning_rate": 2.7887860286716383e-06,
+      "loss": 0.0752,
+      "step": 7791
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.42522926563767893,
+      "learning_rate": 2.7875622035085713e-06,
+      "loss": 0.0707,
+      "step": 7792
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2639977703192523,
+      "learning_rate": 2.786338543140762e-06,
+      "loss": 0.0784,
+      "step": 7793
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3697239805784935,
+      "learning_rate": 2.7851150476593523e-06,
+      "loss": 0.082,
+      "step": 7794
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3751595180440534,
+      "learning_rate": 2.783891717155478e-06,
+      "loss": 0.0393,
+      "step": 7795
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3198995296957473,
+      "learning_rate": 2.782668551720256e-06,
+      "loss": 0.1074,
+      "step": 7796
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5890423476203442,
+      "learning_rate": 2.781445551444799e-06,
+      "loss": 0.0895,
+      "step": 7797
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.26591142817913277,
+      "learning_rate": 2.780222716420201e-06,
+      "loss": 0.0606,
+      "step": 7798
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4395782770188453,
+      "learning_rate": 2.7790000467375467e-06,
+      "loss": 0.104,
+      "step": 7799
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.47830707903245,
+      "learning_rate": 2.777777542487905e-06,
+      "loss": 0.1229,
+      "step": 7800
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3234717114586146,
+      "learning_rate": 2.77655520376234e-06,
+      "loss": 0.0521,
+      "step": 7801
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.1814886558484413,
+      "learning_rate": 2.775333030651895e-06,
+      "loss": 0.0559,
+      "step": 7802
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3798997246875799,
+      "learning_rate": 2.774111023247606e-06,
+      "loss": 0.1199,
+      "step": 7803
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5250078719229171,
+      "learning_rate": 2.7728891816404922e-06,
+      "loss": 0.1141,
+      "step": 7804
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3577445476544192,
+      "learning_rate": 2.771667505921569e-06,
+      "loss": 0.0804,
+      "step": 7805
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.35737388718942253,
+      "learning_rate": 2.77044599618183e-06,
+      "loss": 0.0795,
+      "step": 7806
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4059075186476327,
+      "learning_rate": 2.7692246525122603e-06,
+      "loss": 0.1059,
+      "step": 7807
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.31507643981614714,
+      "learning_rate": 2.7680034750038354e-06,
+      "loss": 0.0838,
+      "step": 7808
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.27236420278011086,
+      "learning_rate": 2.7667824637475137e-06,
+      "loss": 0.045,
+      "step": 7809
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.25166108048761,
+      "learning_rate": 2.7655616188342437e-06,
+      "loss": 0.0696,
+      "step": 7810
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.31330747265961656,
+      "learning_rate": 2.7643409403549587e-06,
+      "loss": 0.1096,
+      "step": 7811
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.12588044884692123,
+      "learning_rate": 2.7631204284005863e-06,
+      "loss": 0.025,
+      "step": 7812
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3687055559235135,
+      "learning_rate": 2.7619000830620347e-06,
+      "loss": 0.0777,
+      "step": 7813
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2656017790739315,
+      "learning_rate": 2.7606799044302013e-06,
+      "loss": 0.0646,
+      "step": 7814
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5328935801261846,
+      "learning_rate": 2.759459892595976e-06,
+      "loss": 0.1435,
+      "step": 7815
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.351131237976295,
+      "learning_rate": 2.7582400476502293e-06,
+      "loss": 0.0767,
+      "step": 7816
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2699508551993985,
+      "learning_rate": 2.757020369683824e-06,
+      "loss": 0.0701,
+      "step": 7817
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2582939422333032,
+      "learning_rate": 2.755800858787605e-06,
+      "loss": 0.092,
+      "step": 7818
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2121190075548633,
+      "learning_rate": 2.7545815150524134e-06,
+      "loss": 0.0625,
+      "step": 7819
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3533572052613979,
+      "learning_rate": 2.7533623385690716e-06,
+      "loss": 0.096,
+      "step": 7820
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4350425122952263,
+      "learning_rate": 2.7521433294283884e-06,
+      "loss": 0.1098,
+      "step": 7821
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.40485789113325527,
+      "learning_rate": 2.7509244877211666e-06,
+      "loss": 0.1183,
+      "step": 7822
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.41907797938320884,
+      "learning_rate": 2.7497058135381903e-06,
+      "loss": 0.0969,
+      "step": 7823
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.41703583241675096,
+      "learning_rate": 2.748487306970233e-06,
+      "loss": 0.1044,
+      "step": 7824
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3705024668974954,
+      "learning_rate": 2.747268968108059e-06,
+      "loss": 0.0806,
+      "step": 7825
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.32594385451834257,
+      "learning_rate": 2.7460507970424143e-06,
+      "loss": 0.1168,
+      "step": 7826
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.1816370898887897,
+      "learning_rate": 2.7448327938640374e-06,
+      "loss": 0.0627,
+      "step": 7827
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5533248363859766,
+      "learning_rate": 2.743614958663649e-06,
+      "loss": 0.0649,
+      "step": 7828
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.29956056363042194,
+      "learning_rate": 2.7423972915319653e-06,
+      "loss": 0.0574,
+      "step": 7829
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4491376535673036,
+      "learning_rate": 2.7411797925596818e-06,
+      "loss": 0.1129,
+      "step": 7830
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2017419965726022,
+      "learning_rate": 2.739962461837485e-06,
+      "loss": 0.0509,
+      "step": 7831
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.48957780482574104,
+      "learning_rate": 2.7387452994560517e-06,
+      "loss": 0.1025,
+      "step": 7832
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.322230990206635,
+      "learning_rate": 2.7375283055060405e-06,
+      "loss": 0.1073,
+      "step": 7833
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2993773628773903,
+      "learning_rate": 2.736311480078101e-06,
+      "loss": 0.0782,
+      "step": 7834
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.24678599431252085,
+      "learning_rate": 2.735094823262868e-06,
+      "loss": 0.0477,
+      "step": 7835
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6238278960056596,
+      "learning_rate": 2.733878335150969e-06,
+      "loss": 0.0616,
+      "step": 7836
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.25164155024049567,
+      "learning_rate": 2.732662015833012e-06,
+      "loss": 0.0567,
+      "step": 7837
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3205467252916991,
+      "learning_rate": 2.7314458653995946e-06,
+      "loss": 0.0891,
+      "step": 7838
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3074138568856897,
+      "learning_rate": 2.7302298839413062e-06,
+      "loss": 0.0908,
+      "step": 7839
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.277509210014652,
+      "learning_rate": 2.7290140715487185e-06,
+      "loss": 0.0896,
+      "step": 7840
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.23960097747581732,
+      "learning_rate": 2.7277984283123906e-06,
+      "loss": 0.0755,
+      "step": 7841
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.36162889905909923,
+      "learning_rate": 2.726582954322874e-06,
+      "loss": 0.0816,
+      "step": 7842
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.26884981010338127,
+      "learning_rate": 2.725367649670702e-06,
+      "loss": 0.0698,
+      "step": 7843
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.285403732899656,
+      "learning_rate": 2.7241525144463987e-06,
+      "loss": 0.0706,
+      "step": 7844
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4037689316534924,
+      "learning_rate": 2.722937548740472e-06,
+      "loss": 0.1042,
+      "step": 7845
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.29525894508852857,
+      "learning_rate": 2.7217227526434237e-06,
+      "loss": 0.0759,
+      "step": 7846
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.1708952397844576,
+      "learning_rate": 2.7205081262457366e-06,
+      "loss": 0.0417,
+      "step": 7847
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.32597083290489226,
+      "learning_rate": 2.719293669637881e-06,
+      "loss": 0.0773,
+      "step": 7848
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.24411555223661527,
+      "learning_rate": 2.7180793829103214e-06,
+      "loss": 0.0655,
+      "step": 7849
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3067400517368821,
+      "learning_rate": 2.716865266153502e-06,
+      "loss": 0.0734,
+      "step": 7850
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3299897782231123,
+      "learning_rate": 2.7156513194578584e-06,
+      "loss": 0.057,
+      "step": 7851
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.28077228179291125,
+      "learning_rate": 2.714437542913809e-06,
+      "loss": 0.0672,
+      "step": 7852
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.17620731993091054,
+      "learning_rate": 2.7132239366117684e-06,
+      "loss": 0.0521,
+      "step": 7853
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.19853743567584312,
+      "learning_rate": 2.712010500642131e-06,
+      "loss": 0.0466,
+      "step": 7854
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3335095358961763,
+      "learning_rate": 2.7107972350952795e-06,
+      "loss": 0.0981,
+      "step": 7855
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.48050156576743913,
+      "learning_rate": 2.709584140061583e-06,
+      "loss": 0.1453,
+      "step": 7856
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.40536760072149425,
+      "learning_rate": 2.708371215631405e-06,
+      "loss": 0.0975,
+      "step": 7857
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.35686882452031177,
+      "learning_rate": 2.7071584618950886e-06,
+      "loss": 0.0902,
+      "step": 7858
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.19574659721337634,
+      "learning_rate": 2.705945878942967e-06,
+      "loss": 0.0587,
+      "step": 7859
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.20187919848150349,
+      "learning_rate": 2.7047334668653574e-06,
+      "loss": 0.0525,
+      "step": 7860
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3547963410962722,
+      "learning_rate": 2.703521225752572e-06,
+      "loss": 0.0825,
+      "step": 7861
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4694460516804668,
+      "learning_rate": 2.7023091556949048e-06,
+      "loss": 0.0938,
+      "step": 7862
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3161041146096953,
+      "learning_rate": 2.7010972567826365e-06,
+      "loss": 0.0854,
+      "step": 7863
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.31826629587531613,
+      "learning_rate": 2.699885529106034e-06,
+      "loss": 0.1001,
+      "step": 7864
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2752697481315665,
+      "learning_rate": 2.698673972755359e-06,
+      "loss": 0.0806,
+      "step": 7865
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4269416297414388,
+      "learning_rate": 2.697462587820852e-06,
+      "loss": 0.0866,
+      "step": 7866
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.34594113775183505,
+      "learning_rate": 2.6962513743927454e-06,
+      "loss": 0.0821,
+      "step": 7867
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3188123008229541,
+      "learning_rate": 2.695040332561256e-06,
+      "loss": 0.0768,
+      "step": 7868
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.29597609788007545,
+      "learning_rate": 2.693829462416589e-06,
+      "loss": 0.0697,
+      "step": 7869
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.23628018380265178,
+      "learning_rate": 2.692618764048939e-06,
+      "loss": 0.0619,
+      "step": 7870
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4242817761984848,
+      "learning_rate": 2.6914082375484863e-06,
+      "loss": 0.112,
+      "step": 7871
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.35032744474092903,
+      "learning_rate": 2.690197883005393e-06,
+      "loss": 0.0639,
+      "step": 7872
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2634822542471033,
+      "learning_rate": 2.68898770050982e-06,
+      "loss": 0.0438,
+      "step": 7873
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4410345109038879,
+      "learning_rate": 2.687777690151906e-06,
+      "loss": 0.1062,
+      "step": 7874
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.29247298539355265,
+      "learning_rate": 2.6865678520217765e-06,
+      "loss": 0.0693,
+      "step": 7875
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5402063674634102,
+      "learning_rate": 2.6853581862095524e-06,
+      "loss": 0.099,
+      "step": 7876
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.1725202969911037,
+      "learning_rate": 2.684148692805335e-06,
+      "loss": 0.0723,
+      "step": 7877
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2177217934113603,
+      "learning_rate": 2.682939371899213e-06,
+      "loss": 0.0622,
+      "step": 7878
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.23646372464505894,
+      "learning_rate": 2.6817302235812625e-06,
+      "loss": 0.0636,
+      "step": 7879
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3756749996670768,
+      "learning_rate": 2.680521247941552e-06,
+      "loss": 0.0995,
+      "step": 7880
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.47637620095847244,
+      "learning_rate": 2.6793124450701314e-06,
+      "loss": 0.1053,
+      "step": 7881
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2656782541565853,
+      "learning_rate": 2.678103815057036e-06,
+      "loss": 0.0471,
+      "step": 7882
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3342829250824342,
+      "learning_rate": 2.6768953579922976e-06,
+      "loss": 0.073,
+      "step": 7883
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.27171851169235217,
+      "learning_rate": 2.6756870739659247e-06,
+      "loss": 0.0741,
+      "step": 7884
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.18468439560027985,
+      "learning_rate": 2.6744789630679198e-06,
+      "loss": 0.0491,
+      "step": 7885
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.18093937977521382,
+      "learning_rate": 2.6732710253882665e-06,
+      "loss": 0.051,
+      "step": 7886
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.44470270813498286,
+      "learning_rate": 2.6720632610169426e-06,
+      "loss": 0.0852,
+      "step": 7887
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.1773735011019568,
+      "learning_rate": 2.6708556700439094e-06,
+      "loss": 0.0438,
+      "step": 7888
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2064305264420676,
+      "learning_rate": 2.6696482525591116e-06,
+      "loss": 0.0488,
+      "step": 7889
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5828310543377022,
+      "learning_rate": 2.6684410086524896e-06,
+      "loss": 0.0909,
+      "step": 7890
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2751868848328496,
+      "learning_rate": 2.6672339384139634e-06,
+      "loss": 0.0484,
+      "step": 7891
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.49140815995877346,
+      "learning_rate": 2.666027041933441e-06,
+      "loss": 0.1199,
+      "step": 7892
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3817822901672961,
+      "learning_rate": 2.6648203193008225e-06,
+      "loss": 0.0905,
+      "step": 7893
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3125312302618564,
+      "learning_rate": 2.6636137706059906e-06,
+      "loss": 0.0552,
+      "step": 7894
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2995485084664994,
+      "learning_rate": 2.6624073959388153e-06,
+      "loss": 0.0932,
+      "step": 7895
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.28106045259664436,
+      "learning_rate": 2.6612011953891526e-06,
+      "loss": 0.0807,
+      "step": 7896
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.22649314143287447,
+      "learning_rate": 2.6599951690468516e-06,
+      "loss": 0.0649,
+      "step": 7897
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2602532925856686,
+      "learning_rate": 2.658789317001742e-06,
+      "loss": 0.0284,
+      "step": 7898
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.25984306927463574,
+      "learning_rate": 2.6575836393436407e-06,
+      "loss": 0.067,
+      "step": 7899
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.24044049077688343,
+      "learning_rate": 2.6563781361623563e-06,
+      "loss": 0.0779,
+      "step": 7900
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2551521840030982,
+      "learning_rate": 2.6551728075476823e-06,
+      "loss": 0.0522,
+      "step": 7901
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3096010478802843,
+      "learning_rate": 2.6539676535893967e-06,
+      "loss": 0.0829,
+      "step": 7902
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.20417451457778238,
+      "learning_rate": 2.652762674377265e-06,
+      "loss": 0.0576,
+      "step": 7903
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.29062435408955417,
+      "learning_rate": 2.651557870001045e-06,
+      "loss": 0.0418,
+      "step": 7904
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2628041823285726,
+      "learning_rate": 2.650353240550475e-06,
+      "loss": 0.0581,
+      "step": 7905
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.37303414114705163,
+      "learning_rate": 2.649148786115282e-06,
+      "loss": 0.1083,
+      "step": 7906
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.30263834980057,
+      "learning_rate": 2.647944506785184e-06,
+      "loss": 0.1086,
+      "step": 7907
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.263391040234311,
+      "learning_rate": 2.6467404026498803e-06,
+      "loss": 0.0674,
+      "step": 7908
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.19306731159863216,
+      "learning_rate": 2.645536473799059e-06,
+      "loss": 0.0579,
+      "step": 7909
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.4793996899831052,
+      "learning_rate": 2.644332720322399e-06,
+      "loss": 0.1167,
+      "step": 7910
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.503266095195825,
+      "learning_rate": 2.6431291423095594e-06,
+      "loss": 0.1261,
+      "step": 7911
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.35496053492939644,
+      "learning_rate": 2.6419257398501923e-06,
+      "loss": 0.1046,
+      "step": 7912
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.49066334216129187,
+      "learning_rate": 2.640722513033931e-06,
+      "loss": 0.0974,
+      "step": 7913
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.43497272531607867,
+      "learning_rate": 2.6395194619504024e-06,
+      "loss": 0.0759,
+      "step": 7914
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.29366808496240576,
+      "learning_rate": 2.6383165866892146e-06,
+      "loss": 0.0611,
+      "step": 7915
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2525665534002413,
+      "learning_rate": 2.6371138873399637e-06,
+      "loss": 0.0414,
+      "step": 7916
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2448144855770156,
+      "learning_rate": 2.6359113639922375e-06,
+      "loss": 0.067,
+      "step": 7917
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3867596272315621,
+      "learning_rate": 2.634709016735604e-06,
+      "loss": 0.1151,
+      "step": 7918
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3210786728905828,
+      "learning_rate": 2.6335068456596223e-06,
+      "loss": 0.0727,
+      "step": 7919
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2020164693887553,
+      "learning_rate": 2.6323048508538353e-06,
+      "loss": 0.0433,
+      "step": 7920
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.4044527713005456,
+      "learning_rate": 2.631103032407777e-06,
+      "loss": 0.1029,
+      "step": 7921
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5106797358987005,
+      "learning_rate": 2.6299013904109645e-06,
+      "loss": 0.0602,
+      "step": 7922
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.36700643800471494,
+      "learning_rate": 2.6286999249529032e-06,
+      "loss": 0.1165,
+      "step": 7923
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3385525640357685,
+      "learning_rate": 2.6274986361230836e-06,
+      "loss": 0.1096,
+      "step": 7924
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.28467652376479996,
+      "learning_rate": 2.626297524010989e-06,
+      "loss": 0.0641,
+      "step": 7925
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.23735562169568133,
+      "learning_rate": 2.625096588706082e-06,
+      "loss": 0.0313,
+      "step": 7926
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.22092309069410704,
+      "learning_rate": 2.6238958302978163e-06,
+      "loss": 0.077,
+      "step": 7927
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7322307021203499,
+      "learning_rate": 2.6226952488756284e-06,
+      "loss": 0.1426,
+      "step": 7928
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.16958910564242013,
+      "learning_rate": 2.62149484452895e-06,
+      "loss": 0.0436,
+      "step": 7929
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5693882229237957,
+      "learning_rate": 2.620294617347191e-06,
+      "loss": 0.1399,
+      "step": 7930
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.32913931125206497,
+      "learning_rate": 2.6190945674197513e-06,
+      "loss": 0.107,
+      "step": 7931
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3925179355450706,
+      "learning_rate": 2.617894694836017e-06,
+      "loss": 0.113,
+      "step": 7932
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3578974717149245,
+      "learning_rate": 2.616694999685364e-06,
+      "loss": 0.1156,
+      "step": 7933
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.38182302238121263,
+      "learning_rate": 2.615495482057151e-06,
+      "loss": 0.1016,
+      "step": 7934
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7766674072775798,
+      "learning_rate": 2.6142961420407252e-06,
+      "loss": 0.1739,
+      "step": 7935
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.39578034441327137,
+      "learning_rate": 2.6130969797254212e-06,
+      "loss": 0.1194,
+      "step": 7936
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3100158836957796,
+      "learning_rate": 2.611897995200556e-06,
+      "loss": 0.0942,
+      "step": 7937
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.28223776458272143,
+      "learning_rate": 2.610699188555442e-06,
+      "loss": 0.0775,
+      "step": 7938
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.351301666317223,
+      "learning_rate": 2.609500559879372e-06,
+      "loss": 0.0881,
+      "step": 7939
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.28547956647131745,
+      "learning_rate": 2.6083021092616235e-06,
+      "loss": 0.091,
+      "step": 7940
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.1851543090787496,
+      "learning_rate": 2.607103836791468e-06,
+      "loss": 0.0452,
+      "step": 7941
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.22685978835876186,
+      "learning_rate": 2.605905742558159e-06,
+      "loss": 0.0562,
+      "step": 7942
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.36176116326103597,
+      "learning_rate": 2.6047078266509372e-06,
+      "loss": 0.096,
+      "step": 7943
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2691616197366879,
+      "learning_rate": 2.6035100891590277e-06,
+      "loss": 0.0503,
+      "step": 7944
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3631583931559883,
+      "learning_rate": 2.602312530171649e-06,
+      "loss": 0.0941,
+      "step": 7945
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.4191126518876571,
+      "learning_rate": 2.601115149778001e-06,
+      "loss": 0.1266,
+      "step": 7946
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2590201740525622,
+      "learning_rate": 2.5999179480672694e-06,
+      "loss": 0.0896,
+      "step": 7947
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2377390387611689,
+      "learning_rate": 2.5987209251286323e-06,
+      "loss": 0.0619,
+      "step": 7948
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.46279872092274565,
+      "learning_rate": 2.597524081051249e-06,
+      "loss": 0.0912,
+      "step": 7949
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2934251277913643,
+      "learning_rate": 2.596327415924266e-06,
+      "loss": 0.0647,
+      "step": 7950
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5327006757661311,
+      "learning_rate": 2.595130929836821e-06,
+      "loss": 0.147,
+      "step": 7951
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.4739423815248122,
+      "learning_rate": 2.593934622878034e-06,
+      "loss": 0.0582,
+      "step": 7952
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.44691232993665647,
+      "learning_rate": 2.592738495137013e-06,
+      "loss": 0.0814,
+      "step": 7953
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.4520183714963539,
+      "learning_rate": 2.5915425467028498e-06,
+      "loss": 0.1168,
+      "step": 7954
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.28150655015612336,
+      "learning_rate": 2.5903467776646297e-06,
+      "loss": 0.0945,
+      "step": 7955
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.42923237418568994,
+      "learning_rate": 2.5891511881114197e-06,
+      "loss": 0.1267,
+      "step": 7956
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.22550780075808677,
+      "learning_rate": 2.5879557781322705e-06,
+      "loss": 0.0406,
+      "step": 7957
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.21929245027387836,
+      "learning_rate": 2.586760547816229e-06,
+      "loss": 0.0618,
+      "step": 7958
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.21636772461612827,
+      "learning_rate": 2.5855654972523196e-06,
+      "loss": 0.064,
+      "step": 7959
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3661217157205062,
+      "learning_rate": 2.5843706265295575e-06,
+      "loss": 0.0972,
+      "step": 7960
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2097512488024827,
+      "learning_rate": 2.583175935736941e-06,
+      "loss": 0.0418,
+      "step": 7961
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.36329489770450357,
+      "learning_rate": 2.581981424963461e-06,
+      "loss": 0.0927,
+      "step": 7962
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.41026926072776804,
+      "learning_rate": 2.5807870942980916e-06,
+      "loss": 0.1117,
+      "step": 7963
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.35886973747634987,
+      "learning_rate": 2.57959294382979e-06,
+      "loss": 0.0965,
+      "step": 7964
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.29296103527102124,
+      "learning_rate": 2.5783989736475077e-06,
+      "loss": 0.0602,
+      "step": 7965
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3690539566551583,
+      "learning_rate": 2.577205183840177e-06,
+      "loss": 0.1113,
+      "step": 7966
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.325298871932655,
+      "learning_rate": 2.576011574496716e-06,
+      "loss": 0.0886,
+      "step": 7967
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.29421768994824826,
+      "learning_rate": 2.574818145706036e-06,
+      "loss": 0.0815,
+      "step": 7968
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.25158215701713493,
+      "learning_rate": 2.5736248975570277e-06,
+      "loss": 0.0941,
+      "step": 7969
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.33778250328556053,
+      "learning_rate": 2.5724318301385714e-06,
+      "loss": 0.1375,
+      "step": 7970
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.433937841332717,
+      "learning_rate": 2.5712389435395324e-06,
+      "loss": 0.0812,
+      "step": 7971
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3119580847235038,
+      "learning_rate": 2.570046237848768e-06,
+      "loss": 0.0917,
+      "step": 7972
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2953838538271283,
+      "learning_rate": 2.5688537131551144e-06,
+      "loss": 0.0653,
+      "step": 7973
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.23440058943056358,
+      "learning_rate": 2.567661369547397e-06,
+      "loss": 0.0741,
+      "step": 7974
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.4332995510552151,
+      "learning_rate": 2.5664692071144326e-06,
+      "loss": 0.1031,
+      "step": 7975
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.4160646948135601,
+      "learning_rate": 2.5652772259450174e-06,
+      "loss": 0.1167,
+      "step": 7976
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.19382790359075167,
+      "learning_rate": 2.5640854261279384e-06,
+      "loss": 0.0606,
+      "step": 7977
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.27681001395100363,
+      "learning_rate": 2.562893807751965e-06,
+      "loss": 0.0837,
+      "step": 7978
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2998379481154383,
+      "learning_rate": 2.5617023709058597e-06,
+      "loss": 0.0537,
+      "step": 7979
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.4277746405934601,
+      "learning_rate": 2.5605111156783663e-06,
+      "loss": 0.1055,
+      "step": 7980
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.36874101247388635,
+      "learning_rate": 2.559320042158214e-06,
+      "loss": 0.1018,
+      "step": 7981
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3061749805245895,
+      "learning_rate": 2.558129150434125e-06,
+      "loss": 0.0684,
+      "step": 7982
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.4120718525427058,
+      "learning_rate": 2.5569384405948015e-06,
+      "loss": 0.0877,
+      "step": 7983
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.28961813172953005,
+      "learning_rate": 2.555747912728933e-06,
+      "loss": 0.0528,
+      "step": 7984
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.255214812442905,
+      "learning_rate": 2.554557566925201e-06,
+      "loss": 0.0538,
+      "step": 7985
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.23039480923391234,
+      "learning_rate": 2.5533674032722665e-06,
+      "loss": 0.0865,
+      "step": 7986
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.22432749672141716,
+      "learning_rate": 2.5521774218587813e-06,
+      "loss": 0.0472,
+      "step": 7987
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.39131026170360134,
+      "learning_rate": 2.5509876227733788e-06,
+      "loss": 0.0822,
+      "step": 7988
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.22949298276261443,
+      "learning_rate": 2.549798006104687e-06,
+      "loss": 0.0746,
+      "step": 7989
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2730336172824576,
+      "learning_rate": 2.5486085719413127e-06,
+      "loss": 0.078,
+      "step": 7990
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5531843830311298,
+      "learning_rate": 2.5474193203718536e-06,
+      "loss": 0.1044,
+      "step": 7991
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.32602872359575064,
+      "learning_rate": 2.5462302514848874e-06,
+      "loss": 0.0774,
+      "step": 7992
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3356118578801749,
+      "learning_rate": 2.5450413653689885e-06,
+      "loss": 0.0889,
+      "step": 7993
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.15671592888727345,
+      "learning_rate": 2.5438526621127108e-06,
+      "loss": 0.0318,
+      "step": 7994
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.24736936035823426,
+      "learning_rate": 2.5426641418045946e-06,
+      "loss": 0.0605,
+      "step": 7995
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.20120185186239548,
+      "learning_rate": 2.5414758045331654e-06,
+      "loss": 0.0697,
+      "step": 7996
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.21887979379456166,
+      "learning_rate": 2.5402876503869432e-06,
+      "loss": 0.0485,
+      "step": 7997
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.24814881072050332,
+      "learning_rate": 2.5390996794544253e-06,
+      "loss": 0.0498,
+      "step": 7998
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.31650711019459493,
+      "learning_rate": 2.5379118918241e-06,
+      "loss": 0.0966,
+      "step": 7999
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.34400013619979486,
+      "learning_rate": 2.5367242875844365e-06,
+      "loss": 0.075,
+      "step": 8000
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.48691739006600604,
+      "learning_rate": 2.5355368668239e-06,
+      "loss": 0.107,
+      "step": 8001
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2548134432087641,
+      "learning_rate": 2.5343496296309345e-06,
+      "loss": 0.0479,
+      "step": 8002
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.37552813887767184,
+      "learning_rate": 2.5331625760939715e-06,
+      "loss": 0.0976,
+      "step": 8003
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2837505775068292,
+      "learning_rate": 2.5319757063014305e-06,
+      "loss": 0.0825,
+      "step": 8004
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2809265562096386,
+      "learning_rate": 2.5307890203417142e-06,
+      "loss": 0.1217,
+      "step": 8005
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.18501569087690484,
+      "learning_rate": 2.529602518303218e-06,
+      "loss": 0.0518,
+      "step": 8006
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3161016690811233,
+      "learning_rate": 2.5284162002743172e-06,
+      "loss": 0.0937,
+      "step": 8007
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.247218317365687,
+      "learning_rate": 2.527230066343374e-06,
+      "loss": 0.0768,
+      "step": 8008
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3240391471159267,
+      "learning_rate": 2.5260441165987426e-06,
+      "loss": 0.0919,
+      "step": 8009
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.42817202855167735,
+      "learning_rate": 2.524858351128757e-06,
+      "loss": 0.1394,
+      "step": 8010
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2516167682932726,
+      "learning_rate": 2.52367277002174e-06,
+      "loss": 0.0602,
+      "step": 8011
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4512517759269975,
+      "learning_rate": 2.5224873733659994e-06,
+      "loss": 0.0806,
+      "step": 8012
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3919637499405157,
+      "learning_rate": 2.5213021612498332e-06,
+      "loss": 0.0766,
+      "step": 8013
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.18140553394668607,
+      "learning_rate": 2.5201171337615216e-06,
+      "loss": 0.0589,
+      "step": 8014
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.12805926326406425,
+      "learning_rate": 2.51893229098933e-06,
+      "loss": 0.032,
+      "step": 8015
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.20761169057330878,
+      "learning_rate": 2.517747633021517e-06,
+      "loss": 0.0453,
+      "step": 8016
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2894182093207002,
+      "learning_rate": 2.516563159946321e-06,
+      "loss": 0.0943,
+      "step": 8017
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.20635762756910017,
+      "learning_rate": 2.5153788718519654e-06,
+      "loss": 0.0468,
+      "step": 8018
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2833224135851099,
+      "learning_rate": 2.514194768826667e-06,
+      "loss": 0.0584,
+      "step": 8019
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.25421539164092294,
+      "learning_rate": 2.5130108509586237e-06,
+      "loss": 0.0701,
+      "step": 8020
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.1964257456031457,
+      "learning_rate": 2.51182711833602e-06,
+      "loss": 0.0507,
+      "step": 8021
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.26858305410470207,
+      "learning_rate": 2.510643571047025e-06,
+      "loss": 0.0605,
+      "step": 8022
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5159884721712533,
+      "learning_rate": 2.5094602091798004e-06,
+      "loss": 0.1366,
+      "step": 8023
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4199788142649705,
+      "learning_rate": 2.508277032822488e-06,
+      "loss": 0.1063,
+      "step": 8024
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.19251075852326502,
+      "learning_rate": 2.5070940420632152e-06,
+      "loss": 0.0311,
+      "step": 8025
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.26765056855751096,
+      "learning_rate": 2.505911236990103e-06,
+      "loss": 0.073,
+      "step": 8026
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3075561279589553,
+      "learning_rate": 2.5047286176912503e-06,
+      "loss": 0.0947,
+      "step": 8027
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.26649119599091714,
+      "learning_rate": 2.503546184254747e-06,
+      "loss": 0.0701,
+      "step": 8028
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.29336334366498784,
+      "learning_rate": 2.502363936768664e-06,
+      "loss": 0.0799,
+      "step": 8029
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.22747500099257587,
+      "learning_rate": 2.501181875321067e-06,
+      "loss": 0.087,
+      "step": 8030
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.20828914005863092,
+      "learning_rate": 2.5000000000000015e-06,
+      "loss": 0.0627,
+      "step": 8031
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.754524687709029,
+      "learning_rate": 2.4988183108934968e-06,
+      "loss": 0.1885,
+      "step": 8032
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3497416774320822,
+      "learning_rate": 2.497636808089577e-06,
+      "loss": 0.0929,
+      "step": 8033
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5020016499156276,
+      "learning_rate": 2.4964554916762446e-06,
+      "loss": 0.1353,
+      "step": 8034
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.333233957396391,
+      "learning_rate": 2.49527436174149e-06,
+      "loss": 0.0955,
+      "step": 8035
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4411361957689018,
+      "learning_rate": 2.4940934183732934e-06,
+      "loss": 0.0812,
+      "step": 8036
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.19724742402346213,
+      "learning_rate": 2.492912661659617e-06,
+      "loss": 0.0377,
+      "step": 8037
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5159963894659594,
+      "learning_rate": 2.4917320916884115e-06,
+      "loss": 0.0791,
+      "step": 8038
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2533603551358582,
+      "learning_rate": 2.490551708547609e-06,
+      "loss": 0.0805,
+      "step": 8039
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2832465978797984,
+      "learning_rate": 2.4893715123251355e-06,
+      "loss": 0.0644,
+      "step": 8040
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3846127429227038,
+      "learning_rate": 2.488191503108897e-06,
+      "loss": 0.0798,
+      "step": 8041
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.32691562379721095,
+      "learning_rate": 2.4870116809867867e-06,
+      "loss": 0.0868,
+      "step": 8042
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.44984869424054613,
+      "learning_rate": 2.4858320460466867e-06,
+      "loss": 0.0904,
+      "step": 8043
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.266892990111272,
+      "learning_rate": 2.4846525983764624e-06,
+      "loss": 0.0493,
+      "step": 8044
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.23035150180528954,
+      "learning_rate": 2.483473338063966e-06,
+      "loss": 0.0765,
+      "step": 8045
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2485047735744902,
+      "learning_rate": 2.4822942651970328e-06,
+      "loss": 0.0852,
+      "step": 8046
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.27373478047833266,
+      "learning_rate": 2.4811153798634915e-06,
+      "loss": 0.0798,
+      "step": 8047
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3131649586130334,
+      "learning_rate": 2.47993668215115e-06,
+      "loss": 0.0993,
+      "step": 8048
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6641047689593959,
+      "learning_rate": 2.478758172147803e-06,
+      "loss": 0.1024,
+      "step": 8049
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.31715245648287627,
+      "learning_rate": 2.477579849941237e-06,
+      "loss": 0.0771,
+      "step": 8050
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.41974305016893426,
+      "learning_rate": 2.4764017156192176e-06,
+      "loss": 0.1167,
+      "step": 8051
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2658635147860032,
+      "learning_rate": 2.4752237692695e-06,
+      "loss": 0.067,
+      "step": 8052
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3075842019827212,
+      "learning_rate": 2.4740460109798216e-06,
+      "loss": 0.0796,
+      "step": 8053
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.18232864950809874,
+      "learning_rate": 2.472868440837913e-06,
+      "loss": 0.0332,
+      "step": 8054
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2668748606444816,
+      "learning_rate": 2.471691058931484e-06,
+      "loss": 0.0637,
+      "step": 8055
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.40018004293384646,
+      "learning_rate": 2.4705138653482323e-06,
+      "loss": 0.131,
+      "step": 8056
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2976738692723799,
+      "learning_rate": 2.469336860175845e-06,
+      "loss": 0.1032,
+      "step": 8057
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.21767646028944168,
+      "learning_rate": 2.46816004350199e-06,
+      "loss": 0.0476,
+      "step": 8058
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2313481084279434,
+      "learning_rate": 2.4669834154143252e-06,
+      "loss": 0.0605,
+      "step": 8059
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.22083637562350358,
+      "learning_rate": 2.465806976000489e-06,
+      "loss": 0.0842,
+      "step": 8060
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.19149052807894268,
+      "learning_rate": 2.464630725348114e-06,
+      "loss": 0.0409,
+      "step": 8061
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2809956347253728,
+      "learning_rate": 2.463454663544812e-06,
+      "loss": 0.0588,
+      "step": 8062
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3466644469789324,
+      "learning_rate": 2.462278790678183e-06,
+      "loss": 0.0929,
+      "step": 8063
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.27576839404603953,
+      "learning_rate": 2.461103106835811e-06,
+      "loss": 0.0584,
+      "step": 8064
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.21884424109249867,
+      "learning_rate": 2.4599276121052724e-06,
+      "loss": 0.0496,
+      "step": 8065
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.21223302770917157,
+      "learning_rate": 2.4587523065741215e-06,
+      "loss": 0.0422,
+      "step": 8066
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3558587317364715,
+      "learning_rate": 2.457577190329903e-06,
+      "loss": 0.0872,
+      "step": 8067
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.35948690772773934,
+      "learning_rate": 2.456402263460144e-06,
+      "loss": 0.1109,
+      "step": 8068
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5142850160991096,
+      "learning_rate": 2.4552275260523643e-06,
+      "loss": 0.0849,
+      "step": 8069
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.44822317913201726,
+      "learning_rate": 2.4540529781940627e-06,
+      "loss": 0.105,
+      "step": 8070
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3587540614966444,
+      "learning_rate": 2.452878619972726e-06,
+      "loss": 0.0767,
+      "step": 8071
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.31999772035977314,
+      "learning_rate": 2.4517044514758283e-06,
+      "loss": 0.0697,
+      "step": 8072
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4832497834881513,
+      "learning_rate": 2.4505304727908256e-06,
+      "loss": 0.0695,
+      "step": 8073
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3605789406964529,
+      "learning_rate": 2.4493566840051675e-06,
+      "loss": 0.0961,
+      "step": 8074
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.35935605366543644,
+      "learning_rate": 2.4481830852062822e-06,
+      "loss": 0.0855,
+      "step": 8075
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.20932573801051588,
+      "learning_rate": 2.4470096764815835e-06,
+      "loss": 0.0447,
+      "step": 8076
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3451073924776638,
+      "learning_rate": 2.445836457918479e-06,
+      "loss": 0.0442,
+      "step": 8077
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.21688131464164123,
+      "learning_rate": 2.444663429604354e-06,
+      "loss": 0.0346,
+      "step": 8078
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4958023049880385,
+      "learning_rate": 2.4434905916265827e-06,
+      "loss": 0.12,
+      "step": 8079
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.32214520519332135,
+      "learning_rate": 2.442317944072523e-06,
+      "loss": 0.0668,
+      "step": 8080
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.348837401589321,
+      "learning_rate": 2.4411454870295243e-06,
+      "loss": 0.0801,
+      "step": 8081
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.31166341805781594,
+      "learning_rate": 2.4399732205849163e-06,
+      "loss": 0.0828,
+      "step": 8082
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4607716011522742,
+      "learning_rate": 2.438801144826014e-06,
+      "loss": 0.1056,
+      "step": 8083
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.23156766433261153,
+      "learning_rate": 2.4376292598401247e-06,
+      "loss": 0.0567,
+      "step": 8084
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2745516870946822,
+      "learning_rate": 2.4364575657145346e-06,
+      "loss": 0.0481,
+      "step": 8085
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6158662316441955,
+      "learning_rate": 2.435286062536519e-06,
+      "loss": 0.0701,
+      "step": 8086
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4416976260609938,
+      "learning_rate": 2.434114750393336e-06,
+      "loss": 0.1559,
+      "step": 8087
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2643541175123459,
+      "learning_rate": 2.4329436293722356e-06,
+      "loss": 0.0506,
+      "step": 8088
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.38772184468999604,
+      "learning_rate": 2.4317726995604486e-06,
+      "loss": 0.0938,
+      "step": 8089
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5929529771801245,
+      "learning_rate": 2.430601961045189e-06,
+      "loss": 0.088,
+      "step": 8090
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.44999434707681407,
+      "learning_rate": 2.429431413913666e-06,
+      "loss": 0.0892,
+      "step": 8091
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6146399970921894,
+      "learning_rate": 2.428261058253065e-06,
+      "loss": 0.1535,
+      "step": 8092
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.27788218486761906,
+      "learning_rate": 2.427090894150561e-06,
+      "loss": 0.0565,
+      "step": 8093
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.32346642998620456,
+      "learning_rate": 2.4259209216933167e-06,
+      "loss": 0.0692,
+      "step": 8094
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3398073967232939,
+      "learning_rate": 2.424751140968478e-06,
+      "loss": 0.1119,
+      "step": 8095
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3529240686077884,
+      "learning_rate": 2.4235815520631755e-06,
+      "loss": 0.0938,
+      "step": 8096
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2844105468958982,
+      "learning_rate": 2.422412155064526e-06,
+      "loss": 0.0577,
+      "step": 8097
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2914077634383241,
+      "learning_rate": 2.421242950059637e-06,
+      "loss": 0.0836,
+      "step": 8098
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.42881957017671174,
+      "learning_rate": 2.4200739371355953e-06,
+      "loss": 0.0733,
+      "step": 8099
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.293584844174575,
+      "learning_rate": 2.4189051163794735e-06,
+      "loss": 0.0625,
+      "step": 8100
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.37657029514672713,
+      "learning_rate": 2.4177364878783372e-06,
+      "loss": 0.0879,
+      "step": 8101
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3552949377056055,
+      "learning_rate": 2.4165680517192298e-06,
+      "loss": 0.0853,
+      "step": 8102
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2790554288012035,
+      "learning_rate": 2.4153998079891837e-06,
+      "loss": 0.0842,
+      "step": 8103
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.21347609155940397,
+      "learning_rate": 2.4142317567752144e-06,
+      "loss": 0.0528,
+      "step": 8104
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3241123585988747,
+      "learning_rate": 2.4130638981643286e-06,
+      "loss": 0.108,
+      "step": 8105
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4993634105141833,
+      "learning_rate": 2.411896232243514e-06,
+      "loss": 0.115,
+      "step": 8106
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2783575675623533,
+      "learning_rate": 2.410728759099743e-06,
+      "loss": 0.0698,
+      "step": 8107
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3468728923999155,
+      "learning_rate": 2.4095614788199796e-06,
+      "loss": 0.0484,
+      "step": 8108
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.40507023077314613,
+      "learning_rate": 2.408394391491167e-06,
+      "loss": 0.1349,
+      "step": 8109
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.43726005770570237,
+      "learning_rate": 2.407227497200236e-06,
+      "loss": 0.0954,
+      "step": 8110
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5280341072157042,
+      "learning_rate": 2.406060796034107e-06,
+      "loss": 0.1425,
+      "step": 8111
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.28028143348750906,
+      "learning_rate": 2.4048942880796805e-06,
+      "loss": 0.0514,
+      "step": 8112
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2928159603820459,
+      "learning_rate": 2.403727973423845e-06,
+      "loss": 0.0732,
+      "step": 8113
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4804918985117611,
+      "learning_rate": 2.402561852153473e-06,
+      "loss": 0.0995,
+      "step": 8114
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3191786734271622,
+      "learning_rate": 2.401395924355427e-06,
+      "loss": 0.0401,
+      "step": 8115
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.28711237609491935,
+      "learning_rate": 2.4002301901165504e-06,
+      "loss": 0.0733,
+      "step": 8116
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.23034313817422605,
+      "learning_rate": 2.3990646495236726e-06,
+      "loss": 0.0502,
+      "step": 8117
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3516996422256878,
+      "learning_rate": 2.3978993026636128e-06,
+      "loss": 0.0765,
+      "step": 8118
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.18382498547990275,
+      "learning_rate": 2.396734149623171e-06,
+      "loss": 0.0125,
+      "step": 8119
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.41475944546263616,
+      "learning_rate": 2.3955691904891355e-06,
+      "loss": 0.0908,
+      "step": 8120
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.28537379126679735,
+      "learning_rate": 2.3944044253482764e-06,
+      "loss": 0.0647,
+      "step": 8121
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.43865571177300045,
+      "learning_rate": 2.3932398542873565e-06,
+      "loss": 0.0733,
+      "step": 8122
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2786524143354487,
+      "learning_rate": 2.3920754773931175e-06,
+      "loss": 0.0646,
+      "step": 8123
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7500231216255687,
+      "learning_rate": 2.390911294752287e-06,
+      "loss": 0.1548,
+      "step": 8124
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.29221400681552423,
+      "learning_rate": 2.3897473064515846e-06,
+      "loss": 0.0664,
+      "step": 8125
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4032963964696591,
+      "learning_rate": 2.388583512577709e-06,
+      "loss": 0.0895,
+      "step": 8126
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2234733384629255,
+      "learning_rate": 2.387419913217346e-06,
+      "loss": 0.0545,
+      "step": 8127
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.39569949841370666,
+      "learning_rate": 2.386256508457165e-06,
+      "loss": 0.0795,
+      "step": 8128
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.26995017957746464,
+      "learning_rate": 2.3850932983838278e-06,
+      "loss": 0.0714,
+      "step": 8129
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.209080128593074,
+      "learning_rate": 2.383930283083974e-06,
+      "loss": 0.034,
+      "step": 8130
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3080747064795015,
+      "learning_rate": 2.3827674626442327e-06,
+      "loss": 0.0812,
+      "step": 8131
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2733800464591378,
+      "learning_rate": 2.3816048371512156e-06,
+      "loss": 0.0715,
+      "step": 8132
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.1711715181922208,
+      "learning_rate": 2.380442406691525e-06,
+      "loss": 0.0434,
+      "step": 8133
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4132237388466646,
+      "learning_rate": 2.3792801713517444e-06,
+      "loss": 0.1072,
+      "step": 8134
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3811966028564031,
+      "learning_rate": 2.378118131218443e-06,
+      "loss": 0.077,
+      "step": 8135
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2529576829270131,
+      "learning_rate": 2.376956286378176e-06,
+      "loss": 0.0868,
+      "step": 8136
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3326983378707,
+      "learning_rate": 2.375794636917486e-06,
+      "loss": 0.0582,
+      "step": 8137
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.38652833085651345,
+      "learning_rate": 2.3746331829228987e-06,
+      "loss": 0.0826,
+      "step": 8138
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3444491147776681,
+      "learning_rate": 2.3734719244809263e-06,
+      "loss": 0.0907,
+      "step": 8139
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3901812724388572,
+      "learning_rate": 2.3723108616780647e-06,
+      "loss": 0.1042,
+      "step": 8140
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.42120335959737143,
+      "learning_rate": 2.3711499946007966e-06,
+      "loss": 0.0822,
+      "step": 8141
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.23353821483913406,
+      "learning_rate": 2.3699893233355924e-06,
+      "loss": 0.0623,
+      "step": 8142
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.1363099328467121,
+      "learning_rate": 2.3688288479689047e-06,
+      "loss": 0.0381,
+      "step": 8143
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3145043414305087,
+      "learning_rate": 2.36766856858717e-06,
+      "loss": 0.0657,
+      "step": 8144
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3764845375555749,
+      "learning_rate": 2.3665084852768166e-06,
+      "loss": 0.0674,
+      "step": 8145
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.22491304031803827,
+      "learning_rate": 2.365348598124253e-06,
+      "loss": 0.041,
+      "step": 8146
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2754475587643003,
+      "learning_rate": 2.364188907215873e-06,
+      "loss": 0.081,
+      "step": 8147
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3851073160924233,
+      "learning_rate": 2.3630294126380564e-06,
+      "loss": 0.0688,
+      "step": 8148
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2812221114577284,
+      "learning_rate": 2.3618701144771726e-06,
+      "loss": 0.0455,
+      "step": 8149
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.29891103456227003,
+      "learning_rate": 2.360711012819571e-06,
+      "loss": 0.0628,
+      "step": 8150
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2664787937025612,
+      "learning_rate": 2.359552107751586e-06,
+      "loss": 0.0832,
+      "step": 8151
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3137430470310707,
+      "learning_rate": 2.3583933993595435e-06,
+      "loss": 0.107,
+      "step": 8152
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.38031404549337133,
+      "learning_rate": 2.3572348877297495e-06,
+      "loss": 0.089,
+      "step": 8153
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.25154042903179935,
+      "learning_rate": 2.3560765729484964e-06,
+      "loss": 0.0808,
+      "step": 8154
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.36703601728832985,
+      "learning_rate": 2.3549184551020597e-06,
+      "loss": 0.1247,
+      "step": 8155
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.28684099655841416,
+      "learning_rate": 2.353760534276708e-06,
+      "loss": 0.0537,
+      "step": 8156
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.25517849069716125,
+      "learning_rate": 2.3526028105586864e-06,
+      "loss": 0.074,
+      "step": 8157
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2974549841593156,
+      "learning_rate": 2.351445284034228e-06,
+      "loss": 0.1157,
+      "step": 8158
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.37723826996287274,
+      "learning_rate": 2.3502879547895557e-06,
+      "loss": 0.0649,
+      "step": 8159
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.15869081585834996,
+      "learning_rate": 2.3491308229108724e-06,
+      "loss": 0.0463,
+      "step": 8160
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.21339357734202838,
+      "learning_rate": 2.347973888484366e-06,
+      "loss": 0.0624,
+      "step": 8161
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3891342095865076,
+      "learning_rate": 2.3468171515962153e-06,
+      "loss": 0.0971,
+      "step": 8162
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4522276696293356,
+      "learning_rate": 2.345660612332579e-06,
+      "loss": 0.0881,
+      "step": 8163
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.19729122357824827,
+      "learning_rate": 2.344504270779603e-06,
+      "loss": 0.0533,
+      "step": 8164
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.34484931590870704,
+      "learning_rate": 2.3433481270234165e-06,
+      "loss": 0.1026,
+      "step": 8165
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5298712150064175,
+      "learning_rate": 2.3421921811501397e-06,
+      "loss": 0.0987,
+      "step": 8166
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.24860946341035167,
+      "learning_rate": 2.3410364332458717e-06,
+      "loss": 0.0674,
+      "step": 8167
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3452771313306856,
+      "learning_rate": 2.3398808833966986e-06,
+      "loss": 0.0551,
+      "step": 8168
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2213633156578429,
+      "learning_rate": 2.3387255316886947e-06,
+      "loss": 0.0482,
+      "step": 8169
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.41450239364038427,
+      "learning_rate": 2.3375703782079167e-06,
+      "loss": 0.1026,
+      "step": 8170
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.31070746665663973,
+      "learning_rate": 2.3364154230404068e-06,
+      "loss": 0.0967,
+      "step": 8171
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3367805468857731,
+      "learning_rate": 2.3352606662721904e-06,
+      "loss": 0.0829,
+      "step": 8172
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.32419958235428303,
+      "learning_rate": 2.334106107989285e-06,
+      "loss": 0.0686,
+      "step": 8173
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.25254632981763864,
+      "learning_rate": 2.332951748277687e-06,
+      "loss": 0.0574,
+      "step": 8174
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.31943237342192005,
+      "learning_rate": 2.331797587223377e-06,
+      "loss": 0.0737,
+      "step": 8175
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2887930862160014,
+      "learning_rate": 2.330643624912328e-06,
+      "loss": 0.0761,
+      "step": 8176
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2952333108657727,
+      "learning_rate": 2.329489861430493e-06,
+      "loss": 0.0601,
+      "step": 8177
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.32231977767869785,
+      "learning_rate": 2.328336296863809e-06,
+      "loss": 0.0951,
+      "step": 8178
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3837727011817731,
+      "learning_rate": 2.327182931298199e-06,
+      "loss": 0.0664,
+      "step": 8179
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3774429133756885,
+      "learning_rate": 2.3260297648195775e-06,
+      "loss": 0.107,
+      "step": 8180
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3480880680120187,
+      "learning_rate": 2.324876797513836e-06,
+      "loss": 0.0643,
+      "step": 8181
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3802512600919708,
+      "learning_rate": 2.3237240294668516e-06,
+      "loss": 0.1233,
+      "step": 8182
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4125467771660437,
+      "learning_rate": 2.3225714607644945e-06,
+      "loss": 0.1076,
+      "step": 8183
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3595899997810304,
+      "learning_rate": 2.321419091492612e-06,
+      "loss": 0.0536,
+      "step": 8184
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.28431198178735206,
+      "learning_rate": 2.320266921737038e-06,
+      "loss": 0.0922,
+      "step": 8185
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.25850827855427566,
+      "learning_rate": 2.319114951583595e-06,
+      "loss": 0.0355,
+      "step": 8186
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.24705841508809565,
+      "learning_rate": 2.3179631811180893e-06,
+      "loss": 0.0625,
+      "step": 8187
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.35221132676035877,
+      "learning_rate": 2.3168116104263087e-06,
+      "loss": 0.0935,
+      "step": 8188
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.16895014936836256,
+      "learning_rate": 2.3156602395940284e-06,
+      "loss": 0.0437,
+      "step": 8189
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.46315374354200667,
+      "learning_rate": 2.314509068707013e-06,
+      "loss": 0.0884,
+      "step": 8190
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4275299754266828,
+      "learning_rate": 2.313358097851007e-06,
+      "loss": 0.1236,
+      "step": 8191
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2254195089672145,
+      "learning_rate": 2.3122073271117377e-06,
+      "loss": 0.0725,
+      "step": 8192
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4673925712183842,
+      "learning_rate": 2.3110567565749266e-06,
+      "loss": 0.0939,
+      "step": 8193
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.22734738873313137,
+      "learning_rate": 2.309906386326272e-06,
+      "loss": 0.076,
+      "step": 8194
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.31884050634619693,
+      "learning_rate": 2.3087562164514616e-06,
+      "loss": 0.0451,
+      "step": 8195
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5955037401392582,
+      "learning_rate": 2.3076062470361633e-06,
+      "loss": 0.1104,
+      "step": 8196
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.333876183468035,
+      "learning_rate": 2.306456478166038e-06,
+      "loss": 0.0599,
+      "step": 8197
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.29679477332703175,
+      "learning_rate": 2.3053069099267258e-06,
+      "loss": 0.0714,
+      "step": 8198
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.24847932574114995,
+      "learning_rate": 2.304157542403852e-06,
+      "loss": 0.0659,
+      "step": 8199
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3730900376316934,
+      "learning_rate": 2.3030083756830273e-06,
+      "loss": 0.1012,
+      "step": 8200
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5050976470812316,
+      "learning_rate": 2.301859409849852e-06,
+      "loss": 0.1144,
+      "step": 8201
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.36682215454047484,
+      "learning_rate": 2.3007106449899057e-06,
+      "loss": 0.0865,
+      "step": 8202
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2465918884401312,
+      "learning_rate": 2.299562081188755e-06,
+      "loss": 0.0711,
+      "step": 8203
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.49510917590764936,
+      "learning_rate": 2.2984137185319493e-06,
+      "loss": 0.1097,
+      "step": 8204
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.21291243920002315,
+      "learning_rate": 2.29726555710503e-06,
+      "loss": 0.0422,
+      "step": 8205
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.22447561988243364,
+      "learning_rate": 2.296117596993517e-06,
+      "loss": 0.0639,
+      "step": 8206
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6643570533675038,
+      "learning_rate": 2.294969838282916e-06,
+      "loss": 0.0904,
+      "step": 8207
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2676314841623326,
+      "learning_rate": 2.2938222810587203e-06,
+      "loss": 0.0847,
+      "step": 8208
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.19901550644355864,
+      "learning_rate": 2.292674925406403e-06,
+      "loss": 0.0711,
+      "step": 8209
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.42956110768676475,
+      "learning_rate": 2.291527771411431e-06,
+      "loss": 0.136,
+      "step": 8210
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3666575379904941,
+      "learning_rate": 2.2903808191592482e-06,
+      "loss": 0.067,
+      "step": 8211
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8460809432169117,
+      "learning_rate": 2.2892340687352864e-06,
+      "loss": 0.1402,
+      "step": 8212
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4397310089339902,
+      "learning_rate": 2.2880875202249616e-06,
+      "loss": 0.101,
+      "step": 8213
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4724799158550494,
+      "learning_rate": 2.2869411737136776e-06,
+      "loss": 0.1226,
+      "step": 8214
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.48398515748068516,
+      "learning_rate": 2.2857950292868196e-06,
+      "loss": 0.1012,
+      "step": 8215
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.31176729534184416,
+      "learning_rate": 2.284649087029758e-06,
+      "loss": 0.1031,
+      "step": 8216
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2725401129390788,
+      "learning_rate": 2.2835033470278515e-06,
+      "loss": 0.0685,
+      "step": 8217
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4255972552172791,
+      "learning_rate": 2.282357809366441e-06,
+      "loss": 0.0896,
+      "step": 8218
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.305404304316731,
+      "learning_rate": 2.2812124741308505e-06,
+      "loss": 0.0691,
+      "step": 8219
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.1918144405130339,
+      "learning_rate": 2.280067341406395e-06,
+      "loss": 0.0355,
+      "step": 8220
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.22301497074716992,
+      "learning_rate": 2.278922411278368e-06,
+      "loss": 0.0489,
+      "step": 8221
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2761686383822466,
+      "learning_rate": 2.277777683832052e-06,
+      "loss": 0.0936,
+      "step": 8222
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.23012506201191857,
+      "learning_rate": 2.2766331591527103e-06,
+      "loss": 0.0692,
+      "step": 8223
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.37612805610696265,
+      "learning_rate": 2.2754888373255975e-06,
+      "loss": 0.0805,
+      "step": 8224
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.21667128895859342,
+      "learning_rate": 2.2743447184359473e-06,
+      "loss": 0.0637,
+      "step": 8225
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2583340176953693,
+      "learning_rate": 2.273200802568979e-06,
+      "loss": 0.0795,
+      "step": 8226
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2348846207661785,
+      "learning_rate": 2.272057089809901e-06,
+      "loss": 0.0469,
+      "step": 8227
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.24174578992106766,
+      "learning_rate": 2.270913580243903e-06,
+      "loss": 0.0517,
+      "step": 8228
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3271962405798819,
+      "learning_rate": 2.2697702739561595e-06,
+      "loss": 0.0711,
+      "step": 8229
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.18671840928034406,
+      "learning_rate": 2.2686271710318287e-06,
+      "loss": 0.0576,
+      "step": 8230
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.18693910168661979,
+      "learning_rate": 2.2674842715560595e-06,
+      "loss": 0.0635,
+      "step": 8231
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4429133842218223,
+      "learning_rate": 2.2663415756139802e-06,
+      "loss": 0.0971,
+      "step": 8232
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.35442171318348153,
+      "learning_rate": 2.2651990832907027e-06,
+      "loss": 0.0911,
+      "step": 8233
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3425757651304039,
+      "learning_rate": 2.264056794671331e-06,
+      "loss": 0.1229,
+      "step": 8234
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2668660403609579,
+      "learning_rate": 2.262914709840947e-06,
+      "loss": 0.0625,
+      "step": 8235
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3564272034094718,
+      "learning_rate": 2.2617728288846185e-06,
+      "loss": 0.1075,
+      "step": 8236
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3213710803396447,
+      "learning_rate": 2.260631151887403e-06,
+      "loss": 0.0561,
+      "step": 8237
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.27268749549507937,
+      "learning_rate": 2.2594896789343372e-06,
+      "loss": 0.0972,
+      "step": 8238
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.42044965318155925,
+      "learning_rate": 2.258348410110445e-06,
+      "loss": 0.1118,
+      "step": 8239
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.30079030768330145,
+      "learning_rate": 2.2572073455007325e-06,
+      "loss": 0.0802,
+      "step": 8240
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.34496689551959236,
+      "learning_rate": 2.256066485190197e-06,
+      "loss": 0.0917,
+      "step": 8241
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3383911450788624,
+      "learning_rate": 2.2549258292638137e-06,
+      "loss": 0.0726,
+      "step": 8242
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.23090556327194298,
+      "learning_rate": 2.2537853778065445e-06,
+      "loss": 0.0648,
+      "step": 8243
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3558244217609913,
+      "learning_rate": 2.2526451309033397e-06,
+      "loss": 0.1193,
+      "step": 8244
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2826079996752124,
+      "learning_rate": 2.2515050886391306e-06,
+      "loss": 0.086,
+      "step": 8245
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3849818201934799,
+      "learning_rate": 2.2503652510988326e-06,
+      "loss": 0.0595,
+      "step": 8246
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.26609201227851337,
+      "learning_rate": 2.2492256183673476e-06,
+      "loss": 0.0876,
+      "step": 8247
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3141802976901484,
+      "learning_rate": 2.248086190529565e-06,
+      "loss": 0.0884,
+      "step": 8248
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4229164929567971,
+      "learning_rate": 2.2469469676703536e-06,
+      "loss": 0.1248,
+      "step": 8249
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4051876021904241,
+      "learning_rate": 2.2458079498745675e-06,
+      "loss": 0.125,
+      "step": 8250
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.25001926485657355,
+      "learning_rate": 2.2446691372270523e-06,
+      "loss": 0.0612,
+      "step": 8251
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.50843339555988,
+      "learning_rate": 2.24353052981263e-06,
+      "loss": 0.0772,
+      "step": 8252
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6155887087397431,
+      "learning_rate": 2.2423921277161103e-06,
+      "loss": 0.042,
+      "step": 8253
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.24807387258629288,
+      "learning_rate": 2.241253931022291e-06,
+      "loss": 0.0678,
+      "step": 8254
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.19562751459739416,
+      "learning_rate": 2.2401159398159492e-06,
+      "loss": 0.041,
+      "step": 8255
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.22540173162042015,
+      "learning_rate": 2.23897815418185e-06,
+      "loss": 0.0435,
+      "step": 8256
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.31509657629465876,
+      "learning_rate": 2.2378405742047406e-06,
+      "loss": 0.0717,
+      "step": 8257
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2968119278186583,
+      "learning_rate": 2.236703199969357e-06,
+      "loss": 0.0834,
+      "step": 8258
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6292634174285717,
+      "learning_rate": 2.2355660315604173e-06,
+      "loss": 0.1524,
+      "step": 8259
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3853349931526165,
+      "learning_rate": 2.234429069062621e-06,
+      "loss": 0.0756,
+      "step": 8260
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3257037756920801,
+      "learning_rate": 2.2332923125606608e-06,
+      "loss": 0.076,
+      "step": 8261
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5049199829859229,
+      "learning_rate": 2.232155762139206e-06,
+      "loss": 0.1096,
+      "step": 8262
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5884114196935599,
+      "learning_rate": 2.231019417882914e-06,
+      "loss": 0.1413,
+      "step": 8263
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4249956825727513,
+      "learning_rate": 2.2298832798764246e-06,
+      "loss": 0.0999,
+      "step": 8264
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4292746325383951,
+      "learning_rate": 2.2287473482043676e-06,
+      "loss": 0.1262,
+      "step": 8265
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3952749160588101,
+      "learning_rate": 2.2276116229513515e-06,
+      "loss": 0.1009,
+      "step": 8266
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2184450788959631,
+      "learning_rate": 2.2264761042019724e-06,
+      "loss": 0.0618,
+      "step": 8267
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.48053085464086254,
+      "learning_rate": 2.2253407920408078e-06,
+      "loss": 0.0836,
+      "step": 8268
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.28941080855400425,
+      "learning_rate": 2.2242056865524263e-06,
+      "loss": 0.1008,
+      "step": 8269
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.22125260822283646,
+      "learning_rate": 2.223070787821376e-06,
+      "loss": 0.0708,
+      "step": 8270
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3188860276328193,
+      "learning_rate": 2.22193609593219e-06,
+      "loss": 0.0738,
+      "step": 8271
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4441931392712277,
+      "learning_rate": 2.2208016109693853e-06,
+      "loss": 0.0469,
+      "step": 8272
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2589215671631317,
+      "learning_rate": 2.219667333017468e-06,
+      "loss": 0.0823,
+      "step": 8273
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.35658879913443536,
+      "learning_rate": 2.2185332621609246e-06,
+      "loss": 0.0828,
+      "step": 8274
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2515164986207159,
+      "learning_rate": 2.2173993984842274e-06,
+      "loss": 0.0793,
+      "step": 8275
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.42418142142400783,
+      "learning_rate": 2.2162657420718325e-06,
+      "loss": 0.1063,
+      "step": 8276
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.26381239013335395,
+      "learning_rate": 2.2151322930081797e-06,
+      "loss": 0.094,
+      "step": 8277
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.40830864053053045,
+      "learning_rate": 2.2139990513776987e-06,
+      "loss": 0.0539,
+      "step": 8278
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.31287292212604134,
+      "learning_rate": 2.2128660172647985e-06,
+      "loss": 0.1005,
+      "step": 8279
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.28493551969664616,
+      "learning_rate": 2.211733190753873e-06,
+      "loss": 0.0838,
+      "step": 8280
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.22407053972757265,
+      "learning_rate": 2.2106005719293007e-06,
+      "loss": 0.0594,
+      "step": 8281
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.21207820772189656,
+      "learning_rate": 2.209468160875449e-06,
+      "loss": 0.0487,
+      "step": 8282
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3221982924551854,
+      "learning_rate": 2.208335957676665e-06,
+      "loss": 0.1005,
+      "step": 8283
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.30321714071028494,
+      "learning_rate": 2.2072039624172793e-06,
+      "loss": 0.077,
+      "step": 8284
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3428388330212152,
+      "learning_rate": 2.206072175181614e-06,
+      "loss": 0.0887,
+      "step": 8285
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2991361152659501,
+      "learning_rate": 2.2049405960539684e-06,
+      "loss": 0.0776,
+      "step": 8286
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2687866617421833,
+      "learning_rate": 2.2038092251186277e-06,
+      "loss": 0.0711,
+      "step": 8287
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.32898224696366285,
+      "learning_rate": 2.2026780624598672e-06,
+      "loss": 0.0729,
+      "step": 8288
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.20414464298871673,
+      "learning_rate": 2.20154710816194e-06,
+      "loss": 0.0491,
+      "step": 8289
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.21836862187615885,
+      "learning_rate": 2.200416362309087e-06,
+      "loss": 0.0362,
+      "step": 8290
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.19247931252578104,
+      "learning_rate": 2.1992858249855293e-06,
+      "loss": 0.0488,
+      "step": 8291
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2245907699571742,
+      "learning_rate": 2.1981554962754813e-06,
+      "loss": 0.0412,
+      "step": 8292
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.33202523241763193,
+      "learning_rate": 2.1970253762631334e-06,
+      "loss": 0.0945,
+      "step": 8293
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.29533226013352354,
+      "learning_rate": 2.1958954650326624e-06,
+      "loss": 0.0866,
+      "step": 8294
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.31270948841039575,
+      "learning_rate": 2.194765762668234e-06,
+      "loss": 0.0851,
+      "step": 8295
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.38024842750123045,
+      "learning_rate": 2.1936362692539936e-06,
+      "loss": 0.052,
+      "step": 8296
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2539527132980576,
+      "learning_rate": 2.192506984874072e-06,
+      "loss": 0.0506,
+      "step": 8297
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3790691511096439,
+      "learning_rate": 2.1913779096125832e-06,
+      "loss": 0.0841,
+      "step": 8298
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.38380322277077483,
+      "learning_rate": 2.1902490435536304e-06,
+      "loss": 0.1123,
+      "step": 8299
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3939305018904942,
+      "learning_rate": 2.1891203867812977e-06,
+      "loss": 0.1244,
+      "step": 8300
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3917614048056084,
+      "learning_rate": 2.1879919393796505e-06,
+      "loss": 0.0886,
+      "step": 8301
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3835769224131753,
+      "learning_rate": 2.1868637014327476e-06,
+      "loss": 0.0996,
+      "step": 8302
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.22923062086987134,
+      "learning_rate": 2.185735673024623e-06,
+      "loss": 0.068,
+      "step": 8303
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2313880447960041,
+      "learning_rate": 2.1846078542393005e-06,
+      "loss": 0.0875,
+      "step": 8304
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.28166880269457767,
+      "learning_rate": 2.183480245160784e-06,
+      "loss": 0.0646,
+      "step": 8305
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.23286866918393045,
+      "learning_rate": 2.1823528458730674e-06,
+      "loss": 0.0764,
+      "step": 8306
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5873112169650783,
+      "learning_rate": 2.181225656460126e-06,
+      "loss": 0.1427,
+      "step": 8307
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.15940495085580306,
+      "learning_rate": 2.1800986770059155e-06,
+      "loss": 0.0498,
+      "step": 8308
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.19668343513785205,
+      "learning_rate": 2.178971907594385e-06,
+      "loss": 0.0604,
+      "step": 8309
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.22044081269960228,
+      "learning_rate": 2.1778453483094604e-06,
+      "loss": 0.0719,
+      "step": 8310
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.276561912619221,
+      "learning_rate": 2.1767189992350523e-06,
+      "loss": 0.0659,
+      "step": 8311
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.40868232965030293,
+      "learning_rate": 2.175592860455062e-06,
+      "loss": 0.079,
+      "step": 8312
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.21590705529561863,
+      "learning_rate": 2.174466932053369e-06,
+      "loss": 0.0688,
+      "step": 8313
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.37969881438766684,
+      "learning_rate": 2.173341214113839e-06,
+      "loss": 0.0864,
+      "step": 8314
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2611057076256042,
+      "learning_rate": 2.17221570672032e-06,
+      "loss": 0.0596,
+      "step": 8315
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3345410625063943,
+      "learning_rate": 2.1710904099566497e-06,
+      "loss": 0.0755,
+      "step": 8316
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.19900779356904622,
+      "learning_rate": 2.1699653239066455e-06,
+      "loss": 0.072,
+      "step": 8317
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2846356919423404,
+      "learning_rate": 2.168840448654108e-06,
+      "loss": 0.0754,
+      "step": 8318
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3521408907321636,
+      "learning_rate": 2.167715784282829e-06,
+      "loss": 0.0705,
+      "step": 8319
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2877522500827389,
+      "learning_rate": 2.1665913308765773e-06,
+      "loss": 0.0762,
+      "step": 8320
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2722918757864807,
+      "learning_rate": 2.165467088519109e-06,
+      "loss": 0.0518,
+      "step": 8321
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3190264944240664,
+      "learning_rate": 2.164343057294163e-06,
+      "loss": 0.1192,
+      "step": 8322
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4833905171420414,
+      "learning_rate": 2.1632192372854667e-06,
+      "loss": 0.1148,
+      "step": 8323
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2637189410013057,
+      "learning_rate": 2.1620956285767268e-06,
+      "loss": 0.094,
+      "step": 8324
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2545600755787577,
+      "learning_rate": 2.160972231251635e-06,
+      "loss": 0.0362,
+      "step": 8325
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.37467484036892373,
+      "learning_rate": 2.1598490453938716e-06,
+      "loss": 0.1164,
+      "step": 8326
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.23520199292378555,
+      "learning_rate": 2.158726071087096e-06,
+      "loss": 0.0546,
+      "step": 8327
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7551659696366472,
+      "learning_rate": 2.1576033084149534e-06,
+      "loss": 0.1224,
+      "step": 8328
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4191249899379254,
+      "learning_rate": 2.1564807574610756e-06,
+      "loss": 0.1131,
+      "step": 8329
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.34532968058970004,
+      "learning_rate": 2.155358418309076e-06,
+      "loss": 0.0825,
+      "step": 8330
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4049522010092695,
+      "learning_rate": 2.1542362910425525e-06,
+      "loss": 0.1173,
+      "step": 8331
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3738414309737468,
+      "learning_rate": 2.153114375745086e-06,
+      "loss": 0.0771,
+      "step": 8332
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3046374813897974,
+      "learning_rate": 2.151992672500247e-06,
+      "loss": 0.0973,
+      "step": 8333
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2520584343762753,
+      "learning_rate": 2.150871181391585e-06,
+      "loss": 0.0963,
+      "step": 8334
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.33102346124299453,
+      "learning_rate": 2.1497499025026348e-06,
+      "loss": 0.0789,
+      "step": 8335
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.1648212026405659,
+      "learning_rate": 2.1486288359169133e-06,
+      "loss": 0.0624,
+      "step": 8336
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.36213414600885024,
+      "learning_rate": 2.147507981717929e-06,
+      "loss": 0.0911,
+      "step": 8337
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.42187564263971966,
+      "learning_rate": 2.146387339989167e-06,
+      "loss": 0.1094,
+      "step": 8338
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.19234902656797517,
+      "learning_rate": 2.1452669108140995e-06,
+      "loss": 0.0628,
+      "step": 8339
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2969239339687833,
+      "learning_rate": 2.1441466942761807e-06,
+      "loss": 0.1153,
+      "step": 8340
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.25015669339859553,
+      "learning_rate": 2.1430266904588544e-06,
+      "loss": 0.057,
+      "step": 8341
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.34756851260200783,
+      "learning_rate": 2.1419068994455435e-06,
+      "loss": 0.0811,
+      "step": 8342
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2181171457248317,
+      "learning_rate": 2.1407873213196562e-06,
+      "loss": 0.0705,
+      "step": 8343
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3405487549680562,
+      "learning_rate": 2.139667956164585e-06,
+      "loss": 0.0852,
+      "step": 8344
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.36686095237676986,
+      "learning_rate": 2.1385488040637057e-06,
+      "loss": 0.1198,
+      "step": 8345
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2813718707575593,
+      "learning_rate": 2.1374298651003826e-06,
+      "loss": 0.0704,
+      "step": 8346
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2584350016154863,
+      "learning_rate": 2.1363111393579587e-06,
+      "loss": 0.0715,
+      "step": 8347
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.23036667172341257,
+      "learning_rate": 2.135192626919763e-06,
+      "loss": 0.0751,
+      "step": 8348
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.22521560760151998,
+      "learning_rate": 2.1340743278691077e-06,
+      "loss": 0.0583,
+      "step": 8349
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5349610050745811,
+      "learning_rate": 2.1329562422892936e-06,
+      "loss": 0.1085,
+      "step": 8350
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3183396778376335,
+      "learning_rate": 2.1318383702636e-06,
+      "loss": 0.0818,
+      "step": 8351
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3943010127877746,
+      "learning_rate": 2.1307207118752917e-06,
+      "loss": 0.0909,
+      "step": 8352
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.48039189209374855,
+      "learning_rate": 2.129603267207621e-06,
+      "loss": 0.0758,
+      "step": 8353
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3885992812700629,
+      "learning_rate": 2.12848603634382e-06,
+      "loss": 0.0684,
+      "step": 8354
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.20430083207197544,
+      "learning_rate": 2.127369019367107e-06,
+      "loss": 0.053,
+      "step": 8355
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.32395637513706593,
+      "learning_rate": 2.1262522163606825e-06,
+      "loss": 0.0886,
+      "step": 8356
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2699451447574738,
+      "learning_rate": 2.125135627407735e-06,
+      "loss": 0.0744,
+      "step": 8357
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.18174907045690855,
+      "learning_rate": 2.1240192525914334e-06,
+      "loss": 0.0629,
+      "step": 8358
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.29084295515623754,
+      "learning_rate": 2.1229030919949295e-06,
+      "loss": 0.0618,
+      "step": 8359
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3164916077501644,
+      "learning_rate": 2.121787145701366e-06,
+      "loss": 0.0808,
+      "step": 8360
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.22405570020793797,
+      "learning_rate": 2.1206714137938624e-06,
+      "loss": 0.0677,
+      "step": 8361
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.35295209069707945,
+      "learning_rate": 2.1195558963555235e-06,
+      "loss": 0.088,
+      "step": 8362
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3310045314634588,
+      "learning_rate": 2.118440593469443e-06,
+      "loss": 0.0708,
+      "step": 8363
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.36240565132332553,
+      "learning_rate": 2.1173255052186934e-06,
+      "loss": 0.0861,
+      "step": 8364
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3638281357565688,
+      "learning_rate": 2.1162106316863333e-06,
+      "loss": 0.108,
+      "step": 8365
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.25470380681193217,
+      "learning_rate": 2.115095972955403e-06,
+      "loss": 0.0767,
+      "step": 8366
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4858707117540823,
+      "learning_rate": 2.1139815291089316e-06,
+      "loss": 0.093,
+      "step": 8367
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.19496605427084243,
+      "learning_rate": 2.112867300229929e-06,
+      "loss": 0.0385,
+      "step": 8368
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.16330483683321675,
+      "learning_rate": 2.111753286401386e-06,
+      "loss": 0.0602,
+      "step": 8369
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4254200884016728,
+      "learning_rate": 2.1106394877062857e-06,
+      "loss": 0.1127,
+      "step": 8370
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2467093187559839,
+      "learning_rate": 2.109525904227588e-06,
+      "loss": 0.0657,
+      "step": 8371
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.18741905357433997,
+      "learning_rate": 2.1084125360482395e-06,
+      "loss": 0.0438,
+      "step": 8372
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3463653946317927,
+      "learning_rate": 2.107299383251168e-06,
+      "loss": 0.081,
+      "step": 8373
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2678088691666406,
+      "learning_rate": 2.1061864459192918e-06,
+      "loss": 0.0781,
+      "step": 8374
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3802184092433274,
+      "learning_rate": 2.105073724135506e-06,
+      "loss": 0.086,
+      "step": 8375
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.47757995524382674,
+      "learning_rate": 2.1039612179826924e-06,
+      "loss": 0.1276,
+      "step": 8376
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.34347790441977466,
+      "learning_rate": 2.10284892754372e-06,
+      "loss": 0.0996,
+      "step": 8377
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7379331150655836,
+      "learning_rate": 2.1017368529014357e-06,
+      "loss": 0.121,
+      "step": 8378
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3288952063018087,
+      "learning_rate": 2.100624994138673e-06,
+      "loss": 0.0724,
+      "step": 8379
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3842620570062215,
+      "learning_rate": 2.0995133513382527e-06,
+      "loss": 0.1049,
+      "step": 8380
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.19194204042663732,
+      "learning_rate": 2.0984019245829745e-06,
+      "loss": 0.0678,
+      "step": 8381
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7171173131565487,
+      "learning_rate": 2.0972907139556237e-06,
+      "loss": 0.122,
+      "step": 8382
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2599427122423824,
+      "learning_rate": 2.096179719538969e-06,
+      "loss": 0.0643,
+      "step": 8383
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.44751356690909866,
+      "learning_rate": 2.095068941415766e-06,
+      "loss": 0.0952,
+      "step": 8384
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3400890017046582,
+      "learning_rate": 2.093958379668751e-06,
+      "loss": 0.0526,
+      "step": 8385
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.20846312971393863,
+      "learning_rate": 2.0928480343806423e-06,
+      "loss": 0.0458,
+      "step": 8386
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5410756770441011,
+      "learning_rate": 2.09173790563415e-06,
+      "loss": 0.1037,
+      "step": 8387
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.248353674784034,
+      "learning_rate": 2.09062799351196e-06,
+      "loss": 0.0568,
+      "step": 8388
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.46795089099270226,
+      "learning_rate": 2.0895182980967453e-06,
+      "loss": 0.1073,
+      "step": 8389
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.19309832299130636,
+      "learning_rate": 2.088408819471161e-06,
+      "loss": 0.0405,
+      "step": 8390
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.270987729281112,
+      "learning_rate": 2.087299557717851e-06,
+      "loss": 0.068,
+      "step": 8391
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.24750933783724813,
+      "learning_rate": 2.0861905129194377e-06,
+      "loss": 0.0639,
+      "step": 8392
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3596252297552899,
+      "learning_rate": 2.085081685158527e-06,
+      "loss": 0.0881,
+      "step": 8393
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.22647214983458508,
+      "learning_rate": 2.083973074517715e-06,
+      "loss": 0.0638,
+      "step": 8394
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4910123388763237,
+      "learning_rate": 2.082864681079576e-06,
+      "loss": 0.1275,
+      "step": 8395
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.23628405386874665,
+      "learning_rate": 2.081756504926667e-06,
+      "loss": 0.0664,
+      "step": 8396
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2909959222530801,
+      "learning_rate": 2.0806485461415347e-06,
+      "loss": 0.0775,
+      "step": 8397
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.26908507252932734,
+      "learning_rate": 2.079540804806706e-06,
+      "loss": 0.0626,
+      "step": 8398
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5802150831002991,
+      "learning_rate": 2.078433281004691e-06,
+      "loss": 0.121,
+      "step": 8399
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.1978838491681722,
+      "learning_rate": 2.0773259748179827e-06,
+      "loss": 0.0429,
+      "step": 8400
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.24581154589215215,
+      "learning_rate": 2.0762188863290632e-06,
+      "loss": 0.0825,
+      "step": 8401
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.21439576309310326,
+      "learning_rate": 2.0751120156203935e-06,
+      "loss": 0.0499,
+      "step": 8402
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.32905432384479627,
+      "learning_rate": 2.0740053627744195e-06,
+      "loss": 0.0612,
+      "step": 8403
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.24921363499672192,
+      "learning_rate": 2.0728989278735693e-06,
+      "loss": 0.0667,
+      "step": 8404
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3682516118629222,
+      "learning_rate": 2.071792711000261e-06,
+      "loss": 0.0814,
+      "step": 8405
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2550398916280299,
+      "learning_rate": 2.0706867122368897e-06,
+      "loss": 0.0821,
+      "step": 8406
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4547458096486892,
+      "learning_rate": 2.069580931665836e-06,
+      "loss": 0.1531,
+      "step": 8407
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.43522368625652363,
+      "learning_rate": 2.0684753693694636e-06,
+      "loss": 0.0713,
+      "step": 8408
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.44752555486562134,
+      "learning_rate": 2.0673700254301253e-06,
+      "loss": 0.1126,
+      "step": 8409
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2698307917708234,
+      "learning_rate": 2.066264899930151e-06,
+      "loss": 0.0706,
+      "step": 8410
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.24850844085048537,
+      "learning_rate": 2.0651599929518574e-06,
+      "loss": 0.0683,
+      "step": 8411
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.29797010470380547,
+      "learning_rate": 2.0640553045775443e-06,
+      "loss": 0.088,
+      "step": 8412
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2008024905706189,
+      "learning_rate": 2.062950834889493e-06,
+      "loss": 0.0544,
+      "step": 8413
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5548691988399708,
+      "learning_rate": 2.061846583969976e-06,
+      "loss": 0.1214,
+      "step": 8414
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2157245605101346,
+      "learning_rate": 2.0607425519012404e-06,
+      "loss": 0.047,
+      "step": 8415
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.43524764830999374,
+      "learning_rate": 2.0596387387655224e-06,
+      "loss": 0.0948,
+      "step": 8416
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.24060152594546,
+      "learning_rate": 2.058535144645038e-06,
+      "loss": 0.0979,
+      "step": 8417
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2636199024725251,
+      "learning_rate": 2.057431769621993e-06,
+      "loss": 0.0883,
+      "step": 8418
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.39978839537230043,
+      "learning_rate": 2.056328613778572e-06,
+      "loss": 0.1136,
+      "step": 8419
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3781309478680478,
+      "learning_rate": 2.055225677196942e-06,
+      "loss": 0.1238,
+      "step": 8420
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3411043132445427,
+      "learning_rate": 2.05412295995926e-06,
+      "loss": 0.1024,
+      "step": 8421
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.32952514850732917,
+      "learning_rate": 2.0530204621476607e-06,
+      "loss": 0.0772,
+      "step": 8422
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3694622923693062,
+      "learning_rate": 2.0519181838442646e-06,
+      "loss": 0.0753,
+      "step": 8423
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4680299330826819,
+      "learning_rate": 2.0508161251311748e-06,
+      "loss": 0.1044,
+      "step": 8424
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3711038438246592,
+      "learning_rate": 2.0497142860904817e-06,
+      "loss": 0.0707,
+      "step": 8425
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4619408187278086,
+      "learning_rate": 2.0486126668042554e-06,
+      "loss": 0.1163,
+      "step": 8426
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.33569376255040817,
+      "learning_rate": 2.0475112673545487e-06,
+      "loss": 0.0673,
+      "step": 8427
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2103920103948312,
+      "learning_rate": 2.046410087823405e-06,
+      "loss": 0.0808,
+      "step": 8428
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4061037085671439,
+      "learning_rate": 2.045309128292843e-06,
+      "loss": 0.0813,
+      "step": 8429
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2829112191875717,
+      "learning_rate": 2.04420838884487e-06,
+      "loss": 0.0635,
+      "step": 8430
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.18993723841079488,
+      "learning_rate": 2.0431078695614728e-06,
+      "loss": 0.0617,
+      "step": 8431
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4829519471029326,
+      "learning_rate": 2.0420075705246283e-06,
+      "loss": 0.1189,
+      "step": 8432
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5709067961081419,
+      "learning_rate": 2.040907491816292e-06,
+      "loss": 0.0759,
+      "step": 8433
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.28283449259845955,
+      "learning_rate": 2.0398076335184015e-06,
+      "loss": 0.0879,
+      "step": 8434
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4228203937879955,
+      "learning_rate": 2.038707995712885e-06,
+      "loss": 0.0972,
+      "step": 8435
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.764629553359395,
+      "learning_rate": 2.0376085784816473e-06,
+      "loss": 0.0694,
+      "step": 8436
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5632052735436225,
+      "learning_rate": 2.0365093819065785e-06,
+      "loss": 0.0648,
+      "step": 8437
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.35578078265866986,
+      "learning_rate": 2.035410406069556e-06,
+      "loss": 0.1156,
+      "step": 8438
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4111706395502615,
+      "learning_rate": 2.034311651052437e-06,
+      "loss": 0.0676,
+      "step": 8439
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2434040680824665,
+      "learning_rate": 2.0332131169370624e-06,
+      "loss": 0.088,
+      "step": 8440
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.21028405553623808,
+      "learning_rate": 2.0321148038052556e-06,
+      "loss": 0.0718,
+      "step": 8441
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.351969601383178,
+      "learning_rate": 2.0310167117388294e-06,
+      "loss": 0.0732,
+      "step": 8442
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.17555885939721344,
+      "learning_rate": 2.0299188408195742e-06,
+      "loss": 0.0485,
+      "step": 8443
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3147388028165598,
+      "learning_rate": 2.028821191129264e-06,
+      "loss": 0.0376,
+      "step": 8444
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2642972871459576,
+      "learning_rate": 2.0277237627496615e-06,
+      "loss": 0.0476,
+      "step": 8445
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5425764197585431,
+      "learning_rate": 2.026626555762508e-06,
+      "loss": 0.1123,
+      "step": 8446
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.27872088794663147,
+      "learning_rate": 2.02552957024953e-06,
+      "loss": 0.0673,
+      "step": 8447
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2565679160096511,
+      "learning_rate": 2.024432806292435e-06,
+      "loss": 0.073,
+      "step": 8448
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3731645511312853,
+      "learning_rate": 2.0233362639729204e-06,
+      "loss": 0.1083,
+      "step": 8449
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.21366848819269063,
+      "learning_rate": 2.0222399433726615e-06,
+      "loss": 0.0711,
+      "step": 8450
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3733688019053652,
+      "learning_rate": 2.021143844573316e-06,
+      "loss": 0.0985,
+      "step": 8451
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.43457777439224915,
+      "learning_rate": 2.0200479676565325e-06,
+      "loss": 0.1098,
+      "step": 8452
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.32355716354670944,
+      "learning_rate": 2.0189523127039355e-06,
+      "loss": 0.0882,
+      "step": 8453
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2907924190503596,
+      "learning_rate": 2.017856879797135e-06,
+      "loss": 0.075,
+      "step": 8454
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.23519417576165608,
+      "learning_rate": 2.0167616690177272e-06,
+      "loss": 0.0602,
+      "step": 8455
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.218485572369199,
+      "learning_rate": 2.0156666804472896e-06,
+      "loss": 0.0583,
+      "step": 8456
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3463550082841253,
+      "learning_rate": 2.0145719141673823e-06,
+      "loss": 0.0648,
+      "step": 8457
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.39630140393255936,
+      "learning_rate": 2.0134773702595484e-06,
+      "loss": 0.0757,
+      "step": 8458
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3866339619249023,
+      "learning_rate": 2.01238304880532e-06,
+      "loss": 0.0777,
+      "step": 8459
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3492133643110339,
+      "learning_rate": 2.011288949886206e-06,
+      "loss": 0.0409,
+      "step": 8460
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5039637790517895,
+      "learning_rate": 2.0101950735836994e-06,
+      "loss": 0.1156,
+      "step": 8461
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.32164155767504793,
+      "learning_rate": 2.009101419979283e-06,
+      "loss": 0.0349,
+      "step": 8462
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3122411688901697,
+      "learning_rate": 2.0080079891544155e-06,
+      "loss": 0.0469,
+      "step": 8463
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.37755826340016413,
+      "learning_rate": 2.006914781190543e-06,
+      "loss": 0.0964,
+      "step": 8464
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2753612623000977,
+      "learning_rate": 2.005821796169092e-06,
+      "loss": 0.0704,
+      "step": 8465
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.24769971796894905,
+      "learning_rate": 2.0047290341714777e-06,
+      "loss": 0.0757,
+      "step": 8466
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.23023768074387638,
+      "learning_rate": 2.0036364952790937e-06,
+      "loss": 0.0662,
+      "step": 8467
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.36002300891044653,
+      "learning_rate": 2.002544179573317e-06,
+      "loss": 0.0756,
+      "step": 8468
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5287592097186977,
+      "learning_rate": 2.0014520871355127e-06,
+      "loss": 0.0806,
+      "step": 8469
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.33091963321383,
+      "learning_rate": 2.0003602180470246e-06,
+      "loss": 0.0513,
+      "step": 8470
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.32636544346329205,
+      "learning_rate": 1.999268572389182e-06,
+      "loss": 0.0761,
+      "step": 8471
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.19175048623923208,
+      "learning_rate": 1.9981771502432947e-06,
+      "loss": 0.0379,
+      "step": 8472
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3044771248200612,
+      "learning_rate": 1.997085951690662e-06,
+      "loss": 0.0837,
+      "step": 8473
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3165734753174872,
+      "learning_rate": 1.995994976812561e-06,
+      "loss": 0.0787,
+      "step": 8474
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2409673826011493,
+      "learning_rate": 1.994904225690254e-06,
+      "loss": 0.0645,
+      "step": 8475
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2629626129129082,
+      "learning_rate": 1.993813698404984e-06,
+      "loss": 0.0809,
+      "step": 8476
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3930171347661737,
+      "learning_rate": 1.9927233950379838e-06,
+      "loss": 0.0757,
+      "step": 8477
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.31669100530274635,
+      "learning_rate": 1.9916333156704642e-06,
+      "loss": 0.0703,
+      "step": 8478
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4598854097267955,
+      "learning_rate": 1.99054346038362e-06,
+      "loss": 0.1017,
+      "step": 8479
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.25319977275121913,
+      "learning_rate": 1.9894538292586304e-06,
+      "loss": 0.0714,
+      "step": 8480
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2545310671076968,
+      "learning_rate": 1.988364422376656e-06,
+      "loss": 0.0718,
+      "step": 8481
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.34467019749411537,
+      "learning_rate": 1.987275239818844e-06,
+      "loss": 0.0626,
+      "step": 8482
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.34897606521787566,
+      "learning_rate": 1.9861862816663234e-06,
+      "loss": 0.0811,
+      "step": 8483
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3451391432867077,
+      "learning_rate": 1.9850975480002057e-06,
+      "loss": 0.0946,
+      "step": 8484
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.18572092932507714,
+      "learning_rate": 1.984009038901583e-06,
+      "loss": 0.0553,
+      "step": 8485
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3626592306908417,
+      "learning_rate": 1.9829207544515387e-06,
+      "loss": 0.108,
+      "step": 8486
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3983146219398756,
+      "learning_rate": 1.9818326947311326e-06,
+      "loss": 0.1087,
+      "step": 8487
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.31522883499755,
+      "learning_rate": 1.980744859821408e-06,
+      "loss": 0.0887,
+      "step": 8488
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2741636112516907,
+      "learning_rate": 1.979657249803396e-06,
+      "loss": 0.0694,
+      "step": 8489
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.39055946050310086,
+      "learning_rate": 1.978569864758107e-06,
+      "loss": 0.0842,
+      "step": 8490
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.34908140403243165,
+      "learning_rate": 1.9774827047665357e-06,
+      "loss": 0.0913,
+      "step": 8491
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.38331206874640184,
+      "learning_rate": 1.976395769909658e-06,
+      "loss": 0.0829,
+      "step": 8492
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2439903850610896,
+      "learning_rate": 1.975309060268439e-06,
+      "loss": 0.0469,
+      "step": 8493
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.4912761390955267,
+      "learning_rate": 1.9742225759238215e-06,
+      "loss": 0.1165,
+      "step": 8494
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.23214921079174586,
+      "learning_rate": 1.9731363169567307e-06,
+      "loss": 0.0441,
+      "step": 8495
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.25573491688857525,
+      "learning_rate": 1.972050283448082e-06,
+      "loss": 0.0617,
+      "step": 8496
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.29143229827524975,
+      "learning_rate": 1.970964475478768e-06,
+      "loss": 0.0564,
+      "step": 8497
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.40077345182736057,
+      "learning_rate": 1.969878893129664e-06,
+      "loss": 0.0602,
+      "step": 8498
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.334789715853125,
+      "learning_rate": 1.968793536481631e-06,
+      "loss": 0.0636,
+      "step": 8499
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.21077862709145756,
+      "learning_rate": 1.9677084056155144e-06,
+      "loss": 0.0547,
+      "step": 8500
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5042201979732444,
+      "learning_rate": 1.96662350061214e-06,
+      "loss": 0.1341,
+      "step": 8501
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.4295493063017846,
+      "learning_rate": 1.9655388215523173e-06,
+      "loss": 0.1111,
+      "step": 8502
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.4620603316204197,
+      "learning_rate": 1.9644543685168418e-06,
+      "loss": 0.117,
+      "step": 8503
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6368737974836071,
+      "learning_rate": 1.9633701415864883e-06,
+      "loss": 0.155,
+      "step": 8504
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6179040138009031,
+      "learning_rate": 1.9622861408420142e-06,
+      "loss": 0.1045,
+      "step": 8505
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.19967687952981883,
+      "learning_rate": 1.9612023663641656e-06,
+      "loss": 0.0538,
+      "step": 8506
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.20942268097242273,
+      "learning_rate": 1.960118818233668e-06,
+      "loss": 0.0615,
+      "step": 8507
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.22525760976935139,
+      "learning_rate": 1.959035496531229e-06,
+      "loss": 0.0623,
+      "step": 8508
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3570799984179504,
+      "learning_rate": 1.9579524013375394e-06,
+      "loss": 0.0985,
+      "step": 8509
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3836991323651682,
+      "learning_rate": 1.9568695327332776e-06,
+      "loss": 0.0712,
+      "step": 8510
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.25641064259957336,
+      "learning_rate": 1.955786890799101e-06,
+      "loss": 0.0692,
+      "step": 8511
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3780333985950308,
+      "learning_rate": 1.9547044756156493e-06,
+      "loss": 0.1008,
+      "step": 8512
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2892969338192432,
+      "learning_rate": 1.953622287263549e-06,
+      "loss": 0.08,
+      "step": 8513
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.38516021365613,
+      "learning_rate": 1.9525403258234084e-06,
+      "loss": 0.0829,
+      "step": 8514
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.42864268732933164,
+      "learning_rate": 1.951458591375816e-06,
+      "loss": 0.1032,
+      "step": 8515
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.16980827386873734,
+      "learning_rate": 1.950377084001346e-06,
+      "loss": 0.0438,
+      "step": 8516
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.32610088354772443,
+      "learning_rate": 1.9492958037805576e-06,
+      "loss": 0.0943,
+      "step": 8517
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.35624047022311983,
+      "learning_rate": 1.948214750793989e-06,
+      "loss": 0.0873,
+      "step": 8518
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.23688965826678443,
+      "learning_rate": 1.947133925122162e-06,
+      "loss": 0.0688,
+      "step": 8519
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.32540265060456125,
+      "learning_rate": 1.946053326845587e-06,
+      "loss": 0.0847,
+      "step": 8520
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.34916619318427433,
+      "learning_rate": 1.94497295604475e-06,
+      "loss": 0.0965,
+      "step": 8521
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.32179700827909685,
+      "learning_rate": 1.943892812800123e-06,
+      "loss": 0.0788,
+      "step": 8522
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2426318719346473,
+      "learning_rate": 1.9428128971921633e-06,
+      "loss": 0.0608,
+      "step": 8523
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2432053021198872,
+      "learning_rate": 1.941733209301309e-06,
+      "loss": 0.0427,
+      "step": 8524
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.34656210228519874,
+      "learning_rate": 1.9406537492079815e-06,
+      "loss": 0.0916,
+      "step": 8525
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.46448863782728345,
+      "learning_rate": 1.9395745169925823e-06,
+      "loss": 0.0996,
+      "step": 8526
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2374105775600577,
+      "learning_rate": 1.9384955127355032e-06,
+      "loss": 0.0595,
+      "step": 8527
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.368170653000627,
+      "learning_rate": 1.937416736517113e-06,
+      "loss": 0.1103,
+      "step": 8528
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5461100414797615,
+      "learning_rate": 1.9363381884177635e-06,
+      "loss": 0.1054,
+      "step": 8529
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.25884766353785843,
+      "learning_rate": 1.9352598685177942e-06,
+      "loss": 0.0772,
+      "step": 8530
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.31664226038978116,
+      "learning_rate": 1.934181776897524e-06,
+      "loss": 0.1167,
+      "step": 8531
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2720614345815531,
+      "learning_rate": 1.933103913637254e-06,
+      "loss": 0.0712,
+      "step": 8532
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.37683024052490904,
+      "learning_rate": 1.9320262788172696e-06,
+      "loss": 0.1149,
+      "step": 8533
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.45275804524832103,
+      "learning_rate": 1.9309488725178415e-06,
+      "loss": 0.0867,
+      "step": 8534
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2576651796996571,
+      "learning_rate": 1.9298716948192197e-06,
+      "loss": 0.0732,
+      "step": 8535
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.4300314333953212,
+      "learning_rate": 1.9287947458016374e-06,
+      "loss": 0.1012,
+      "step": 8536
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2754475262068441,
+      "learning_rate": 1.9277180255453154e-06,
+      "loss": 0.0675,
+      "step": 8537
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6744490503627898,
+      "learning_rate": 1.9266415341304524e-06,
+      "loss": 0.072,
+      "step": 8538
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.24537260077628936,
+      "learning_rate": 1.9255652716372307e-06,
+      "loss": 0.0799,
+      "step": 8539
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2332707268434878,
+      "learning_rate": 1.9244892381458164e-06,
+      "loss": 0.0525,
+      "step": 8540
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3244815322951511,
+      "learning_rate": 1.9234134337363613e-06,
+      "loss": 0.0882,
+      "step": 8541
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3890515872322682,
+      "learning_rate": 1.922337858488997e-06,
+      "loss": 0.094,
+      "step": 8542
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.29159352975622804,
+      "learning_rate": 1.9212625124838363e-06,
+      "loss": 0.0866,
+      "step": 8543
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.4266420484417009,
+      "learning_rate": 1.9201873958009776e-06,
+      "loss": 0.0899,
+      "step": 8544
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5099198120632253,
+      "learning_rate": 1.9191125085205047e-06,
+      "loss": 0.1197,
+      "step": 8545
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.4252373406429376,
+      "learning_rate": 1.9180378507224797e-06,
+      "loss": 0.1089,
+      "step": 8546
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.24882572374304812,
+      "learning_rate": 1.916963422486949e-06,
+      "loss": 0.0672,
+      "step": 8547
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.29076623430755955,
+      "learning_rate": 1.915889223893943e-06,
+      "loss": 0.0692,
+      "step": 8548
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2526994460435599,
+      "learning_rate": 1.9148152550234715e-06,
+      "loss": 0.073,
+      "step": 8549
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.49834985927370773,
+      "learning_rate": 1.913741515955535e-06,
+      "loss": 0.0782,
+      "step": 8550
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3531671118218845,
+      "learning_rate": 1.9126680067701085e-06,
+      "loss": 0.0798,
+      "step": 8551
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.304935035676818,
+      "learning_rate": 1.911594727547154e-06,
+      "loss": 0.0939,
+      "step": 8552
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2885637580433055,
+      "learning_rate": 1.9105216783666143e-06,
+      "loss": 0.0687,
+      "step": 8553
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3490227025641991,
+      "learning_rate": 1.909448859308418e-06,
+      "loss": 0.0825,
+      "step": 8554
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3019769104046824,
+      "learning_rate": 1.9083762704524756e-06,
+      "loss": 0.0686,
+      "step": 8555
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3140347874557103,
+      "learning_rate": 1.9073039118786786e-06,
+      "loss": 0.0641,
+      "step": 8556
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2921808536725846,
+      "learning_rate": 1.9062317836669002e-06,
+      "loss": 0.0409,
+      "step": 8557
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.25815241897086005,
+      "learning_rate": 1.9051598858970033e-06,
+      "loss": 0.1007,
+      "step": 8558
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.41246333032623084,
+      "learning_rate": 1.9040882186488263e-06,
+      "loss": 0.1049,
+      "step": 8559
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.1966014976410521,
+      "learning_rate": 1.9030167820021927e-06,
+      "loss": 0.042,
+      "step": 8560
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.18197116647764305,
+      "learning_rate": 1.9019455760369115e-06,
+      "loss": 0.0442,
+      "step": 8561
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.25935200573564243,
+      "learning_rate": 1.9008746008327722e-06,
+      "loss": 0.0707,
+      "step": 8562
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.39449134136279485,
+      "learning_rate": 1.8998038564695436e-06,
+      "loss": 0.0925,
+      "step": 8563
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.13576304778052012,
+      "learning_rate": 1.8987333430269861e-06,
+      "loss": 0.043,
+      "step": 8564
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.25640560295247133,
+      "learning_rate": 1.8976630605848357e-06,
+      "loss": 0.0821,
+      "step": 8565
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.34270795130679027,
+      "learning_rate": 1.8965930092228124e-06,
+      "loss": 0.098,
+      "step": 8566
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.31316471176577015,
+      "learning_rate": 1.8955231890206187e-06,
+      "loss": 0.0779,
+      "step": 8567
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.18601313957211163,
+      "learning_rate": 1.894453600057945e-06,
+      "loss": 0.0611,
+      "step": 8568
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.25663875456757806,
+      "learning_rate": 1.8933842424144583e-06,
+      "loss": 0.0623,
+      "step": 8569
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5170634353782716,
+      "learning_rate": 1.8923151161698084e-06,
+      "loss": 0.1217,
+      "step": 8570
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2782315748594582,
+      "learning_rate": 1.8912462214036343e-06,
+      "loss": 0.079,
+      "step": 8571
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.285729215681705,
+      "learning_rate": 1.8901775581955518e-06,
+      "loss": 0.1064,
+      "step": 8572
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.43080402499791876,
+      "learning_rate": 1.88910912662516e-06,
+      "loss": 0.0879,
+      "step": 8573
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3603952642305913,
+      "learning_rate": 1.8880409267720417e-06,
+      "loss": 0.0869,
+      "step": 8574
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.21097104183887597,
+      "learning_rate": 1.8869729587157647e-06,
+      "loss": 0.0508,
+      "step": 8575
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3247021848064528,
+      "learning_rate": 1.885905222535877e-06,
+      "loss": 0.129,
+      "step": 8576
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.19651041413473003,
+      "learning_rate": 1.8848377183119076e-06,
+      "loss": 0.0489,
+      "step": 8577
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3265273261981132,
+      "learning_rate": 1.8837704461233735e-06,
+      "loss": 0.0674,
+      "step": 8578
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2217765585260521,
+      "learning_rate": 1.88270340604977e-06,
+      "loss": 0.0571,
+      "step": 8579
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6220350260843028,
+      "learning_rate": 1.8816365981705748e-06,
+      "loss": 0.1044,
+      "step": 8580
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.29388577121007736,
+      "learning_rate": 1.8805700225652534e-06,
+      "loss": 0.0833,
+      "step": 8581
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3096248031484029,
+      "learning_rate": 1.8795036793132487e-06,
+      "loss": 0.0708,
+      "step": 8582
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3596016465172378,
+      "learning_rate": 1.8784375684939881e-06,
+      "loss": 0.1037,
+      "step": 8583
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2806453467013293,
+      "learning_rate": 1.8773716901868805e-06,
+      "loss": 0.0803,
+      "step": 8584
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2722194069570301,
+      "learning_rate": 1.876306044471321e-06,
+      "loss": 0.0619,
+      "step": 8585
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.500368177699109,
+      "learning_rate": 1.8752406314266848e-06,
+      "loss": 0.1074,
+      "step": 8586
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.19191877413248948,
+      "learning_rate": 1.8741754511323273e-06,
+      "loss": 0.0604,
+      "step": 8587
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.22010401451192818,
+      "learning_rate": 1.8731105036675939e-06,
+      "loss": 0.0579,
+      "step": 8588
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.17140897248582623,
+      "learning_rate": 1.872045789111805e-06,
+      "loss": 0.0509,
+      "step": 8589
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.16405010773083645,
+      "learning_rate": 1.8709813075442678e-06,
+      "loss": 0.0293,
+      "step": 8590
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.528591413495145,
+      "learning_rate": 1.8699170590442683e-06,
+      "loss": 0.0988,
+      "step": 8591
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.38761348223789505,
+      "learning_rate": 1.868853043691083e-06,
+      "loss": 0.0784,
+      "step": 8592
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.22520792456417432,
+      "learning_rate": 1.8677892615639626e-06,
+      "loss": 0.0525,
+      "step": 8593
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.17809455928089962,
+      "learning_rate": 1.866725712742143e-06,
+      "loss": 0.046,
+      "step": 8594
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.41192993824968416,
+      "learning_rate": 1.8656623973048466e-06,
+      "loss": 0.1019,
+      "step": 8595
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3836326776477548,
+      "learning_rate": 1.8645993153312736e-06,
+      "loss": 0.0816,
+      "step": 8596
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2956545262143723,
+      "learning_rate": 1.8635364669006073e-06,
+      "loss": 0.0467,
+      "step": 8597
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.46303551666203485,
+      "learning_rate": 1.8624738520920176e-06,
+      "loss": 0.1137,
+      "step": 8598
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.25992743855380407,
+      "learning_rate": 1.8614114709846531e-06,
+      "loss": 0.0693,
+      "step": 8599
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2823570899252034,
+      "learning_rate": 1.8603493236576458e-06,
+      "loss": 0.087,
+      "step": 8600
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3077436998283426,
+      "learning_rate": 1.8592874101901087e-06,
+      "loss": 0.0693,
+      "step": 8601
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5004601074837179,
+      "learning_rate": 1.8582257306611435e-06,
+      "loss": 0.1339,
+      "step": 8602
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.374641622616434,
+      "learning_rate": 1.8571642851498278e-06,
+      "loss": 0.0952,
+      "step": 8603
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.24605204811060066,
+      "learning_rate": 1.8561030737352226e-06,
+      "loss": 0.0795,
+      "step": 8604
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.17920773535033102,
+      "learning_rate": 1.8550420964963772e-06,
+      "loss": 0.0293,
+      "step": 8605
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.19454660686575073,
+      "learning_rate": 1.8539813535123168e-06,
+      "loss": 0.0468,
+      "step": 8606
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.27688784662274873,
+      "learning_rate": 1.8529208448620522e-06,
+      "loss": 0.0554,
+      "step": 8607
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2860967402599092,
+      "learning_rate": 1.8518605706245745e-06,
+      "loss": 0.0617,
+      "step": 8608
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.34831612536026135,
+      "learning_rate": 1.8508005308788623e-06,
+      "loss": 0.0905,
+      "step": 8609
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.21994747701689207,
+      "learning_rate": 1.8497407257038724e-06,
+      "loss": 0.0537,
+      "step": 8610
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6481709251188935,
+      "learning_rate": 1.8486811551785445e-06,
+      "loss": 0.1338,
+      "step": 8611
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.16854281677463817,
+      "learning_rate": 1.8476218193818007e-06,
+      "loss": 0.0389,
+      "step": 8612
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.22451520637281053,
+      "learning_rate": 1.8465627183925488e-06,
+      "loss": 0.0728,
+      "step": 8613
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3524275447506338,
+      "learning_rate": 1.8455038522896761e-06,
+      "loss": 0.0812,
+      "step": 8614
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.41652203151895745,
+      "learning_rate": 1.8444452211520531e-06,
+      "loss": 0.0986,
+      "step": 8615
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.35620612459749806,
+      "learning_rate": 1.8433868250585318e-06,
+      "loss": 0.0788,
+      "step": 8616
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.31026363685158903,
+      "learning_rate": 1.8423286640879474e-06,
+      "loss": 0.0682,
+      "step": 8617
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.41123393727912216,
+      "learning_rate": 1.8412707383191203e-06,
+      "loss": 0.0775,
+      "step": 8618
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2898005388635064,
+      "learning_rate": 1.8402130478308495e-06,
+      "loss": 0.0823,
+      "step": 8619
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.36717247211938964,
+      "learning_rate": 1.8391555927019177e-06,
+      "loss": 0.0836,
+      "step": 8620
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.4549120776121371,
+      "learning_rate": 1.8380983730110895e-06,
+      "loss": 0.1216,
+      "step": 8621
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.21057728343910734,
+      "learning_rate": 1.8370413888371152e-06,
+      "loss": 0.0511,
+      "step": 8622
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3138541833302261,
+      "learning_rate": 1.8359846402587246e-06,
+      "loss": 0.0622,
+      "step": 8623
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.23827953589534256,
+      "learning_rate": 1.8349281273546288e-06,
+      "loss": 0.0747,
+      "step": 8624
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.4342755359960006,
+      "learning_rate": 1.833871850203523e-06,
+      "loss": 0.0949,
+      "step": 8625
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.14727810370518662,
+      "learning_rate": 1.832815808884087e-06,
+      "loss": 0.0428,
+      "step": 8626
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.22125256958175882,
+      "learning_rate": 1.83176000347498e-06,
+      "loss": 0.0544,
+      "step": 8627
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3765734030478906,
+      "learning_rate": 1.8307044340548425e-06,
+      "loss": 0.1002,
+      "step": 8628
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.4268078333894236,
+      "learning_rate": 1.8296491007023032e-06,
+      "loss": 0.0767,
+      "step": 8629
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.47729681115747163,
+      "learning_rate": 1.8285940034959675e-06,
+      "loss": 0.1061,
+      "step": 8630
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.34937699841822933,
+      "learning_rate": 1.8275391425144234e-06,
+      "loss": 0.0783,
+      "step": 8631
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2177383747992872,
+      "learning_rate": 1.8264845178362467e-06,
+      "loss": 0.0549,
+      "step": 8632
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2527299874051213,
+      "learning_rate": 1.8254301295399901e-06,
+      "loss": 0.0816,
+      "step": 8633
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2699610776027627,
+      "learning_rate": 1.8243759777041908e-06,
+      "loss": 0.0787,
+      "step": 8634
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.25467081786197815,
+      "learning_rate": 1.823322062407366e-06,
+      "loss": 0.0493,
+      "step": 8635
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.29735273146432556,
+      "learning_rate": 1.822268383728022e-06,
+      "loss": 0.0821,
+      "step": 8636
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3053433320474836,
+      "learning_rate": 1.8212149417446395e-06,
+      "loss": 0.0884,
+      "step": 8637
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3045493640132282,
+      "learning_rate": 1.8201617365356849e-06,
+      "loss": 0.0677,
+      "step": 8638
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.21505617293362075,
+      "learning_rate": 1.8191087681796094e-06,
+      "loss": 0.0416,
+      "step": 8639
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3081263273269807,
+      "learning_rate": 1.8180560367548433e-06,
+      "loss": 0.0886,
+      "step": 8640
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3350906921722642,
+      "learning_rate": 1.8170035423397996e-06,
+      "loss": 0.0951,
+      "step": 8641
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2904368106063874,
+      "learning_rate": 1.8159512850128724e-06,
+      "loss": 0.0933,
+      "step": 8642
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.36128543941868674,
+      "learning_rate": 1.8148992648524439e-06,
+      "loss": 0.0918,
+      "step": 8643
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.26085713603674,
+      "learning_rate": 1.813847481936873e-06,
+      "loss": 0.0551,
+      "step": 8644
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3249519606778011,
+      "learning_rate": 1.8127959363445003e-06,
+      "loss": 0.0611,
+      "step": 8645
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2057574035486612,
+      "learning_rate": 1.811744628153655e-06,
+      "loss": 0.0727,
+      "step": 8646
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.23028273606829058,
+      "learning_rate": 1.8106935574426426e-06,
+      "loss": 0.0618,
+      "step": 8647
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2585961472052786,
+      "learning_rate": 1.8096427242897513e-06,
+      "loss": 0.0587,
+      "step": 8648
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6153779768180376,
+      "learning_rate": 1.8085921287732572e-06,
+      "loss": 0.1088,
+      "step": 8649
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.4221687483075273,
+      "learning_rate": 1.807541770971412e-06,
+      "loss": 0.0647,
+      "step": 8650
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.28487048551357025,
+      "learning_rate": 1.8064916509624537e-06,
+      "loss": 0.0896,
+      "step": 8651
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3081947861834471,
+      "learning_rate": 1.805441768824599e-06,
+      "loss": 0.0573,
+      "step": 8652
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.40059624030232976,
+      "learning_rate": 1.804392124636053e-06,
+      "loss": 0.1021,
+      "step": 8653
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2751407282957148,
+      "learning_rate": 1.803342718474997e-06,
+      "loss": 0.0556,
+      "step": 8654
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.514292756242048,
+      "learning_rate": 1.8022935504195953e-06,
+      "loss": 0.1442,
+      "step": 8655
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3485153096511675,
+      "learning_rate": 1.801244620548e-06,
+      "loss": 0.0689,
+      "step": 8656
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.36193338219730853,
+      "learning_rate": 1.8001959289383397e-06,
+      "loss": 0.0938,
+      "step": 8657
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.30244050086938523,
+      "learning_rate": 1.7991474756687266e-06,
+      "loss": 0.1069,
+      "step": 8658
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.22308022945755457,
+      "learning_rate": 1.7980992608172543e-06,
+      "loss": 0.0751,
+      "step": 8659
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2871274660145017,
+      "learning_rate": 1.7970512844620031e-06,
+      "loss": 0.0801,
+      "step": 8660
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.34817401355808386,
+      "learning_rate": 1.7960035466810306e-06,
+      "loss": 0.1108,
+      "step": 8661
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3875303957814468,
+      "learning_rate": 1.794956047552377e-06,
+      "loss": 0.0896,
+      "step": 8662
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.1725560357197752,
+      "learning_rate": 1.7939087871540695e-06,
+      "loss": 0.0378,
+      "step": 8663
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3513066545607775,
+      "learning_rate": 1.7928617655641122e-06,
+      "loss": 0.066,
+      "step": 8664
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3779562358907929,
+      "learning_rate": 1.791814982860494e-06,
+      "loss": 0.066,
+      "step": 8665
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3379868295796861,
+      "learning_rate": 1.7907684391211834e-06,
+      "loss": 0.1065,
+      "step": 8666
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.284081879878797,
+      "learning_rate": 1.7897221344241355e-06,
+      "loss": 0.0737,
+      "step": 8667
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2921253828260448,
+      "learning_rate": 1.7886760688472849e-06,
+      "loss": 0.0789,
+      "step": 8668
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.26074279692059793,
+      "learning_rate": 1.787630242468546e-06,
+      "loss": 0.0532,
+      "step": 8669
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.24148583940991145,
+      "learning_rate": 1.7865846553658222e-06,
+      "loss": 0.0501,
+      "step": 8670
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.4316766538791286,
+      "learning_rate": 1.7855393076169924e-06,
+      "loss": 0.0717,
+      "step": 8671
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.10586896885396005,
+      "learning_rate": 1.7844941992999187e-06,
+      "loss": 0.0142,
+      "step": 8672
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.42154239011994626,
+      "learning_rate": 1.7834493304924511e-06,
+      "loss": 0.1098,
+      "step": 8673
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2027515746861528,
+      "learning_rate": 1.7824047012724144e-06,
+      "loss": 0.0517,
+      "step": 8674
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.48287025620246105,
+      "learning_rate": 1.7813603117176204e-06,
+      "loss": 0.1156,
+      "step": 8675
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2524719521194105,
+      "learning_rate": 1.780316161905858e-06,
+      "loss": 0.0654,
+      "step": 8676
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.390376371558901,
+      "learning_rate": 1.7792722519149054e-06,
+      "loss": 0.1255,
+      "step": 8677
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.35560524744515615,
+      "learning_rate": 1.7782285818225181e-06,
+      "loss": 0.0979,
+      "step": 8678
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2861444678733224,
+      "learning_rate": 1.7771851517064337e-06,
+      "loss": 0.0663,
+      "step": 8679
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5039509023997834,
+      "learning_rate": 1.776141961644372e-06,
+      "loss": 0.1548,
+      "step": 8680
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.29094961281556475,
+      "learning_rate": 1.7750990117140393e-06,
+      "loss": 0.0761,
+      "step": 8681
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.1977818371205157,
+      "learning_rate": 1.7740563019931185e-06,
+      "loss": 0.0482,
+      "step": 8682
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.1685375988936261,
+      "learning_rate": 1.7730138325592772e-06,
+      "loss": 0.0513,
+      "step": 8683
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.31627008190400774,
+      "learning_rate": 1.7719716034901635e-06,
+      "loss": 0.0709,
+      "step": 8684
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.21724756290984287,
+      "learning_rate": 1.7709296148634081e-06,
+      "loss": 0.0521,
+      "step": 8685
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.19879558195739336,
+      "learning_rate": 1.7698878667566277e-06,
+      "loss": 0.062,
+      "step": 8686
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2932456994401182,
+      "learning_rate": 1.7688463592474154e-06,
+      "loss": 0.0728,
+      "step": 8687
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2491461840708174,
+      "learning_rate": 1.7678050924133488e-06,
+      "loss": 0.0768,
+      "step": 8688
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3128929919532564,
+      "learning_rate": 1.7667640663319868e-06,
+      "loss": 0.069,
+      "step": 8689
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6108429328082423,
+      "learning_rate": 1.7657232810808738e-06,
+      "loss": 0.16,
+      "step": 8690
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.28882902424628315,
+      "learning_rate": 1.7646827367375314e-06,
+      "loss": 0.0727,
+      "step": 8691
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3390837493679277,
+      "learning_rate": 1.7636424333794661e-06,
+      "loss": 0.1063,
+      "step": 8692
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2842551280277906,
+      "learning_rate": 1.7626023710841638e-06,
+      "loss": 0.0846,
+      "step": 8693
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.31923131279803996,
+      "learning_rate": 1.7615625499290984e-06,
+      "loss": 0.0769,
+      "step": 8694
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3640303854454691,
+      "learning_rate": 1.7605229699917186e-06,
+      "loss": 0.0959,
+      "step": 8695
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.46707786235956106,
+      "learning_rate": 1.7594836313494583e-06,
+      "loss": 0.1248,
+      "step": 8696
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3106428675461031,
+      "learning_rate": 1.7584445340797362e-06,
+      "loss": 0.1002,
+      "step": 8697
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.22159452429851254,
+      "learning_rate": 1.7574056782599486e-06,
+      "loss": 0.047,
+      "step": 8698
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3232279840228523,
+      "learning_rate": 1.7563670639674756e-06,
+      "loss": 0.0842,
+      "step": 8699
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.30923461729718876,
+      "learning_rate": 1.7553286912796775e-06,
+      "loss": 0.0617,
+      "step": 8700
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3398334293264469,
+      "learning_rate": 1.7542905602739018e-06,
+      "loss": 0.0844,
+      "step": 8701
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3085803356767103,
+      "learning_rate": 1.7532526710274728e-06,
+      "loss": 0.1051,
+      "step": 8702
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.31046214472357836,
+      "learning_rate": 1.7522150236176966e-06,
+      "loss": 0.0962,
+      "step": 8703
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2768882979956396,
+      "learning_rate": 1.751177618121867e-06,
+      "loss": 0.0782,
+      "step": 8704
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3326129294764262,
+      "learning_rate": 1.7501404546172545e-06,
+      "loss": 0.0812,
+      "step": 8705
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.27416905221469967,
+      "learning_rate": 1.7491035331811112e-06,
+      "loss": 0.0657,
+      "step": 8706
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.35078918052918373,
+      "learning_rate": 1.7480668538906759e-06,
+      "loss": 0.0914,
+      "step": 8707
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.23242424127272066,
+      "learning_rate": 1.7470304168231656e-06,
+      "loss": 0.0521,
+      "step": 8708
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.25873811728404233,
+      "learning_rate": 1.7459942220557791e-06,
+      "loss": 0.0878,
+      "step": 8709
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.26841697766911105,
+      "learning_rate": 1.744958269665698e-06,
+      "loss": 0.0765,
+      "step": 8710
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.24867793403145752,
+      "learning_rate": 1.743922559730088e-06,
+      "loss": 0.052,
+      "step": 8711
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.4176886883940249,
+      "learning_rate": 1.7428870923260944e-06,
+      "loss": 0.0908,
+      "step": 8712
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.37593505594656645,
+      "learning_rate": 1.741851867530842e-06,
+      "loss": 0.0687,
+      "step": 8713
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5328837487002965,
+      "learning_rate": 1.7408168854214451e-06,
+      "loss": 0.1433,
+      "step": 8714
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.22322119893185083,
+      "learning_rate": 1.739782146074992e-06,
+      "loss": 0.0717,
+      "step": 8715
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.19109203210180206,
+      "learning_rate": 1.7387476495685574e-06,
+      "loss": 0.0384,
+      "step": 8716
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6718769446273014,
+      "learning_rate": 1.7377133959791941e-06,
+      "loss": 0.101,
+      "step": 8717
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.22730740276550823,
+      "learning_rate": 1.7366793853839426e-06,
+      "loss": 0.0629,
+      "step": 8718
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.23334270942185734,
+      "learning_rate": 1.7356456178598214e-06,
+      "loss": 0.0667,
+      "step": 8719
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2985597337975208,
+      "learning_rate": 1.734612093483829e-06,
+      "loss": 0.0715,
+      "step": 8720
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.4439876551516199,
+      "learning_rate": 1.7335788123329517e-06,
+      "loss": 0.1155,
+      "step": 8721
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3012377339851268,
+      "learning_rate": 1.7325457744841523e-06,
+      "loss": 0.079,
+      "step": 8722
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.20970008898597461,
+      "learning_rate": 1.731512980014377e-06,
+      "loss": 0.0637,
+      "step": 8723
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.25080807724361265,
+      "learning_rate": 1.7304804290005568e-06,
+      "loss": 0.045,
+      "step": 8724
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.4761017383300039,
+      "learning_rate": 1.7294481215196002e-06,
+      "loss": 0.1229,
+      "step": 8725
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.21185000446656518,
+      "learning_rate": 1.7284160576484005e-06,
+      "loss": 0.0657,
+      "step": 8726
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3552183574719446,
+      "learning_rate": 1.727384237463829e-06,
+      "loss": 0.1029,
+      "step": 8727
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.380019654986308,
+      "learning_rate": 1.7263526610427455e-06,
+      "loss": 0.0766,
+      "step": 8728
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3148135389349021,
+      "learning_rate": 1.725321328461987e-06,
+      "loss": 0.0792,
+      "step": 8729
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.22413238574563393,
+      "learning_rate": 1.7242902397983703e-06,
+      "loss": 0.0844,
+      "step": 8730
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.364385506699739,
+      "learning_rate": 1.7232593951287002e-06,
+      "loss": 0.0759,
+      "step": 8731
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.32918589855198443,
+      "learning_rate": 1.7222287945297594e-06,
+      "loss": 0.071,
+      "step": 8732
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3533364453809993,
+      "learning_rate": 1.721198438078312e-06,
+      "loss": 0.0779,
+      "step": 8733
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3475817541011987,
+      "learning_rate": 1.7201683258511036e-06,
+      "loss": 0.0844,
+      "step": 8734
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.30900219663427225,
+      "learning_rate": 1.7191384579248666e-06,
+      "loss": 0.0779,
+      "step": 8735
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3754553406086458,
+      "learning_rate": 1.7181088343763096e-06,
+      "loss": 0.0994,
+      "step": 8736
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.33187186455592665,
+      "learning_rate": 1.7170794552821235e-06,
+      "loss": 0.098,
+      "step": 8737
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.4973758225921344,
+      "learning_rate": 1.716050320718986e-06,
+      "loss": 0.1078,
+      "step": 8738
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.302222666763661,
+      "learning_rate": 1.7150214307635503e-06,
+      "loss": 0.0645,
+      "step": 8739
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.26073419589042646,
+      "learning_rate": 1.7139927854924537e-06,
+      "loss": 0.0625,
+      "step": 8740
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3014173484670803,
+      "learning_rate": 1.7129643849823186e-06,
+      "loss": 0.062,
+      "step": 8741
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3334597613426135,
+      "learning_rate": 1.7119362293097442e-06,
+      "loss": 0.0651,
+      "step": 8742
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.4658117011075586,
+      "learning_rate": 1.7109083185513137e-06,
+      "loss": 0.1199,
+      "step": 8743
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.24402236638300986,
+      "learning_rate": 1.7098806527835904e-06,
+      "loss": 0.0603,
+      "step": 8744
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.31070114321252706,
+      "learning_rate": 1.7088532320831247e-06,
+      "loss": 0.0907,
+      "step": 8745
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3755196523547807,
+      "learning_rate": 1.707826056526442e-06,
+      "loss": 0.0969,
+      "step": 8746
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.21531389187109795,
+      "learning_rate": 1.706799126190053e-06,
+      "loss": 0.0644,
+      "step": 8747
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.46027226049894004,
+      "learning_rate": 1.7057724411504478e-06,
+      "loss": 0.1275,
+      "step": 8748
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3551730442805734,
+      "learning_rate": 1.7047460014841033e-06,
+      "loss": 0.0778,
+      "step": 8749
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3638717755514339,
+      "learning_rate": 1.7037198072674727e-06,
+      "loss": 0.0648,
+      "step": 8750
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2549676718095175,
+      "learning_rate": 1.702693858576993e-06,
+      "loss": 0.0486,
+      "step": 8751
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.4794863435420585,
+      "learning_rate": 1.7016681554890835e-06,
+      "loss": 0.1008,
+      "step": 8752
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2674691611776805,
+      "learning_rate": 1.7006426980801416e-06,
+      "loss": 0.0412,
+      "step": 8753
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.24301743057900513,
+      "learning_rate": 1.699617486426554e-06,
+      "loss": 0.0737,
+      "step": 8754
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5365760752924508,
+      "learning_rate": 1.6985925206046817e-06,
+      "loss": 0.1111,
+      "step": 8755
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.25325189925172464,
+      "learning_rate": 1.6975678006908707e-06,
+      "loss": 0.0708,
+      "step": 8756
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.30387032501115085,
+      "learning_rate": 1.6965433267614466e-06,
+      "loss": 0.0751,
+      "step": 8757
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.48228192104702716,
+      "learning_rate": 1.6955190988927212e-06,
+      "loss": 0.1358,
+      "step": 8758
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3297876811440097,
+      "learning_rate": 1.694495117160983e-06,
+      "loss": 0.0742,
+      "step": 8759
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2444151792730252,
+      "learning_rate": 1.6934713816425048e-06,
+      "loss": 0.0638,
+      "step": 8760
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.26093111820011816,
+      "learning_rate": 1.692447892413538e-06,
+      "loss": 0.0783,
+      "step": 8761
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2046438594919639,
+      "learning_rate": 1.6914246495503224e-06,
+      "loss": 0.0559,
+      "step": 8762
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2454665511297569,
+      "learning_rate": 1.6904016531290718e-06,
+      "loss": 0.0638,
+      "step": 8763
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3658237098586329,
+      "learning_rate": 1.6893789032259849e-06,
+      "loss": 0.1023,
+      "step": 8764
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3951852791666201,
+      "learning_rate": 1.6883563999172447e-06,
+      "loss": 0.0849,
+      "step": 8765
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.4265537722389651,
+      "learning_rate": 1.6873341432790114e-06,
+      "loss": 0.0953,
+      "step": 8766
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3279769070756954,
+      "learning_rate": 1.6863121333874294e-06,
+      "loss": 0.0899,
+      "step": 8767
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3823382305434872,
+      "learning_rate": 1.6852903703186208e-06,
+      "loss": 0.0872,
+      "step": 8768
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.35788709796349316,
+      "learning_rate": 1.6842688541486973e-06,
+      "loss": 0.0423,
+      "step": 8769
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.29422093668250787,
+      "learning_rate": 1.683247584953745e-06,
+      "loss": 0.0622,
+      "step": 8770
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2708870773010524,
+      "learning_rate": 1.6822265628098323e-06,
+      "loss": 0.0584,
+      "step": 8771
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3566076802147964,
+      "learning_rate": 1.6812057877930143e-06,
+      "loss": 0.0875,
+      "step": 8772
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.33692850483435854,
+      "learning_rate": 1.6801852599793228e-06,
+      "loss": 0.0736,
+      "step": 8773
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.24828768418671482,
+      "learning_rate": 1.6791649794447707e-06,
+      "loss": 0.0674,
+      "step": 8774
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2914861070819317,
+      "learning_rate": 1.6781449462653581e-06,
+      "loss": 0.0685,
+      "step": 8775
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.4039476693451992,
+      "learning_rate": 1.6771251605170607e-06,
+      "loss": 0.1261,
+      "step": 8776
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2599833385056515,
+      "learning_rate": 1.6761056222758392e-06,
+      "loss": 0.0799,
+      "step": 8777
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.17469350611586879,
+      "learning_rate": 1.675086331617632e-06,
+      "loss": 0.037,
+      "step": 8778
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.48777290571931775,
+      "learning_rate": 1.6740672886183657e-06,
+      "loss": 0.1124,
+      "step": 8779
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3876369724582235,
+      "learning_rate": 1.673048493353942e-06,
+      "loss": 0.0799,
+      "step": 8780
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.279589861323937,
+      "learning_rate": 1.6720299459002465e-06,
+      "loss": 0.078,
+      "step": 8781
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.43792598401609445,
+      "learning_rate": 1.6710116463331494e-06,
+      "loss": 0.0721,
+      "step": 8782
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.34440930256098673,
+      "learning_rate": 1.6699935947284968e-06,
+      "loss": 0.1047,
+      "step": 8783
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.22879595501186198,
+      "learning_rate": 1.6689757911621208e-06,
+      "loss": 0.0484,
+      "step": 8784
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2653312986755267,
+      "learning_rate": 1.6679582357098305e-06,
+      "loss": 0.0772,
+      "step": 8785
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.23865071920660855,
+      "learning_rate": 1.666940928447423e-06,
+      "loss": 0.0607,
+      "step": 8786
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.25499019530608713,
+      "learning_rate": 1.6659238694506718e-06,
+      "loss": 0.1018,
+      "step": 8787
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.442939863632077,
+      "learning_rate": 1.6649070587953315e-06,
+      "loss": 0.0846,
+      "step": 8788
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.23052421375233845,
+      "learning_rate": 1.6638904965571434e-06,
+      "loss": 0.0408,
+      "step": 8789
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5262021594960135,
+      "learning_rate": 1.6628741828118255e-06,
+      "loss": 0.1312,
+      "step": 8790
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.28748731992576276,
+      "learning_rate": 1.6618581176350783e-06,
+      "loss": 0.0891,
+      "step": 8791
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.36642481646238256,
+      "learning_rate": 1.6608423011025838e-06,
+      "loss": 0.0844,
+      "step": 8792
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.24318038904527697,
+      "learning_rate": 1.6598267332900076e-06,
+      "loss": 0.0679,
+      "step": 8793
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2906128558843935,
+      "learning_rate": 1.6588114142729945e-06,
+      "loss": 0.0711,
+      "step": 8794
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.25204536125244803,
+      "learning_rate": 1.6577963441271694e-06,
+      "loss": 0.0751,
+      "step": 8795
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2940313094974745,
+      "learning_rate": 1.6567815229281442e-06,
+      "loss": 0.0827,
+      "step": 8796
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3862467664450845,
+      "learning_rate": 1.6557669507515068e-06,
+      "loss": 0.0838,
+      "step": 8797
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3543092068951903,
+      "learning_rate": 1.654752627672827e-06,
+      "loss": 0.1055,
+      "step": 8798
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3525550025002627,
+      "learning_rate": 1.6537385537676604e-06,
+      "loss": 0.1013,
+      "step": 8799
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.22784324146758156,
+      "learning_rate": 1.65272472911154e-06,
+      "loss": 0.0465,
+      "step": 8800
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.24767915408011168,
+      "learning_rate": 1.6517111537799807e-06,
+      "loss": 0.0902,
+      "step": 8801
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.40542789502037624,
+      "learning_rate": 1.6506978278484786e-06,
+      "loss": 0.0858,
+      "step": 8802
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.4057641943239592,
+      "learning_rate": 1.6496847513925146e-06,
+      "loss": 0.1013,
+      "step": 8803
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.27192793059634984,
+      "learning_rate": 1.6486719244875477e-06,
+      "loss": 0.0946,
+      "step": 8804
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2540788002373435,
+      "learning_rate": 1.6476593472090174e-06,
+      "loss": 0.0618,
+      "step": 8805
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.31236944112448795,
+      "learning_rate": 1.6466470196323487e-06,
+      "loss": 0.0732,
+      "step": 8806
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.30085462504372423,
+      "learning_rate": 1.6456349418329453e-06,
+      "loss": 0.069,
+      "step": 8807
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3095148700451659,
+      "learning_rate": 1.6446231138861917e-06,
+      "loss": 0.0653,
+      "step": 8808
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3365481333236518,
+      "learning_rate": 1.6436115358674537e-06,
+      "loss": 0.0836,
+      "step": 8809
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.24706560873289923,
+      "learning_rate": 1.6426002078520825e-06,
+      "loss": 0.0641,
+      "step": 8810
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2857532907945194,
+      "learning_rate": 1.6415891299154062e-06,
+      "loss": 0.0839,
+      "step": 8811
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.39810543980973123,
+      "learning_rate": 1.6405783021327337e-06,
+      "loss": 0.102,
+      "step": 8812
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5595514057121227,
+      "learning_rate": 1.6395677245793612e-06,
+      "loss": 0.1063,
+      "step": 8813
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.28743196867553805,
+      "learning_rate": 1.6385573973305608e-06,
+      "loss": 0.0652,
+      "step": 8814
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3249305774448179,
+      "learning_rate": 1.6375473204615871e-06,
+      "loss": 0.063,
+      "step": 8815
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.1794434629666851,
+      "learning_rate": 1.6365374940476748e-06,
+      "loss": 0.0558,
+      "step": 8816
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.24525414621512315,
+      "learning_rate": 1.6355279181640454e-06,
+      "loss": 0.0911,
+      "step": 8817
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.39460153155299793,
+      "learning_rate": 1.6345185928858964e-06,
+      "loss": 0.0539,
+      "step": 8818
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5604901117794143,
+      "learning_rate": 1.6335095182884074e-06,
+      "loss": 0.0836,
+      "step": 8819
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.25226085137623727,
+      "learning_rate": 1.6325006944467415e-06,
+      "loss": 0.0809,
+      "step": 8820
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.27183695482667647,
+      "learning_rate": 1.6314921214360385e-06,
+      "loss": 0.0806,
+      "step": 8821
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2262437959587454,
+      "learning_rate": 1.6304837993314276e-06,
+      "loss": 0.0564,
+      "step": 8822
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.29919211156628095,
+      "learning_rate": 1.629475728208012e-06,
+      "loss": 0.0599,
+      "step": 8823
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.19684777812558601,
+      "learning_rate": 1.628467908140879e-06,
+      "loss": 0.065,
+      "step": 8824
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.28108651643597476,
+      "learning_rate": 1.6274603392050954e-06,
+      "loss": 0.0902,
+      "step": 8825
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.33640273344399413,
+      "learning_rate": 1.6264530214757136e-06,
+      "loss": 0.0626,
+      "step": 8826
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5746747114343644,
+      "learning_rate": 1.6254459550277629e-06,
+      "loss": 0.1159,
+      "step": 8827
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.29607922246677104,
+      "learning_rate": 1.6244391399362559e-06,
+      "loss": 0.0701,
+      "step": 8828
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.28716385949240214,
+      "learning_rate": 1.6234325762761844e-06,
+      "loss": 0.0716,
+      "step": 8829
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2831341975382209,
+      "learning_rate": 1.6224262641225258e-06,
+      "loss": 0.0761,
+      "step": 8830
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2524898251032297,
+      "learning_rate": 1.6214202035502353e-06,
+      "loss": 0.0799,
+      "step": 8831
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2459502549435526,
+      "learning_rate": 1.6204143946342476e-06,
+      "loss": 0.0502,
+      "step": 8832
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3257782433241818,
+      "learning_rate": 1.6194088374494853e-06,
+      "loss": 0.0778,
+      "step": 8833
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.43897238548525785,
+      "learning_rate": 1.6184035320708464e-06,
+      "loss": 0.1223,
+      "step": 8834
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.19824750425729723,
+      "learning_rate": 1.617398478573211e-06,
+      "loss": 0.0569,
+      "step": 8835
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3836055503811429,
+      "learning_rate": 1.616393677031441e-06,
+      "loss": 0.0952,
+      "step": 8836
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.19337413493663652,
+      "learning_rate": 1.6153891275203826e-06,
+      "loss": 0.0409,
+      "step": 8837
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2828506589726233,
+      "learning_rate": 1.6143848301148585e-06,
+      "loss": 0.0672,
+      "step": 8838
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.218640198905327,
+      "learning_rate": 1.6133807848896733e-06,
+      "loss": 0.0768,
+      "step": 8839
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.4332412141416325,
+      "learning_rate": 1.6123769919196174e-06,
+      "loss": 0.1091,
+      "step": 8840
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.34063336551874746,
+      "learning_rate": 1.6113734512794576e-06,
+      "loss": 0.0955,
+      "step": 8841
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.21132869066053817,
+      "learning_rate": 1.6103701630439434e-06,
+      "loss": 0.0569,
+      "step": 8842
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2833890841688146,
+      "learning_rate": 1.6093671272878037e-06,
+      "loss": 0.062,
+      "step": 8843
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2447811069010782,
+      "learning_rate": 1.6083643440857538e-06,
+      "loss": 0.0756,
+      "step": 8844
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3565920962407,
+      "learning_rate": 1.6073618135124858e-06,
+      "loss": 0.0756,
+      "step": 8845
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2698247726912429,
+      "learning_rate": 1.6063595356426715e-06,
+      "loss": 0.0676,
+      "step": 8846
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3307486771866237,
+      "learning_rate": 1.60535751055097e-06,
+      "loss": 0.1003,
+      "step": 8847
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.1972372387957145,
+      "learning_rate": 1.6043557383120162e-06,
+      "loss": 0.0607,
+      "step": 8848
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.303169237828924,
+      "learning_rate": 1.603354219000427e-06,
+      "loss": 0.0916,
+      "step": 8849
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.26847628526120426,
+      "learning_rate": 1.6023529526908038e-06,
+      "loss": 0.0759,
+      "step": 8850
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.24337545445596412,
+      "learning_rate": 1.6013519394577255e-06,
+      "loss": 0.0705,
+      "step": 8851
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.21101594700922224,
+      "learning_rate": 1.6003511793757531e-06,
+      "loss": 0.0454,
+      "step": 8852
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.38200480660497776,
+      "learning_rate": 1.599350672519428e-06,
+      "loss": 0.0769,
+      "step": 8853
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.24251602895997362,
+      "learning_rate": 1.5983504189632765e-06,
+      "loss": 0.0639,
+      "step": 8854
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3762614575062755,
+      "learning_rate": 1.5973504187818017e-06,
+      "loss": 0.1168,
+      "step": 8855
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.37452503390720765,
+      "learning_rate": 1.5963506720494882e-06,
+      "loss": 0.1018,
+      "step": 8856
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3489453989048087,
+      "learning_rate": 1.5953511788408061e-06,
+      "loss": 0.0945,
+      "step": 8857
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4270716464059508,
+      "learning_rate": 1.5943519392302015e-06,
+      "loss": 0.0894,
+      "step": 8858
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.28228543702750797,
+      "learning_rate": 1.5933529532921037e-06,
+      "loss": 0.081,
+      "step": 8859
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.546144731510037,
+      "learning_rate": 1.5923542211009223e-06,
+      "loss": 0.1082,
+      "step": 8860
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.28662677108352147,
+      "learning_rate": 1.5913557427310505e-06,
+      "loss": 0.0993,
+      "step": 8861
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.41688201531795727,
+      "learning_rate": 1.5903575182568599e-06,
+      "loss": 0.092,
+      "step": 8862
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4051923595311536,
+      "learning_rate": 1.5893595477527024e-06,
+      "loss": 0.099,
+      "step": 8863
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.500516793418049,
+      "learning_rate": 1.5883618312929155e-06,
+      "loss": 0.0774,
+      "step": 8864
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.49813121010719147,
+      "learning_rate": 1.5873643689518143e-06,
+      "loss": 0.0781,
+      "step": 8865
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.40327672688718613,
+      "learning_rate": 1.5863671608036929e-06,
+      "loss": 0.0891,
+      "step": 8866
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.23448886084922121,
+      "learning_rate": 1.5853702069228333e-06,
+      "loss": 0.0806,
+      "step": 8867
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.43689397347979536,
+      "learning_rate": 1.584373507383492e-06,
+      "loss": 0.112,
+      "step": 8868
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2610799586128663,
+      "learning_rate": 1.5833770622599093e-06,
+      "loss": 0.0544,
+      "step": 8869
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.33747247474257175,
+      "learning_rate": 1.582380871626305e-06,
+      "loss": 0.0794,
+      "step": 8870
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2989105804839214,
+      "learning_rate": 1.5813849355568838e-06,
+      "loss": 0.069,
+      "step": 8871
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2656174964752859,
+      "learning_rate": 1.5803892541258275e-06,
+      "loss": 0.084,
+      "step": 8872
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3253857189320341,
+      "learning_rate": 1.579393827407299e-06,
+      "loss": 0.1227,
+      "step": 8873
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7008567579998054,
+      "learning_rate": 1.578398655475446e-06,
+      "loss": 0.1095,
+      "step": 8874
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3727563355250439,
+      "learning_rate": 1.5774037384043938e-06,
+      "loss": 0.0696,
+      "step": 8875
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6964088532215039,
+      "learning_rate": 1.5764090762682487e-06,
+      "loss": 0.1012,
+      "step": 8876
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.248231884366039,
+      "learning_rate": 1.5754146691410983e-06,
+      "loss": 0.0807,
+      "step": 8877
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.33429576489806867,
+      "learning_rate": 1.5744205170970144e-06,
+      "loss": 0.0855,
+      "step": 8878
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.34994428128221283,
+      "learning_rate": 1.5734266202100462e-06,
+      "loss": 0.0833,
+      "step": 8879
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3930967493132058,
+      "learning_rate": 1.5724329785542231e-06,
+      "loss": 0.1044,
+      "step": 8880
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.23409488069020024,
+      "learning_rate": 1.5714395922035597e-06,
+      "loss": 0.0782,
+      "step": 8881
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.33718353637663323,
+      "learning_rate": 1.5704464612320492e-06,
+      "loss": 0.069,
+      "step": 8882
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.39258527549023753,
+      "learning_rate": 1.5694535857136638e-06,
+      "loss": 0.0951,
+      "step": 8883
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3074733950824599,
+      "learning_rate": 1.5684609657223593e-06,
+      "loss": 0.0741,
+      "step": 8884
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6471691362871344,
+      "learning_rate": 1.5674686013320733e-06,
+      "loss": 0.1396,
+      "step": 8885
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.41914622854438677,
+      "learning_rate": 1.566476492616722e-06,
+      "loss": 0.1023,
+      "step": 8886
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.31994196846466866,
+      "learning_rate": 1.5654846396502028e-06,
+      "loss": 0.081,
+      "step": 8887
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.39626803824151124,
+      "learning_rate": 1.5644930425063959e-06,
+      "loss": 0.0826,
+      "step": 8888
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.23452047918847987,
+      "learning_rate": 1.5635017012591585e-06,
+      "loss": 0.0553,
+      "step": 8889
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4192790173107763,
+      "learning_rate": 1.562510615982335e-06,
+      "loss": 0.1181,
+      "step": 8890
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.19699957994707024,
+      "learning_rate": 1.5615197867497455e-06,
+      "loss": 0.0372,
+      "step": 8891
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.25547081768775104,
+      "learning_rate": 1.5605292136351935e-06,
+      "loss": 0.0779,
+      "step": 8892
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.336014373188817,
+      "learning_rate": 1.5595388967124603e-06,
+      "loss": 0.06,
+      "step": 8893
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7111542374022048,
+      "learning_rate": 1.5585488360553137e-06,
+      "loss": 0.1225,
+      "step": 8894
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.315145443973822,
+      "learning_rate": 1.5575590317374979e-06,
+      "loss": 0.0829,
+      "step": 8895
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.30793845527860425,
+      "learning_rate": 1.5565694838327388e-06,
+      "loss": 0.0936,
+      "step": 8896
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.31420607087977354,
+      "learning_rate": 1.5555801924147424e-06,
+      "loss": 0.0867,
+      "step": 8897
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3301332980904552,
+      "learning_rate": 1.5545911575572004e-06,
+      "loss": 0.1192,
+      "step": 8898
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.41096934109322686,
+      "learning_rate": 1.5536023793337796e-06,
+      "loss": 0.0771,
+      "step": 8899
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4980888735141741,
+      "learning_rate": 1.5526138578181304e-06,
+      "loss": 0.1049,
+      "step": 8900
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3430972377184634,
+      "learning_rate": 1.5516255930838824e-06,
+      "loss": 0.086,
+      "step": 8901
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.24458587448416488,
+      "learning_rate": 1.5506375852046496e-06,
+      "loss": 0.0724,
+      "step": 8902
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.24939762100495907,
+      "learning_rate": 1.549649834254024e-06,
+      "loss": 0.0572,
+      "step": 8903
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3798712461011829,
+      "learning_rate": 1.5486623403055767e-06,
+      "loss": 0.073,
+      "step": 8904
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2501315416702115,
+      "learning_rate": 1.5476751034328652e-06,
+      "loss": 0.068,
+      "step": 8905
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.24825002836392268,
+      "learning_rate": 1.5466881237094233e-06,
+      "loss": 0.0749,
+      "step": 8906
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2590045369833802,
+      "learning_rate": 1.5457014012087656e-06,
+      "loss": 0.0584,
+      "step": 8907
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.29426838902260505,
+      "learning_rate": 1.5447149360043916e-06,
+      "loss": 0.0493,
+      "step": 8908
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.43912112515181195,
+      "learning_rate": 1.5437287281697782e-06,
+      "loss": 0.1079,
+      "step": 8909
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4188582599245124,
+      "learning_rate": 1.5427427777783832e-06,
+      "loss": 0.1125,
+      "step": 8910
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.36847501602279475,
+      "learning_rate": 1.5417570849036446e-06,
+      "loss": 0.0842,
+      "step": 8911
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.26760948973492155,
+      "learning_rate": 1.540771649618985e-06,
+      "loss": 0.0692,
+      "step": 8912
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.36870775380330195,
+      "learning_rate": 1.539786471997805e-06,
+      "loss": 0.0992,
+      "step": 8913
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4425000435844594,
+      "learning_rate": 1.5388015521134847e-06,
+      "loss": 0.0845,
+      "step": 8914
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.23521823695791527,
+      "learning_rate": 1.5378168900393887e-06,
+      "loss": 0.0705,
+      "step": 8915
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3792702305667949,
+      "learning_rate": 1.5368324858488592e-06,
+      "loss": 0.0898,
+      "step": 8916
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.18818566710247914,
+      "learning_rate": 1.5358483396152213e-06,
+      "loss": 0.0509,
+      "step": 8917
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.23797397866573497,
+      "learning_rate": 1.5348644514117773e-06,
+      "loss": 0.0684,
+      "step": 8918
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.33105986330207154,
+      "learning_rate": 1.5338808213118162e-06,
+      "loss": 0.078,
+      "step": 8919
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3053456167927793,
+      "learning_rate": 1.5328974493886034e-06,
+      "loss": 0.0818,
+      "step": 8920
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3360753563199587,
+      "learning_rate": 1.5319143357153842e-06,
+      "loss": 0.0979,
+      "step": 8921
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.22995699377649478,
+      "learning_rate": 1.5309314803653902e-06,
+      "loss": 0.0709,
+      "step": 8922
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.302926561231055,
+      "learning_rate": 1.5299488834118286e-06,
+      "loss": 0.0796,
+      "step": 8923
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.32656444825675474,
+      "learning_rate": 1.5289665449278868e-06,
+      "loss": 0.0717,
+      "step": 8924
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3240604299529103,
+      "learning_rate": 1.5279844649867381e-06,
+      "loss": 0.094,
+      "step": 8925
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.31300044832674584,
+      "learning_rate": 1.5270026436615333e-06,
+      "loss": 0.0465,
+      "step": 8926
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.27855691232989227,
+      "learning_rate": 1.5260210810254027e-06,
+      "loss": 0.0674,
+      "step": 8927
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.33436809153151714,
+      "learning_rate": 1.525039777151458e-06,
+      "loss": 0.083,
+      "step": 8928
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.36502621963203413,
+      "learning_rate": 1.524058732112796e-06,
+      "loss": 0.1081,
+      "step": 8929
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.25531291528520716,
+      "learning_rate": 1.5230779459824873e-06,
+      "loss": 0.0639,
+      "step": 8930
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.30426986069774775,
+      "learning_rate": 1.5220974188335869e-06,
+      "loss": 0.0613,
+      "step": 8931
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.37713613168457427,
+      "learning_rate": 1.5211171507391325e-06,
+      "loss": 0.0714,
+      "step": 8932
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3177241694915858,
+      "learning_rate": 1.520137141772139e-06,
+      "loss": 0.0925,
+      "step": 8933
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.30137627234537917,
+      "learning_rate": 1.5191573920056025e-06,
+      "loss": 0.0544,
+      "step": 8934
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3253107905617309,
+      "learning_rate": 1.5181779015124993e-06,
+      "loss": 0.1209,
+      "step": 8935
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2399610077682187,
+      "learning_rate": 1.5171986703657909e-06,
+      "loss": 0.076,
+      "step": 8936
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.29479676246693537,
+      "learning_rate": 1.5162196986384143e-06,
+      "loss": 0.0469,
+      "step": 8937
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4098417012827499,
+      "learning_rate": 1.5152409864032874e-06,
+      "loss": 0.1013,
+      "step": 8938
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2281331913886879,
+      "learning_rate": 1.5142625337333139e-06,
+      "loss": 0.0481,
+      "step": 8939
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.37941146186776964,
+      "learning_rate": 1.5132843407013726e-06,
+      "loss": 0.0812,
+      "step": 8940
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2364200786859717,
+      "learning_rate": 1.5123064073803235e-06,
+      "loss": 0.0661,
+      "step": 8941
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.22248878846635126,
+      "learning_rate": 1.511328733843012e-06,
+      "loss": 0.0555,
+      "step": 8942
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.35169934283765586,
+      "learning_rate": 1.5103513201622593e-06,
+      "loss": 0.0836,
+      "step": 8943
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.35351434787551783,
+      "learning_rate": 1.5093741664108696e-06,
+      "loss": 0.0812,
+      "step": 8944
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2599288693323992,
+      "learning_rate": 1.508397272661624e-06,
+      "loss": 0.0705,
+      "step": 8945
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4714786781672482,
+      "learning_rate": 1.507420638987292e-06,
+      "loss": 0.0788,
+      "step": 8946
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.20741640414773405,
+      "learning_rate": 1.5064442654606166e-06,
+      "loss": 0.0671,
+      "step": 8947
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3253731635206642,
+      "learning_rate": 1.505468152154322e-06,
+      "loss": 0.1075,
+      "step": 8948
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.423574081166215,
+      "learning_rate": 1.5044922991411177e-06,
+      "loss": 0.1274,
+      "step": 8949
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.32180403940704294,
+      "learning_rate": 1.50351670649369e-06,
+      "loss": 0.0839,
+      "step": 8950
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3011156844593209,
+      "learning_rate": 1.5025413742847068e-06,
+      "loss": 0.0726,
+      "step": 8951
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.20731735088658268,
+      "learning_rate": 1.5015663025868144e-06,
+      "loss": 0.07,
+      "step": 8952
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2707908493862437,
+      "learning_rate": 1.5005914914726455e-06,
+      "loss": 0.0444,
+      "step": 8953
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.37517716786688143,
+      "learning_rate": 1.4996169410148076e-06,
+      "loss": 0.0811,
+      "step": 8954
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4888746616625594,
+      "learning_rate": 1.4986426512858916e-06,
+      "loss": 0.0728,
+      "step": 8955
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3775610464651205,
+      "learning_rate": 1.4976686223584675e-06,
+      "loss": 0.0593,
+      "step": 8956
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.35699133759534624,
+      "learning_rate": 1.4966948543050851e-06,
+      "loss": 0.0868,
+      "step": 8957
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3382146069450258,
+      "learning_rate": 1.49572134719828e-06,
+      "loss": 0.0969,
+      "step": 8958
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.37089649232226474,
+      "learning_rate": 1.4947481011105636e-06,
+      "loss": 0.0949,
+      "step": 8959
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.29673773801225556,
+      "learning_rate": 1.4937751161144275e-06,
+      "loss": 0.0717,
+      "step": 8960
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.22574776702747865,
+      "learning_rate": 1.4928023922823442e-06,
+      "loss": 0.0627,
+      "step": 8961
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2897026898360463,
+      "learning_rate": 1.4918299296867723e-06,
+      "loss": 0.0842,
+      "step": 8962
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.29366634619067306,
+      "learning_rate": 1.490857728400143e-06,
+      "loss": 0.0813,
+      "step": 8963
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3087150040995132,
+      "learning_rate": 1.4898857884948725e-06,
+      "loss": 0.0841,
+      "step": 8964
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.4590410761728625,
+      "learning_rate": 1.4889141100433545e-06,
+      "loss": 0.0977,
+      "step": 8965
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.21076252658550823,
+      "learning_rate": 1.4879426931179696e-06,
+      "loss": 0.0431,
+      "step": 8966
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3564162063175528,
+      "learning_rate": 1.4869715377910715e-06,
+      "loss": 0.0781,
+      "step": 8967
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.1922290865265554,
+      "learning_rate": 1.4860006441349983e-06,
+      "loss": 0.0213,
+      "step": 8968
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2736661817969576,
+      "learning_rate": 1.4850300122220662e-06,
+      "loss": 0.0737,
+      "step": 8969
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.312284192143055,
+      "learning_rate": 1.4840596421245768e-06,
+      "loss": 0.0782,
+      "step": 8970
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6388991512648625,
+      "learning_rate": 1.4830895339148067e-06,
+      "loss": 0.1367,
+      "step": 8971
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3002977774249973,
+      "learning_rate": 1.482119687665014e-06,
+      "loss": 0.0746,
+      "step": 8972
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2312226188086535,
+      "learning_rate": 1.4811501034474418e-06,
+      "loss": 0.0841,
+      "step": 8973
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.276817738945182,
+      "learning_rate": 1.480180781334309e-06,
+      "loss": 0.0958,
+      "step": 8974
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.24069314005872958,
+      "learning_rate": 1.4792117213978148e-06,
+      "loss": 0.0671,
+      "step": 8975
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.4794788338636877,
+      "learning_rate": 1.4782429237101425e-06,
+      "loss": 0.0642,
+      "step": 8976
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.23906200921909176,
+      "learning_rate": 1.477274388343453e-06,
+      "loss": 0.0707,
+      "step": 8977
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.369460659514181,
+      "learning_rate": 1.4763061153698887e-06,
+      "loss": 0.1115,
+      "step": 8978
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.24985494002295278,
+      "learning_rate": 1.4753381048615706e-06,
+      "loss": 0.057,
+      "step": 8979
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.267243101255548,
+      "learning_rate": 1.4743703568906048e-06,
+      "loss": 0.0697,
+      "step": 8980
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7098557760666024,
+      "learning_rate": 1.4734028715290728e-06,
+      "loss": 0.1321,
+      "step": 8981
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.22441670507516367,
+      "learning_rate": 1.472435648849037e-06,
+      "loss": 0.046,
+      "step": 8982
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.4174382399838925,
+      "learning_rate": 1.471468688922546e-06,
+      "loss": 0.1098,
+      "step": 8983
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.19723608682021473,
+      "learning_rate": 1.4705019918216218e-06,
+      "loss": 0.0593,
+      "step": 8984
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.25685606449167836,
+      "learning_rate": 1.46953555761827e-06,
+      "loss": 0.0816,
+      "step": 8985
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.4907315977765806,
+      "learning_rate": 1.4685693863844752e-06,
+      "loss": 0.1247,
+      "step": 8986
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.4285027523258741,
+      "learning_rate": 1.4676034781922056e-06,
+      "loss": 0.0966,
+      "step": 8987
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.20539011021613662,
+      "learning_rate": 1.4666378331134073e-06,
+      "loss": 0.0516,
+      "step": 8988
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2625599215059745,
+      "learning_rate": 1.4656724512200044e-06,
+      "loss": 0.0754,
+      "step": 8989
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.24539203258290104,
+      "learning_rate": 1.4647073325839084e-06,
+      "loss": 0.05,
+      "step": 8990
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.28239754359414915,
+      "learning_rate": 1.4637424772770042e-06,
+      "loss": 0.0867,
+      "step": 8991
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3037797468278806,
+      "learning_rate": 1.4627778853711594e-06,
+      "loss": 0.053,
+      "step": 8992
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.40775203700654555,
+      "learning_rate": 1.4618135569382253e-06,
+      "loss": 0.0754,
+      "step": 8993
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.27011957574466455,
+      "learning_rate": 1.4608494920500287e-06,
+      "loss": 0.0676,
+      "step": 8994
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5092248284917518,
+      "learning_rate": 1.4598856907783792e-06,
+      "loss": 0.0935,
+      "step": 8995
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3305601578414505,
+      "learning_rate": 1.4589221531950642e-06,
+      "loss": 0.0914,
+      "step": 8996
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.22118126080798028,
+      "learning_rate": 1.4579588793718575e-06,
+      "loss": 0.0637,
+      "step": 8997
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.32247275133620557,
+      "learning_rate": 1.456995869380507e-06,
+      "loss": 0.074,
+      "step": 8998
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5017056623997849,
+      "learning_rate": 1.456033123292742e-06,
+      "loss": 0.1036,
+      "step": 8999
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.33382550794309285,
+      "learning_rate": 1.4550706411802774e-06,
+      "loss": 0.0924,
+      "step": 9000
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6789040251250273,
+      "learning_rate": 1.4541084231148017e-06,
+      "loss": 0.1309,
+      "step": 9001
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2665076247915018,
+      "learning_rate": 1.453146469167987e-06,
+      "loss": 0.0922,
+      "step": 9002
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.23213898008908482,
+      "learning_rate": 1.4521847794114835e-06,
+      "loss": 0.0576,
+      "step": 9003
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2248932638434717,
+      "learning_rate": 1.4512233539169268e-06,
+      "loss": 0.0438,
+      "step": 9004
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2835690276813325,
+      "learning_rate": 1.450262192755928e-06,
+      "loss": 0.0771,
+      "step": 9005
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.29322717058529885,
+      "learning_rate": 1.4493012960000784e-06,
+      "loss": 0.0441,
+      "step": 9006
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.4052261028393808,
+      "learning_rate": 1.448340663720954e-06,
+      "loss": 0.1048,
+      "step": 9007
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2770794207009056,
+      "learning_rate": 1.4473802959901069e-06,
+      "loss": 0.1124,
+      "step": 9008
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.4429745291276455,
+      "learning_rate": 1.4464201928790694e-06,
+      "loss": 0.0862,
+      "step": 9009
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.22072899415649624,
+      "learning_rate": 1.4454603544593588e-06,
+      "loss": 0.0634,
+      "step": 9010
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2267164504348625,
+      "learning_rate": 1.4445007808024675e-06,
+      "loss": 0.0928,
+      "step": 9011
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2110049235768872,
+      "learning_rate": 1.4435414719798702e-06,
+      "loss": 0.0523,
+      "step": 9012
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.27275130621687793,
+      "learning_rate": 1.4425824280630207e-06,
+      "loss": 0.0884,
+      "step": 9013
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.30593084470245585,
+      "learning_rate": 1.441623649123357e-06,
+      "loss": 0.0692,
+      "step": 9014
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.27342559207893546,
+      "learning_rate": 1.4406651352322932e-06,
+      "loss": 0.1003,
+      "step": 9015
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.18338588799638847,
+      "learning_rate": 1.4397068864612223e-06,
+      "loss": 0.0336,
+      "step": 9016
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5076698254301681,
+      "learning_rate": 1.4387489028815254e-06,
+      "loss": 0.1263,
+      "step": 9017
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3031893217642408,
+      "learning_rate": 1.4377911845645553e-06,
+      "loss": 0.0581,
+      "step": 9018
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.44022376570033506,
+      "learning_rate": 1.4368337315816495e-06,
+      "loss": 0.1223,
+      "step": 9019
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.41652171225266876,
+      "learning_rate": 1.4358765440041229e-06,
+      "loss": 0.1048,
+      "step": 9020
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2664166462358083,
+      "learning_rate": 1.4349196219032752e-06,
+      "loss": 0.0628,
+      "step": 9021
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.375706908063255,
+      "learning_rate": 1.4339629653503823e-06,
+      "loss": 0.1013,
+      "step": 9022
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3649409874473625,
+      "learning_rate": 1.433006574416701e-06,
+      "loss": 0.0666,
+      "step": 9023
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.303482617666949,
+      "learning_rate": 1.43205044917347e-06,
+      "loss": 0.0727,
+      "step": 9024
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3931415657806582,
+      "learning_rate": 1.4310945896919043e-06,
+      "loss": 0.0993,
+      "step": 9025
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.37629458331307347,
+      "learning_rate": 1.430138996043206e-06,
+      "loss": 0.1054,
+      "step": 9026
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.27913000940133625,
+      "learning_rate": 1.429183668298551e-06,
+      "loss": 0.1055,
+      "step": 9027
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3211784614554781,
+      "learning_rate": 1.4282286065290978e-06,
+      "loss": 0.1261,
+      "step": 9028
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.49791522267033933,
+      "learning_rate": 1.4272738108059836e-06,
+      "loss": 0.0914,
+      "step": 9029
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.35242131679930966,
+      "learning_rate": 1.4263192812003302e-06,
+      "loss": 0.0816,
+      "step": 9030
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2960467544019307,
+      "learning_rate": 1.425365017783235e-06,
+      "loss": 0.0619,
+      "step": 9031
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.4752485813627897,
+      "learning_rate": 1.4244110206257772e-06,
+      "loss": 0.0896,
+      "step": 9032
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.23606901469735783,
+      "learning_rate": 1.4234572897990146e-06,
+      "loss": 0.0649,
+      "step": 9033
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.44986533527721007,
+      "learning_rate": 1.4225038253739887e-06,
+      "loss": 0.1064,
+      "step": 9034
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.33491455372354395,
+      "learning_rate": 1.4215506274217189e-06,
+      "loss": 0.1008,
+      "step": 9035
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3167114358511054,
+      "learning_rate": 1.4205976960132045e-06,
+      "loss": 0.0877,
+      "step": 9036
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.28092484899191167,
+      "learning_rate": 1.4196450312194232e-06,
+      "loss": 0.0765,
+      "step": 9037
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.4751160830065029,
+      "learning_rate": 1.4186926331113392e-06,
+      "loss": 0.1232,
+      "step": 9038
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2703189887778611,
+      "learning_rate": 1.41774050175989e-06,
+      "loss": 0.069,
+      "step": 9039
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.506201789950593,
+      "learning_rate": 1.4167886372359958e-06,
+      "loss": 0.1004,
+      "step": 9040
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5566671157665424,
+      "learning_rate": 1.4158370396105581e-06,
+      "loss": 0.0776,
+      "step": 9041
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.22019667106651766,
+      "learning_rate": 1.4148857089544577e-06,
+      "loss": 0.0658,
+      "step": 9042
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.36813301464474457,
+      "learning_rate": 1.413934645338554e-06,
+      "loss": 0.0966,
+      "step": 9043
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.27196913329447064,
+      "learning_rate": 1.4129838488336872e-06,
+      "loss": 0.08,
+      "step": 9044
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3308766565428638,
+      "learning_rate": 1.4120333195106807e-06,
+      "loss": 0.0701,
+      "step": 9045
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.34774797564339244,
+      "learning_rate": 1.4110830574403339e-06,
+      "loss": 0.1055,
+      "step": 9046
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3179537533855218,
+      "learning_rate": 1.4101330626934262e-06,
+      "loss": 0.0775,
+      "step": 9047
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.26022121076761734,
+      "learning_rate": 1.4091833353407224e-06,
+      "loss": 0.0829,
+      "step": 9048
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2856584692743634,
+      "learning_rate": 1.4082338754529618e-06,
+      "loss": 0.0825,
+      "step": 9049
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3280682834203328,
+      "learning_rate": 1.4072846831008641e-06,
+      "loss": 0.0642,
+      "step": 9050
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.242506963688167,
+      "learning_rate": 1.4063357583551341e-06,
+      "loss": 0.0506,
+      "step": 9051
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.32885076680305075,
+      "learning_rate": 1.4053871012864511e-06,
+      "loss": 0.091,
+      "step": 9052
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.34717192053713514,
+      "learning_rate": 1.4044387119654774e-06,
+      "loss": 0.1118,
+      "step": 9053
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3344377698357655,
+      "learning_rate": 1.403490590462852e-06,
+      "loss": 0.111,
+      "step": 9054
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.28658875547217344,
+      "learning_rate": 1.4025427368492006e-06,
+      "loss": 0.0758,
+      "step": 9055
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.25315245928228775,
+      "learning_rate": 1.4015951511951227e-06,
+      "loss": 0.0668,
+      "step": 9056
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2055357957307144,
+      "learning_rate": 1.4006478335711987e-06,
+      "loss": 0.0669,
+      "step": 9057
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.32148181635280576,
+      "learning_rate": 1.3997007840479937e-06,
+      "loss": 0.0895,
+      "step": 9058
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.40570641281715086,
+      "learning_rate": 1.3987540026960473e-06,
+      "loss": 0.0946,
+      "step": 9059
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.18528781036277567,
+      "learning_rate": 1.3978074895858818e-06,
+      "loss": 0.0563,
+      "step": 9060
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2979438698244309,
+      "learning_rate": 1.396861244787997e-06,
+      "loss": 0.0877,
+      "step": 9061
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3233573299270331,
+      "learning_rate": 1.3959152683728776e-06,
+      "loss": 0.1171,
+      "step": 9062
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.24922987460857152,
+      "learning_rate": 1.3949695604109854e-06,
+      "loss": 0.0646,
+      "step": 9063
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2914190294519474,
+      "learning_rate": 1.394024120972759e-06,
+      "loss": 0.0786,
+      "step": 9064
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2988407277398744,
+      "learning_rate": 1.393078950128624e-06,
+      "loss": 0.1035,
+      "step": 9065
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.1876797214361556,
+      "learning_rate": 1.392134047948981e-06,
+      "loss": 0.0581,
+      "step": 9066
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.34075821483212215,
+      "learning_rate": 1.3911894145042093e-06,
+      "loss": 0.0977,
+      "step": 9067
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.24360904545632936,
+      "learning_rate": 1.3902450498646753e-06,
+      "loss": 0.0586,
+      "step": 9068
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.35727690519778,
+      "learning_rate": 1.389300954100718e-06,
+      "loss": 0.0693,
+      "step": 9069
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.23223491557066642,
+      "learning_rate": 1.388357127282659e-06,
+      "loss": 0.0559,
+      "step": 9070
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.31778657430077106,
+      "learning_rate": 1.3874135694807994e-06,
+      "loss": 0.0852,
+      "step": 9071
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2911285208593612,
+      "learning_rate": 1.386470280765424e-06,
+      "loss": 0.0588,
+      "step": 9072
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.4002003470793741,
+      "learning_rate": 1.3855272612067917e-06,
+      "loss": 0.1063,
+      "step": 9073
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.4373750741467451,
+      "learning_rate": 1.3845845108751438e-06,
+      "loss": 0.1128,
+      "step": 9074
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3090594402600677,
+      "learning_rate": 1.3836420298407044e-06,
+      "loss": 0.0491,
+      "step": 9075
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3259196951884869,
+      "learning_rate": 1.3826998181736728e-06,
+      "loss": 0.0587,
+      "step": 9076
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5512487342771755,
+      "learning_rate": 1.381757875944232e-06,
+      "loss": 0.122,
+      "step": 9077
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2575720598354923,
+      "learning_rate": 1.3808162032225403e-06,
+      "loss": 0.0876,
+      "step": 9078
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.21443003769851024,
+      "learning_rate": 1.3798748000787427e-06,
+      "loss": 0.0794,
+      "step": 9079
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2561444915863631,
+      "learning_rate": 1.378933666582959e-06,
+      "loss": 0.058,
+      "step": 9080
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.28728586866455486,
+      "learning_rate": 1.3779928028052885e-06,
+      "loss": 0.0733,
+      "step": 9081
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3745453796970236,
+      "learning_rate": 1.377052208815815e-06,
+      "loss": 0.1394,
+      "step": 9082
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6595915030224648,
+      "learning_rate": 1.3761118846845989e-06,
+      "loss": 0.1048,
+      "step": 9083
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3234201462504862,
+      "learning_rate": 1.375171830481678e-06,
+      "loss": 0.0848,
+      "step": 9084
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.1976262481838455,
+      "learning_rate": 1.3742320462770775e-06,
+      "loss": 0.0634,
+      "step": 9085
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.208103518678512,
+      "learning_rate": 1.3732925321407958e-06,
+      "loss": 0.0363,
+      "step": 9086
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2975968535939927,
+      "learning_rate": 1.372353288142813e-06,
+      "loss": 0.0994,
+      "step": 9087
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5200256443960833,
+      "learning_rate": 1.371414314353089e-06,
+      "loss": 0.1527,
+      "step": 9088
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.21033096410764265,
+      "learning_rate": 1.3704756108415657e-06,
+      "loss": 0.0391,
+      "step": 9089
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2809555751680728,
+      "learning_rate": 1.369537177678163e-06,
+      "loss": 0.0802,
+      "step": 9090
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8239302190455301,
+      "learning_rate": 1.36859901493278e-06,
+      "loss": 0.1815,
+      "step": 9091
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.31557592179690763,
+      "learning_rate": 1.3676611226752973e-06,
+      "loss": 0.0675,
+      "step": 9092
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2549647237244421,
+      "learning_rate": 1.366723500975572e-06,
+      "loss": 0.0728,
+      "step": 9093
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3716165060551907,
+      "learning_rate": 1.3657861499034474e-06,
+      "loss": 0.0965,
+      "step": 9094
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3536941715085643,
+      "learning_rate": 1.364849069528742e-06,
+      "loss": 0.0634,
+      "step": 9095
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.40249659657989595,
+      "learning_rate": 1.3639122599212534e-06,
+      "loss": 0.1064,
+      "step": 9096
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.26664910689930643,
+      "learning_rate": 1.3629757211507598e-06,
+      "loss": 0.0606,
+      "step": 9097
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.28032972595210925,
+      "learning_rate": 1.362039453287024e-06,
+      "loss": 0.076,
+      "step": 9098
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.24761656936911136,
+      "learning_rate": 1.3611034563997816e-06,
+      "loss": 0.0857,
+      "step": 9099
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.28718012479688754,
+      "learning_rate": 1.3601677305587524e-06,
+      "loss": 0.0794,
+      "step": 9100
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.48448441228178585,
+      "learning_rate": 1.359232275833633e-06,
+      "loss": 0.0779,
+      "step": 9101
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.1887399522682956,
+      "learning_rate": 1.3582970922941036e-06,
+      "loss": 0.0413,
+      "step": 9102
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.26410487586101533,
+      "learning_rate": 1.3573621800098213e-06,
+      "loss": 0.0644,
+      "step": 9103
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.30879792360296227,
+      "learning_rate": 1.3564275390504244e-06,
+      "loss": 0.1012,
+      "step": 9104
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3307359260223385,
+      "learning_rate": 1.3554931694855278e-06,
+      "loss": 0.054,
+      "step": 9105
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.24141346817593878,
+      "learning_rate": 1.3545590713847323e-06,
+      "loss": 0.06,
+      "step": 9106
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.22715120002187913,
+      "learning_rate": 1.3536252448176135e-06,
+      "loss": 0.07,
+      "step": 9107
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2146055360311639,
+      "learning_rate": 1.3526916898537262e-06,
+      "loss": 0.0561,
+      "step": 9108
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.18162295880046728,
+      "learning_rate": 1.351758406562611e-06,
+      "loss": 0.0491,
+      "step": 9109
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2947989667758957,
+      "learning_rate": 1.3508253950137822e-06,
+      "loss": 0.0871,
+      "step": 9110
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.28819258608365494,
+      "learning_rate": 1.3498926552767362e-06,
+      "loss": 0.0311,
+      "step": 9111
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.1701447588430931,
+      "learning_rate": 1.3489601874209468e-06,
+      "loss": 0.0332,
+      "step": 9112
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.18461017934954646,
+      "learning_rate": 1.3480279915158728e-06,
+      "loss": 0.0512,
+      "step": 9113
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2948992444065201,
+      "learning_rate": 1.3470960676309491e-06,
+      "loss": 0.07,
+      "step": 9114
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.23553140670887057,
+      "learning_rate": 1.346164415835588e-06,
+      "loss": 0.0367,
+      "step": 9115
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2808188165900945,
+      "learning_rate": 1.3452330361991877e-06,
+      "loss": 0.0512,
+      "step": 9116
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3347905601974998,
+      "learning_rate": 1.3443019287911218e-06,
+      "loss": 0.0891,
+      "step": 9117
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.4357652473059663,
+      "learning_rate": 1.3433710936807426e-06,
+      "loss": 0.1338,
+      "step": 9118
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.24090227958287116,
+      "learning_rate": 1.3424405309373873e-06,
+      "loss": 0.0771,
+      "step": 9119
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3876223704845664,
+      "learning_rate": 1.3415102406303681e-06,
+      "loss": 0.1145,
+      "step": 9120
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.31469508362196214,
+      "learning_rate": 1.3405802228289788e-06,
+      "loss": 0.0673,
+      "step": 9121
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.27908118014739175,
+      "learning_rate": 1.33965047760249e-06,
+      "loss": 0.0745,
+      "step": 9122
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2367200269629889,
+      "learning_rate": 1.3387210050201588e-06,
+      "loss": 0.0529,
+      "step": 9123
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3037619408955906,
+      "learning_rate": 1.3377918051512157e-06,
+      "loss": 0.0689,
+      "step": 9124
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2136408276494688,
+      "learning_rate": 1.3368628780648712e-06,
+      "loss": 0.0557,
+      "step": 9125
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.39559527285243673,
+      "learning_rate": 1.3359342238303203e-06,
+      "loss": 0.1235,
+      "step": 9126
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2982668715143579,
+      "learning_rate": 1.3350058425167334e-06,
+      "loss": 0.0854,
+      "step": 9127
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3249028934435638,
+      "learning_rate": 1.3340777341932616e-06,
+      "loss": 0.0907,
+      "step": 9128
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.39201012760442144,
+      "learning_rate": 1.3331498989290337e-06,
+      "loss": 0.0988,
+      "step": 9129
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.33211321089772666,
+      "learning_rate": 1.3322223367931643e-06,
+      "loss": 0.0883,
+      "step": 9130
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.31617862146563375,
+      "learning_rate": 1.3312950478547416e-06,
+      "loss": 0.0764,
+      "step": 9131
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.36963105389555884,
+      "learning_rate": 1.3303680321828332e-06,
+      "loss": 0.0905,
+      "step": 9132
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.29613206935387587,
+      "learning_rate": 1.3294412898464937e-06,
+      "loss": 0.0451,
+      "step": 9133
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.313363258168235,
+      "learning_rate": 1.3285148209147486e-06,
+      "loss": 0.1025,
+      "step": 9134
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.18470489519287475,
+      "learning_rate": 1.3275886254566068e-06,
+      "loss": 0.0358,
+      "step": 9135
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.45664323715895416,
+      "learning_rate": 1.3266627035410584e-06,
+      "loss": 0.0869,
+      "step": 9136
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.21957774998844323,
+      "learning_rate": 1.3257370552370713e-06,
+      "loss": 0.051,
+      "step": 9137
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.21910337263325588,
+      "learning_rate": 1.324811680613592e-06,
+      "loss": 0.0512,
+      "step": 9138
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.42484304280494495,
+      "learning_rate": 1.3238865797395466e-06,
+      "loss": 0.1014,
+      "step": 9139
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.315009704950275,
+      "learning_rate": 1.3229617526838457e-06,
+      "loss": 0.0835,
+      "step": 9140
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3325039849436051,
+      "learning_rate": 1.3220371995153735e-06,
+      "loss": 0.079,
+      "step": 9141
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3588680149133279,
+      "learning_rate": 1.3211129203029953e-06,
+      "loss": 0.0825,
+      "step": 9142
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.4209612026699083,
+      "learning_rate": 1.320188915115559e-06,
+      "loss": 0.1119,
+      "step": 9143
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.23459897043118016,
+      "learning_rate": 1.319265184021889e-06,
+      "loss": 0.0746,
+      "step": 9144
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.4809473992240087,
+      "learning_rate": 1.3183417270907894e-06,
+      "loss": 0.129,
+      "step": 9145
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.30819298183566163,
+      "learning_rate": 1.3174185443910442e-06,
+      "loss": 0.1015,
+      "step": 9146
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.4470678667944507,
+      "learning_rate": 1.3164956359914198e-06,
+      "loss": 0.102,
+      "step": 9147
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.24118226356446548,
+      "learning_rate": 1.3155730019606588e-06,
+      "loss": 0.0757,
+      "step": 9148
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.35950195381662287,
+      "learning_rate": 1.314650642367482e-06,
+      "loss": 0.0841,
+      "step": 9149
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.49782855569517626,
+      "learning_rate": 1.3137285572805957e-06,
+      "loss": 0.1508,
+      "step": 9150
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.4145138392072528,
+      "learning_rate": 1.3128067467686812e-06,
+      "loss": 0.0867,
+      "step": 9151
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.16381227587322522,
+      "learning_rate": 1.3118852109003988e-06,
+      "loss": 0.0573,
+      "step": 9152
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3255996905238742,
+      "learning_rate": 1.31096394974439e-06,
+      "loss": 0.0913,
+      "step": 9153
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.34773943274912794,
+      "learning_rate": 1.3100429633692774e-06,
+      "loss": 0.0996,
+      "step": 9154
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.4416283287016242,
+      "learning_rate": 1.3091222518436607e-06,
+      "loss": 0.0966,
+      "step": 9155
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5321065695552621,
+      "learning_rate": 1.3082018152361182e-06,
+      "loss": 0.1166,
+      "step": 9156
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.20566322082248967,
+      "learning_rate": 1.307281653615212e-06,
+      "loss": 0.0526,
+      "step": 9157
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2884705810384233,
+      "learning_rate": 1.3063617670494799e-06,
+      "loss": 0.0688,
+      "step": 9158
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2814858366153959,
+      "learning_rate": 1.305442155607441e-06,
+      "loss": 0.0588,
+      "step": 9159
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.21303761314197944,
+      "learning_rate": 1.3045228193575922e-06,
+      "loss": 0.0538,
+      "step": 9160
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.36587992532023206,
+      "learning_rate": 1.3036037583684102e-06,
+      "loss": 0.0966,
+      "step": 9161
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3936324947936362,
+      "learning_rate": 1.3026849727083552e-06,
+      "loss": 0.0784,
+      "step": 9162
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.33937712973154177,
+      "learning_rate": 1.3017664624458615e-06,
+      "loss": 0.0716,
+      "step": 9163
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3208734756274665,
+      "learning_rate": 1.3008482276493457e-06,
+      "loss": 0.0861,
+      "step": 9164
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.31395310269000604,
+      "learning_rate": 1.2999302683872011e-06,
+      "loss": 0.0975,
+      "step": 9165
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.41125509611979827,
+      "learning_rate": 1.2990125847278068e-06,
+      "loss": 0.1026,
+      "step": 9166
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.34437017223147487,
+      "learning_rate": 1.2980951767395144e-06,
+      "loss": 0.0754,
+      "step": 9167
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.25752273104505297,
+      "learning_rate": 1.2971780444906585e-06,
+      "loss": 0.0587,
+      "step": 9168
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.35324211165188757,
+      "learning_rate": 1.296261188049553e-06,
+      "loss": 0.0838,
+      "step": 9169
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.35892680823859235,
+      "learning_rate": 1.2953446074844879e-06,
+      "loss": 0.0987,
+      "step": 9170
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.22923565241839478,
+      "learning_rate": 1.2944283028637394e-06,
+      "loss": 0.0646,
+      "step": 9171
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3184439509659246,
+      "learning_rate": 1.2935122742555577e-06,
+      "loss": 0.0971,
+      "step": 9172
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.34755485185078694,
+      "learning_rate": 1.2925965217281717e-06,
+      "loss": 0.0544,
+      "step": 9173
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.268085972182783,
+      "learning_rate": 1.2916810453497958e-06,
+      "loss": 0.0731,
+      "step": 9174
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2683416658219828,
+      "learning_rate": 1.2907658451886184e-06,
+      "loss": 0.0638,
+      "step": 9175
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2530225295580043,
+      "learning_rate": 1.2898509213128068e-06,
+      "loss": 0.0648,
+      "step": 9176
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.21962678440856956,
+      "learning_rate": 1.2889362737905136e-06,
+      "loss": 0.0821,
+      "step": 9177
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.27066394516084225,
+      "learning_rate": 1.288021902689865e-06,
+      "loss": 0.0637,
+      "step": 9178
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3184093914824342,
+      "learning_rate": 1.287107808078969e-06,
+      "loss": 0.0648,
+      "step": 9179
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.37685532950081446,
+      "learning_rate": 1.2861939900259112e-06,
+      "loss": 0.0861,
+      "step": 9180
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2838812393048947,
+      "learning_rate": 1.2852804485987613e-06,
+      "loss": 0.0412,
+      "step": 9181
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.24668458375793811,
+      "learning_rate": 1.2843671838655635e-06,
+      "loss": 0.0644,
+      "step": 9182
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.223491805500163,
+      "learning_rate": 1.2834541958943408e-06,
+      "loss": 0.046,
+      "step": 9183
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.4674453226110058,
+      "learning_rate": 1.2825414847531026e-06,
+      "loss": 0.1145,
+      "step": 9184
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3542464639609329,
+      "learning_rate": 1.2816290505098294e-06,
+      "loss": 0.0781,
+      "step": 9185
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.4728523806385116,
+      "learning_rate": 1.280716893232486e-06,
+      "loss": 0.094,
+      "step": 9186
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3154872018629003,
+      "learning_rate": 1.2798050129890139e-06,
+      "loss": 0.0666,
+      "step": 9187
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3053727599756188,
+      "learning_rate": 1.2788934098473365e-06,
+      "loss": 0.0938,
+      "step": 9188
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.41722895534651366,
+      "learning_rate": 1.277982083875356e-06,
+      "loss": 0.0877,
+      "step": 9189
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3560267057030059,
+      "learning_rate": 1.2770710351409504e-06,
+      "loss": 0.0835,
+      "step": 9190
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5332085312168945,
+      "learning_rate": 1.276160263711983e-06,
+      "loss": 0.1016,
+      "step": 9191
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6967215794854511,
+      "learning_rate": 1.2752497696562927e-06,
+      "loss": 0.144,
+      "step": 9192
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5367054611624333,
+      "learning_rate": 1.2743395530416959e-06,
+      "loss": 0.1107,
+      "step": 9193
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3672236201648932,
+      "learning_rate": 1.2734296139359942e-06,
+      "loss": 0.0808,
+      "step": 9194
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.4027181556175779,
+      "learning_rate": 1.2725199524069642e-06,
+      "loss": 0.0809,
+      "step": 9195
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3246616688708011,
+      "learning_rate": 1.2716105685223617e-06,
+      "loss": 0.0715,
+      "step": 9196
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3479525139436896,
+      "learning_rate": 1.2707014623499226e-06,
+      "loss": 0.0861,
+      "step": 9197
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.4270912548981726,
+      "learning_rate": 1.2697926339573647e-06,
+      "loss": 0.099,
+      "step": 9198
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.32769536113991343,
+      "learning_rate": 1.2688840834123812e-06,
+      "loss": 0.0908,
+      "step": 9199
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3344201919727719,
+      "learning_rate": 1.267975810782645e-06,
+      "loss": 0.0936,
+      "step": 9200
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.4994865853553957,
+      "learning_rate": 1.2670678161358129e-06,
+      "loss": 0.115,
+      "step": 9201
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5327470747544643,
+      "learning_rate": 1.2661600995395158e-06,
+      "loss": 0.1174,
+      "step": 9202
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.283788195718664,
+      "learning_rate": 1.2652526610613657e-06,
+      "loss": 0.0755,
+      "step": 9203
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.34880036782477436,
+      "learning_rate": 1.2643455007689526e-06,
+      "loss": 0.0793,
+      "step": 9204
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3276318482962535,
+      "learning_rate": 1.2634386187298493e-06,
+      "loss": 0.0734,
+      "step": 9205
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2283586728070119,
+      "learning_rate": 1.262532015011605e-06,
+      "loss": 0.0301,
+      "step": 9206
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3618973209143697,
+      "learning_rate": 1.2616256896817474e-06,
+      "loss": 0.08,
+      "step": 9207
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.4024403584861021,
+      "learning_rate": 1.2607196428077873e-06,
+      "loss": 0.1147,
+      "step": 9208
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.4073580010558178,
+      "learning_rate": 1.2598138744572107e-06,
+      "loss": 0.1262,
+      "step": 9209
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3276037349107275,
+      "learning_rate": 1.2589083846974843e-06,
+      "loss": 0.0776,
+      "step": 9210
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3585427443266532,
+      "learning_rate": 1.2580031735960556e-06,
+      "loss": 0.0879,
+      "step": 9211
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3405047620428224,
+      "learning_rate": 1.2570982412203498e-06,
+      "loss": 0.0964,
+      "step": 9212
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2501209030875898,
+      "learning_rate": 1.2561935876377706e-06,
+      "loss": 0.0512,
+      "step": 9213
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.25599665798845916,
+      "learning_rate": 1.2552892129157012e-06,
+      "loss": 0.0581,
+      "step": 9214
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3103634267271324,
+      "learning_rate": 1.2543851171215072e-06,
+      "loss": 0.0854,
+      "step": 9215
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2320527994653225,
+      "learning_rate": 1.2534813003225293e-06,
+      "loss": 0.0664,
+      "step": 9216
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3151179113093802,
+      "learning_rate": 1.252577762586088e-06,
+      "loss": 0.0988,
+      "step": 9217
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.549686231503436,
+      "learning_rate": 1.2516745039794864e-06,
+      "loss": 0.116,
+      "step": 9218
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.23671336435756016,
+      "learning_rate": 1.2507715245700036e-06,
+      "loss": 0.0606,
+      "step": 9219
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3860078526211724,
+      "learning_rate": 1.2498688244248985e-06,
+      "loss": 0.0888,
+      "step": 9220
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2924911321483093,
+      "learning_rate": 1.2489664036114079e-06,
+      "loss": 0.0998,
+      "step": 9221
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2864676631900646,
+      "learning_rate": 1.2480642621967525e-06,
+      "loss": 0.0735,
+      "step": 9222
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6617321428723307,
+      "learning_rate": 1.2471624002481276e-06,
+      "loss": 0.1301,
+      "step": 9223
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2187410370201043,
+      "learning_rate": 1.2462608178327074e-06,
+      "loss": 0.0528,
+      "step": 9224
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3264560369202226,
+      "learning_rate": 1.24535951501765e-06,
+      "loss": 0.0771,
+      "step": 9225
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2871133741514348,
+      "learning_rate": 1.2444584918700885e-06,
+      "loss": 0.0725,
+      "step": 9226
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6245343908210043,
+      "learning_rate": 1.2435577484571366e-06,
+      "loss": 0.1023,
+      "step": 9227
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.276183308165265,
+      "learning_rate": 1.2426572848458857e-06,
+      "loss": 0.1067,
+      "step": 9228
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3394828686624172,
+      "learning_rate": 1.2417571011034075e-06,
+      "loss": 0.0745,
+      "step": 9229
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.46067134800713977,
+      "learning_rate": 1.2408571972967547e-06,
+      "loss": 0.1118,
+      "step": 9230
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.17730063589145664,
+      "learning_rate": 1.239957573492957e-06,
+      "loss": 0.0257,
+      "step": 9231
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2116527428913274,
+      "learning_rate": 1.2390582297590225e-06,
+      "loss": 0.0589,
+      "step": 9232
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3254278097208631,
+      "learning_rate": 1.238159166161939e-06,
+      "loss": 0.0542,
+      "step": 9233
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.26352865096776623,
+      "learning_rate": 1.237260382768677e-06,
+      "loss": 0.0488,
+      "step": 9234
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.17625151302969838,
+      "learning_rate": 1.2363618796461807e-06,
+      "loss": 0.0566,
+      "step": 9235
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.28049986475212957,
+      "learning_rate": 1.2354636568613766e-06,
+      "loss": 0.0959,
+      "step": 9236
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2145859191594404,
+      "learning_rate": 1.2345657144811695e-06,
+      "loss": 0.0606,
+      "step": 9237
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.46653354540853426,
+      "learning_rate": 1.2336680525724415e-06,
+      "loss": 0.0602,
+      "step": 9238
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.230994084036041,
+      "learning_rate": 1.2327706712020594e-06,
+      "loss": 0.0648,
+      "step": 9239
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.4335279778063143,
+      "learning_rate": 1.2318735704368634e-06,
+      "loss": 0.1039,
+      "step": 9240
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3564292176371483,
+      "learning_rate": 1.2309767503436731e-06,
+      "loss": 0.0649,
+      "step": 9241
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3133147494815925,
+      "learning_rate": 1.2300802109892918e-06,
+      "loss": 0.0685,
+      "step": 9242
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.4414189658587401,
+      "learning_rate": 1.229183952440498e-06,
+      "loss": 0.0981,
+      "step": 9243
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.22359238565160444,
+      "learning_rate": 1.228287974764049e-06,
+      "loss": 0.0335,
+      "step": 9244
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3510933683235041,
+      "learning_rate": 1.2273922780266845e-06,
+      "loss": 0.0646,
+      "step": 9245
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.21996907851273628,
+      "learning_rate": 1.2264968622951208e-06,
+      "loss": 0.036,
+      "step": 9246
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.4398368183051803,
+      "learning_rate": 1.225601727636052e-06,
+      "loss": 0.0989,
+      "step": 9247
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.25834230864058244,
+      "learning_rate": 1.2247068741161533e-06,
+      "loss": 0.0855,
+      "step": 9248
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3094501529566432,
+      "learning_rate": 1.2238123018020808e-06,
+      "loss": 0.0679,
+      "step": 9249
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2053615124491915,
+      "learning_rate": 1.222918010760466e-06,
+      "loss": 0.0627,
+      "step": 9250
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3007431007395312,
+      "learning_rate": 1.2220240010579193e-06,
+      "loss": 0.083,
+      "step": 9251
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3087149869198408,
+      "learning_rate": 1.2211302727610347e-06,
+      "loss": 0.0436,
+      "step": 9252
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2215414533531904,
+      "learning_rate": 1.2202368259363812e-06,
+      "loss": 0.0604,
+      "step": 9253
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.43182902303850856,
+      "learning_rate": 1.2193436606505072e-06,
+      "loss": 0.1244,
+      "step": 9254
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.19047223059665688,
+      "learning_rate": 1.21845077696994e-06,
+      "loss": 0.0568,
+      "step": 9255
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.30423573990868696,
+      "learning_rate": 1.2175581749611898e-06,
+      "loss": 0.0728,
+      "step": 9256
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5251447613032193,
+      "learning_rate": 1.2166658546907411e-06,
+      "loss": 0.0776,
+      "step": 9257
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.465648460561916,
+      "learning_rate": 1.2157738162250577e-06,
+      "loss": 0.0872,
+      "step": 9258
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3260908016750977,
+      "learning_rate": 1.214882059630586e-06,
+      "loss": 0.1006,
+      "step": 9259
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5012418158960525,
+      "learning_rate": 1.2139905849737487e-06,
+      "loss": 0.0946,
+      "step": 9260
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.364099262400626,
+      "learning_rate": 1.2130993923209467e-06,
+      "loss": 0.0668,
+      "step": 9261
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.39582601520545835,
+      "learning_rate": 1.2122084817385638e-06,
+      "loss": 0.0755,
+      "step": 9262
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.32905043591367644,
+      "learning_rate": 1.2113178532929582e-06,
+      "loss": 0.0686,
+      "step": 9263
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2627850880012583,
+      "learning_rate": 1.21042750705047e-06,
+      "loss": 0.0835,
+      "step": 9264
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.4961173326127303,
+      "learning_rate": 1.2095374430774154e-06,
+      "loss": 0.0997,
+      "step": 9265
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5857777439196745,
+      "learning_rate": 1.2086476614400943e-06,
+      "loss": 0.1383,
+      "step": 9266
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.49850260175246414,
+      "learning_rate": 1.2077581622047818e-06,
+      "loss": 0.104,
+      "step": 9267
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.28579229557836094,
+      "learning_rate": 1.2068689454377313e-06,
+      "loss": 0.0399,
+      "step": 9268
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2604370825992014,
+      "learning_rate": 1.20598001120518e-06,
+      "loss": 0.0637,
+      "step": 9269
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.30371635232132094,
+      "learning_rate": 1.2050913595733392e-06,
+      "loss": 0.0857,
+      "step": 9270
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2928447481850043,
+      "learning_rate": 1.2042029906084012e-06,
+      "loss": 0.0631,
+      "step": 9271
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.49681443954098764,
+      "learning_rate": 1.2033149043765346e-06,
+      "loss": 0.1116,
+      "step": 9272
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.544656570669009,
+      "learning_rate": 1.2024271009438932e-06,
+      "loss": 0.0924,
+      "step": 9273
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3663706898653954,
+      "learning_rate": 1.2015395803766034e-06,
+      "loss": 0.0748,
+      "step": 9274
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3213708679614775,
+      "learning_rate": 1.2006523427407724e-06,
+      "loss": 0.0367,
+      "step": 9275
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.19040594980044329,
+      "learning_rate": 1.1997653881024884e-06,
+      "loss": 0.0582,
+      "step": 9276
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2837855617236932,
+      "learning_rate": 1.198878716527817e-06,
+      "loss": 0.0672,
+      "step": 9277
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.4637737504449664,
+      "learning_rate": 1.1979923280828016e-06,
+      "loss": 0.1137,
+      "step": 9278
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2865487749627345,
+      "learning_rate": 1.1971062228334646e-06,
+      "loss": 0.0667,
+      "step": 9279
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.22232684739949388,
+      "learning_rate": 1.196220400845811e-06,
+      "loss": 0.066,
+      "step": 9280
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2594550711372454,
+      "learning_rate": 1.1953348621858207e-06,
+      "loss": 0.067,
+      "step": 9281
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.32639688506198294,
+      "learning_rate": 1.1944496069194523e-06,
+      "loss": 0.0922,
+      "step": 9282
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.4181197692851612,
+      "learning_rate": 1.1935646351126478e-06,
+      "loss": 0.1064,
+      "step": 9283
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.33395983062068574,
+      "learning_rate": 1.1926799468313227e-06,
+      "loss": 0.064,
+      "step": 9284
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3839744243208467,
+      "learning_rate": 1.1917955421413735e-06,
+      "loss": 0.0679,
+      "step": 9285
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3463911864700188,
+      "learning_rate": 1.190911421108678e-06,
+      "loss": 0.1059,
+      "step": 9286
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.40826901514633357,
+      "learning_rate": 1.1900275837990899e-06,
+      "loss": 0.0953,
+      "step": 9287
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.44036851412061884,
+      "learning_rate": 1.1891440302784418e-06,
+      "loss": 0.1145,
+      "step": 9288
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3649675073626109,
+      "learning_rate": 1.1882607606125447e-06,
+      "loss": 0.0628,
+      "step": 9289
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3263009285481507,
+      "learning_rate": 1.1873777748671922e-06,
+      "loss": 0.0714,
+      "step": 9290
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5055607782042301,
+      "learning_rate": 1.1864950731081536e-06,
+      "loss": 0.106,
+      "step": 9291
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5152625006457674,
+      "learning_rate": 1.185612655401176e-06,
+      "loss": 0.0843,
+      "step": 9292
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5653781212651706,
+      "learning_rate": 1.1847305218119898e-06,
+      "loss": 0.1353,
+      "step": 9293
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.37880194789335686,
+      "learning_rate": 1.1838486724062992e-06,
+      "loss": 0.0972,
+      "step": 9294
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.4308745733146846,
+      "learning_rate": 1.1829671072497906e-06,
+      "loss": 0.0874,
+      "step": 9295
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.39757064981961926,
+      "learning_rate": 1.1820858264081281e-06,
+      "loss": 0.0918,
+      "step": 9296
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.29906010198248384,
+      "learning_rate": 1.1812048299469526e-06,
+      "loss": 0.0659,
+      "step": 9297
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3784480265726095,
+      "learning_rate": 1.1803241179318886e-06,
+      "loss": 0.0988,
+      "step": 9298
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6050789612783471,
+      "learning_rate": 1.1794436904285355e-06,
+      "loss": 0.093,
+      "step": 9299
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.30023379946434486,
+      "learning_rate": 1.1785635475024731e-06,
+      "loss": 0.0902,
+      "step": 9300
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.30685317374439186,
+      "learning_rate": 1.1776836892192578e-06,
+      "loss": 0.0837,
+      "step": 9301
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.20348060964221507,
+      "learning_rate": 1.1768041156444293e-06,
+      "loss": 0.0794,
+      "step": 9302
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2671707214757036,
+      "learning_rate": 1.175924826843502e-06,
+      "loss": 0.0526,
+      "step": 9303
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.22407515347863294,
+      "learning_rate": 1.17504582288197e-06,
+      "loss": 0.082,
+      "step": 9304
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.46308565975101823,
+      "learning_rate": 1.1741671038253077e-06,
+      "loss": 0.1088,
+      "step": 9305
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.29804296510829625,
+      "learning_rate": 1.173288669738965e-06,
+      "loss": 0.0588,
+      "step": 9306
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3496570055014727,
+      "learning_rate": 1.1724105206883763e-06,
+      "loss": 0.0816,
+      "step": 9307
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.43357340188482607,
+      "learning_rate": 1.171532656738949e-06,
+      "loss": 0.0857,
+      "step": 9308
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.37015159066891573,
+      "learning_rate": 1.1706550779560705e-06,
+      "loss": 0.1109,
+      "step": 9309
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.40703606575616763,
+      "learning_rate": 1.1697777844051105e-06,
+      "loss": 0.0656,
+      "step": 9310
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.39581708565725676,
+      "learning_rate": 1.1689007761514143e-06,
+      "loss": 0.1043,
+      "step": 9311
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.41312565780093147,
+      "learning_rate": 1.1680240532603055e-06,
+      "loss": 0.1044,
+      "step": 9312
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3341608806589965,
+      "learning_rate": 1.1671476157970873e-06,
+      "loss": 0.0681,
+      "step": 9313
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.19637090260168866,
+      "learning_rate": 1.166271463827044e-06,
+      "loss": 0.0394,
+      "step": 9314
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.43301115687759745,
+      "learning_rate": 1.1653955974154353e-06,
+      "loss": 0.07,
+      "step": 9315
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.32825181940964554,
+      "learning_rate": 1.164520016627499e-06,
+      "loss": 0.0679,
+      "step": 9316
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.316220326367159,
+      "learning_rate": 1.1636447215284563e-06,
+      "loss": 0.0787,
+      "step": 9317
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.46761205507602793,
+      "learning_rate": 1.1627697121835035e-06,
+      "loss": 0.1006,
+      "step": 9318
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.19645546475225717,
+      "learning_rate": 1.1618949886578145e-06,
+      "loss": 0.0591,
+      "step": 9319
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.34900732390252254,
+      "learning_rate": 1.1610205510165473e-06,
+      "loss": 0.1121,
+      "step": 9320
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5355660971225606,
+      "learning_rate": 1.1601463993248329e-06,
+      "loss": 0.1237,
+      "step": 9321
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2886544909843859,
+      "learning_rate": 1.1592725336477833e-06,
+      "loss": 0.0707,
+      "step": 9322
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2902073720988028,
+      "learning_rate": 1.158398954050488e-06,
+      "loss": 0.0926,
+      "step": 9323
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.36037309962481706,
+      "learning_rate": 1.1575256605980189e-06,
+      "loss": 0.109,
+      "step": 9324
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2708472773171315,
+      "learning_rate": 1.1566526533554228e-06,
+      "loss": 0.0882,
+      "step": 9325
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.44194293310601235,
+      "learning_rate": 1.155779932387725e-06,
+      "loss": 0.11,
+      "step": 9326
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.26089661809459846,
+      "learning_rate": 1.1549074977599333e-06,
+      "loss": 0.0586,
+      "step": 9327
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.47339368871831516,
+      "learning_rate": 1.1540353495370304e-06,
+      "loss": 0.096,
+      "step": 9328
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.21961815733042944,
+      "learning_rate": 1.1531634877839792e-06,
+      "loss": 0.0358,
+      "step": 9329
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.45987807831637395,
+      "learning_rate": 1.152291912565719e-06,
+      "loss": 0.0832,
+      "step": 9330
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2347152976061844,
+      "learning_rate": 1.1514206239471737e-06,
+      "loss": 0.0445,
+      "step": 9331
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2968466021735074,
+      "learning_rate": 1.1505496219932399e-06,
+      "loss": 0.0911,
+      "step": 9332
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2340864911808723,
+      "learning_rate": 1.1496789067687935e-06,
+      "loss": 0.0709,
+      "step": 9333
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.28987640847842494,
+      "learning_rate": 1.1488084783386933e-06,
+      "loss": 0.0681,
+      "step": 9334
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5341363955376943,
+      "learning_rate": 1.147938336767772e-06,
+      "loss": 0.1337,
+      "step": 9335
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.29881233165587984,
+      "learning_rate": 1.1470684821208428e-06,
+      "loss": 0.0965,
+      "step": 9336
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.30295929237120667,
+      "learning_rate": 1.146198914462699e-06,
+      "loss": 0.0934,
+      "step": 9337
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3824552462172366,
+      "learning_rate": 1.14532963385811e-06,
+      "loss": 0.0656,
+      "step": 9338
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2830981094687693,
+      "learning_rate": 1.144460640371825e-06,
+      "loss": 0.0703,
+      "step": 9339
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3180052576142439,
+      "learning_rate": 1.1435919340685702e-06,
+      "loss": 0.0696,
+      "step": 9340
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.29043449646022373,
+      "learning_rate": 1.142723515013055e-06,
+      "loss": 0.0599,
+      "step": 9341
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.22981331829063498,
+      "learning_rate": 1.1418553832699624e-06,
+      "loss": 0.0535,
+      "step": 9342
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.29175778970561117,
+      "learning_rate": 1.140987538903955e-06,
+      "loss": 0.0707,
+      "step": 9343
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3211786055903773,
+      "learning_rate": 1.1401199819796777e-06,
+      "loss": 0.0903,
+      "step": 9344
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.516524402508203,
+      "learning_rate": 1.139252712561749e-06,
+      "loss": 0.1199,
+      "step": 9345
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2929054470081414,
+      "learning_rate": 1.1383857307147695e-06,
+      "loss": 0.0884,
+      "step": 9346
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3427345335274688,
+      "learning_rate": 1.1375190365033145e-06,
+      "loss": 0.1023,
+      "step": 9347
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.32549223324032595,
+      "learning_rate": 1.1366526299919438e-06,
+      "loss": 0.0626,
+      "step": 9348
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.32181859855254363,
+      "learning_rate": 1.135786511245191e-06,
+      "loss": 0.0783,
+      "step": 9349
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.23688833559680336,
+      "learning_rate": 1.1349206803275682e-06,
+      "loss": 0.0619,
+      "step": 9350
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.4372714170111681,
+      "learning_rate": 1.1340551373035701e-06,
+      "loss": 0.0986,
+      "step": 9351
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.20914236746793263,
+      "learning_rate": 1.1331898822376664e-06,
+      "loss": 0.0488,
+      "step": 9352
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5617233165118068,
+      "learning_rate": 1.1323249151943045e-06,
+      "loss": 0.1539,
+      "step": 9353
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.4582047601348013,
+      "learning_rate": 1.131460236237915e-06,
+      "loss": 0.132,
+      "step": 9354
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.29709266278517865,
+      "learning_rate": 1.1305958454329036e-06,
+      "loss": 0.0928,
+      "step": 9355
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.27799174256557124,
+      "learning_rate": 1.1297317428436545e-06,
+      "loss": 0.0757,
+      "step": 9356
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6172203818227928,
+      "learning_rate": 1.1288679285345288e-06,
+      "loss": 0.1081,
+      "step": 9357
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.31712321927323595,
+      "learning_rate": 1.1280044025698733e-06,
+      "loss": 0.0719,
+      "step": 9358
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.36042875625158505,
+      "learning_rate": 1.1271411650140052e-06,
+      "loss": 0.0791,
+      "step": 9359
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2150978497167032,
+      "learning_rate": 1.1262782159312236e-06,
+      "loss": 0.043,
+      "step": 9360
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.25335583963677155,
+      "learning_rate": 1.125415555385807e-06,
+      "loss": 0.0723,
+      "step": 9361
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.22182092215421234,
+      "learning_rate": 1.1245531834420115e-06,
+      "loss": 0.0281,
+      "step": 9362
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.43031058521395876,
+      "learning_rate": 1.123691100164071e-06,
+      "loss": 0.093,
+      "step": 9363
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2539893310142868,
+      "learning_rate": 1.1228293056161987e-06,
+      "loss": 0.0653,
+      "step": 9364
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3199911626174709,
+      "learning_rate": 1.1219677998625845e-06,
+      "loss": 0.0815,
+      "step": 9365
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.4732875326076397,
+      "learning_rate": 1.1211065829674007e-06,
+      "loss": 0.1319,
+      "step": 9366
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2207601332851445,
+      "learning_rate": 1.1202456549947954e-06,
+      "loss": 0.0752,
+      "step": 9367
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.38324574821991314,
+      "learning_rate": 1.1193850160088948e-06,
+      "loss": 0.0999,
+      "step": 9368
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.36410890361579085,
+      "learning_rate": 1.1185246660738025e-06,
+      "loss": 0.0888,
+      "step": 9369
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.4077372457861351,
+      "learning_rate": 1.1176646052536067e-06,
+      "loss": 0.0973,
+      "step": 9370
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2874286157242859,
+      "learning_rate": 1.1168048336123666e-06,
+      "loss": 0.0816,
+      "step": 9371
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.27483479656349746,
+      "learning_rate": 1.1159453512141238e-06,
+      "loss": 0.0678,
+      "step": 9372
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.4328440612785833,
+      "learning_rate": 1.1150861581228978e-06,
+      "loss": 0.129,
+      "step": 9373
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.16645049465152567,
+      "learning_rate": 1.114227254402685e-06,
+      "loss": 0.0431,
+      "step": 9374
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.36472653749834893,
+      "learning_rate": 1.113368640117463e-06,
+      "loss": 0.0938,
+      "step": 9375
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.24997845642505542,
+      "learning_rate": 1.1125103153311873e-06,
+      "loss": 0.0609,
+      "step": 9376
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.38933672911441664,
+      "learning_rate": 1.1116522801077873e-06,
+      "loss": 0.0717,
+      "step": 9377
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.4833171667719454,
+      "learning_rate": 1.1107945345111781e-06,
+      "loss": 0.1001,
+      "step": 9378
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.32797178174366665,
+      "learning_rate": 1.1099370786052489e-06,
+      "loss": 0.0967,
+      "step": 9379
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3053390590459215,
+      "learning_rate": 1.109079912453867e-06,
+      "loss": 0.0913,
+      "step": 9380
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.35632643637962647,
+      "learning_rate": 1.108223036120878e-06,
+      "loss": 0.095,
+      "step": 9381
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.29643629164448054,
+      "learning_rate": 1.1073664496701103e-06,
+      "loss": 0.0865,
+      "step": 9382
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.24622018429467024,
+      "learning_rate": 1.1065101531653654e-06,
+      "loss": 0.0506,
+      "step": 9383
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2145037311033517,
+      "learning_rate": 1.105654146670424e-06,
+      "loss": 0.0571,
+      "step": 9384
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3297512004467472,
+      "learning_rate": 1.1047984302490495e-06,
+      "loss": 0.0641,
+      "step": 9385
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.48553134972174045,
+      "learning_rate": 1.1039430039649795e-06,
+      "loss": 0.1001,
+      "step": 9386
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.22196605932519084,
+      "learning_rate": 1.1030878678819307e-06,
+      "loss": 0.0671,
+      "step": 9387
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.4530690811035253,
+      "learning_rate": 1.1022330220635969e-06,
+      "loss": 0.0837,
+      "step": 9388
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.25768154944476174,
+      "learning_rate": 1.1013784665736554e-06,
+      "loss": 0.049,
+      "step": 9389
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.38111299082343153,
+      "learning_rate": 1.1005242014757567e-06,
+      "loss": 0.0901,
+      "step": 9390
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2726035560376019,
+      "learning_rate": 1.0996702268335308e-06,
+      "loss": 0.0579,
+      "step": 9391
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2721059596813946,
+      "learning_rate": 1.0988165427105885e-06,
+      "loss": 0.0635,
+      "step": 9392
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3614775055241465,
+      "learning_rate": 1.0979631491705162e-06,
+      "loss": 0.0854,
+      "step": 9393
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.23025476417282853,
+      "learning_rate": 1.097110046276878e-06,
+      "loss": 0.0787,
+      "step": 9394
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.25897108429739835,
+      "learning_rate": 1.096257234093221e-06,
+      "loss": 0.0837,
+      "step": 9395
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.412858913285288,
+      "learning_rate": 1.0954047126830663e-06,
+      "loss": 0.1279,
+      "step": 9396
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.26515248655763857,
+      "learning_rate": 1.0945524821099147e-06,
+      "loss": 0.0595,
+      "step": 9397
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.35337904666777814,
+      "learning_rate": 1.0937005424372431e-06,
+      "loss": 0.0966,
+      "step": 9398
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3444547127862197,
+      "learning_rate": 1.092848893728513e-06,
+      "loss": 0.0751,
+      "step": 9399
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.45933753487575285,
+      "learning_rate": 1.091997536047158e-06,
+      "loss": 0.1194,
+      "step": 9400
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.25231411626635764,
+      "learning_rate": 1.091146469456591e-06,
+      "loss": 0.0591,
+      "step": 9401
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.20793834255996602,
+      "learning_rate": 1.090295694020207e-06,
+      "loss": 0.043,
+      "step": 9402
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.374460615764852,
+      "learning_rate": 1.0894452098013758e-06,
+      "loss": 0.0895,
+      "step": 9403
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.4684536880803592,
+      "learning_rate": 1.0885950168634457e-06,
+      "loss": 0.1639,
+      "step": 9404
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.26704351179738445,
+      "learning_rate": 1.0877451152697428e-06,
+      "loss": 0.0655,
+      "step": 9405
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2640189603666819,
+      "learning_rate": 1.0868955050835755e-06,
+      "loss": 0.0636,
+      "step": 9406
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2912415975300197,
+      "learning_rate": 1.086046186368227e-06,
+      "loss": 0.0785,
+      "step": 9407
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.4176305951783421,
+      "learning_rate": 1.0851971591869575e-06,
+      "loss": 0.0815,
+      "step": 9408
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.44509235245635737,
+      "learning_rate": 1.0843484236030105e-06,
+      "loss": 0.0883,
+      "step": 9409
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3256122554546421,
+      "learning_rate": 1.083499979679603e-06,
+      "loss": 0.0714,
+      "step": 9410
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.28567074769986633,
+      "learning_rate": 1.0826518274799313e-06,
+      "loss": 0.0631,
+      "step": 9411
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.28451706882348354,
+      "learning_rate": 1.0818039670671725e-06,
+      "loss": 0.0414,
+      "step": 9412
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.44350728196451367,
+      "learning_rate": 1.0809563985044796e-06,
+      "loss": 0.1002,
+      "step": 9413
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.34426239445685,
+      "learning_rate": 1.0801091218549847e-06,
+      "loss": 0.0714,
+      "step": 9414
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.36844961118504277,
+      "learning_rate": 1.0792621371817957e-06,
+      "loss": 0.0865,
+      "step": 9415
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.49899936843600834,
+      "learning_rate": 1.0784154445480043e-06,
+      "loss": 0.0888,
+      "step": 9416
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.23109702330938567,
+      "learning_rate": 1.0775690440166753e-06,
+      "loss": 0.0423,
+      "step": 9417
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.34440355863584776,
+      "learning_rate": 1.0767229356508524e-06,
+      "loss": 0.0808,
+      "step": 9418
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.35611727418842365,
+      "learning_rate": 1.0758771195135614e-06,
+      "loss": 0.1075,
+      "step": 9419
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.4896959184577979,
+      "learning_rate": 1.0750315956678025e-06,
+      "loss": 0.1127,
+      "step": 9420
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3277182256566695,
+      "learning_rate": 1.0741863641765548e-06,
+      "loss": 0.0632,
+      "step": 9421
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.38255715330385964,
+      "learning_rate": 1.0733414251027751e-06,
+      "loss": 0.1369,
+      "step": 9422
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.29490165596071827,
+      "learning_rate": 1.0724967785094014e-06,
+      "loss": 0.0388,
+      "step": 9423
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2831306258948654,
+      "learning_rate": 1.0716524244593473e-06,
+      "loss": 0.1043,
+      "step": 9424
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.36518045132182364,
+      "learning_rate": 1.0708083630155036e-06,
+      "loss": 0.0751,
+      "step": 9425
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5797122499433458,
+      "learning_rate": 1.069964594240744e-06,
+      "loss": 0.1318,
+      "step": 9426
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.27735092485057206,
+      "learning_rate": 1.069121118197915e-06,
+      "loss": 0.0579,
+      "step": 9427
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.22515569090773913,
+      "learning_rate": 1.0682779349498428e-06,
+      "loss": 0.0728,
+      "step": 9428
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.248802714047661,
+      "learning_rate": 1.0674350445593357e-06,
+      "loss": 0.0567,
+      "step": 9429
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2321232648998306,
+      "learning_rate": 1.0665924470891753e-06,
+      "loss": 0.085,
+      "step": 9430
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.38019342045157584,
+      "learning_rate": 1.0657501426021233e-06,
+      "loss": 0.0992,
+      "step": 9431
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3023057906984438,
+      "learning_rate": 1.0649081311609188e-06,
+      "loss": 0.0664,
+      "step": 9432
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2961175596293181,
+      "learning_rate": 1.0640664128282797e-06,
+      "loss": 0.084,
+      "step": 9433
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.39992940364904905,
+      "learning_rate": 1.0632249876669037e-06,
+      "loss": 0.0906,
+      "step": 9434
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2050863137714559,
+      "learning_rate": 1.062383855739464e-06,
+      "loss": 0.0747,
+      "step": 9435
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5692649664640083,
+      "learning_rate": 1.0615430171086127e-06,
+      "loss": 0.1,
+      "step": 9436
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3718390881509671,
+      "learning_rate": 1.0607024718369802e-06,
+      "loss": 0.1043,
+      "step": 9437
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2930905590612547,
+      "learning_rate": 1.059862219987176e-06,
+      "loss": 0.0807,
+      "step": 9438
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.38051474016125814,
+      "learning_rate": 1.0590222616217877e-06,
+      "loss": 0.0935,
+      "step": 9439
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3255129636430382,
+      "learning_rate": 1.0581825968033788e-06,
+      "loss": 0.0671,
+      "step": 9440
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7213886319785722,
+      "learning_rate": 1.0573432255944926e-06,
+      "loss": 0.14,
+      "step": 9441
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4026173774861173,
+      "learning_rate": 1.0565041480576493e-06,
+      "loss": 0.0765,
+      "step": 9442
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.29046274420787477,
+      "learning_rate": 1.0556653642553517e-06,
+      "loss": 0.0629,
+      "step": 9443
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4015107018020187,
+      "learning_rate": 1.0548268742500745e-06,
+      "loss": 0.1216,
+      "step": 9444
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2945006087799275,
+      "learning_rate": 1.053988678104273e-06,
+      "loss": 0.0869,
+      "step": 9445
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2515773430322903,
+      "learning_rate": 1.0531507758803827e-06,
+      "loss": 0.0829,
+      "step": 9446
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4900612012575581,
+      "learning_rate": 1.0523131676408154e-06,
+      "loss": 0.1599,
+      "step": 9447
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3091266708599519,
+      "learning_rate": 1.0514758534479601e-06,
+      "loss": 0.0811,
+      "step": 9448
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.31748186443582654,
+      "learning_rate": 1.0506388333641836e-06,
+      "loss": 0.117,
+      "step": 9449
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.22563766672562718,
+      "learning_rate": 1.0498021074518355e-06,
+      "loss": 0.0611,
+      "step": 9450
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5330579468842649,
+      "learning_rate": 1.0489656757732375e-06,
+      "loss": 0.1079,
+      "step": 9451
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.37375961770527794,
+      "learning_rate": 1.0481295383906914e-06,
+      "loss": 0.0789,
+      "step": 9452
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.39041849706724446,
+      "learning_rate": 1.0472936953664798e-06,
+      "loss": 0.0983,
+      "step": 9453
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.30707174061388903,
+      "learning_rate": 1.0464581467628597e-06,
+      "loss": 0.096,
+      "step": 9454
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4458638824848113,
+      "learning_rate": 1.0456228926420682e-06,
+      "loss": 0.1091,
+      "step": 9455
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4239362180324649,
+      "learning_rate": 1.0447879330663185e-06,
+      "loss": 0.0947,
+      "step": 9456
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.21481715140697885,
+      "learning_rate": 1.0439532680978054e-06,
+      "loss": 0.0691,
+      "step": 9457
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.09852494353072135,
+      "learning_rate": 1.0431188977986983e-06,
+      "loss": 0.017,
+      "step": 9458
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.34761684295440354,
+      "learning_rate": 1.0422848222311455e-06,
+      "loss": 0.0793,
+      "step": 9459
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4151762958884645,
+      "learning_rate": 1.0414510414572755e-06,
+      "loss": 0.1006,
+      "step": 9460
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3349420872015461,
+      "learning_rate": 1.0406175555391923e-06,
+      "loss": 0.0842,
+      "step": 9461
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.38084513280878435,
+      "learning_rate": 1.039784364538977e-06,
+      "loss": 0.0831,
+      "step": 9462
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2971313501240971,
+      "learning_rate": 1.038951468518694e-06,
+      "loss": 0.0745,
+      "step": 9463
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4084980181400941,
+      "learning_rate": 1.0381188675403803e-06,
+      "loss": 0.1094,
+      "step": 9464
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4114010564745562,
+      "learning_rate": 1.0372865616660532e-06,
+      "loss": 0.1103,
+      "step": 9465
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.22641981586402618,
+      "learning_rate": 1.0364545509577057e-06,
+      "loss": 0.0937,
+      "step": 9466
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.35285123482097275,
+      "learning_rate": 1.035622835477314e-06,
+      "loss": 0.0719,
+      "step": 9467
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.281639337429761,
+      "learning_rate": 1.0347914152868283e-06,
+      "loss": 0.0553,
+      "step": 9468
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.1922242880042777,
+      "learning_rate": 1.033960290448175e-06,
+      "loss": 0.0475,
+      "step": 9469
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.32858527518905223,
+      "learning_rate": 1.033129461023265e-06,
+      "loss": 0.0998,
+      "step": 9470
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4202812320405979,
+      "learning_rate": 1.0322989270739814e-06,
+      "loss": 0.0787,
+      "step": 9471
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.24740082052772397,
+      "learning_rate": 1.0314686886621872e-06,
+      "loss": 0.0568,
+      "step": 9472
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.49258359512008126,
+      "learning_rate": 1.0306387458497218e-06,
+      "loss": 0.0719,
+      "step": 9473
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.43993148823945727,
+      "learning_rate": 1.0298090986984077e-06,
+      "loss": 0.0891,
+      "step": 9474
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4448497054580627,
+      "learning_rate": 1.0289797472700397e-06,
+      "loss": 0.0787,
+      "step": 9475
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2674762105525837,
+      "learning_rate": 1.0281506916263918e-06,
+      "loss": 0.0629,
+      "step": 9476
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.31276316195512316,
+      "learning_rate": 1.0273219318292193e-06,
+      "loss": 0.0791,
+      "step": 9477
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3003362768657269,
+      "learning_rate": 1.0264934679402523e-06,
+      "loss": 0.0874,
+      "step": 9478
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.24181030501496012,
+      "learning_rate": 1.0256653000211969e-06,
+      "loss": 0.0527,
+      "step": 9479
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.344260843065536,
+      "learning_rate": 1.0248374281337441e-06,
+      "loss": 0.0731,
+      "step": 9480
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3786440009143864,
+      "learning_rate": 1.0240098523395569e-06,
+      "loss": 0.1014,
+      "step": 9481
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3003477594172609,
+      "learning_rate": 1.0231825727002776e-06,
+      "loss": 0.0839,
+      "step": 9482
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.36990139911029807,
+      "learning_rate": 1.0223555892775256e-06,
+      "loss": 0.0868,
+      "step": 9483
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.29915541545008373,
+      "learning_rate": 1.0215289021329023e-06,
+      "loss": 0.0727,
+      "step": 9484
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4263460358796248,
+      "learning_rate": 1.0207025113279827e-06,
+      "loss": 0.0922,
+      "step": 9485
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.22455781156748722,
+      "learning_rate": 1.0198764169243198e-06,
+      "loss": 0.048,
+      "step": 9486
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.24250498892971267,
+      "learning_rate": 1.0190506189834482e-06,
+      "loss": 0.0599,
+      "step": 9487
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.35149428487455425,
+      "learning_rate": 1.0182251175668779e-06,
+      "loss": 0.0814,
+      "step": 9488
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3375633843484217,
+      "learning_rate": 1.0173999127360957e-06,
+      "loss": 0.1013,
+      "step": 9489
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3857001936935201,
+      "learning_rate": 1.0165750045525673e-06,
+      "loss": 0.0844,
+      "step": 9490
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.24759213148661807,
+      "learning_rate": 1.015750393077739e-06,
+      "loss": 0.0445,
+      "step": 9491
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2892582334898657,
+      "learning_rate": 1.014926078373032e-06,
+      "loss": 0.0558,
+      "step": 9492
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4304169615966122,
+      "learning_rate": 1.0141020604998437e-06,
+      "loss": 0.0988,
+      "step": 9493
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2634242748307677,
+      "learning_rate": 1.0132783395195545e-06,
+      "loss": 0.0939,
+      "step": 9494
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.572803060882255,
+      "learning_rate": 1.0124549154935192e-06,
+      "loss": 0.0801,
+      "step": 9495
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3375868878036133,
+      "learning_rate": 1.0116317884830695e-06,
+      "loss": 0.0914,
+      "step": 9496
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.35469303096046023,
+      "learning_rate": 1.01080895854952e-06,
+      "loss": 0.0691,
+      "step": 9497
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.37002450046140567,
+      "learning_rate": 1.0099864257541575e-06,
+      "loss": 0.1044,
+      "step": 9498
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.33804472847245937,
+      "learning_rate": 1.00916419015825e-06,
+      "loss": 0.0706,
+      "step": 9499
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3608794011457335,
+      "learning_rate": 1.0083422518230412e-06,
+      "loss": 0.0586,
+      "step": 9500
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2131082898312194,
+      "learning_rate": 1.007520610809754e-06,
+      "loss": 0.0473,
+      "step": 9501
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.22556929941496373,
+      "learning_rate": 1.0066992671795905e-06,
+      "loss": 0.0572,
+      "step": 9502
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.33109044504824375,
+      "learning_rate": 1.0058782209937284e-06,
+      "loss": 0.1041,
+      "step": 9503
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2912627081265232,
+      "learning_rate": 1.0050574723133245e-06,
+      "loss": 0.0641,
+      "step": 9504
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.41218942592048113,
+      "learning_rate": 1.0042370211995101e-06,
+      "loss": 0.102,
+      "step": 9505
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4096836015680222,
+      "learning_rate": 1.0034168677134015e-06,
+      "loss": 0.118,
+      "step": 9506
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.43869659020070123,
+      "learning_rate": 1.0025970119160854e-06,
+      "loss": 0.1113,
+      "step": 9507
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7797877869711737,
+      "learning_rate": 1.0017774538686314e-06,
+      "loss": 0.1413,
+      "step": 9508
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.25897759129555037,
+      "learning_rate": 1.0009581936320838e-06,
+      "loss": 0.0626,
+      "step": 9509
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.35017725681712436,
+      "learning_rate": 1.0001392312674645e-06,
+      "loss": 0.062,
+      "step": 9510
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.16520783092672198,
+      "learning_rate": 9.993205668357775e-07,
+      "loss": 0.0397,
+      "step": 9511
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.31679494400956504,
+      "learning_rate": 9.985022003980004e-07,
+      "loss": 0.0793,
+      "step": 9512
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.24074401494415626,
+      "learning_rate": 9.9768413201509e-07,
+      "loss": 0.0784,
+      "step": 9513
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.27580190533815174,
+      "learning_rate": 9.968663617479796e-07,
+      "loss": 0.0609,
+      "step": 9514
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.19535047074178918,
+      "learning_rate": 9.96048889657583e-07,
+      "loss": 0.0719,
+      "step": 9515
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2872448038897539,
+      "learning_rate": 9.952317158047908e-07,
+      "loss": 0.0733,
+      "step": 9516
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.47575822851132854,
+      "learning_rate": 9.944148402504676e-07,
+      "loss": 0.0871,
+      "step": 9517
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.1884864409574822,
+      "learning_rate": 9.935982630554636e-07,
+      "loss": 0.0528,
+      "step": 9518
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3458148469916603,
+      "learning_rate": 9.927819842805997e-07,
+      "loss": 0.0994,
+      "step": 9519
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.1853699595890618,
+      "learning_rate": 9.919660039866757e-07,
+      "loss": 0.0487,
+      "step": 9520
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3398138573946735,
+      "learning_rate": 9.911503222344743e-07,
+      "loss": 0.1055,
+      "step": 9521
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2986193298725853,
+      "learning_rate": 9.903349390847493e-07,
+      "loss": 0.0567,
+      "step": 9522
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4304820281903488,
+      "learning_rate": 9.895198545982365e-07,
+      "loss": 0.1113,
+      "step": 9523
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.32418751394090867,
+      "learning_rate": 9.887050688356464e-07,
+      "loss": 0.0761,
+      "step": 9524
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.40507284821579265,
+      "learning_rate": 9.878905818576718e-07,
+      "loss": 0.086,
+      "step": 9525
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2520521625248654,
+      "learning_rate": 9.87076393724979e-07,
+      "loss": 0.0778,
+      "step": 9526
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.267784860256976,
+      "learning_rate": 9.862625044982116e-07,
+      "loss": 0.0898,
+      "step": 9527
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.24375663959375155,
+      "learning_rate": 9.854489142379958e-07,
+      "loss": 0.0603,
+      "step": 9528
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.26066480031165973,
+      "learning_rate": 9.84635623004932e-07,
+      "loss": 0.0724,
+      "step": 9529
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2799189474938784,
+      "learning_rate": 9.838226308595978e-07,
+      "loss": 0.0391,
+      "step": 9530
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.26333539435902226,
+      "learning_rate": 9.83009937862549e-07,
+      "loss": 0.0538,
+      "step": 9531
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.18994937182771404,
+      "learning_rate": 9.82197544074322e-07,
+      "loss": 0.0615,
+      "step": 9532
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2911355980643953,
+      "learning_rate": 9.813854495554271e-07,
+      "loss": 0.095,
+      "step": 9533
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2823999234235467,
+      "learning_rate": 9.805736543663529e-07,
+      "loss": 0.1044,
+      "step": 9534
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.37222280819664677,
+      "learning_rate": 9.797621585675687e-07,
+      "loss": 0.0971,
+      "step": 9535
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.22530633357904414,
+      "learning_rate": 9.789509622195192e-07,
+      "loss": 0.0469,
+      "step": 9536
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.18876721135174265,
+      "learning_rate": 9.781400653826244e-07,
+      "loss": 0.0672,
+      "step": 9537
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.22955872181301532,
+      "learning_rate": 9.773294681172886e-07,
+      "loss": 0.0725,
+      "step": 9538
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.22469469241823972,
+      "learning_rate": 9.765191704838879e-07,
+      "loss": 0.0543,
+      "step": 9539
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3934945196532428,
+      "learning_rate": 9.757091725427769e-07,
+      "loss": 0.0876,
+      "step": 9540
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.30483661984977134,
+      "learning_rate": 9.748994743542895e-07,
+      "loss": 0.0717,
+      "step": 9541
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.29198721554322654,
+      "learning_rate": 9.740900759787381e-07,
+      "loss": 0.0869,
+      "step": 9542
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4217519720494303,
+      "learning_rate": 9.732809774764107e-07,
+      "loss": 0.0812,
+      "step": 9543
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2615473542211019,
+      "learning_rate": 9.724721789075726e-07,
+      "loss": 0.0656,
+      "step": 9544
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3428016120404283,
+      "learning_rate": 9.716636803324692e-07,
+      "loss": 0.0714,
+      "step": 9545
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3221118511929357,
+      "learning_rate": 9.70855481811322e-07,
+      "loss": 0.0747,
+      "step": 9546
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.33802306437093976,
+      "learning_rate": 9.700475834043299e-07,
+      "loss": 0.0471,
+      "step": 9547
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3275720590014501,
+      "learning_rate": 9.692399851716689e-07,
+      "loss": 0.0749,
+      "step": 9548
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.28056324360057766,
+      "learning_rate": 9.68432687173496e-07,
+      "loss": 0.0719,
+      "step": 9549
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2963831161567858,
+      "learning_rate": 9.676256894699421e-07,
+      "loss": 0.0702,
+      "step": 9550
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2954622161252651,
+      "learning_rate": 9.668189921211158e-07,
+      "loss": 0.0719,
+      "step": 9551
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.21810156831998798,
+      "learning_rate": 9.660125951871075e-07,
+      "loss": 0.071,
+      "step": 9552
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5269466745788222,
+      "learning_rate": 9.652064987279807e-07,
+      "loss": 0.1023,
+      "step": 9553
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3050697099109935,
+      "learning_rate": 9.644007028037771e-07,
+      "loss": 0.1161,
+      "step": 9554
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.27508259709916427,
+      "learning_rate": 9.635952074745191e-07,
+      "loss": 0.0493,
+      "step": 9555
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6101469256763596,
+      "learning_rate": 9.627900128002044e-07,
+      "loss": 0.1157,
+      "step": 9556
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2420013315025714,
+      "learning_rate": 9.61985118840808e-07,
+      "loss": 0.0878,
+      "step": 9557
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.4374001692686658,
+      "learning_rate": 9.611805256562811e-07,
+      "loss": 0.1099,
+      "step": 9558
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2688766849098861,
+      "learning_rate": 9.603762333065586e-07,
+      "loss": 0.0872,
+      "step": 9559
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.36511916332830674,
+      "learning_rate": 9.595722418515469e-07,
+      "loss": 0.0732,
+      "step": 9560
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.31871974683446297,
+      "learning_rate": 9.587685513511297e-07,
+      "loss": 0.0611,
+      "step": 9561
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3491156997031344,
+      "learning_rate": 9.579651618651748e-07,
+      "loss": 0.0512,
+      "step": 9562
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.4384981931791563,
+      "learning_rate": 9.571620734535208e-07,
+      "loss": 0.1168,
+      "step": 9563
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.22598879952824824,
+      "learning_rate": 9.563592861759867e-07,
+      "loss": 0.0728,
+      "step": 9564
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.42204102673121263,
+      "learning_rate": 9.55556800092368e-07,
+      "loss": 0.0865,
+      "step": 9565
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.21609389454811964,
+      "learning_rate": 9.54754615262441e-07,
+      "loss": 0.0637,
+      "step": 9566
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2211024456573294,
+      "learning_rate": 9.53952731745955e-07,
+      "loss": 0.0298,
+      "step": 9567
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3588321453206782,
+      "learning_rate": 9.531511496026397e-07,
+      "loss": 0.1004,
+      "step": 9568
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.4431574812526109,
+      "learning_rate": 9.523498688922e-07,
+      "loss": 0.1096,
+      "step": 9569
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.4132679353807532,
+      "learning_rate": 9.51548889674323e-07,
+      "loss": 0.0921,
+      "step": 9570
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.33711368885112064,
+      "learning_rate": 9.507482120086686e-07,
+      "loss": 0.0876,
+      "step": 9571
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6457169768546224,
+      "learning_rate": 9.499478359548758e-07,
+      "loss": 0.0804,
+      "step": 9572
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.4370573508670499,
+      "learning_rate": 9.491477615725603e-07,
+      "loss": 0.1121,
+      "step": 9573
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.44020553545066887,
+      "learning_rate": 9.483479889213187e-07,
+      "loss": 0.1049,
+      "step": 9574
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.282153746580046,
+      "learning_rate": 9.475485180607213e-07,
+      "loss": 0.0744,
+      "step": 9575
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.290839404306481,
+      "learning_rate": 9.467493490503177e-07,
+      "loss": 0.0637,
+      "step": 9576
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.13818578751209026,
+      "learning_rate": 9.459504819496345e-07,
+      "loss": 0.0175,
+      "step": 9577
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.40812002917388324,
+      "learning_rate": 9.45151916818175e-07,
+      "loss": 0.0955,
+      "step": 9578
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3484525975033104,
+      "learning_rate": 9.44353653715423e-07,
+      "loss": 0.0504,
+      "step": 9579
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2567767141875635,
+      "learning_rate": 9.435556927008371e-07,
+      "loss": 0.0497,
+      "step": 9580
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2824550615456103,
+      "learning_rate": 9.427580338338532e-07,
+      "loss": 0.0768,
+      "step": 9581
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3380780173080835,
+      "learning_rate": 9.419606771738854e-07,
+      "loss": 0.0899,
+      "step": 9582
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3792074620002089,
+      "learning_rate": 9.411636227803267e-07,
+      "loss": 0.1188,
+      "step": 9583
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3060343305488842,
+      "learning_rate": 9.403668707125463e-07,
+      "loss": 0.083,
+      "step": 9584
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2841543855123277,
+      "learning_rate": 9.395704210298895e-07,
+      "loss": 0.0879,
+      "step": 9585
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3502374473609649,
+      "learning_rate": 9.387742737916822e-07,
+      "loss": 0.0962,
+      "step": 9586
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.25842695784572767,
+      "learning_rate": 9.379784290572258e-07,
+      "loss": 0.0656,
+      "step": 9587
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.4540296494458854,
+      "learning_rate": 9.371828868857974e-07,
+      "loss": 0.0712,
+      "step": 9588
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3432389174598196,
+      "learning_rate": 9.363876473366568e-07,
+      "loss": 0.0949,
+      "step": 9589
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.22945750950744656,
+      "learning_rate": 9.355927104690366e-07,
+      "loss": 0.0465,
+      "step": 9590
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.353773220103368,
+      "learning_rate": 9.347980763421482e-07,
+      "loss": 0.0738,
+      "step": 9591
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.339449248143001,
+      "learning_rate": 9.340037450151789e-07,
+      "loss": 0.0845,
+      "step": 9592
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.48394329738078856,
+      "learning_rate": 9.332097165472986e-07,
+      "loss": 0.0953,
+      "step": 9593
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3023242834286774,
+      "learning_rate": 9.324159909976493e-07,
+      "loss": 0.0978,
+      "step": 9594
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5094623349513073,
+      "learning_rate": 9.316225684253511e-07,
+      "loss": 0.1073,
+      "step": 9595
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2784479387712475,
+      "learning_rate": 9.308294488895053e-07,
+      "loss": 0.0776,
+      "step": 9596
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6926111362898754,
+      "learning_rate": 9.300366324491872e-07,
+      "loss": 0.1343,
+      "step": 9597
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.4406050064398591,
+      "learning_rate": 9.292441191634494e-07,
+      "loss": 0.106,
+      "step": 9598
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.4480923296057095,
+      "learning_rate": 9.284519090913225e-07,
+      "loss": 0.128,
+      "step": 9599
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5023795495712815,
+      "learning_rate": 9.276600022918176e-07,
+      "loss": 0.0817,
+      "step": 9600
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5075426578494602,
+      "learning_rate": 9.268683988239186e-07,
+      "loss": 0.0985,
+      "step": 9601
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3027121547749337,
+      "learning_rate": 9.260770987465873e-07,
+      "loss": 0.0854,
+      "step": 9602
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.22173060992373933,
+      "learning_rate": 9.252861021187676e-07,
+      "loss": 0.0442,
+      "step": 9603
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.31549545033445914,
+      "learning_rate": 9.244954089993762e-07,
+      "loss": 0.095,
+      "step": 9604
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5168590521469552,
+      "learning_rate": 9.237050194473068e-07,
+      "loss": 0.0288,
+      "step": 9605
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3615744731190038,
+      "learning_rate": 9.229149335214349e-07,
+      "loss": 0.0597,
+      "step": 9606
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3767879314837329,
+      "learning_rate": 9.221251512806095e-07,
+      "loss": 0.1055,
+      "step": 9607
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.4639846995779717,
+      "learning_rate": 9.213356727836587e-07,
+      "loss": 0.107,
+      "step": 9608
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.4010130271982931,
+      "learning_rate": 9.205464980893852e-07,
+      "loss": 0.0542,
+      "step": 9609
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.4685865165614232,
+      "learning_rate": 9.197576272565744e-07,
+      "loss": 0.1397,
+      "step": 9610
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6573985325570642,
+      "learning_rate": 9.18969060343985e-07,
+      "loss": 0.1402,
+      "step": 9611
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.310333291169207,
+      "learning_rate": 9.181807974103524e-07,
+      "loss": 0.0808,
+      "step": 9612
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2746032677666339,
+      "learning_rate": 9.173928385143932e-07,
+      "loss": 0.0734,
+      "step": 9613
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.4903263931213346,
+      "learning_rate": 9.166051837147988e-07,
+      "loss": 0.0958,
+      "step": 9614
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.1866404093254978,
+      "learning_rate": 9.158178330702378e-07,
+      "loss": 0.0565,
+      "step": 9615
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.46527643199952196,
+      "learning_rate": 9.150307866393554e-07,
+      "loss": 0.084,
+      "step": 9616
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3095889462978008,
+      "learning_rate": 9.142440444807782e-07,
+      "loss": 0.06,
+      "step": 9617
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3267215213323229,
+      "learning_rate": 9.134576066531054e-07,
+      "loss": 0.0721,
+      "step": 9618
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.30233497404182386,
+      "learning_rate": 9.12671473214915e-07,
+      "loss": 0.0731,
+      "step": 9619
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.31526999086031327,
+      "learning_rate": 9.118856442247648e-07,
+      "loss": 0.0793,
+      "step": 9620
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2683763604434798,
+      "learning_rate": 9.111001197411867e-07,
+      "loss": 0.0754,
+      "step": 9621
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.36007606888687704,
+      "learning_rate": 9.103148998226902e-07,
+      "loss": 0.0874,
+      "step": 9622
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3556495163568835,
+      "learning_rate": 9.095299845277655e-07,
+      "loss": 0.1077,
+      "step": 9623
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.401886522684654,
+      "learning_rate": 9.087453739148761e-07,
+      "loss": 0.1367,
+      "step": 9624
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.21200826045047594,
+      "learning_rate": 9.07961068042465e-07,
+      "loss": 0.0569,
+      "step": 9625
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.4233207165730966,
+      "learning_rate": 9.071770669689495e-07,
+      "loss": 0.1008,
+      "step": 9626
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2700683691074575,
+      "learning_rate": 9.063933707527305e-07,
+      "loss": 0.0752,
+      "step": 9627
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.4493900028978525,
+      "learning_rate": 9.0560997945218e-07,
+      "loss": 0.1133,
+      "step": 9628
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.1854490994459956,
+      "learning_rate": 9.048268931256482e-07,
+      "loss": 0.0394,
+      "step": 9629
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.23903462925591118,
+      "learning_rate": 9.040441118314669e-07,
+      "loss": 0.0865,
+      "step": 9630
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.14879659256639505,
+      "learning_rate": 9.032616356279412e-07,
+      "loss": 0.0462,
+      "step": 9631
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.20321658335036927,
+      "learning_rate": 9.024794645733543e-07,
+      "loss": 0.0421,
+      "step": 9632
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.33960720971075176,
+      "learning_rate": 9.016975987259647e-07,
+      "loss": 0.0838,
+      "step": 9633
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.21702327514100297,
+      "learning_rate": 9.009160381440141e-07,
+      "loss": 0.0601,
+      "step": 9634
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.34677334025746076,
+      "learning_rate": 9.00134782885716e-07,
+      "loss": 0.0777,
+      "step": 9635
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3129532603409501,
+      "learning_rate": 8.993538330092627e-07,
+      "loss": 0.1071,
+      "step": 9636
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3728818018838984,
+      "learning_rate": 8.985731885728221e-07,
+      "loss": 0.0772,
+      "step": 9637
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6312124071541648,
+      "learning_rate": 8.97792849634545e-07,
+      "loss": 0.1224,
+      "step": 9638
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3024944979026759,
+      "learning_rate": 8.970128162525532e-07,
+      "loss": 0.0592,
+      "step": 9639
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2905629021377918,
+      "learning_rate": 8.962330884849485e-07,
+      "loss": 0.0723,
+      "step": 9640
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3177480164517761,
+      "learning_rate": 8.954536663898084e-07,
+      "loss": 0.0724,
+      "step": 9641
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.27600637740816836,
+      "learning_rate": 8.946745500251913e-07,
+      "loss": 0.0629,
+      "step": 9642
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.25054918317991387,
+      "learning_rate": 8.938957394491293e-07,
+      "loss": 0.0713,
+      "step": 9643
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.28250026721585486,
+      "learning_rate": 8.931172347196321e-07,
+      "loss": 0.0901,
+      "step": 9644
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.26270272271383177,
+      "learning_rate": 8.923390358946876e-07,
+      "loss": 0.0458,
+      "step": 9645
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5098695170702431,
+      "learning_rate": 8.915611430322591e-07,
+      "loss": 0.0979,
+      "step": 9646
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.31923681846017427,
+      "learning_rate": 8.907835561902916e-07,
+      "loss": 0.0709,
+      "step": 9647
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.34218251105136327,
+      "learning_rate": 8.900062754267031e-07,
+      "loss": 0.0542,
+      "step": 9648
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.41589290847448607,
+      "learning_rate": 8.892293007993891e-07,
+      "loss": 0.096,
+      "step": 9649
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.4092742837999429,
+      "learning_rate": 8.884526323662229e-07,
+      "loss": 0.0756,
+      "step": 9650
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2707721423005706,
+      "learning_rate": 8.876762701850572e-07,
+      "loss": 0.0564,
+      "step": 9651
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3601308893178154,
+      "learning_rate": 8.869002143137196e-07,
+      "loss": 0.0636,
+      "step": 9652
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3615333984680277,
+      "learning_rate": 8.861244648100126e-07,
+      "loss": 0.0842,
+      "step": 9653
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2408710141193774,
+      "learning_rate": 8.853490217317223e-07,
+      "loss": 0.0721,
+      "step": 9654
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.48325883789266844,
+      "learning_rate": 8.845738851366059e-07,
+      "loss": 0.1152,
+      "step": 9655
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2105906020620536,
+      "learning_rate": 8.83799055082401e-07,
+      "loss": 0.0506,
+      "step": 9656
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0781274809704673,
+      "learning_rate": 8.830245316268198e-07,
+      "loss": 0.1456,
+      "step": 9657
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2580514923924286,
+      "learning_rate": 8.822503148275562e-07,
+      "loss": 0.0539,
+      "step": 9658
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.39631906910048026,
+      "learning_rate": 8.814764047422769e-07,
+      "loss": 0.0977,
+      "step": 9659
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.4855048831608224,
+      "learning_rate": 8.807028014286257e-07,
+      "loss": 0.1217,
+      "step": 9660
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2794252456682048,
+      "learning_rate": 8.799295049442275e-07,
+      "loss": 0.064,
+      "step": 9661
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.38782355425491977,
+      "learning_rate": 8.791565153466807e-07,
+      "loss": 0.0965,
+      "step": 9662
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2740909304583038,
+      "learning_rate": 8.783838326935617e-07,
+      "loss": 0.0786,
+      "step": 9663
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3865933825942271,
+      "learning_rate": 8.77611457042426e-07,
+      "loss": 0.0912,
+      "step": 9664
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2947881885094054,
+      "learning_rate": 8.768393884508042e-07,
+      "loss": 0.0796,
+      "step": 9665
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.4047619633106657,
+      "learning_rate": 8.760676269762031e-07,
+      "loss": 0.1295,
+      "step": 9666
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5082922812068742,
+      "learning_rate": 8.752961726761084e-07,
+      "loss": 0.1151,
+      "step": 9667
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3597204793766933,
+      "learning_rate": 8.745250256079835e-07,
+      "loss": 0.0664,
+      "step": 9668
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.32292411416653366,
+      "learning_rate": 8.737541858292675e-07,
+      "loss": 0.1,
+      "step": 9669
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3590337515021964,
+      "learning_rate": 8.729836533973757e-07,
+      "loss": 0.094,
+      "step": 9670
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3800466392767771,
+      "learning_rate": 8.722134283697048e-07,
+      "loss": 0.0994,
+      "step": 9671
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.25841107649520534,
+      "learning_rate": 8.714435108036235e-07,
+      "loss": 0.0568,
+      "step": 9672
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.4184436492346018,
+      "learning_rate": 8.706739007564796e-07,
+      "loss": 0.1136,
+      "step": 9673
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.45465173069109066,
+      "learning_rate": 8.69904598285598e-07,
+      "loss": 0.0447,
+      "step": 9674
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.20034083049690538,
+      "learning_rate": 8.691356034482828e-07,
+      "loss": 0.059,
+      "step": 9675
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.35598775528681215,
+      "learning_rate": 8.683669163018116e-07,
+      "loss": 0.0667,
+      "step": 9676
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2120594666942776,
+      "learning_rate": 8.675985369034401e-07,
+      "loss": 0.0767,
+      "step": 9677
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.26480759831642103,
+      "learning_rate": 8.668304653104037e-07,
+      "loss": 0.0903,
+      "step": 9678
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3187643372457018,
+      "learning_rate": 8.660627015799116e-07,
+      "loss": 0.0904,
+      "step": 9679
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.29471329751798747,
+      "learning_rate": 8.652952457691505e-07,
+      "loss": 0.0729,
+      "step": 9680
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.44756574075472416,
+      "learning_rate": 8.645280979352871e-07,
+      "loss": 0.0971,
+      "step": 9681
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2583459264812097,
+      "learning_rate": 8.637612581354615e-07,
+      "loss": 0.0568,
+      "step": 9682
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3131310293250757,
+      "learning_rate": 8.629947264267935e-07,
+      "loss": 0.0754,
+      "step": 9683
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.24431402058383406,
+      "learning_rate": 8.622285028663762e-07,
+      "loss": 0.058,
+      "step": 9684
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2477901129582285,
+      "learning_rate": 8.614625875112859e-07,
+      "loss": 0.0617,
+      "step": 9685
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2780079542352381,
+      "learning_rate": 8.606969804185706e-07,
+      "loss": 0.0882,
+      "step": 9686
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.24351735002130515,
+      "learning_rate": 8.599316816452563e-07,
+      "loss": 0.0745,
+      "step": 9687
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3022830131179911,
+      "learning_rate": 8.591666912483493e-07,
+      "loss": 0.0844,
+      "step": 9688
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2783109160581392,
+      "learning_rate": 8.584020092848289e-07,
+      "loss": 0.0593,
+      "step": 9689
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.27278334289045836,
+      "learning_rate": 8.57637635811654e-07,
+      "loss": 0.0727,
+      "step": 9690
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6362226072160814,
+      "learning_rate": 8.568735708857573e-07,
+      "loss": 0.1543,
+      "step": 9691
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.22384616884927755,
+      "learning_rate": 8.561098145640546e-07,
+      "loss": 0.0672,
+      "step": 9692
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5007086434591281,
+      "learning_rate": 8.553463669034317e-07,
+      "loss": 0.11,
+      "step": 9693
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2319049436059855,
+      "learning_rate": 8.545832279607552e-07,
+      "loss": 0.0724,
+      "step": 9694
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.48532287309033173,
+      "learning_rate": 8.538203977928699e-07,
+      "loss": 0.0996,
+      "step": 9695
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2548642543218809,
+      "learning_rate": 8.53057876456595e-07,
+      "loss": 0.0686,
+      "step": 9696
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5413586283716167,
+      "learning_rate": 8.522956640087254e-07,
+      "loss": 0.1078,
+      "step": 9697
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2726265262432998,
+      "learning_rate": 8.515337605060386e-07,
+      "loss": 0.0845,
+      "step": 9698
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.27484376516497994,
+      "learning_rate": 8.507721660052837e-07,
+      "loss": 0.0883,
+      "step": 9699
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5342369722961724,
+      "learning_rate": 8.500108805631885e-07,
+      "loss": 0.1306,
+      "step": 9700
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.22049548489486223,
+      "learning_rate": 8.492499042364578e-07,
+      "loss": 0.0579,
+      "step": 9701
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5504009026710616,
+      "learning_rate": 8.484892370817749e-07,
+      "loss": 0.1113,
+      "step": 9702
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.21938906435708538,
+      "learning_rate": 8.477288791557986e-07,
+      "loss": 0.0654,
+      "step": 9703
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.4846058905822086,
+      "learning_rate": 8.469688305151635e-07,
+      "loss": 0.1148,
+      "step": 9704
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.28037140948560935,
+      "learning_rate": 8.462090912164822e-07,
+      "loss": 0.0813,
+      "step": 9705
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5062864225993273,
+      "learning_rate": 8.454496613163465e-07,
+      "loss": 0.1096,
+      "step": 9706
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.4932077781360014,
+      "learning_rate": 8.446905408713219e-07,
+      "loss": 0.1038,
+      "step": 9707
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.39720690186103114,
+      "learning_rate": 8.439317299379524e-07,
+      "loss": 0.0952,
+      "step": 9708
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6527065422606126,
+      "learning_rate": 8.431732285727579e-07,
+      "loss": 0.1233,
+      "step": 9709
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3527215027975017,
+      "learning_rate": 8.424150368322371e-07,
+      "loss": 0.0934,
+      "step": 9710
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2517436879589864,
+      "learning_rate": 8.416571547728641e-07,
+      "loss": 0.0596,
+      "step": 9711
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.28916597944144945,
+      "learning_rate": 8.408995824510907e-07,
+      "loss": 0.0753,
+      "step": 9712
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.23550095085639877,
+      "learning_rate": 8.401423199233449e-07,
+      "loss": 0.0788,
+      "step": 9713
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3846497286855851,
+      "learning_rate": 8.393853672460306e-07,
+      "loss": 0.1013,
+      "step": 9714
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3449298251335863,
+      "learning_rate": 8.386287244755331e-07,
+      "loss": 0.0828,
+      "step": 9715
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2957821328828303,
+      "learning_rate": 8.378723916682096e-07,
+      "loss": 0.0518,
+      "step": 9716
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.4045435212096391,
+      "learning_rate": 8.371163688803968e-07,
+      "loss": 0.1151,
+      "step": 9717
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.21285318599741368,
+      "learning_rate": 8.363606561684062e-07,
+      "loss": 0.0651,
+      "step": 9718
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.4037082978472771,
+      "learning_rate": 8.356052535885307e-07,
+      "loss": 0.0841,
+      "step": 9719
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.402477557504167,
+      "learning_rate": 8.348501611970345e-07,
+      "loss": 0.0704,
+      "step": 9720
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.29674799902788374,
+      "learning_rate": 8.34095379050161e-07,
+      "loss": 0.0922,
+      "step": 9721
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2524671372615894,
+      "learning_rate": 8.333409072041332e-07,
+      "loss": 0.0532,
+      "step": 9722
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.35078220262974263,
+      "learning_rate": 8.325867457151471e-07,
+      "loss": 0.0736,
+      "step": 9723
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2892492010107794,
+      "learning_rate": 8.318328946393772e-07,
+      "loss": 0.0667,
+      "step": 9724
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3746811910499197,
+      "learning_rate": 8.310793540329737e-07,
+      "loss": 0.0821,
+      "step": 9725
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.4362615617310372,
+      "learning_rate": 8.303261239520665e-07,
+      "loss": 0.0819,
+      "step": 9726
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.24611361178384084,
+      "learning_rate": 8.295732044527599e-07,
+      "loss": 0.064,
+      "step": 9727
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2986013022063484,
+      "learning_rate": 8.288205955911344e-07,
+      "loss": 0.0407,
+      "step": 9728
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2784203450073894,
+      "learning_rate": 8.280682974232512e-07,
+      "loss": 0.0459,
+      "step": 9729
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.35718324062087736,
+      "learning_rate": 8.273163100051445e-07,
+      "loss": 0.1046,
+      "step": 9730
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.36596675772184795,
+      "learning_rate": 8.265646333928251e-07,
+      "loss": 0.0826,
+      "step": 9731
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.30634599996325323,
+      "learning_rate": 8.258132676422853e-07,
+      "loss": 0.0757,
+      "step": 9732
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2176533675054571,
+      "learning_rate": 8.250622128094899e-07,
+      "loss": 0.0395,
+      "step": 9733
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.31737432489957107,
+      "learning_rate": 8.243114689503817e-07,
+      "loss": 0.086,
+      "step": 9734
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.30669122203703475,
+      "learning_rate": 8.23561036120879e-07,
+      "loss": 0.0854,
+      "step": 9735
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2679540839682881,
+      "learning_rate": 8.228109143768815e-07,
+      "loss": 0.0942,
+      "step": 9736
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.29047022727739147,
+      "learning_rate": 8.220611037742604e-07,
+      "loss": 0.0619,
+      "step": 9737
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6248323078678272,
+      "learning_rate": 8.213116043688657e-07,
+      "loss": 0.094,
+      "step": 9738
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.22436947384991715,
+      "learning_rate": 8.205624162165271e-07,
+      "loss": 0.0794,
+      "step": 9739
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3052104152800874,
+      "learning_rate": 8.198135393730461e-07,
+      "loss": 0.0767,
+      "step": 9740
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.21375665884849118,
+      "learning_rate": 8.190649738942041e-07,
+      "loss": 0.0801,
+      "step": 9741
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5298953966148979,
+      "learning_rate": 8.183167198357578e-07,
+      "loss": 0.0812,
+      "step": 9742
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.31951539857083106,
+      "learning_rate": 8.17568777253443e-07,
+      "loss": 0.0606,
+      "step": 9743
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.35860945363601293,
+      "learning_rate": 8.168211462029707e-07,
+      "loss": 0.0822,
+      "step": 9744
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.20310627136584156,
+      "learning_rate": 8.160738267400265e-07,
+      "loss": 0.0372,
+      "step": 9745
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.34219474493738816,
+      "learning_rate": 8.153268189202785e-07,
+      "loss": 0.1077,
+      "step": 9746
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.24892287379396272,
+      "learning_rate": 8.145801227993661e-07,
+      "loss": 0.0434,
+      "step": 9747
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2989440114665037,
+      "learning_rate": 8.138337384329087e-07,
+      "loss": 0.0569,
+      "step": 9748
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3841773142884086,
+      "learning_rate": 8.130876658764986e-07,
+      "loss": 0.1023,
+      "step": 9749
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.4970036268749609,
+      "learning_rate": 8.123419051857118e-07,
+      "loss": 0.0982,
+      "step": 9750
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3378261401661527,
+      "learning_rate": 8.115964564160944e-07,
+      "loss": 0.1019,
+      "step": 9751
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.4060210129193685,
+      "learning_rate": 8.10851319623171e-07,
+      "loss": 0.0815,
+      "step": 9752
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.22267634863716657,
+      "learning_rate": 8.10106494862446e-07,
+      "loss": 0.0632,
+      "step": 9753
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3521131553401419,
+      "learning_rate": 8.093619821893972e-07,
+      "loss": 0.0934,
+      "step": 9754
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5570641999382414,
+      "learning_rate": 8.086177816594792e-07,
+      "loss": 0.14,
+      "step": 9755
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.30863703812741766,
+      "learning_rate": 8.078738933281266e-07,
+      "loss": 0.0919,
+      "step": 9756
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2969089564892958,
+      "learning_rate": 8.071303172507472e-07,
+      "loss": 0.0754,
+      "step": 9757
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5727093831307662,
+      "learning_rate": 8.063870534827273e-07,
+      "loss": 0.1493,
+      "step": 9758
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5449620745945319,
+      "learning_rate": 8.056441020794275e-07,
+      "loss": 0.1182,
+      "step": 9759
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2351460603334859,
+      "learning_rate": 8.049014630961905e-07,
+      "loss": 0.0371,
+      "step": 9760
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.32908295144057753,
+      "learning_rate": 8.041591365883312e-07,
+      "loss": 0.119,
+      "step": 9761
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.32139705546893066,
+      "learning_rate": 8.034171226111404e-07,
+      "loss": 0.0875,
+      "step": 9762
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.39166941152477774,
+      "learning_rate": 8.0267542121989e-07,
+      "loss": 0.1213,
+      "step": 9763
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5944846542499271,
+      "learning_rate": 8.019340324698261e-07,
+      "loss": 0.085,
+      "step": 9764
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2622798464119927,
+      "learning_rate": 8.011929564161708e-07,
+      "loss": 0.0816,
+      "step": 9765
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.22659022794020647,
+      "learning_rate": 8.004521931141223e-07,
+      "loss": 0.0566,
+      "step": 9766
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.23989737438444092,
+      "learning_rate": 7.997117426188606e-07,
+      "loss": 0.051,
+      "step": 9767
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3828085526897713,
+      "learning_rate": 7.989716049855362e-07,
+      "loss": 0.1027,
+      "step": 9768
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.22102970608373645,
+      "learning_rate": 7.982317802692785e-07,
+      "loss": 0.0315,
+      "step": 9769
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.45992663464078987,
+      "learning_rate": 7.974922685251962e-07,
+      "loss": 0.0965,
+      "step": 9770
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3044792386492315,
+      "learning_rate": 7.967530698083715e-07,
+      "loss": 0.0566,
+      "step": 9771
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.21138835304350942,
+      "learning_rate": 7.960141841738633e-07,
+      "loss": 0.0492,
+      "step": 9772
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6716361210464431,
+      "learning_rate": 7.952756116767074e-07,
+      "loss": 0.1559,
+      "step": 9773
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.21035813725060956,
+      "learning_rate": 7.945373523719196e-07,
+      "loss": 0.0524,
+      "step": 9774
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.21383125815789836,
+      "learning_rate": 7.937994063144888e-07,
+      "loss": 0.0411,
+      "step": 9775
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3068827279937841,
+      "learning_rate": 7.93061773559381e-07,
+      "loss": 0.0944,
+      "step": 9776
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2448350614156476,
+      "learning_rate": 7.923244541615383e-07,
+      "loss": 0.0784,
+      "step": 9777
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.41737729715777283,
+      "learning_rate": 7.915874481758829e-07,
+      "loss": 0.1093,
+      "step": 9778
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3533633694578978,
+      "learning_rate": 7.908507556573103e-07,
+      "loss": 0.0727,
+      "step": 9779
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.30464384816873197,
+      "learning_rate": 7.901143766606933e-07,
+      "loss": 0.1024,
+      "step": 9780
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3386405752903696,
+      "learning_rate": 7.893783112408821e-07,
+      "loss": 0.0966,
+      "step": 9781
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.28067305965953726,
+      "learning_rate": 7.886425594527014e-07,
+      "loss": 0.0716,
+      "step": 9782
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.31493557318688786,
+      "learning_rate": 7.879071213509576e-07,
+      "loss": 0.0696,
+      "step": 9783
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.31860440494469616,
+      "learning_rate": 7.871719969904284e-07,
+      "loss": 0.061,
+      "step": 9784
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.36253219616062227,
+      "learning_rate": 7.864371864258702e-07,
+      "loss": 0.0458,
+      "step": 9785
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2659507727252553,
+      "learning_rate": 7.857026897120151e-07,
+      "loss": 0.0386,
+      "step": 9786
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.25039686848856685,
+      "learning_rate": 7.84968506903575e-07,
+      "loss": 0.0631,
+      "step": 9787
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2507733380653957,
+      "learning_rate": 7.842346380552351e-07,
+      "loss": 0.0519,
+      "step": 9788
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5127571888203863,
+      "learning_rate": 7.835010832216567e-07,
+      "loss": 0.1039,
+      "step": 9789
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.4018434322536532,
+      "learning_rate": 7.827678424574819e-07,
+      "loss": 0.0787,
+      "step": 9790
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5212942016039955,
+      "learning_rate": 7.820349158173252e-07,
+      "loss": 0.0997,
+      "step": 9791
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.18456911570363677,
+      "learning_rate": 7.813023033557793e-07,
+      "loss": 0.0777,
+      "step": 9792
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.23803338174440045,
+      "learning_rate": 7.805700051274123e-07,
+      "loss": 0.054,
+      "step": 9793
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.23918196853071208,
+      "learning_rate": 7.798380211867729e-07,
+      "loss": 0.0566,
+      "step": 9794
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.2736766113941927,
+      "learning_rate": 7.791063515883817e-07,
+      "loss": 0.0999,
+      "step": 9795
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.2720499951863573,
+      "learning_rate": 7.783749963867366e-07,
+      "loss": 0.0712,
+      "step": 9796
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.313492814100594,
+      "learning_rate": 7.776439556363158e-07,
+      "loss": 0.0783,
+      "step": 9797
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.4635126441496074,
+      "learning_rate": 7.769132293915705e-07,
+      "loss": 0.0977,
+      "step": 9798
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5340604764904753,
+      "learning_rate": 7.761828177069292e-07,
+      "loss": 0.1237,
+      "step": 9799
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.30593321770160103,
+      "learning_rate": 7.754527206367957e-07,
+      "loss": 0.0683,
+      "step": 9800
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.40228543747457834,
+      "learning_rate": 7.747229382355547e-07,
+      "loss": 0.0699,
+      "step": 9801
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.41585589051746025,
+      "learning_rate": 7.739934705575636e-07,
+      "loss": 0.0817,
+      "step": 9802
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.2988102249065971,
+      "learning_rate": 7.732643176571553e-07,
+      "loss": 0.0688,
+      "step": 9803
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3286184791322606,
+      "learning_rate": 7.725354795886448e-07,
+      "loss": 0.0757,
+      "step": 9804
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.301890543929499,
+      "learning_rate": 7.718069564063185e-07,
+      "loss": 0.086,
+      "step": 9805
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.1906780526197197,
+      "learning_rate": 7.710787481644399e-07,
+      "loss": 0.035,
+      "step": 9806
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.39161367682091747,
+      "learning_rate": 7.703508549172528e-07,
+      "loss": 0.0807,
+      "step": 9807
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.25080202890267445,
+      "learning_rate": 7.696232767189732e-07,
+      "loss": 0.0416,
+      "step": 9808
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.27709767937567886,
+      "learning_rate": 7.68896013623796e-07,
+      "loss": 0.0632,
+      "step": 9809
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.41341307646340014,
+      "learning_rate": 7.681690656858904e-07,
+      "loss": 0.1034,
+      "step": 9810
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5104878095236919,
+      "learning_rate": 7.674424329594066e-07,
+      "loss": 0.1363,
+      "step": 9811
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.26515473658048555,
+      "learning_rate": 7.667161154984665e-07,
+      "loss": 0.0652,
+      "step": 9812
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3805648670636288,
+      "learning_rate": 7.659901133571695e-07,
+      "loss": 0.0869,
+      "step": 9813
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.2675557247717521,
+      "learning_rate": 7.652644265895953e-07,
+      "loss": 0.0663,
+      "step": 9814
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5210774658131873,
+      "learning_rate": 7.645390552497956e-07,
+      "loss": 0.1031,
+      "step": 9815
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.30290395059655223,
+      "learning_rate": 7.638139993918003e-07,
+      "loss": 0.083,
+      "step": 9816
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3237611713567306,
+      "learning_rate": 7.630892590696148e-07,
+      "loss": 0.1044,
+      "step": 9817
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.21539807141489517,
+      "learning_rate": 7.623648343372242e-07,
+      "loss": 0.0444,
+      "step": 9818
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.40962766860635247,
+      "learning_rate": 7.616407252485874e-07,
+      "loss": 0.1135,
+      "step": 9819
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.23448944523820134,
+      "learning_rate": 7.609169318576376e-07,
+      "loss": 0.0623,
+      "step": 9820
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.43859832638155555,
+      "learning_rate": 7.601934542182909e-07,
+      "loss": 0.0786,
+      "step": 9821
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3681232908285159,
+      "learning_rate": 7.594702923844344e-07,
+      "loss": 0.0893,
+      "step": 9822
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3566704486693897,
+      "learning_rate": 7.587474464099326e-07,
+      "loss": 0.119,
+      "step": 9823
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.34094763129153904,
+      "learning_rate": 7.580249163486286e-07,
+      "loss": 0.1016,
+      "step": 9824
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.32306606375173197,
+      "learning_rate": 7.573027022543406e-07,
+      "loss": 0.0919,
+      "step": 9825
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3790513005152477,
+      "learning_rate": 7.565808041808625e-07,
+      "loss": 0.0916,
+      "step": 9826
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.4311829312743003,
+      "learning_rate": 7.558592221819649e-07,
+      "loss": 0.1083,
+      "step": 9827
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5437097447163345,
+      "learning_rate": 7.551379563113981e-07,
+      "loss": 0.1096,
+      "step": 9828
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3571798483590183,
+      "learning_rate": 7.544170066228835e-07,
+      "loss": 0.1276,
+      "step": 9829
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.2575823702220296,
+      "learning_rate": 7.536963731701219e-07,
+      "loss": 0.0645,
+      "step": 9830
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.41718105386456633,
+      "learning_rate": 7.529760560067922e-07,
+      "loss": 0.086,
+      "step": 9831
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.29223269215470365,
+      "learning_rate": 7.522560551865465e-07,
+      "loss": 0.0719,
+      "step": 9832
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.49086716805133845,
+      "learning_rate": 7.515363707630146e-07,
+      "loss": 0.1481,
+      "step": 9833
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.2943607707451245,
+      "learning_rate": 7.50817002789802e-07,
+      "loss": 0.0788,
+      "step": 9834
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.2859534975837805,
+      "learning_rate": 7.500979513204937e-07,
+      "loss": 0.0709,
+      "step": 9835
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.43894889342180987,
+      "learning_rate": 7.493792164086472e-07,
+      "loss": 0.0913,
+      "step": 9836
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.4162640333569657,
+      "learning_rate": 7.48660798107797e-07,
+      "loss": 0.0671,
+      "step": 9837
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.21052968988779,
+      "learning_rate": 7.479426964714582e-07,
+      "loss": 0.0619,
+      "step": 9838
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.27478217175932546,
+      "learning_rate": 7.472249115531166e-07,
+      "loss": 0.1014,
+      "step": 9839
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3802474396411029,
+      "learning_rate": 7.465074434062386e-07,
+      "loss": 0.0632,
+      "step": 9840
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.26968163989390265,
+      "learning_rate": 7.45790292084263e-07,
+      "loss": 0.0846,
+      "step": 9841
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.22018793290799799,
+      "learning_rate": 7.450734576406104e-07,
+      "loss": 0.0438,
+      "step": 9842
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.38354668013316706,
+      "learning_rate": 7.443569401286737e-07,
+      "loss": 0.069,
+      "step": 9843
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.1972513442228845,
+      "learning_rate": 7.436407396018225e-07,
+      "loss": 0.0641,
+      "step": 9844
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.22380100172955078,
+      "learning_rate": 7.429248561134034e-07,
+      "loss": 0.0615,
+      "step": 9845
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.2590908847573438,
+      "learning_rate": 7.422092897167416e-07,
+      "loss": 0.0633,
+      "step": 9846
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.18487375380366977,
+      "learning_rate": 7.41494040465135e-07,
+      "loss": 0.0294,
+      "step": 9847
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.17920872061455342,
+      "learning_rate": 7.407791084118598e-07,
+      "loss": 0.0482,
+      "step": 9848
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3486444139550612,
+      "learning_rate": 7.400644936101676e-07,
+      "loss": 0.0726,
+      "step": 9849
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.410032500694567,
+      "learning_rate": 7.393501961132887e-07,
+      "loss": 0.095,
+      "step": 9850
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.45223181720855704,
+      "learning_rate": 7.386362159744275e-07,
+      "loss": 0.1232,
+      "step": 9851
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.38715021314705456,
+      "learning_rate": 7.379225532467654e-07,
+      "loss": 0.0902,
+      "step": 9852
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.19027227517917547,
+      "learning_rate": 7.372092079834597e-07,
+      "loss": 0.0409,
+      "step": 9853
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.2041203105785427,
+      "learning_rate": 7.364961802376435e-07,
+      "loss": 0.0623,
+      "step": 9854
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.364313219372423,
+      "learning_rate": 7.3578347006243e-07,
+      "loss": 0.073,
+      "step": 9855
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3593433607435564,
+      "learning_rate": 7.350710775109043e-07,
+      "loss": 0.0696,
+      "step": 9856
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3380728338896378,
+      "learning_rate": 7.343590026361291e-07,
+      "loss": 0.0732,
+      "step": 9857
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.46046863139736105,
+      "learning_rate": 7.336472454911453e-07,
+      "loss": 0.1318,
+      "step": 9858
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.4047601325960266,
+      "learning_rate": 7.329358061289682e-07,
+      "loss": 0.1155,
+      "step": 9859
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3591799970801823,
+      "learning_rate": 7.322246846025899e-07,
+      "loss": 0.1032,
+      "step": 9860
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.34244510451419896,
+      "learning_rate": 7.315138809649768e-07,
+      "loss": 0.1013,
+      "step": 9861
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.2432773049895317,
+      "learning_rate": 7.308033952690774e-07,
+      "loss": 0.0763,
+      "step": 9862
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3290926710902916,
+      "learning_rate": 7.300932275678113e-07,
+      "loss": 0.0724,
+      "step": 9863
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.19390099621854553,
+      "learning_rate": 7.293833779140741e-07,
+      "loss": 0.0379,
+      "step": 9864
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.4248593085786264,
+      "learning_rate": 7.286738463607423e-07,
+      "loss": 0.0943,
+      "step": 9865
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.2679217027790947,
+      "learning_rate": 7.279646329606648e-07,
+      "loss": 0.0847,
+      "step": 9866
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.22919174370667947,
+      "learning_rate": 7.272557377666678e-07,
+      "loss": 0.0651,
+      "step": 9867
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6873300176874477,
+      "learning_rate": 7.265471608315527e-07,
+      "loss": 0.1084,
+      "step": 9868
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.4720270836099034,
+      "learning_rate": 7.258389022081014e-07,
+      "loss": 0.0951,
+      "step": 9869
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.29445532994096063,
+      "learning_rate": 7.251309619490671e-07,
+      "loss": 0.1028,
+      "step": 9870
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.31028921464432496,
+      "learning_rate": 7.244233401071804e-07,
+      "loss": 0.0809,
+      "step": 9871
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.30929135798891627,
+      "learning_rate": 7.23716036735152e-07,
+      "loss": 0.0802,
+      "step": 9872
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5115663003663974,
+      "learning_rate": 7.230090518856641e-07,
+      "loss": 0.1042,
+      "step": 9873
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3640175786699371,
+      "learning_rate": 7.223023856113776e-07,
+      "loss": 0.0899,
+      "step": 9874
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.2922998888202173,
+      "learning_rate": 7.215960379649273e-07,
+      "loss": 0.0742,
+      "step": 9875
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.25922278889898753,
+      "learning_rate": 7.208900089989291e-07,
+      "loss": 0.0932,
+      "step": 9876
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3158416145902828,
+      "learning_rate": 7.201842987659707e-07,
+      "loss": 0.0909,
+      "step": 9877
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3420751749927588,
+      "learning_rate": 7.194789073186159e-07,
+      "loss": 0.0873,
+      "step": 9878
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.22508741871126006,
+      "learning_rate": 7.187738347094097e-07,
+      "loss": 0.0721,
+      "step": 9879
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.40556355748776063,
+      "learning_rate": 7.180690809908681e-07,
+      "loss": 0.0757,
+      "step": 9880
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.32428246207605577,
+      "learning_rate": 7.173646462154838e-07,
+      "loss": 0.0884,
+      "step": 9881
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.39158887210753257,
+      "learning_rate": 7.166605304357305e-07,
+      "loss": 0.0615,
+      "step": 9882
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.4153514475728676,
+      "learning_rate": 7.159567337040535e-07,
+      "loss": 0.084,
+      "step": 9883
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.39279057423817826,
+      "learning_rate": 7.152532560728748e-07,
+      "loss": 0.0655,
+      "step": 9884
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.34062529008343856,
+      "learning_rate": 7.145500975945929e-07,
+      "loss": 0.1003,
+      "step": 9885
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.40964575679890336,
+      "learning_rate": 7.138472583215855e-07,
+      "loss": 0.1153,
+      "step": 9886
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.4048773126769146,
+      "learning_rate": 7.131447383062034e-07,
+      "loss": 0.0956,
+      "step": 9887
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.35514762466309213,
+      "learning_rate": 7.124425376007726e-07,
+      "loss": 0.0867,
+      "step": 9888
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.29531222056047096,
+      "learning_rate": 7.117406562575995e-07,
+      "loss": 0.0648,
+      "step": 9889
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3588619924644892,
+      "learning_rate": 7.110390943289636e-07,
+      "loss": 0.0937,
+      "step": 9890
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.38244855700186453,
+      "learning_rate": 7.103378518671205e-07,
+      "loss": 0.1186,
+      "step": 9891
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.22513882271751143,
+      "learning_rate": 7.096369289243026e-07,
+      "loss": 0.0489,
+      "step": 9892
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.2971890069441585,
+      "learning_rate": 7.089363255527204e-07,
+      "loss": 0.054,
+      "step": 9893
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.14545636220958127,
+      "learning_rate": 7.082360418045581e-07,
+      "loss": 0.0369,
+      "step": 9894
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5198452838122398,
+      "learning_rate": 7.075360777319756e-07,
+      "loss": 0.1027,
+      "step": 9895
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.18100923321806167,
+      "learning_rate": 7.068364333871125e-07,
+      "loss": 0.0344,
+      "step": 9896
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.30341355434640155,
+      "learning_rate": 7.061371088220814e-07,
+      "loss": 0.0595,
+      "step": 9897
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.40472857429239506,
+      "learning_rate": 7.054381040889713e-07,
+      "loss": 0.0802,
+      "step": 9898
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.28743750799965895,
+      "learning_rate": 7.047394192398493e-07,
+      "loss": 0.0617,
+      "step": 9899
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5333154439294154,
+      "learning_rate": 7.040410543267573e-07,
+      "loss": 0.1103,
+      "step": 9900
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3613073134694714,
+      "learning_rate": 7.033430094017135e-07,
+      "loss": 0.0841,
+      "step": 9901
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3274965841964515,
+      "learning_rate": 7.026452845167115e-07,
+      "loss": 0.0656,
+      "step": 9902
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3965812791183081,
+      "learning_rate": 7.019478797237233e-07,
+      "loss": 0.0857,
+      "step": 9903
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.2068028937966365,
+      "learning_rate": 7.012507950746949e-07,
+      "loss": 0.0481,
+      "step": 9904
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.25877586207386494,
+      "learning_rate": 7.005540306215486e-07,
+      "loss": 0.0683,
+      "step": 9905
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3144415533135142,
+      "learning_rate": 6.998575864161855e-07,
+      "loss": 0.0703,
+      "step": 9906
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.33881020182870164,
+      "learning_rate": 6.991614625104792e-07,
+      "loss": 0.1008,
+      "step": 9907
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.38532765669259733,
+      "learning_rate": 6.984656589562816e-07,
+      "loss": 0.1182,
+      "step": 9908
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6137803620133151,
+      "learning_rate": 6.977701758054184e-07,
+      "loss": 0.0685,
+      "step": 9909
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3028415874968051,
+      "learning_rate": 6.970750131096965e-07,
+      "loss": 0.1129,
+      "step": 9910
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.35994530508844846,
+      "learning_rate": 6.96380170920894e-07,
+      "loss": 0.1149,
+      "step": 9911
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4367432821588394,
+      "learning_rate": 6.956856492907665e-07,
+      "loss": 0.0937,
+      "step": 9912
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3450504373792145,
+      "learning_rate": 6.949914482710452e-07,
+      "loss": 0.0828,
+      "step": 9913
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0803155264973436,
+      "learning_rate": 6.942975679134406e-07,
+      "loss": 0.1037,
+      "step": 9914
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.30500559154768886,
+      "learning_rate": 6.93604008269636e-07,
+      "loss": 0.0691,
+      "step": 9915
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.36903167958889754,
+      "learning_rate": 6.929107693912912e-07,
+      "loss": 0.0739,
+      "step": 9916
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.24279262336888582,
+      "learning_rate": 6.92217851330042e-07,
+      "loss": 0.0592,
+      "step": 9917
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.31138986809214103,
+      "learning_rate": 6.915252541375029e-07,
+      "loss": 0.0856,
+      "step": 9918
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3204645184146256,
+      "learning_rate": 6.908329778652617e-07,
+      "loss": 0.1103,
+      "step": 9919
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5590846089069286,
+      "learning_rate": 6.901410225648825e-07,
+      "loss": 0.119,
+      "step": 9920
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3962538260118628,
+      "learning_rate": 6.894493882879072e-07,
+      "loss": 0.111,
+      "step": 9921
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8932587199170292,
+      "learning_rate": 6.887580750858514e-07,
+      "loss": 0.1055,
+      "step": 9922
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6734518554147418,
+      "learning_rate": 6.880670830102099e-07,
+      "loss": 0.1393,
+      "step": 9923
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.38651345904532425,
+      "learning_rate": 6.87376412112451e-07,
+      "loss": 0.0869,
+      "step": 9924
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.23371898323045504,
+      "learning_rate": 6.866860624440197e-07,
+      "loss": 0.0672,
+      "step": 9925
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3608194670820361,
+      "learning_rate": 6.859960340563359e-07,
+      "loss": 0.075,
+      "step": 9926
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.21348011690748372,
+      "learning_rate": 6.853063270007998e-07,
+      "loss": 0.0654,
+      "step": 9927
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5428174568424596,
+      "learning_rate": 6.84616941328784e-07,
+      "loss": 0.127,
+      "step": 9928
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4149427480287126,
+      "learning_rate": 6.839278770916358e-07,
+      "loss": 0.0945,
+      "step": 9929
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4890960349298005,
+      "learning_rate": 6.83239134340683e-07,
+      "loss": 0.0864,
+      "step": 9930
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3484796547155267,
+      "learning_rate": 6.825507131272269e-07,
+      "loss": 0.1159,
+      "step": 9931
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2468165205071676,
+      "learning_rate": 6.818626135025436e-07,
+      "loss": 0.0614,
+      "step": 9932
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2932159993254829,
+      "learning_rate": 6.811748355178888e-07,
+      "loss": 0.0593,
+      "step": 9933
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.17937781591431232,
+      "learning_rate": 6.804873792244915e-07,
+      "loss": 0.0551,
+      "step": 9934
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3473795470819226,
+      "learning_rate": 6.798002446735569e-07,
+      "loss": 0.1106,
+      "step": 9935
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2613062861364457,
+      "learning_rate": 6.791134319162662e-07,
+      "loss": 0.0489,
+      "step": 9936
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.32659553785753825,
+      "learning_rate": 6.784269410037792e-07,
+      "loss": 0.1134,
+      "step": 9937
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5194624281863731,
+      "learning_rate": 6.777407719872286e-07,
+      "loss": 0.1138,
+      "step": 9938
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.35260135765109557,
+      "learning_rate": 6.770549249177233e-07,
+      "loss": 0.101,
+      "step": 9939
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.21181658763747813,
+      "learning_rate": 6.76369399846351e-07,
+      "loss": 0.0565,
+      "step": 9940
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.33144567198397046,
+      "learning_rate": 6.756841968241734e-07,
+      "loss": 0.085,
+      "step": 9941
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.1789964089003555,
+      "learning_rate": 6.749993159022273e-07,
+      "loss": 0.0554,
+      "step": 9942
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.20497789437728106,
+      "learning_rate": 6.743147571315261e-07,
+      "loss": 0.0464,
+      "step": 9943
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3536453341132406,
+      "learning_rate": 6.736305205630623e-07,
+      "loss": 0.096,
+      "step": 9944
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2700831444384398,
+      "learning_rate": 6.729466062478001e-07,
+      "loss": 0.0761,
+      "step": 9945
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2351775697950334,
+      "learning_rate": 6.722630142366804e-07,
+      "loss": 0.0813,
+      "step": 9946
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.29596166157228354,
+      "learning_rate": 6.715797445806233e-07,
+      "loss": 0.0901,
+      "step": 9947
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.28979324368630516,
+      "learning_rate": 6.708967973305219e-07,
+      "loss": 0.1037,
+      "step": 9948
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3977331231328276,
+      "learning_rate": 6.702141725372446e-07,
+      "loss": 0.097,
+      "step": 9949
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2682830549857594,
+      "learning_rate": 6.695318702516401e-07,
+      "loss": 0.0634,
+      "step": 9950
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.267912643128008,
+      "learning_rate": 6.688498905245288e-07,
+      "loss": 0.0763,
+      "step": 9951
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2878486540449759,
+      "learning_rate": 6.68168233406708e-07,
+      "loss": 0.0794,
+      "step": 9952
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.39476888375283403,
+      "learning_rate": 6.674868989489513e-07,
+      "loss": 0.0945,
+      "step": 9953
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4796810571018395,
+      "learning_rate": 6.6680588720201e-07,
+      "loss": 0.0881,
+      "step": 9954
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3891566341049095,
+      "learning_rate": 6.66125198216609e-07,
+      "loss": 0.0946,
+      "step": 9955
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3538794755193433,
+      "learning_rate": 6.654448320434487e-07,
+      "loss": 0.0962,
+      "step": 9956
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3407869839406074,
+      "learning_rate": 6.647647887332092e-07,
+      "loss": 0.0651,
+      "step": 9957
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4724830041498625,
+      "learning_rate": 6.640850683365424e-07,
+      "loss": 0.0924,
+      "step": 9958
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2988038615242455,
+      "learning_rate": 6.634056709040787e-07,
+      "loss": 0.0664,
+      "step": 9959
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2786867952125467,
+      "learning_rate": 6.627265964864222e-07,
+      "loss": 0.0869,
+      "step": 9960
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.338381956720589,
+      "learning_rate": 6.620478451341561e-07,
+      "loss": 0.0796,
+      "step": 9961
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.37141617871926136,
+      "learning_rate": 6.613694168978374e-07,
+      "loss": 0.0991,
+      "step": 9962
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3699427130549968,
+      "learning_rate": 6.606913118279973e-07,
+      "loss": 0.1045,
+      "step": 9963
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.306013040079827,
+      "learning_rate": 6.600135299751481e-07,
+      "loss": 0.0613,
+      "step": 9964
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2752507605540219,
+      "learning_rate": 6.593360713897729e-07,
+      "loss": 0.084,
+      "step": 9965
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2955830427664039,
+      "learning_rate": 6.586589361223328e-07,
+      "loss": 0.0765,
+      "step": 9966
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.24715143348720378,
+      "learning_rate": 6.579821242232665e-07,
+      "loss": 0.0793,
+      "step": 9967
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3220376085684528,
+      "learning_rate": 6.573056357429853e-07,
+      "loss": 0.1095,
+      "step": 9968
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5265670743370436,
+      "learning_rate": 6.566294707318782e-07,
+      "loss": 0.1246,
+      "step": 9969
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.41214130609218913,
+      "learning_rate": 6.559536292403096e-07,
+      "loss": 0.1358,
+      "step": 9970
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3591485915869745,
+      "learning_rate": 6.552781113186213e-07,
+      "loss": 0.0485,
+      "step": 9971
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.27695386441168834,
+      "learning_rate": 6.546029170171297e-07,
+      "loss": 0.0729,
+      "step": 9972
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9920329242533058,
+      "learning_rate": 6.539280463861252e-07,
+      "loss": 0.1357,
+      "step": 9973
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3378754690999391,
+      "learning_rate": 6.532534994758788e-07,
+      "loss": 0.1101,
+      "step": 9974
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4077153172114649,
+      "learning_rate": 6.525792763366329e-07,
+      "loss": 0.0736,
+      "step": 9975
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.43715877417675025,
+      "learning_rate": 6.519053770186084e-07,
+      "loss": 0.1074,
+      "step": 9976
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2808069240758969,
+      "learning_rate": 6.512318015720004e-07,
+      "loss": 0.0811,
+      "step": 9977
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.23595589614730264,
+      "learning_rate": 6.505585500469819e-07,
+      "loss": 0.0611,
+      "step": 9978
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.25544977764727195,
+      "learning_rate": 6.498856224937e-07,
+      "loss": 0.0678,
+      "step": 9979
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3036858745935714,
+      "learning_rate": 6.492130189622781e-07,
+      "loss": 0.0954,
+      "step": 9980
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3969699432531286,
+      "learning_rate": 6.485407395028148e-07,
+      "loss": 0.0674,
+      "step": 9981
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.844057351929389,
+      "learning_rate": 6.478687841653874e-07,
+      "loss": 0.1174,
+      "step": 9982
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.18800533116579318,
+      "learning_rate": 6.471971530000459e-07,
+      "loss": 0.0468,
+      "step": 9983
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.20522865614953167,
+      "learning_rate": 6.465258460568174e-07,
+      "loss": 0.0559,
+      "step": 9984
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.25691264030770145,
+      "learning_rate": 6.458548633857036e-07,
+      "loss": 0.0664,
+      "step": 9985
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.21670358768985237,
+      "learning_rate": 6.451842050366858e-07,
+      "loss": 0.0456,
+      "step": 9986
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.324863784374698,
+      "learning_rate": 6.445138710597165e-07,
+      "loss": 0.0991,
+      "step": 9987
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3516150256372846,
+      "learning_rate": 6.438438615047271e-07,
+      "loss": 0.1219,
+      "step": 9988
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.16975888050080762,
+      "learning_rate": 6.431741764216237e-07,
+      "loss": 0.06,
+      "step": 9989
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4161984267397283,
+      "learning_rate": 6.425048158602865e-07,
+      "loss": 0.0945,
+      "step": 9990
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4043672873691549,
+      "learning_rate": 6.41835779870576e-07,
+      "loss": 0.0849,
+      "step": 9991
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.46843555497686135,
+      "learning_rate": 6.411670685023247e-07,
+      "loss": 0.0666,
+      "step": 9992
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.26065090527405504,
+      "learning_rate": 6.404986818053421e-07,
+      "loss": 0.0712,
+      "step": 9993
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.37799962792487357,
+      "learning_rate": 6.39830619829413e-07,
+      "loss": 0.0633,
+      "step": 9994
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.33301947051005987,
+      "learning_rate": 6.391628826243001e-07,
+      "loss": 0.0958,
+      "step": 9995
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3574224160882498,
+      "learning_rate": 6.384954702397394e-07,
+      "loss": 0.0831,
+      "step": 9996
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.35024048521073536,
+      "learning_rate": 6.378283827254427e-07,
+      "loss": 0.0725,
+      "step": 9997
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.31861449675038783,
+      "learning_rate": 6.371616201311009e-07,
+      "loss": 0.0857,
+      "step": 9998
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.30967611099697434,
+      "learning_rate": 6.364951825063765e-07,
+      "loss": 0.0723,
+      "step": 9999
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.35809143861575465,
+      "learning_rate": 6.358290699009106e-07,
+      "loss": 0.1042,
+      "step": 10000
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.40626894229806215,
+      "learning_rate": 6.351632823643178e-07,
+      "loss": 0.077,
+      "step": 10001
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3686656245731218,
+      "learning_rate": 6.344978199461915e-07,
+      "loss": 0.0991,
+      "step": 10002
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4820689579500976,
+      "learning_rate": 6.338326826960989e-07,
+      "loss": 0.1197,
+      "step": 10003
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2330912853072014,
+      "learning_rate": 6.331678706635813e-07,
+      "loss": 0.0572,
+      "step": 10004
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.37731167924736797,
+      "learning_rate": 6.325033838981609e-07,
+      "loss": 0.0605,
+      "step": 10005
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2523130116950927,
+      "learning_rate": 6.318392224493309e-07,
+      "loss": 0.071,
+      "step": 10006
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4025894834448294,
+      "learning_rate": 6.311753863665603e-07,
+      "loss": 0.0987,
+      "step": 10007
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.27599719809877776,
+      "learning_rate": 6.30511875699299e-07,
+      "loss": 0.0733,
+      "step": 10008
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4919532154506478,
+      "learning_rate": 6.298486904969675e-07,
+      "loss": 0.0965,
+      "step": 10009
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.20638605968772533,
+      "learning_rate": 6.291858308089633e-07,
+      "loss": 0.0523,
+      "step": 10010
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.47135193749683096,
+      "learning_rate": 6.285232966846589e-07,
+      "loss": 0.1041,
+      "step": 10011
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.43289601190114324,
+      "learning_rate": 6.278610881734065e-07,
+      "loss": 0.0963,
+      "step": 10012
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.435381571819881,
+      "learning_rate": 6.271992053245296e-07,
+      "loss": 0.1405,
+      "step": 10013
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5126033372432839,
+      "learning_rate": 6.265376481873287e-07,
+      "loss": 0.1195,
+      "step": 10014
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.21779021357190934,
+      "learning_rate": 6.25876416811082e-07,
+      "loss": 0.0576,
+      "step": 10015
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2989099413506579,
+      "learning_rate": 6.252155112450409e-07,
+      "loss": 0.0919,
+      "step": 10016
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2958396235730252,
+      "learning_rate": 6.24554931538433e-07,
+      "loss": 0.0753,
+      "step": 10017
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2753224658757622,
+      "learning_rate": 6.238946777404619e-07,
+      "loss": 0.0567,
+      "step": 10018
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2925403621564145,
+      "learning_rate": 6.232347499003094e-07,
+      "loss": 0.0845,
+      "step": 10019
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.33754052775932514,
+      "learning_rate": 6.225751480671288e-07,
+      "loss": 0.0894,
+      "step": 10020
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.23248067214408336,
+      "learning_rate": 6.219158722900509e-07,
+      "loss": 0.0775,
+      "step": 10021
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.15912741345557144,
+      "learning_rate": 6.212569226181836e-07,
+      "loss": 0.0488,
+      "step": 10022
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2656976245486098,
+      "learning_rate": 6.205982991006093e-07,
+      "loss": 0.0743,
+      "step": 10023
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.340150623186326,
+      "learning_rate": 6.199400017863844e-07,
+      "loss": 0.0998,
+      "step": 10024
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.36717487028904194,
+      "learning_rate": 6.192820307245445e-07,
+      "loss": 0.0834,
+      "step": 10025
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3698375951686579,
+      "learning_rate": 6.186243859640989e-07,
+      "loss": 0.0872,
+      "step": 10026
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3151516640981602,
+      "learning_rate": 6.179670675540322e-07,
+      "loss": 0.0901,
+      "step": 10027
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.27648733119213587,
+      "learning_rate": 6.173100755433048e-07,
+      "loss": 0.063,
+      "step": 10028
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2909117653166107,
+      "learning_rate": 6.166534099808552e-07,
+      "loss": 0.0562,
+      "step": 10029
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.35965534577714,
+      "learning_rate": 6.159970709155943e-07,
+      "loss": 0.0729,
+      "step": 10030
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3959655871915268,
+      "learning_rate": 6.153410583964092e-07,
+      "loss": 0.0995,
+      "step": 10031
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.34862531830459653,
+      "learning_rate": 6.146853724721652e-07,
+      "loss": 0.0914,
+      "step": 10032
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.30899925620861934,
+      "learning_rate": 6.140300131917015e-07,
+      "loss": 0.0709,
+      "step": 10033
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.21019894135443865,
+      "learning_rate": 6.133749806038325e-07,
+      "loss": 0.0382,
+      "step": 10034
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.32840265695534265,
+      "learning_rate": 6.12720274757348e-07,
+      "loss": 0.1019,
+      "step": 10035
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2920827945890285,
+      "learning_rate": 6.120658957010162e-07,
+      "loss": 0.0854,
+      "step": 10036
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.4476076160764897,
+      "learning_rate": 6.114118434835781e-07,
+      "loss": 0.091,
+      "step": 10037
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.44075988335182503,
+      "learning_rate": 6.107581181537503e-07,
+      "loss": 0.1357,
+      "step": 10038
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2621923594633043,
+      "learning_rate": 6.101047197602278e-07,
+      "loss": 0.0844,
+      "step": 10039
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.22935029461028514,
+      "learning_rate": 6.094516483516794e-07,
+      "loss": 0.055,
+      "step": 10040
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.22211277565660706,
+      "learning_rate": 6.087989039767472e-07,
+      "loss": 0.0866,
+      "step": 10041
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3697892472230128,
+      "learning_rate": 6.081464866840552e-07,
+      "loss": 0.0678,
+      "step": 10042
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.4057106054291832,
+      "learning_rate": 6.074943965221969e-07,
+      "loss": 0.0858,
+      "step": 10043
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.45883441864136787,
+      "learning_rate": 6.068426335397442e-07,
+      "loss": 0.0768,
+      "step": 10044
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6617785743403138,
+      "learning_rate": 6.06191197785243e-07,
+      "loss": 0.1263,
+      "step": 10045
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.22899207271752697,
+      "learning_rate": 6.055400893072183e-07,
+      "loss": 0.0715,
+      "step": 10046
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2816184833941627,
+      "learning_rate": 6.048893081541674e-07,
+      "loss": 0.0782,
+      "step": 10047
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5064648873609169,
+      "learning_rate": 6.042388543745642e-07,
+      "loss": 0.1331,
+      "step": 10048
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.18797318740207142,
+      "learning_rate": 6.035887280168579e-07,
+      "loss": 0.0626,
+      "step": 10049
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2341344234682925,
+      "learning_rate": 6.029389291294747e-07,
+      "loss": 0.0598,
+      "step": 10050
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3512452207778412,
+      "learning_rate": 6.022894577608146e-07,
+      "loss": 0.1098,
+      "step": 10051
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.34607291364100073,
+      "learning_rate": 6.016403139592541e-07,
+      "loss": 0.0858,
+      "step": 10052
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3123602129253076,
+      "learning_rate": 6.009914977731446e-07,
+      "loss": 0.0687,
+      "step": 10053
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.251331360341203,
+      "learning_rate": 6.003430092508156e-07,
+      "loss": 0.0877,
+      "step": 10054
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.4344539389887402,
+      "learning_rate": 5.99694848440569e-07,
+      "loss": 0.094,
+      "step": 10055
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.40992625273747296,
+      "learning_rate": 5.990470153906841e-07,
+      "loss": 0.1006,
+      "step": 10056
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.35065072571167394,
+      "learning_rate": 5.98399510149415e-07,
+      "loss": 0.1073,
+      "step": 10057
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.531683677164515,
+      "learning_rate": 5.977523327649903e-07,
+      "loss": 0.0987,
+      "step": 10058
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.505513606779016,
+      "learning_rate": 5.971054832856177e-07,
+      "loss": 0.1252,
+      "step": 10059
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.33564076128811127,
+      "learning_rate": 5.964589617594774e-07,
+      "loss": 0.1035,
+      "step": 10060
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.20162835599461876,
+      "learning_rate": 5.958127682347264e-07,
+      "loss": 0.0549,
+      "step": 10061
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3038869763699886,
+      "learning_rate": 5.951669027594958e-07,
+      "loss": 0.1041,
+      "step": 10062
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3753884917034024,
+      "learning_rate": 5.945213653818954e-07,
+      "loss": 0.0886,
+      "step": 10063
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.23551272816867713,
+      "learning_rate": 5.938761561500073e-07,
+      "loss": 0.0607,
+      "step": 10064
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3479306454516724,
+      "learning_rate": 5.932312751118896e-07,
+      "loss": 0.0943,
+      "step": 10065
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3836962601323056,
+      "learning_rate": 5.92586722315579e-07,
+      "loss": 0.0988,
+      "step": 10066
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.4125365898913317,
+      "learning_rate": 5.919424978090843e-07,
+      "loss": 0.0678,
+      "step": 10067
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3022674371688516,
+      "learning_rate": 5.91298601640391e-07,
+      "loss": 0.0835,
+      "step": 10068
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.24855599509345083,
+      "learning_rate": 5.906550338574596e-07,
+      "loss": 0.0416,
+      "step": 10069
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2366265294150871,
+      "learning_rate": 5.90011794508229e-07,
+      "loss": 0.0624,
+      "step": 10070
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3054822460605164,
+      "learning_rate": 5.893688836406098e-07,
+      "loss": 0.1003,
+      "step": 10071
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5693754429143377,
+      "learning_rate": 5.887263013024886e-07,
+      "loss": 0.0913,
+      "step": 10072
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3590839416763244,
+      "learning_rate": 5.880840475417315e-07,
+      "loss": 0.0875,
+      "step": 10073
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.26328879378723563,
+      "learning_rate": 5.874421224061755e-07,
+      "loss": 0.0596,
+      "step": 10074
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3039715264159054,
+      "learning_rate": 5.868005259436343e-07,
+      "loss": 0.0902,
+      "step": 10075
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3998908034187788,
+      "learning_rate": 5.861592582018998e-07,
+      "loss": 0.0668,
+      "step": 10076
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7110174508168057,
+      "learning_rate": 5.855183192287367e-07,
+      "loss": 0.1365,
+      "step": 10077
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3998245029451963,
+      "learning_rate": 5.848777090718849e-07,
+      "loss": 0.0711,
+      "step": 10078
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3227790174012581,
+      "learning_rate": 5.842374277790608e-07,
+      "loss": 0.054,
+      "step": 10079
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3390945760519033,
+      "learning_rate": 5.835974753979573e-07,
+      "loss": 0.0841,
+      "step": 10080
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.18067998683394357,
+      "learning_rate": 5.829578519762413e-07,
+      "loss": 0.0439,
+      "step": 10081
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2446286218041023,
+      "learning_rate": 5.823185575615553e-07,
+      "loss": 0.0511,
+      "step": 10082
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.21025385259522128,
+      "learning_rate": 5.816795922015184e-07,
+      "loss": 0.0371,
+      "step": 10083
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.21919923061775645,
+      "learning_rate": 5.810409559437241e-07,
+      "loss": 0.0793,
+      "step": 10084
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.580564164850924,
+      "learning_rate": 5.804026488357423e-07,
+      "loss": 0.0916,
+      "step": 10085
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.35026425716209636,
+      "learning_rate": 5.797646709251153e-07,
+      "loss": 0.1088,
+      "step": 10086
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.354811859883803,
+      "learning_rate": 5.79127022259367e-07,
+      "loss": 0.1056,
+      "step": 10087
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2678649803196242,
+      "learning_rate": 5.784897028859915e-07,
+      "loss": 0.0621,
+      "step": 10088
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.32969620933248084,
+      "learning_rate": 5.778527128524591e-07,
+      "loss": 0.0879,
+      "step": 10089
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2798721662342978,
+      "learning_rate": 5.772160522062187e-07,
+      "loss": 0.0849,
+      "step": 10090
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.25182569151114936,
+      "learning_rate": 5.765797209946916e-07,
+      "loss": 0.0648,
+      "step": 10091
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.21865477899744626,
+      "learning_rate": 5.759437192652739e-07,
+      "loss": 0.0634,
+      "step": 10092
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2850319371447458,
+      "learning_rate": 5.753080470653411e-07,
+      "loss": 0.0778,
+      "step": 10093
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.23589932933515276,
+      "learning_rate": 5.746727044422401e-07,
+      "loss": 0.0651,
+      "step": 10094
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3019480784331829,
+      "learning_rate": 5.74037691443296e-07,
+      "loss": 0.0857,
+      "step": 10095
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3154725287574297,
+      "learning_rate": 5.734030081158071e-07,
+      "loss": 0.0724,
+      "step": 10096
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.21602922609034222,
+      "learning_rate": 5.727686545070493e-07,
+      "loss": 0.072,
+      "step": 10097
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.4329351899219665,
+      "learning_rate": 5.721346306642733e-07,
+      "loss": 0.1119,
+      "step": 10098
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.18544762084551014,
+      "learning_rate": 5.715009366347029e-07,
+      "loss": 0.0458,
+      "step": 10099
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5563154679794391,
+      "learning_rate": 5.708675724655416e-07,
+      "loss": 0.1264,
+      "step": 10100
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2485403851592536,
+      "learning_rate": 5.702345382039653e-07,
+      "loss": 0.0717,
+      "step": 10101
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.32432658645935014,
+      "learning_rate": 5.696018338971259e-07,
+      "loss": 0.0731,
+      "step": 10102
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.45674140230988364,
+      "learning_rate": 5.689694595921497e-07,
+      "loss": 0.098,
+      "step": 10103
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.4899105897067363,
+      "learning_rate": 5.683374153361421e-07,
+      "loss": 0.1272,
+      "step": 10104
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.30639314046242133,
+      "learning_rate": 5.6770570117618e-07,
+      "loss": 0.0668,
+      "step": 10105
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.4964217596715763,
+      "learning_rate": 5.670743171593157e-07,
+      "loss": 0.0821,
+      "step": 10106
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.23361283372302385,
+      "learning_rate": 5.664432633325817e-07,
+      "loss": 0.0528,
+      "step": 10107
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.24142385203668879,
+      "learning_rate": 5.658125397429809e-07,
+      "loss": 0.0412,
+      "step": 10108
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.21254369781288915,
+      "learning_rate": 5.651821464374918e-07,
+      "loss": 0.061,
+      "step": 10109
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.24200911900799923,
+      "learning_rate": 5.645520834630718e-07,
+      "loss": 0.0643,
+      "step": 10110
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5600097345438724,
+      "learning_rate": 5.639223508666519e-07,
+      "loss": 0.0942,
+      "step": 10111
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3491473062639947,
+      "learning_rate": 5.632929486951372e-07,
+      "loss": 0.0919,
+      "step": 10112
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.24315213421564805,
+      "learning_rate": 5.626638769954079e-07,
+      "loss": 0.0585,
+      "step": 10113
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.41236815141197186,
+      "learning_rate": 5.620351358143239e-07,
+      "loss": 0.0717,
+      "step": 10114
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.33504781720963045,
+      "learning_rate": 5.614067251987154e-07,
+      "loss": 0.0731,
+      "step": 10115
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3128590469777471,
+      "learning_rate": 5.607786451953912e-07,
+      "loss": 0.0691,
+      "step": 10116
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.23033080064507722,
+      "learning_rate": 5.601508958511331e-07,
+      "loss": 0.075,
+      "step": 10117
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.19658909336726813,
+      "learning_rate": 5.595234772127011e-07,
+      "loss": 0.0646,
+      "step": 10118
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2510375988849471,
+      "learning_rate": 5.588963893268279e-07,
+      "loss": 0.067,
+      "step": 10119
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2573015565067798,
+      "learning_rate": 5.582696322402237e-07,
+      "loss": 0.0823,
+      "step": 10120
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5561555205062403,
+      "learning_rate": 5.576432059995707e-07,
+      "loss": 0.1126,
+      "step": 10121
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.30365844221003885,
+      "learning_rate": 5.570171106515315e-07,
+      "loss": 0.0466,
+      "step": 10122
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.32601208151312766,
+      "learning_rate": 5.563913462427401e-07,
+      "loss": 0.0858,
+      "step": 10123
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.20664641937100506,
+      "learning_rate": 5.557659128198072e-07,
+      "loss": 0.0408,
+      "step": 10124
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2057668485448398,
+      "learning_rate": 5.551408104293188e-07,
+      "loss": 0.0464,
+      "step": 10125
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.275302541041108,
+      "learning_rate": 5.54516039117835e-07,
+      "loss": 0.0977,
+      "step": 10126
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.22596811227362334,
+      "learning_rate": 5.538915989318943e-07,
+      "loss": 0.0695,
+      "step": 10127
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.23848898256759685,
+      "learning_rate": 5.532674899180079e-07,
+      "loss": 0.0806,
+      "step": 10128
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.20674249425206906,
+      "learning_rate": 5.526437121226629e-07,
+      "loss": 0.034,
+      "step": 10129
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.4611595113157456,
+      "learning_rate": 5.520202655923212e-07,
+      "loss": 0.0811,
+      "step": 10130
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3339739097495186,
+      "learning_rate": 5.513971503734223e-07,
+      "loss": 0.118,
+      "step": 10131
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.34023055318300904,
+      "learning_rate": 5.507743665123788e-07,
+      "loss": 0.0807,
+      "step": 10132
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.4529090994019407,
+      "learning_rate": 5.501519140555784e-07,
+      "loss": 0.1131,
+      "step": 10133
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2959287939753219,
+      "learning_rate": 5.495297930493864e-07,
+      "loss": 0.0639,
+      "step": 10134
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3083781397633366,
+      "learning_rate": 5.489080035401412e-07,
+      "loss": 0.0918,
+      "step": 10135
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2658363308666795,
+      "learning_rate": 5.482865455741576e-07,
+      "loss": 0.08,
+      "step": 10136
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5614104857624876,
+      "learning_rate": 5.476654191977243e-07,
+      "loss": 0.1071,
+      "step": 10137
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.26866241445030264,
+      "learning_rate": 5.470446244571081e-07,
+      "loss": 0.0712,
+      "step": 10138
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.22916407226265162,
+      "learning_rate": 5.464241613985494e-07,
+      "loss": 0.0732,
+      "step": 10139
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.23361139735585032,
+      "learning_rate": 5.458040300682616e-07,
+      "loss": 0.0576,
+      "step": 10140
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.45699667092551394,
+      "learning_rate": 5.451842305124377e-07,
+      "loss": 0.0861,
+      "step": 10141
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3779438131504885,
+      "learning_rate": 5.445647627772443e-07,
+      "loss": 0.104,
+      "step": 10142
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.21099429836653283,
+      "learning_rate": 5.439456269088222e-07,
+      "loss": 0.0649,
+      "step": 10143
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.28044221210124665,
+      "learning_rate": 5.433268229532867e-07,
+      "loss": 0.0828,
+      "step": 10144
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.36199221607048093,
+      "learning_rate": 5.427083509567327e-07,
+      "loss": 0.0831,
+      "step": 10145
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.39216804062376454,
+      "learning_rate": 5.420902109652265e-07,
+      "loss": 0.1029,
+      "step": 10146
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3780502826396717,
+      "learning_rate": 5.414724030248098e-07,
+      "loss": 0.0718,
+      "step": 10147
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3050551022564213,
+      "learning_rate": 5.408549271815023e-07,
+      "loss": 0.0809,
+      "step": 10148
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.36436516543996395,
+      "learning_rate": 5.402377834812961e-07,
+      "loss": 0.0618,
+      "step": 10149
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3092014068121197,
+      "learning_rate": 5.396209719701584e-07,
+      "loss": 0.0929,
+      "step": 10150
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6360406986906038,
+      "learning_rate": 5.390044926940358e-07,
+      "loss": 0.1222,
+      "step": 10151
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.35744604704974464,
+      "learning_rate": 5.383883456988459e-07,
+      "loss": 0.1151,
+      "step": 10152
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2826134725793427,
+      "learning_rate": 5.377725310304827e-07,
+      "loss": 0.083,
+      "step": 10153
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5205982078654979,
+      "learning_rate": 5.371570487348149e-07,
+      "loss": 0.1365,
+      "step": 10154
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.33780526807175987,
+      "learning_rate": 5.365418988576887e-07,
+      "loss": 0.0943,
+      "step": 10155
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.26999103018681314,
+      "learning_rate": 5.359270814449241e-07,
+      "loss": 0.0555,
+      "step": 10156
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2775509996516515,
+      "learning_rate": 5.353125965423139e-07,
+      "loss": 0.0621,
+      "step": 10157
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3208658967614774,
+      "learning_rate": 5.346984441956316e-07,
+      "loss": 0.0879,
+      "step": 10158
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.10915325703474758,
+      "learning_rate": 5.340846244506214e-07,
+      "loss": 0.0289,
+      "step": 10159
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.35765169206162944,
+      "learning_rate": 5.334711373530043e-07,
+      "loss": 0.0959,
+      "step": 10160
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2810410289620671,
+      "learning_rate": 5.328579829484754e-07,
+      "loss": 0.0808,
+      "step": 10161
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3166000050568358,
+      "learning_rate": 5.322451612827078e-07,
+      "loss": 0.073,
+      "step": 10162
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.30449514646004555,
+      "learning_rate": 5.316326724013477e-07,
+      "loss": 0.0806,
+      "step": 10163
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.29423842962256364,
+      "learning_rate": 5.310205163500148e-07,
+      "loss": 0.0756,
+      "step": 10164
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.327304131929898,
+      "learning_rate": 5.304086931743085e-07,
+      "loss": 0.0869,
+      "step": 10165
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3675276971314589,
+      "learning_rate": 5.297972029198006e-07,
+      "loss": 0.0442,
+      "step": 10166
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.22519923978457085,
+      "learning_rate": 5.291860456320363e-07,
+      "loss": 0.0391,
+      "step": 10167
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.35575681309286683,
+      "learning_rate": 5.28575221356541e-07,
+      "loss": 0.0773,
+      "step": 10168
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.30400740819080263,
+      "learning_rate": 5.279647301388107e-07,
+      "loss": 0.0844,
+      "step": 10169
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.36732024366900184,
+      "learning_rate": 5.273545720243195e-07,
+      "loss": 0.1245,
+      "step": 10170
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.1960655751678503,
+      "learning_rate": 5.267447470585135e-07,
+      "loss": 0.0555,
+      "step": 10171
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3030606164429105,
+      "learning_rate": 5.261352552868187e-07,
+      "loss": 0.095,
+      "step": 10172
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3085420021645029,
+      "learning_rate": 5.255260967546321e-07,
+      "loss": 0.0785,
+      "step": 10173
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.24766138873802865,
+      "learning_rate": 5.249172715073264e-07,
+      "loss": 0.0612,
+      "step": 10174
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.19946730276298946,
+      "learning_rate": 5.24308779590253e-07,
+      "loss": 0.0317,
+      "step": 10175
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.4221255888641955,
+      "learning_rate": 5.237006210487339e-07,
+      "loss": 0.1098,
+      "step": 10176
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.4302044304689142,
+      "learning_rate": 5.230927959280696e-07,
+      "loss": 0.0631,
+      "step": 10177
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2328379140151972,
+      "learning_rate": 5.224853042735323e-07,
+      "loss": 0.0679,
+      "step": 10178
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2699513452362352,
+      "learning_rate": 5.21878146130374e-07,
+      "loss": 0.0952,
+      "step": 10179
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.45347490184686484,
+      "learning_rate": 5.212713215438187e-07,
+      "loss": 0.0865,
+      "step": 10180
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2559295798724765,
+      "learning_rate": 5.206648305590645e-07,
+      "loss": 0.0547,
+      "step": 10181
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.40675486523356985,
+      "learning_rate": 5.200586732212892e-07,
+      "loss": 0.0772,
+      "step": 10182
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.4073187418841058,
+      "learning_rate": 5.194528495756413e-07,
+      "loss": 0.1447,
+      "step": 10183
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.40045063472907716,
+      "learning_rate": 5.18847359667246e-07,
+      "loss": 0.0858,
+      "step": 10184
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6488649111846287,
+      "learning_rate": 5.182422035412033e-07,
+      "loss": 0.0952,
+      "step": 10185
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.18427024002170167,
+      "learning_rate": 5.176373812425906e-07,
+      "loss": 0.0311,
+      "step": 10186
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5901415861337089,
+      "learning_rate": 5.170328928164569e-07,
+      "loss": 0.0727,
+      "step": 10187
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3615338297571059,
+      "learning_rate": 5.164287383078287e-07,
+      "loss": 0.1257,
+      "step": 10188
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3277681420696021,
+      "learning_rate": 5.158249177617064e-07,
+      "loss": 0.1026,
+      "step": 10189
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.21689988665360102,
+      "learning_rate": 5.152214312230668e-07,
+      "loss": 0.0569,
+      "step": 10190
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3572332178946842,
+      "learning_rate": 5.146182787368609e-07,
+      "loss": 0.1172,
+      "step": 10191
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.22588690189929764,
+      "learning_rate": 5.140154603480152e-07,
+      "loss": 0.0654,
+      "step": 10192
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2695550597192518,
+      "learning_rate": 5.134129761014306e-07,
+      "loss": 0.0643,
+      "step": 10193
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.25010907600136456,
+      "learning_rate": 5.128108260419828e-07,
+      "loss": 0.0612,
+      "step": 10194
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.23556562080626983,
+      "learning_rate": 5.122090102145255e-07,
+      "loss": 0.0616,
+      "step": 10195
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3621675350765161,
+      "learning_rate": 5.116075286638844e-07,
+      "loss": 0.0704,
+      "step": 10196
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3131261952714749,
+      "learning_rate": 5.110063814348614e-07,
+      "loss": 0.071,
+      "step": 10197
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.38354680874174724,
+      "learning_rate": 5.104055685722326e-07,
+      "loss": 0.1007,
+      "step": 10198
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.353198951561074,
+      "learning_rate": 5.098050901207519e-07,
+      "loss": 0.0649,
+      "step": 10199
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2408579198957943,
+      "learning_rate": 5.092049461251453e-07,
+      "loss": 0.0674,
+      "step": 10200
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2773421566768113,
+      "learning_rate": 5.086051366301143e-07,
+      "loss": 0.0524,
+      "step": 10201
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.31780623767360566,
+      "learning_rate": 5.080056616803375e-07,
+      "loss": 0.0873,
+      "step": 10202
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.22842394481054543,
+      "learning_rate": 5.074065213204677e-07,
+      "loss": 0.0393,
+      "step": 10203
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.48957375136074877,
+      "learning_rate": 5.068077155951306e-07,
+      "loss": 0.1083,
+      "step": 10204
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.31720943528062123,
+      "learning_rate": 5.062092445489292e-07,
+      "loss": 0.0897,
+      "step": 10205
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3934655126248802,
+      "learning_rate": 5.05611108226442e-07,
+      "loss": 0.1043,
+      "step": 10206
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.27299753882075806,
+      "learning_rate": 5.050133066722218e-07,
+      "loss": 0.0599,
+      "step": 10207
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.16080844450749188,
+      "learning_rate": 5.044158399307941e-07,
+      "loss": 0.0428,
+      "step": 10208
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.29585174910428724,
+      "learning_rate": 5.03818708046665e-07,
+      "loss": 0.0887,
+      "step": 10209
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.4756540658564396,
+      "learning_rate": 5.032219110643105e-07,
+      "loss": 0.0765,
+      "step": 10210
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3831573844688135,
+      "learning_rate": 5.026254490281835e-07,
+      "loss": 0.0965,
+      "step": 10211
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.41443214000563344,
+      "learning_rate": 5.020293219827116e-07,
+      "loss": 0.0876,
+      "step": 10212
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3922449204842818,
+      "learning_rate": 5.014335299722989e-07,
+      "loss": 0.0875,
+      "step": 10213
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.4280807004114287,
+      "learning_rate": 5.008380730413231e-07,
+      "loss": 0.1074,
+      "step": 10214
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.30198797463437604,
+      "learning_rate": 5.002429512341367e-07,
+      "loss": 0.0629,
+      "step": 10215
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.25042853511453145,
+      "learning_rate": 4.996481645950691e-07,
+      "loss": 0.0905,
+      "step": 10216
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.22591671629981414,
+      "learning_rate": 4.99053713168422e-07,
+      "loss": 0.078,
+      "step": 10217
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.281593486909036,
+      "learning_rate": 4.98459596998474e-07,
+      "loss": 0.0731,
+      "step": 10218
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3699086140036204,
+      "learning_rate": 4.978658161294797e-07,
+      "loss": 0.1002,
+      "step": 10219
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2635814628460597,
+      "learning_rate": 4.97272370605666e-07,
+      "loss": 0.0724,
+      "step": 10220
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7727226390361319,
+      "learning_rate": 4.966792604712362e-07,
+      "loss": 0.2018,
+      "step": 10221
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3386638076236892,
+      "learning_rate": 4.960864857703684e-07,
+      "loss": 0.07,
+      "step": 10222
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.4258233614013364,
+      "learning_rate": 4.954940465472169e-07,
+      "loss": 0.0969,
+      "step": 10223
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.16571779248145949,
+      "learning_rate": 4.949019428459101e-07,
+      "loss": 0.0265,
+      "step": 10224
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.28982450174778257,
+      "learning_rate": 4.943101747105494e-07,
+      "loss": 0.0864,
+      "step": 10225
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3438381130784451,
+      "learning_rate": 4.937187421852152e-07,
+      "loss": 0.0823,
+      "step": 10226
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.26235028237812424,
+      "learning_rate": 4.931276453139605e-07,
+      "loss": 0.0672,
+      "step": 10227
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2970159247063792,
+      "learning_rate": 4.925368841408129e-07,
+      "loss": 0.0498,
+      "step": 10228
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2997101783544131,
+      "learning_rate": 4.919464587097756e-07,
+      "loss": 0.0923,
+      "step": 10229
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.24796452119856754,
+      "learning_rate": 4.913563690648282e-07,
+      "loss": 0.054,
+      "step": 10230
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.4638822164242629,
+      "learning_rate": 4.907666152499229e-07,
+      "loss": 0.0874,
+      "step": 10231
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.26648054083531814,
+      "learning_rate": 4.901771973089875e-07,
+      "loss": 0.0651,
+      "step": 10232
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.59773389861169,
+      "learning_rate": 4.895881152859272e-07,
+      "loss": 0.1319,
+      "step": 10233
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.47661722187376643,
+      "learning_rate": 4.889993692246192e-07,
+      "loss": 0.0929,
+      "step": 10234
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.16138837852530785,
+      "learning_rate": 4.884109591689168e-07,
+      "loss": 0.0357,
+      "step": 10235
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2949496298846048,
+      "learning_rate": 4.878228851626465e-07,
+      "loss": 0.0758,
+      "step": 10236
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2543849112834512,
+      "learning_rate": 4.872351472496146e-07,
+      "loss": 0.0376,
+      "step": 10237
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2145955663391157,
+      "learning_rate": 4.866477454735979e-07,
+      "loss": 0.0478,
+      "step": 10238
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.24434501939954376,
+      "learning_rate": 4.860606798783479e-07,
+      "loss": 0.0644,
+      "step": 10239
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2937302254917918,
+      "learning_rate": 4.854739505075956e-07,
+      "loss": 0.0818,
+      "step": 10240
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2553345241588614,
+      "learning_rate": 4.848875574050421e-07,
+      "loss": 0.0657,
+      "step": 10241
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5721737199276666,
+      "learning_rate": 4.843015006143648e-07,
+      "loss": 0.0912,
+      "step": 10242
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2733376399076147,
+      "learning_rate": 4.837157801792186e-07,
+      "loss": 0.0597,
+      "step": 10243
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.30212893104126315,
+      "learning_rate": 4.831303961432304e-07,
+      "loss": 0.0396,
+      "step": 10244
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3208961585722439,
+      "learning_rate": 4.825453485500032e-07,
+      "loss": 0.0804,
+      "step": 10245
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5206776666001215,
+      "learning_rate": 4.819606374431135e-07,
+      "loss": 0.0807,
+      "step": 10246
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2508554646560112,
+      "learning_rate": 4.813762628661162e-07,
+      "loss": 0.0597,
+      "step": 10247
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.308374542677377,
+      "learning_rate": 4.807922248625374e-07,
+      "loss": 0.0474,
+      "step": 10248
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.235376590980154,
+      "learning_rate": 4.802085234758796e-07,
+      "loss": 0.0439,
+      "step": 10249
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8773818973956543,
+      "learning_rate": 4.796251587496214e-07,
+      "loss": 0.1012,
+      "step": 10250
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5156017019552253,
+      "learning_rate": 4.790421307272141e-07,
+      "loss": 0.0968,
+      "step": 10251
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.21238967928582178,
+      "learning_rate": 4.784594394520858e-07,
+      "loss": 0.0777,
+      "step": 10252
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3666093768411957,
+      "learning_rate": 4.778770849676373e-07,
+      "loss": 0.1054,
+      "step": 10253
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.43457962060793637,
+      "learning_rate": 4.772950673172483e-07,
+      "loss": 0.0447,
+      "step": 10254
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2579033526879363,
+      "learning_rate": 4.767133865442686e-07,
+      "loss": 0.0627,
+      "step": 10255
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.36003414050401195,
+      "learning_rate": 4.761320426920263e-07,
+      "loss": 0.0745,
+      "step": 10256
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.32375235312733147,
+      "learning_rate": 4.7555103580382234e-07,
+      "loss": 0.0979,
+      "step": 10257
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.23507560639990824,
+      "learning_rate": 4.749703659229343e-07,
+      "loss": 0.0743,
+      "step": 10258
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.30075282359524214,
+      "learning_rate": 4.7439003309261433e-07,
+      "loss": 0.0701,
+      "step": 10259
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2760026121332858,
+      "learning_rate": 4.738100373560883e-07,
+      "loss": 0.0609,
+      "step": 10260
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.20152794464550844,
+      "learning_rate": 4.732303787565573e-07,
+      "loss": 0.0504,
+      "step": 10261
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2959931002283258,
+      "learning_rate": 4.726510573371973e-07,
+      "loss": 0.0734,
+      "step": 10262
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3486099193423995,
+      "learning_rate": 4.7207207314116097e-07,
+      "loss": 0.0891,
+      "step": 10263
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.386772936605961,
+      "learning_rate": 4.7149342621157447e-07,
+      "loss": 0.0605,
+      "step": 10264
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7764557068990194,
+      "learning_rate": 4.709151165915371e-07,
+      "loss": 0.1019,
+      "step": 10265
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4156304549160289,
+      "learning_rate": 4.703371443241256e-07,
+      "loss": 0.0741,
+      "step": 10266
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.31056722056849706,
+      "learning_rate": 4.6975950945239157e-07,
+      "loss": 0.0858,
+      "step": 10267
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4013091395394022,
+      "learning_rate": 4.691822120193595e-07,
+      "loss": 0.0832,
+      "step": 10268
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.28671706496106414,
+      "learning_rate": 4.6860525206803066e-07,
+      "loss": 0.0793,
+      "step": 10269
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.23547596273598484,
+      "learning_rate": 4.6802862964137885e-07,
+      "loss": 0.0645,
+      "step": 10270
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.25677223175567787,
+      "learning_rate": 4.674523447823565e-07,
+      "loss": 0.0576,
+      "step": 10271
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3884967273802368,
+      "learning_rate": 4.668763975338875e-07,
+      "loss": 0.027,
+      "step": 10272
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.28648261467681785,
+      "learning_rate": 4.663007879388709e-07,
+      "loss": 0.097,
+      "step": 10273
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2776977282929443,
+      "learning_rate": 4.65725516040183e-07,
+      "loss": 0.0838,
+      "step": 10274
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3389796810483693,
+      "learning_rate": 4.6515058188067286e-07,
+      "loss": 0.1075,
+      "step": 10275
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6206385845292867,
+      "learning_rate": 4.64575985503164e-07,
+      "loss": 0.0969,
+      "step": 10276
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.35733929444124096,
+      "learning_rate": 4.6400172695045767e-07,
+      "loss": 0.0899,
+      "step": 10277
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.29254147699566185,
+      "learning_rate": 4.6342780626532693e-07,
+      "loss": 0.0823,
+      "step": 10278
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.35023298831837707,
+      "learning_rate": 4.6285422349052034e-07,
+      "loss": 0.0563,
+      "step": 10279
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.16485257246911403,
+      "learning_rate": 4.62280978668761e-07,
+      "loss": 0.0341,
+      "step": 10280
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.428637768013447,
+      "learning_rate": 4.617080718427497e-07,
+      "loss": 0.0715,
+      "step": 10281
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.19768073472783207,
+      "learning_rate": 4.611355030551584e-07,
+      "loss": 0.0571,
+      "step": 10282
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2940470223077482,
+      "learning_rate": 4.6056327234863527e-07,
+      "loss": 0.0735,
+      "step": 10283
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3868099699550164,
+      "learning_rate": 4.5999137976580456e-07,
+      "loss": 0.1057,
+      "step": 10284
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3421544007731041,
+      "learning_rate": 4.594198253492632e-07,
+      "loss": 0.0455,
+      "step": 10285
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2943359566913793,
+      "learning_rate": 4.5884860914158393e-07,
+      "loss": 0.0643,
+      "step": 10286
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.26324892211029793,
+      "learning_rate": 4.5827773118531374e-07,
+      "loss": 0.0635,
+      "step": 10287
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3062664386470111,
+      "learning_rate": 4.5770719152297637e-07,
+      "loss": 0.0726,
+      "step": 10288
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.23420903247841512,
+      "learning_rate": 4.571369901970685e-07,
+      "loss": 0.0716,
+      "step": 10289
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.32220686253937486,
+      "learning_rate": 4.5656712725006046e-07,
+      "loss": 0.1211,
+      "step": 10290
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5089778524708687,
+      "learning_rate": 4.559976027244012e-07,
+      "loss": 0.0834,
+      "step": 10291
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2605011411550036,
+      "learning_rate": 4.554284166625117e-07,
+      "loss": 0.0616,
+      "step": 10292
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.37348259657523597,
+      "learning_rate": 4.548595691067864e-07,
+      "loss": 0.1023,
+      "step": 10293
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.30545091060130575,
+      "learning_rate": 4.542910600995992e-07,
+      "loss": 0.0888,
+      "step": 10294
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.41881365674578086,
+      "learning_rate": 4.5372288968329403e-07,
+      "loss": 0.0789,
+      "step": 10295
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.30162096523054843,
+      "learning_rate": 4.5315505790019255e-07,
+      "loss": 0.0518,
+      "step": 10296
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.323327871858337,
+      "learning_rate": 4.5258756479258817e-07,
+      "loss": 0.0714,
+      "step": 10297
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.29800942606073044,
+      "learning_rate": 4.5202041040275427e-07,
+      "loss": 0.0633,
+      "step": 10298
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3234224270431746,
+      "learning_rate": 4.514535947729337e-07,
+      "loss": 0.0775,
+      "step": 10299
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.22619306297448796,
+      "learning_rate": 4.508871179453456e-07,
+      "loss": 0.0551,
+      "step": 10300
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.33216494321211715,
+      "learning_rate": 4.503209799621866e-07,
+      "loss": 0.0718,
+      "step": 10301
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.1866277227643434,
+      "learning_rate": 4.4975518086562475e-07,
+      "loss": 0.0551,
+      "step": 10302
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2520963126410752,
+      "learning_rate": 4.4918972069780406e-07,
+      "loss": 0.0535,
+      "step": 10303
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.37960338059646304,
+      "learning_rate": 4.486245995008426e-07,
+      "loss": 0.1,
+      "step": 10304
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.27147910553472443,
+      "learning_rate": 4.48059817316836e-07,
+      "loss": 0.0888,
+      "step": 10305
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.26035636758196845,
+      "learning_rate": 4.4749537418785074e-07,
+      "loss": 0.0779,
+      "step": 10306
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3319508060053674,
+      "learning_rate": 4.469312701559292e-07,
+      "loss": 0.0672,
+      "step": 10307
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.35643497786630396,
+      "learning_rate": 4.463675052630917e-07,
+      "loss": 0.1215,
+      "step": 10308
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.32137436731792407,
+      "learning_rate": 4.458040795513291e-07,
+      "loss": 0.0509,
+      "step": 10309
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.30131576243462616,
+      "learning_rate": 4.4524099306260784e-07,
+      "loss": 0.0693,
+      "step": 10310
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3834484791028972,
+      "learning_rate": 4.4467824583887154e-07,
+      "loss": 0.0879,
+      "step": 10311
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.39777282093582,
+      "learning_rate": 4.4411583792203673e-07,
+      "loss": 0.1457,
+      "step": 10312
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.1225991836815811,
+      "learning_rate": 4.4355376935399374e-07,
+      "loss": 0.0326,
+      "step": 10313
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2721349251688276,
+      "learning_rate": 4.4299204017660914e-07,
+      "loss": 0.0695,
+      "step": 10314
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.29951396540538383,
+      "learning_rate": 4.4243065043172385e-07,
+      "loss": 0.0399,
+      "step": 10315
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2832713752554066,
+      "learning_rate": 4.4186960016115443e-07,
+      "loss": 0.0845,
+      "step": 10316
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.27521846147019713,
+      "learning_rate": 4.413088894066886e-07,
+      "loss": 0.0632,
+      "step": 10317
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3860538624801942,
+      "learning_rate": 4.4074851821009456e-07,
+      "loss": 0.0764,
+      "step": 10318
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.35581482484988686,
+      "learning_rate": 4.4018848661311007e-07,
+      "loss": 0.101,
+      "step": 10319
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.48670616574989567,
+      "learning_rate": 4.3962879465745013e-07,
+      "loss": 0.0854,
+      "step": 10320
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4266729338674842,
+      "learning_rate": 4.390694423848024e-07,
+      "loss": 0.1025,
+      "step": 10321
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.19311682858007226,
+      "learning_rate": 4.3851042983683313e-07,
+      "loss": 0.0388,
+      "step": 10322
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.14655881812517357,
+      "learning_rate": 4.3795175705518e-07,
+      "loss": 0.0444,
+      "step": 10323
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3521108718058451,
+      "learning_rate": 4.373934240814559e-07,
+      "loss": 0.0694,
+      "step": 10324
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.27749808980474194,
+      "learning_rate": 4.368354309572476e-07,
+      "loss": 0.0795,
+      "step": 10325
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.35796360683785816,
+      "learning_rate": 4.362777777241195e-07,
+      "loss": 0.1099,
+      "step": 10326
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.36608619782556956,
+      "learning_rate": 4.357204644236085e-07,
+      "loss": 0.1077,
+      "step": 10327
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.20890109456707825,
+      "learning_rate": 4.351634910972258e-07,
+      "loss": 0.0581,
+      "step": 10328
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5230488917594488,
+      "learning_rate": 4.3460685778645874e-07,
+      "loss": 0.1232,
+      "step": 10329
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3156965371421897,
+      "learning_rate": 4.340505645327675e-07,
+      "loss": 0.083,
+      "step": 10330
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.43787747246516584,
+      "learning_rate": 4.3349461137758955e-07,
+      "loss": 0.0608,
+      "step": 10331
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3731609959315829,
+      "learning_rate": 4.32938998362335e-07,
+      "loss": 0.1247,
+      "step": 10332
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2928238435665105,
+      "learning_rate": 4.323837255283886e-07,
+      "loss": 0.0608,
+      "step": 10333
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.42056105913323677,
+      "learning_rate": 4.3182879291711e-07,
+      "loss": 0.1039,
+      "step": 10334
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3312379779744026,
+      "learning_rate": 4.31274200569835e-07,
+      "loss": 0.0792,
+      "step": 10335
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6524435471606699,
+      "learning_rate": 4.307199485278729e-07,
+      "loss": 0.1117,
+      "step": 10336
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.33222670650635105,
+      "learning_rate": 4.3016603683250666e-07,
+      "loss": 0.0672,
+      "step": 10337
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.31374329740436335,
+      "learning_rate": 4.2961246552499445e-07,
+      "loss": 0.0985,
+      "step": 10338
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.26720376827635073,
+      "learning_rate": 4.290592346465711e-07,
+      "loss": 0.0717,
+      "step": 10339
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.30285381406600864,
+      "learning_rate": 4.2850634423844405e-07,
+      "loss": 0.0965,
+      "step": 10340
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.23224558256617409,
+      "learning_rate": 4.279537943417944e-07,
+      "loss": 0.0533,
+      "step": 10341
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.23451275860202472,
+      "learning_rate": 4.2740158499778083e-07,
+      "loss": 0.0871,
+      "step": 10342
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3001457108440875,
+      "learning_rate": 4.268497162475349e-07,
+      "loss": 0.0856,
+      "step": 10343
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3451959377218374,
+      "learning_rate": 4.262981881321615e-07,
+      "loss": 0.0702,
+      "step": 10344
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2313840663757002,
+      "learning_rate": 4.2574700069274386e-07,
+      "loss": 0.0515,
+      "step": 10345
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4223668888691095,
+      "learning_rate": 4.251961539703364e-07,
+      "loss": 0.099,
+      "step": 10346
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2741831456052615,
+      "learning_rate": 4.2464564800596965e-07,
+      "loss": 0.0563,
+      "step": 10347
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2444429906158703,
+      "learning_rate": 4.240954828406479e-07,
+      "loss": 0.045,
+      "step": 10348
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4088940220193649,
+      "learning_rate": 4.235456585153519e-07,
+      "loss": 0.0766,
+      "step": 10349
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.27409980545021667,
+      "learning_rate": 4.2299617507103484e-07,
+      "loss": 0.0878,
+      "step": 10350
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3629484461627276,
+      "learning_rate": 4.2244703254862515e-07,
+      "loss": 0.1002,
+      "step": 10351
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.31260928812599686,
+      "learning_rate": 4.218982309890274e-07,
+      "loss": 0.0884,
+      "step": 10352
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.22388083489538424,
+      "learning_rate": 4.2134977043311877e-07,
+      "loss": 0.0607,
+      "step": 10353
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.43421735852025695,
+      "learning_rate": 4.2080165092175217e-07,
+      "loss": 0.1026,
+      "step": 10354
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3258412201122716,
+      "learning_rate": 4.2025387249575333e-07,
+      "loss": 0.0858,
+      "step": 10355
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3054051857494564,
+      "learning_rate": 4.197064351959257e-07,
+      "loss": 0.0979,
+      "step": 10356
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2370452226499411,
+      "learning_rate": 4.1915933906304497e-07,
+      "loss": 0.0617,
+      "step": 10357
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4369980372747044,
+      "learning_rate": 4.186125841378613e-07,
+      "loss": 0.0947,
+      "step": 10358
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.22333325226976525,
+      "learning_rate": 4.180661704611022e-07,
+      "loss": 0.0758,
+      "step": 10359
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3566997906528589,
+      "learning_rate": 4.1752009807346617e-07,
+      "loss": 0.1123,
+      "step": 10360
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.34485280690394426,
+      "learning_rate": 4.169743670156279e-07,
+      "loss": 0.0689,
+      "step": 10361
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8119810328765202,
+      "learning_rate": 4.164289773282365e-07,
+      "loss": 0.1581,
+      "step": 10362
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4982987666757085,
+      "learning_rate": 4.1588392905191624e-07,
+      "loss": 0.0827,
+      "step": 10363
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6304857869085873,
+      "learning_rate": 4.153392222272662e-07,
+      "loss": 0.0779,
+      "step": 10364
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3363888844140371,
+      "learning_rate": 4.1479485689485734e-07,
+      "loss": 0.0907,
+      "step": 10365
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.33172137610364805,
+      "learning_rate": 4.1425083309523885e-07,
+      "loss": 0.0911,
+      "step": 10366
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9081140970498471,
+      "learning_rate": 4.137071508689328e-07,
+      "loss": 0.0861,
+      "step": 10367
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.32942256797708763,
+      "learning_rate": 4.131638102564339e-07,
+      "loss": 0.051,
+      "step": 10368
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5921407601034936,
+      "learning_rate": 4.126208112982161e-07,
+      "loss": 0.1118,
+      "step": 10369
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2270469660376063,
+      "learning_rate": 4.1207815403472353e-07,
+      "loss": 0.0504,
+      "step": 10370
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3538272142474261,
+      "learning_rate": 4.1153583850637666e-07,
+      "loss": 0.1056,
+      "step": 10371
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2508500543248445,
+      "learning_rate": 4.109938647535694e-07,
+      "loss": 0.0643,
+      "step": 10372
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.28308932111462787,
+      "learning_rate": 4.1045223281667324e-07,
+      "loss": 0.0917,
+      "step": 10373
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5064738236898707,
+      "learning_rate": 4.0991094273603036e-07,
+      "loss": 0.0449,
+      "step": 10374
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.38821339350090184,
+      "learning_rate": 4.093699945519591e-07,
+      "loss": 0.1037,
+      "step": 10375
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.35831034477868734,
+      "learning_rate": 4.088293883047539e-07,
+      "loss": 0.1113,
+      "step": 10376
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.19645772333719635,
+      "learning_rate": 4.0828912403468146e-07,
+      "loss": 0.0544,
+      "step": 10377
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4092757305235048,
+      "learning_rate": 4.077492017819834e-07,
+      "loss": 0.1019,
+      "step": 10378
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.38899190655504556,
+      "learning_rate": 4.0720962158687594e-07,
+      "loss": 0.1219,
+      "step": 10379
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7656656374359254,
+      "learning_rate": 4.066703834895519e-07,
+      "loss": 0.1325,
+      "step": 10380
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2607572446676187,
+      "learning_rate": 4.0613148753017586e-07,
+      "loss": 0.0403,
+      "step": 10381
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.22854638385844114,
+      "learning_rate": 4.0559293374888673e-07,
+      "loss": 0.0698,
+      "step": 10382
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3032568377489978,
+      "learning_rate": 4.0505472218580146e-07,
+      "loss": 0.0676,
+      "step": 10383
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3450014623218361,
+      "learning_rate": 4.0451685288100785e-07,
+      "loss": 0.0917,
+      "step": 10384
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.4098843339799118,
+      "learning_rate": 4.039793258745689e-07,
+      "loss": 0.0725,
+      "step": 10385
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3021890462077491,
+      "learning_rate": 4.0344214120652426e-07,
+      "loss": 0.0904,
+      "step": 10386
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.315233855832079,
+      "learning_rate": 4.0290529891688637e-07,
+      "loss": 0.0861,
+      "step": 10387
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.38616058761388317,
+      "learning_rate": 4.0236879904564153e-07,
+      "loss": 0.0635,
+      "step": 10388
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3209315872051834,
+      "learning_rate": 4.018326416327506e-07,
+      "loss": 0.0823,
+      "step": 10389
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.18676843195558632,
+      "learning_rate": 4.0129682671815164e-07,
+      "loss": 0.0612,
+      "step": 10390
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.22725537879387447,
+      "learning_rate": 4.007613543417549e-07,
+      "loss": 0.0687,
+      "step": 10391
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2899317988800566,
+      "learning_rate": 4.0022622454344516e-07,
+      "loss": 0.0699,
+      "step": 10392
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.24421775988496255,
+      "learning_rate": 3.996914373630806e-07,
+      "loss": 0.0818,
+      "step": 10393
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.21490426332050236,
+      "learning_rate": 3.991569928404976e-07,
+      "loss": 0.06,
+      "step": 10394
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.48883729935748477,
+      "learning_rate": 3.986228910155032e-07,
+      "loss": 0.069,
+      "step": 10395
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.35525939724063726,
+      "learning_rate": 3.980891319278818e-07,
+      "loss": 0.0804,
+      "step": 10396
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.32388192670503696,
+      "learning_rate": 3.975557156173893e-07,
+      "loss": 0.0724,
+      "step": 10397
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3914793411707041,
+      "learning_rate": 3.9702264212375775e-07,
+      "loss": 0.0655,
+      "step": 10398
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.24412864501660192,
+      "learning_rate": 3.9648991148669493e-07,
+      "loss": 0.0705,
+      "step": 10399
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2495225936013205,
+      "learning_rate": 3.959575237458807e-07,
+      "loss": 0.0809,
+      "step": 10400
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3864750342929013,
+      "learning_rate": 3.954254789409706e-07,
+      "loss": 0.0844,
+      "step": 10401
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.25403259807413925,
+      "learning_rate": 3.9489377711159405e-07,
+      "loss": 0.0671,
+      "step": 10402
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2659199612816149,
+      "learning_rate": 3.94362418297356e-07,
+      "loss": 0.0829,
+      "step": 10403
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8307169870072073,
+      "learning_rate": 3.9383140253783535e-07,
+      "loss": 0.178,
+      "step": 10404
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6153387894325313,
+      "learning_rate": 3.933007298725849e-07,
+      "loss": 0.1184,
+      "step": 10405
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3062710098602297,
+      "learning_rate": 3.927704003411309e-07,
+      "loss": 0.0681,
+      "step": 10406
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2611274756209047,
+      "learning_rate": 3.9224041398297717e-07,
+      "loss": 0.0863,
+      "step": 10407
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.35101871736208623,
+      "learning_rate": 3.917107708376e-07,
+      "loss": 0.0832,
+      "step": 10408
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3788759457930212,
+      "learning_rate": 3.911814709444489e-07,
+      "loss": 0.0806,
+      "step": 10409
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.4995479574979192,
+      "learning_rate": 3.906525143429513e-07,
+      "loss": 0.1479,
+      "step": 10410
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2014908255005349,
+      "learning_rate": 3.901239010725055e-07,
+      "loss": 0.0474,
+      "step": 10411
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3435530161666291,
+      "learning_rate": 3.895956311724863e-07,
+      "loss": 0.112,
+      "step": 10412
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3134306382494912,
+      "learning_rate": 3.890677046822411e-07,
+      "loss": 0.0731,
+      "step": 10413
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.28343788372598033,
+      "learning_rate": 3.8854012164109443e-07,
+      "loss": 0.084,
+      "step": 10414
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.26248043903083584,
+      "learning_rate": 3.8801288208834386e-07,
+      "loss": 0.058,
+      "step": 10415
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.36292430607520376,
+      "learning_rate": 3.874859860632596e-07,
+      "loss": 0.079,
+      "step": 10416
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.22181257518377587,
+      "learning_rate": 3.869594336050897e-07,
+      "loss": 0.074,
+      "step": 10417
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6261155179138623,
+      "learning_rate": 3.864332247530539e-07,
+      "loss": 0.1016,
+      "step": 10418
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2375704756094497,
+      "learning_rate": 3.8590735954634694e-07,
+      "loss": 0.066,
+      "step": 10419
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.42351872738735774,
+      "learning_rate": 3.853818380241403e-07,
+      "loss": 0.0707,
+      "step": 10420
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6119375908436636,
+      "learning_rate": 3.84856660225576e-07,
+      "loss": 0.0901,
+      "step": 10421
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.20252568521423933,
+      "learning_rate": 3.843318261897727e-07,
+      "loss": 0.054,
+      "step": 10422
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3367252189959483,
+      "learning_rate": 3.8380733595582254e-07,
+      "loss": 0.0777,
+      "step": 10423
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.1790652678058368,
+      "learning_rate": 3.832831895627942e-07,
+      "loss": 0.0544,
+      "step": 10424
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.23785935808743974,
+      "learning_rate": 3.8275938704972816e-07,
+      "loss": 0.0695,
+      "step": 10425
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.23772999168644451,
+      "learning_rate": 3.8223592845563984e-07,
+      "loss": 0.0688,
+      "step": 10426
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.29237650213745126,
+      "learning_rate": 3.817128138195203e-07,
+      "loss": 0.0436,
+      "step": 10427
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3648829015912419,
+      "learning_rate": 3.8119004318033446e-07,
+      "loss": 0.0587,
+      "step": 10428
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.26562577490611833,
+      "learning_rate": 3.8066761657702067e-07,
+      "loss": 0.0821,
+      "step": 10429
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2520389770112586,
+      "learning_rate": 3.8014553404849107e-07,
+      "loss": 0.0659,
+      "step": 10430
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5312096551999453,
+      "learning_rate": 3.796237956336363e-07,
+      "loss": 0.0797,
+      "step": 10431
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3062944973342576,
+      "learning_rate": 3.7910240137131624e-07,
+      "loss": 0.0536,
+      "step": 10432
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3131665355180927,
+      "learning_rate": 3.7858135130036775e-07,
+      "loss": 0.098,
+      "step": 10433
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.4318152023368978,
+      "learning_rate": 3.780606454596025e-07,
+      "loss": 0.0587,
+      "step": 10434
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3501077846583156,
+      "learning_rate": 3.7754028388780504e-07,
+      "loss": 0.0875,
+      "step": 10435
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.45633262914274236,
+      "learning_rate": 3.7702026662373493e-07,
+      "loss": 0.085,
+      "step": 10436
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2964466053852528,
+      "learning_rate": 3.765005937061261e-07,
+      "loss": 0.0657,
+      "step": 10437
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.1380071861219292,
+      "learning_rate": 3.759812651736877e-07,
+      "loss": 0.02,
+      "step": 10438
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.32894179408353524,
+      "learning_rate": 3.7546228106510096e-07,
+      "loss": 0.0752,
+      "step": 10439
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.27178268975586095,
+      "learning_rate": 3.749436414190227e-07,
+      "loss": 0.0448,
+      "step": 10440
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.33157059734022054,
+      "learning_rate": 3.744253462740866e-07,
+      "loss": 0.0619,
+      "step": 10441
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3311711518329391,
+      "learning_rate": 3.73907395668896e-07,
+      "loss": 0.0767,
+      "step": 10442
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.43539908643047226,
+      "learning_rate": 3.7338978964203076e-07,
+      "loss": 0.0914,
+      "step": 10443
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.39211911672779165,
+      "learning_rate": 3.728725282320467e-07,
+      "loss": 0.0861,
+      "step": 10444
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5349911915057601,
+      "learning_rate": 3.723556114774718e-07,
+      "loss": 0.0887,
+      "step": 10445
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.1502168439262821,
+      "learning_rate": 3.718390394168092e-07,
+      "loss": 0.0468,
+      "step": 10446
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.42889618283909337,
+      "learning_rate": 3.713228120885348e-07,
+      "loss": 0.0961,
+      "step": 10447
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.4453911928834456,
+      "learning_rate": 3.708069295311023e-07,
+      "loss": 0.0668,
+      "step": 10448
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2679866598277611,
+      "learning_rate": 3.7029139178293704e-07,
+      "loss": 0.0554,
+      "step": 10449
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3779248824586545,
+      "learning_rate": 3.6977619888243833e-07,
+      "loss": 0.0999,
+      "step": 10450
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.20518766866395668,
+      "learning_rate": 3.6926135086798154e-07,
+      "loss": 0.0488,
+      "step": 10451
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7027321346839837,
+      "learning_rate": 3.68746847777916e-07,
+      "loss": 0.1048,
+      "step": 10452
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.25032200553723244,
+      "learning_rate": 3.682326896505628e-07,
+      "loss": 0.0725,
+      "step": 10453
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.44088445575936414,
+      "learning_rate": 3.6771887652422235e-07,
+      "loss": 0.1062,
+      "step": 10454
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2891506316173497,
+      "learning_rate": 3.672054084371646e-07,
+      "loss": 0.0882,
+      "step": 10455
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.32008525694190726,
+      "learning_rate": 3.666922854276361e-07,
+      "loss": 0.0788,
+      "step": 10456
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.4521141391172543,
+      "learning_rate": 3.6617950753385646e-07,
+      "loss": 0.1224,
+      "step": 10457
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.30370150536975143,
+      "learning_rate": 3.656670747940222e-07,
+      "loss": 0.0821,
+      "step": 10458
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3951520324096447,
+      "learning_rate": 3.651549872463006e-07,
+      "loss": 0.0902,
+      "step": 10459
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.43360259034236925,
+      "learning_rate": 3.646432449288362e-07,
+      "loss": 0.0927,
+      "step": 10460
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5755670973663171,
+      "learning_rate": 3.6413184787974456e-07,
+      "loss": 0.1544,
+      "step": 10461
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.32136463616518385,
+      "learning_rate": 3.6362079613711965e-07,
+      "loss": 0.0803,
+      "step": 10462
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.35873482886870334,
+      "learning_rate": 3.6311008973902717e-07,
+      "loss": 0.092,
+      "step": 10463
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2670205864331704,
+      "learning_rate": 3.625997287235067e-07,
+      "loss": 0.0741,
+      "step": 10464
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.1892564032663384,
+      "learning_rate": 3.6208971312857335e-07,
+      "loss": 0.0667,
+      "step": 10465
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.24514540197921247,
+      "learning_rate": 3.6158004299221506e-07,
+      "loss": 0.0649,
+      "step": 10466
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.24853166695287798,
+      "learning_rate": 3.6107071835239705e-07,
+      "loss": 0.0709,
+      "step": 10467
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3328433017063389,
+      "learning_rate": 3.6056173924705564e-07,
+      "loss": 0.0678,
+      "step": 10468
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.30700522231253863,
+      "learning_rate": 3.600531057141021e-07,
+      "loss": 0.096,
+      "step": 10469
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.4421505798983173,
+      "learning_rate": 3.595448177914229e-07,
+      "loss": 0.127,
+      "step": 10470
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6520325413532139,
+      "learning_rate": 3.5903687551687884e-07,
+      "loss": 0.0838,
+      "step": 10471
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2030842161751872,
+      "learning_rate": 3.585292789283035e-07,
+      "loss": 0.052,
+      "step": 10472
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.24212380855308172,
+      "learning_rate": 3.5802202806350664e-07,
+      "loss": 0.0678,
+      "step": 10473
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3263595315933203,
+      "learning_rate": 3.575151229602697e-07,
+      "loss": 0.0893,
+      "step": 10474
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.352126983671568,
+      "learning_rate": 3.570085636563514e-07,
+      "loss": 0.1094,
+      "step": 10475
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3658986824898356,
+      "learning_rate": 3.5650235018948264e-07,
+      "loss": 0.0741,
+      "step": 10476
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.26298437411133213,
+      "learning_rate": 3.5599648259736883e-07,
+      "loss": 0.0589,
+      "step": 10477
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2521789956647028,
+      "learning_rate": 3.5549096091769096e-07,
+      "loss": 0.0919,
+      "step": 10478
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.40582955334202075,
+      "learning_rate": 3.5498578518810276e-07,
+      "loss": 0.1223,
+      "step": 10479
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.18308185199039048,
+      "learning_rate": 3.5448095544623194e-07,
+      "loss": 0.047,
+      "step": 10480
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.4666904672104624,
+      "learning_rate": 3.539764717296812e-07,
+      "loss": 0.1083,
+      "step": 10481
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3950306175657269,
+      "learning_rate": 3.534723340760288e-07,
+      "loss": 0.0806,
+      "step": 10482
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3391062142550214,
+      "learning_rate": 3.529685425228252e-07,
+      "loss": 0.0898,
+      "step": 10483
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.37638710972000183,
+      "learning_rate": 3.5246509710759434e-07,
+      "loss": 0.0547,
+      "step": 10484
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.17248565962104662,
+      "learning_rate": 3.519619978678379e-07,
+      "loss": 0.0406,
+      "step": 10485
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.25024692474359556,
+      "learning_rate": 3.5145924484102866e-07,
+      "loss": 0.0745,
+      "step": 10486
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.33672913772656693,
+      "learning_rate": 3.50956838064615e-07,
+      "loss": 0.072,
+      "step": 10487
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.47644391097858924,
+      "learning_rate": 3.5045477757601754e-07,
+      "loss": 0.0965,
+      "step": 10488
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.26358756306344006,
+      "learning_rate": 3.4995306341263414e-07,
+      "loss": 0.0537,
+      "step": 10489
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.21333784941205572,
+      "learning_rate": 3.49451695611836e-07,
+      "loss": 0.0794,
+      "step": 10490
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3468605773610321,
+      "learning_rate": 3.489506742109655e-07,
+      "loss": 0.1292,
+      "step": 10491
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2532132240303655,
+      "learning_rate": 3.4844999924734444e-07,
+      "loss": 0.0549,
+      "step": 10492
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.33763460805989276,
+      "learning_rate": 3.479496707582647e-07,
+      "loss": 0.0646,
+      "step": 10493
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.47471329047445865,
+      "learning_rate": 3.474496887809925e-07,
+      "loss": 0.0944,
+      "step": 10494
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5184970627391564,
+      "learning_rate": 3.4695005335277196e-07,
+      "loss": 0.0927,
+      "step": 10495
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5482564924865132,
+      "learning_rate": 3.464507645108167e-07,
+      "loss": 0.1072,
+      "step": 10496
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.26146433915154516,
+      "learning_rate": 3.4595182229231805e-07,
+      "loss": 0.0765,
+      "step": 10497
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2833952828881233,
+      "learning_rate": 3.4545322673443846e-07,
+      "loss": 0.06,
+      "step": 10498
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.25964675330088915,
+      "learning_rate": 3.449549778743183e-07,
+      "loss": 0.0904,
+      "step": 10499
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.24525051661574676,
+      "learning_rate": 3.4445707574906896e-07,
+      "loss": 0.0665,
+      "step": 10500
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.28054263243408406,
+      "learning_rate": 3.439595203957763e-07,
+      "loss": 0.0832,
+      "step": 10501
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2846810692005072,
+      "learning_rate": 3.4346231185150237e-07,
+      "loss": 0.0796,
+      "step": 10502
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.406924815608045,
+      "learning_rate": 3.4296545015328254e-07,
+      "loss": 0.0666,
+      "step": 10503
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3325068894870213,
+      "learning_rate": 3.42468935338125e-07,
+      "loss": 0.1097,
+      "step": 10504
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.37001391262801203,
+      "learning_rate": 3.4197276744301233e-07,
+      "loss": 0.0703,
+      "step": 10505
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.4522199526074526,
+      "learning_rate": 3.4147694650490337e-07,
+      "loss": 0.0744,
+      "step": 10506
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3460647939827886,
+      "learning_rate": 3.409814725607302e-07,
+      "loss": 0.0583,
+      "step": 10507
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.18811806538613207,
+      "learning_rate": 3.4048634564739615e-07,
+      "loss": 0.0569,
+      "step": 10508
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3294954360897352,
+      "learning_rate": 3.3999156580178384e-07,
+      "loss": 0.0894,
+      "step": 10509
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.27252655235928647,
+      "learning_rate": 3.3949713306074663e-07,
+      "loss": 0.0613,
+      "step": 10510
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.33323910518383537,
+      "learning_rate": 3.390030474611111e-07,
+      "loss": 0.0759,
+      "step": 10511
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.20062127867840693,
+      "learning_rate": 3.385093090396818e-07,
+      "loss": 0.0528,
+      "step": 10512
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3390222959342926,
+      "learning_rate": 3.3801591783323427e-07,
+      "loss": 0.0677,
+      "step": 10513
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2473870775536169,
+      "learning_rate": 3.3752287387851966e-07,
+      "loss": 0.0448,
+      "step": 10514
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.34236027484224807,
+      "learning_rate": 3.370301772122614e-07,
+      "loss": 0.0825,
+      "step": 10515
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3583592990098601,
+      "learning_rate": 3.3653782787116007e-07,
+      "loss": 0.0979,
+      "step": 10516
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.31999527877070755,
+      "learning_rate": 3.3604582589188804e-07,
+      "loss": 0.0758,
+      "step": 10517
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.25868058460947924,
+      "learning_rate": 3.355541713110916e-07,
+      "loss": 0.0526,
+      "step": 10518
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.30531092742352656,
+      "learning_rate": 3.3506286416539356e-07,
+      "loss": 0.0913,
+      "step": 10519
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2514835573253783,
+      "learning_rate": 3.345719044913887e-07,
+      "loss": 0.0589,
+      "step": 10520
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2554346321368242,
+      "learning_rate": 3.3408129232564643e-07,
+      "loss": 0.0941,
+      "step": 10521
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.42332144022705254,
+      "learning_rate": 3.3359102770470995e-07,
+      "loss": 0.0787,
+      "step": 10522
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.25057039178563384,
+      "learning_rate": 3.331011106650983e-07,
+      "loss": 0.0734,
+      "step": 10523
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.22888472576557964,
+      "learning_rate": 3.326115412433023e-07,
+      "loss": 0.0696,
+      "step": 10524
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2837018169304391,
+      "learning_rate": 3.3212231947578776e-07,
+      "loss": 0.0596,
+      "step": 10525
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.33671498312772125,
+      "learning_rate": 3.316334453989961e-07,
+      "loss": 0.0938,
+      "step": 10526
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.19046463705906086,
+      "learning_rate": 3.31144919049341e-07,
+      "loss": 0.0367,
+      "step": 10527
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.26463219555433937,
+      "learning_rate": 3.306567404632099e-07,
+      "loss": 0.0739,
+      "step": 10528
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.4454770489519749,
+      "learning_rate": 3.3016890967696505e-07,
+      "loss": 0.0548,
+      "step": 10529
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.31359564436429543,
+      "learning_rate": 3.296814267269444e-07,
+      "loss": 0.1005,
+      "step": 10530
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3273724667288739,
+      "learning_rate": 3.2919429164945727e-07,
+      "loss": 0.087,
+      "step": 10531
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.4518487269253093,
+      "learning_rate": 3.287075044807891e-07,
+      "loss": 0.0822,
+      "step": 10532
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.29733008655387794,
+      "learning_rate": 3.2822106525719876e-07,
+      "loss": 0.0719,
+      "step": 10533
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.28975007308871087,
+      "learning_rate": 3.277349740149172e-07,
+      "loss": 0.0764,
+      "step": 10534
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.15990816976795472,
+      "learning_rate": 3.272492307901537e-07,
+      "loss": 0.0346,
+      "step": 10535
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3947060466676695,
+      "learning_rate": 3.267638356190883e-07,
+      "loss": 0.0663,
+      "step": 10536
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.584502766580984,
+      "learning_rate": 3.2627878853787596e-07,
+      "loss": 0.0748,
+      "step": 10537
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2926138660466728,
+      "learning_rate": 3.2579408958264555e-07,
+      "loss": 0.0641,
+      "step": 10538
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3054040774427007,
+      "learning_rate": 3.25309738789501e-07,
+      "loss": 0.1105,
+      "step": 10539
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.4410149104236629,
+      "learning_rate": 3.2482573619451897e-07,
+      "loss": 0.1148,
+      "step": 10540
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5433760761701039,
+      "learning_rate": 3.2434208183375114e-07,
+      "loss": 0.1131,
+      "step": 10541
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.49555734878324664,
+      "learning_rate": 3.238587757432221e-07,
+      "loss": 0.0833,
+      "step": 10542
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2662690571406797,
+      "learning_rate": 3.2337581795893304e-07,
+      "loss": 0.0708,
+      "step": 10543
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.33803000823296375,
+      "learning_rate": 3.228932085168562e-07,
+      "loss": 0.077,
+      "step": 10544
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.4042093060129459,
+      "learning_rate": 3.2241094745293857e-07,
+      "loss": 0.0862,
+      "step": 10545
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2393566917912546,
+      "learning_rate": 3.2192903480310287e-07,
+      "loss": 0.0778,
+      "step": 10546
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3668246107966983,
+      "learning_rate": 3.21447470603245e-07,
+      "loss": 0.08,
+      "step": 10547
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.34951635990558216,
+      "learning_rate": 3.2096625488923394e-07,
+      "loss": 0.0836,
+      "step": 10548
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.35209366397886566,
+      "learning_rate": 3.2048538769691274e-07,
+      "loss": 0.1049,
+      "step": 10549
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.678524513533921,
+      "learning_rate": 3.2000486906210105e-07,
+      "loss": 0.1071,
+      "step": 10550
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2741064983489201,
+      "learning_rate": 3.1952469902058967e-07,
+      "loss": 0.0708,
+      "step": 10551
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.288369291183368,
+      "learning_rate": 3.190448776081434e-07,
+      "loss": 0.0771,
+      "step": 10552
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.37482059852323857,
+      "learning_rate": 3.1856540486050414e-07,
+      "loss": 0.0979,
+      "step": 10553
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.30145012252113956,
+      "learning_rate": 3.18086280813385e-07,
+      "loss": 0.0708,
+      "step": 10554
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3064269380223725,
+      "learning_rate": 3.176075055024741e-07,
+      "loss": 0.0834,
+      "step": 10555
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.26874681344461454,
+      "learning_rate": 3.171290789634318e-07,
+      "loss": 0.0674,
+      "step": 10556
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.4059155035424254,
+      "learning_rate": 3.166510012318963e-07,
+      "loss": 0.0843,
+      "step": 10557
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.30266201892715733,
+      "learning_rate": 3.1617327234347685e-07,
+      "loss": 0.0751,
+      "step": 10558
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.27986848896118965,
+      "learning_rate": 3.156958923337572e-07,
+      "loss": 0.0642,
+      "step": 10559
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.4828619497745637,
+      "learning_rate": 3.152188612382956e-07,
+      "loss": 0.1315,
+      "step": 10560
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.4761244133583572,
+      "learning_rate": 3.1474217909262415e-07,
+      "loss": 0.133,
+      "step": 10561
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3280095053415452,
+      "learning_rate": 3.142658459322484e-07,
+      "loss": 0.074,
+      "step": 10562
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3092142711676505,
+      "learning_rate": 3.137898617926494e-07,
+      "loss": 0.1016,
+      "step": 10563
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.4611464566894081,
+      "learning_rate": 3.133142267092809e-07,
+      "loss": 0.142,
+      "step": 10564
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.4871804063392949,
+      "learning_rate": 3.1283894071757027e-07,
+      "loss": 0.0952,
+      "step": 10565
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.26152097126464496,
+      "learning_rate": 3.1236400385291966e-07,
+      "loss": 0.0602,
+      "step": 10566
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.27470637997284963,
+      "learning_rate": 3.1188941615070635e-07,
+      "loss": 0.0848,
+      "step": 10567
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6272577386750017,
+      "learning_rate": 3.1141517764627983e-07,
+      "loss": 0.0734,
+      "step": 10568
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.30903804672318314,
+      "learning_rate": 3.10941288374963e-07,
+      "loss": 0.0534,
+      "step": 10569
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.702414945729997,
+      "learning_rate": 3.104677483720553e-07,
+      "loss": 0.0815,
+      "step": 10570
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3330389424097082,
+      "learning_rate": 3.09994557672828e-07,
+      "loss": 0.0719,
+      "step": 10571
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3240480399560202,
+      "learning_rate": 3.09521716312528e-07,
+      "loss": 0.0727,
+      "step": 10572
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3895906169163102,
+      "learning_rate": 3.0904922432637373e-07,
+      "loss": 0.0893,
+      "step": 10573
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.1592286078402547,
+      "learning_rate": 3.0857708174956034e-07,
+      "loss": 0.0506,
+      "step": 10574
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2580433824725031,
+      "learning_rate": 3.0810528861725586e-07,
+      "loss": 0.0389,
+      "step": 10575
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6450703521510163,
+      "learning_rate": 3.0763384496460104e-07,
+      "loss": 0.1391,
+      "step": 10576
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5987838311500814,
+      "learning_rate": 3.0716275082671276e-07,
+      "loss": 0.1338,
+      "step": 10577
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.42137137186709683,
+      "learning_rate": 3.0669200623868136e-07,
+      "loss": 0.1084,
+      "step": 10578
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.26046920342369423,
+      "learning_rate": 3.062216112355687e-07,
+      "loss": 0.0508,
+      "step": 10579
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.22880540521923015,
+      "learning_rate": 3.0575156585241463e-07,
+      "loss": 0.0431,
+      "step": 10580
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.45656405730575056,
+      "learning_rate": 3.052818701242294e-07,
+      "loss": 0.0759,
+      "step": 10581
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.42194495348293215,
+      "learning_rate": 3.0481252408600003e-07,
+      "loss": 0.1288,
+      "step": 10582
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.32962733916510273,
+      "learning_rate": 3.043435277726842e-07,
+      "loss": 0.0695,
+      "step": 10583
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3801706104726801,
+      "learning_rate": 3.0387488121921716e-07,
+      "loss": 0.1055,
+      "step": 10584
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.19630267906230492,
+      "learning_rate": 3.0340658446050665e-07,
+      "loss": 0.0407,
+      "step": 10585
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3308721668865176,
+      "learning_rate": 3.0293863753143193e-07,
+      "loss": 0.1011,
+      "step": 10586
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.4662723227989085,
+      "learning_rate": 3.0247104046685073e-07,
+      "loss": 0.0999,
+      "step": 10587
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.31103857353863984,
+      "learning_rate": 3.0200379330159177e-07,
+      "loss": 0.09,
+      "step": 10588
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.4577575370164702,
+      "learning_rate": 3.015368960704584e-07,
+      "loss": 0.0968,
+      "step": 10589
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2570775377556307,
+      "learning_rate": 3.0107034880822674e-07,
+      "loss": 0.0781,
+      "step": 10590
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.34312737924216574,
+      "learning_rate": 3.0060415154964937e-07,
+      "loss": 0.0818,
+      "step": 10591
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.14871748548183075,
+      "learning_rate": 3.0013830432945037e-07,
+      "loss": 0.045,
+      "step": 10592
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3386365312809842,
+      "learning_rate": 2.9967280718232903e-07,
+      "loss": 0.1087,
+      "step": 10593
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.33828358721708823,
+      "learning_rate": 2.9920766014295886e-07,
+      "loss": 0.0746,
+      "step": 10594
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2738901649331908,
+      "learning_rate": 2.9874286324598657e-07,
+      "loss": 0.0482,
+      "step": 10595
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2712945454395805,
+      "learning_rate": 2.9827841652603273e-07,
+      "loss": 0.0851,
+      "step": 10596
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.31652130956199204,
+      "learning_rate": 2.9781432001769086e-07,
+      "loss": 0.1016,
+      "step": 10597
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3625633838994426,
+      "learning_rate": 2.9735057375553154e-07,
+      "loss": 0.0592,
+      "step": 10598
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.23771259681559842,
+      "learning_rate": 2.968871777740967e-07,
+      "loss": 0.0363,
+      "step": 10599
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3101318905081017,
+      "learning_rate": 2.9642413210790253e-07,
+      "loss": 0.0695,
+      "step": 10600
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.41914793798237976,
+      "learning_rate": 2.959614367914387e-07,
+      "loss": 0.1039,
+      "step": 10601
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.24704702339604917,
+      "learning_rate": 2.9549909185916993e-07,
+      "loss": 0.0674,
+      "step": 10602
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3022853090130656,
+      "learning_rate": 2.950370973455352e-07,
+      "loss": 0.0835,
+      "step": 10603
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.22099288861972136,
+      "learning_rate": 2.9457545328494596e-07,
+      "loss": 0.0634,
+      "step": 10604
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.26939161956845403,
+      "learning_rate": 2.94114159711788e-07,
+      "loss": 0.0413,
+      "step": 10605
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3837592934915364,
+      "learning_rate": 2.936532166604206e-07,
+      "loss": 0.1,
+      "step": 10606
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.36370660818135603,
+      "learning_rate": 2.931926241651789e-07,
+      "loss": 0.0913,
+      "step": 10607
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.33039888658643185,
+      "learning_rate": 2.9273238226036947e-07,
+      "loss": 0.0942,
+      "step": 10608
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.31153450001509664,
+      "learning_rate": 2.9227249098027434e-07,
+      "loss": 0.0571,
+      "step": 10609
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.18881085738850661,
+      "learning_rate": 2.918129503591477e-07,
+      "loss": 0.0398,
+      "step": 10610
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3336975147296537,
+      "learning_rate": 2.9135376043122056e-07,
+      "loss": 0.0896,
+      "step": 10611
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.1571289423943912,
+      "learning_rate": 2.9089492123069497e-07,
+      "loss": 0.0432,
+      "step": 10612
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2689176169190801,
+      "learning_rate": 2.904364327917486e-07,
+      "loss": 0.0602,
+      "step": 10613
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.32584866666490875,
+      "learning_rate": 2.8997829514853137e-07,
+      "loss": 0.0491,
+      "step": 10614
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3557088237261066,
+      "learning_rate": 2.8952050833516877e-07,
+      "loss": 0.0562,
+      "step": 10615
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.30025664800341817,
+      "learning_rate": 2.890630723857596e-07,
+      "loss": 0.0413,
+      "step": 10616
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2457728564681372,
+      "learning_rate": 2.886059873343755e-07,
+      "loss": 0.078,
+      "step": 10617
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.4243062270469183,
+      "learning_rate": 2.881492532150637e-07,
+      "loss": 0.0539,
+      "step": 10618
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.24500580977027087,
+      "learning_rate": 2.876928700618442e-07,
+      "loss": 0.0627,
+      "step": 10619
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.23459753992949234,
+      "learning_rate": 2.872368379087104e-07,
+      "loss": 0.0539,
+      "step": 10620
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.26208057601175905,
+      "learning_rate": 2.8678115678963115e-07,
+      "loss": 0.0565,
+      "step": 10621
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2877380084663504,
+      "learning_rate": 2.8632582673854825e-07,
+      "loss": 0.0624,
+      "step": 10622
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.36095301786344913,
+      "learning_rate": 2.858708477893768e-07,
+      "loss": 0.0905,
+      "step": 10623
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.37438226972767596,
+      "learning_rate": 2.8541621997600575e-07,
+      "loss": 0.0897,
+      "step": 10624
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2753598538508495,
+      "learning_rate": 2.8496194333229975e-07,
+      "loss": 0.0974,
+      "step": 10625
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.21850999250467382,
+      "learning_rate": 2.845080178920956e-07,
+      "loss": 0.0416,
+      "step": 10626
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3214391115033729,
+      "learning_rate": 2.840544436892029e-07,
+      "loss": 0.0806,
+      "step": 10627
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.25545571585626325,
+      "learning_rate": 2.836012207574085e-07,
+      "loss": 0.0599,
+      "step": 10628
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.36986655432666005,
+      "learning_rate": 2.831483491304704e-07,
+      "loss": 0.0555,
+      "step": 10629
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.28101986268397056,
+      "learning_rate": 2.826958288421211e-07,
+      "loss": 0.0713,
+      "step": 10630
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2659055513523229,
+      "learning_rate": 2.8224365992606584e-07,
+      "loss": 0.0758,
+      "step": 10631
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.26775905010852646,
+      "learning_rate": 2.8179184241598603e-07,
+      "loss": 0.0812,
+      "step": 10632
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3012004691828637,
+      "learning_rate": 2.8134037634553633e-07,
+      "loss": 0.0736,
+      "step": 10633
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5405324750854406,
+      "learning_rate": 2.808892617483422e-07,
+      "loss": 0.0912,
+      "step": 10634
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3046402271099459,
+      "learning_rate": 2.8043849865800774e-07,
+      "loss": 0.0733,
+      "step": 10635
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.39863037753583463,
+      "learning_rate": 2.799880871081073e-07,
+      "loss": 0.079,
+      "step": 10636
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.40206253885735577,
+      "learning_rate": 2.7953802713218957e-07,
+      "loss": 0.1257,
+      "step": 10637
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3489190058449629,
+      "learning_rate": 2.7908831876377885e-07,
+      "loss": 0.0939,
+      "step": 10638
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.20489123775638518,
+      "learning_rate": 2.7863896203637165e-07,
+      "loss": 0.0461,
+      "step": 10639
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2824713187504084,
+      "learning_rate": 2.781899569834384e-07,
+      "loss": 0.0803,
+      "step": 10640
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.20221908958629,
+      "learning_rate": 2.77741303638423e-07,
+      "loss": 0.0467,
+      "step": 10641
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.21658574694668653,
+      "learning_rate": 2.772930020347453e-07,
+      "loss": 0.0407,
+      "step": 10642
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.337935501775752,
+      "learning_rate": 2.768450522057964e-07,
+      "loss": 0.076,
+      "step": 10643
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.21048115227383948,
+      "learning_rate": 2.7639745418494233e-07,
+      "loss": 0.0593,
+      "step": 10644
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.40982669711008407,
+      "learning_rate": 2.759502080055232e-07,
+      "loss": 0.0843,
+      "step": 10645
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.38198004624061155,
+      "learning_rate": 2.755033137008517e-07,
+      "loss": 0.0457,
+      "step": 10646
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2859408319962393,
+      "learning_rate": 2.7505677130421624e-07,
+      "loss": 0.098,
+      "step": 10647
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2509317574694417,
+      "learning_rate": 2.746105808488764e-07,
+      "loss": 0.0614,
+      "step": 10648
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.51728116132231,
+      "learning_rate": 2.741647423680688e-07,
+      "loss": 0.0898,
+      "step": 10649
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2986396436562258,
+      "learning_rate": 2.7371925589500137e-07,
+      "loss": 0.0564,
+      "step": 10650
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.268150260584324,
+      "learning_rate": 2.7327412146285536e-07,
+      "loss": 0.0677,
+      "step": 10651
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.20202805153272946,
+      "learning_rate": 2.728293391047887e-07,
+      "loss": 0.0428,
+      "step": 10652
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.37578281823150106,
+      "learning_rate": 2.723849088539304e-07,
+      "loss": 0.138,
+      "step": 10653
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4570903178962846,
+      "learning_rate": 2.7194083074338405e-07,
+      "loss": 0.0854,
+      "step": 10654
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.41159020157343845,
+      "learning_rate": 2.714971048062287e-07,
+      "loss": 0.0745,
+      "step": 10655
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.30365278470017293,
+      "learning_rate": 2.71053731075514e-07,
+      "loss": 0.0634,
+      "step": 10656
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.30157206353722177,
+      "learning_rate": 2.7061070958426585e-07,
+      "loss": 0.0834,
+      "step": 10657
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2015676863928254,
+      "learning_rate": 2.7016804036548214e-07,
+      "loss": 0.0466,
+      "step": 10658
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.33018156256166303,
+      "learning_rate": 2.697257234521367e-07,
+      "loss": 0.0816,
+      "step": 10659
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.461212713954877,
+      "learning_rate": 2.6928375887717516e-07,
+      "loss": 0.092,
+      "step": 10660
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.38600741659420473,
+      "learning_rate": 2.6884214667351694e-07,
+      "loss": 0.0932,
+      "step": 10661
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2911092057464604,
+      "learning_rate": 2.684008868740573e-07,
+      "loss": 0.0587,
+      "step": 10662
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.34345628494224295,
+      "learning_rate": 2.6795997951166333e-07,
+      "loss": 0.0667,
+      "step": 10663
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.15689201314123802,
+      "learning_rate": 2.6751942461917646e-07,
+      "loss": 0.0454,
+      "step": 10664
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3155251662219744,
+      "learning_rate": 2.670792222294105e-07,
+      "loss": 0.0597,
+      "step": 10665
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.27330919797660425,
+      "learning_rate": 2.6663937237515646e-07,
+      "loss": 0.0651,
+      "step": 10666
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5820658352893192,
+      "learning_rate": 2.661998750891759e-07,
+      "loss": 0.1526,
+      "step": 10667
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4455780710192732,
+      "learning_rate": 2.657607304042048e-07,
+      "loss": 0.0891,
+      "step": 10668
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.39254660622402754,
+      "learning_rate": 2.6532193835295374e-07,
+      "loss": 0.0928,
+      "step": 10669
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5070593765650299,
+      "learning_rate": 2.648834989681054e-07,
+      "loss": 0.0949,
+      "step": 10670
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.29509843400299524,
+      "learning_rate": 2.6444541228231867e-07,
+      "loss": 0.0723,
+      "step": 10671
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3779101798218891,
+      "learning_rate": 2.6400767832822414e-07,
+      "loss": 0.107,
+      "step": 10672
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5315043888151257,
+      "learning_rate": 2.635702971384274e-07,
+      "loss": 0.1013,
+      "step": 10673
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4889985446644459,
+      "learning_rate": 2.631332687455057e-07,
+      "loss": 0.1362,
+      "step": 10674
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.23095392097665712,
+      "learning_rate": 2.62696593182013e-07,
+      "loss": 0.0551,
+      "step": 10675
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3154703360187384,
+      "learning_rate": 2.62260270480475e-07,
+      "loss": 0.0802,
+      "step": 10676
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2169419453313571,
+      "learning_rate": 2.618243006733917e-07,
+      "loss": 0.03,
+      "step": 10677
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.32075604273952646,
+      "learning_rate": 2.613886837932356e-07,
+      "loss": 0.0708,
+      "step": 10678
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.35938608771554126,
+      "learning_rate": 2.6095341987245503e-07,
+      "loss": 0.11,
+      "step": 10679
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2851445840072855,
+      "learning_rate": 2.605185089434714e-07,
+      "loss": 0.0567,
+      "step": 10680
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.252652873899567,
+      "learning_rate": 2.6008395103867825e-07,
+      "loss": 0.0555,
+      "step": 10681
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.34964815149536616,
+      "learning_rate": 2.5964974619044403e-07,
+      "loss": 0.0552,
+      "step": 10682
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2577347229942725,
+      "learning_rate": 2.592158944311118e-07,
+      "loss": 0.0992,
+      "step": 10683
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.23998755551313036,
+      "learning_rate": 2.587823957929969e-07,
+      "loss": 0.0367,
+      "step": 10684
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.28547270389760815,
+      "learning_rate": 2.583492503083884e-07,
+      "loss": 0.0833,
+      "step": 10685
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3640939669729773,
+      "learning_rate": 2.5791645800955055e-07,
+      "loss": 0.1208,
+      "step": 10686
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7762202913151485,
+      "learning_rate": 2.5748401892871923e-07,
+      "loss": 0.0861,
+      "step": 10687
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.33739724603665006,
+      "learning_rate": 2.5705193309810475e-07,
+      "loss": 0.0698,
+      "step": 10688
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.276461230913421,
+      "learning_rate": 2.56620200549893e-07,
+      "loss": 0.0789,
+      "step": 10689
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6546707167915755,
+      "learning_rate": 2.561888213162406e-07,
+      "loss": 0.1289,
+      "step": 10690
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2609747241904819,
+      "learning_rate": 2.5575779542928e-07,
+      "loss": 0.0476,
+      "step": 10691
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.35659619315624297,
+      "learning_rate": 2.553271229211146e-07,
+      "loss": 0.0675,
+      "step": 10692
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.363098873056821,
+      "learning_rate": 2.548968038238264e-07,
+      "loss": 0.076,
+      "step": 10693
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.25278872524319185,
+      "learning_rate": 2.5446683816946593e-07,
+      "loss": 0.0801,
+      "step": 10694
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3907987234237108,
+      "learning_rate": 2.5403722599005976e-07,
+      "loss": 0.0482,
+      "step": 10695
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.26626351353033323,
+      "learning_rate": 2.536079673176084e-07,
+      "loss": 0.0411,
+      "step": 10696
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2631643809052591,
+      "learning_rate": 2.531790621840863e-07,
+      "loss": 0.0688,
+      "step": 10697
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.31159268215396135,
+      "learning_rate": 2.527505106214395e-07,
+      "loss": 0.0862,
+      "step": 10698
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5037018353706154,
+      "learning_rate": 2.523223126615887e-07,
+      "loss": 0.1016,
+      "step": 10699
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.32808332314128585,
+      "learning_rate": 2.5189446833642997e-07,
+      "loss": 0.0687,
+      "step": 10700
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5471854355134675,
+      "learning_rate": 2.5146697767783114e-07,
+      "loss": 0.0987,
+      "step": 10701
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2644053146360624,
+      "learning_rate": 2.510398407176329e-07,
+      "loss": 0.0722,
+      "step": 10702
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2259187429182303,
+      "learning_rate": 2.506130574876531e-07,
+      "loss": 0.0705,
+      "step": 10703
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2240345005194627,
+      "learning_rate": 2.5018662801968016e-07,
+      "loss": 0.0626,
+      "step": 10704
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2299818547519316,
+      "learning_rate": 2.49760552345476e-07,
+      "loss": 0.0748,
+      "step": 10705
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.282624728294897,
+      "learning_rate": 2.493348304967791e-07,
+      "loss": 0.047,
+      "step": 10706
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.18010589559773116,
+      "learning_rate": 2.4890946250529844e-07,
+      "loss": 0.0523,
+      "step": 10707
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2095888690906835,
+      "learning_rate": 2.484844484027182e-07,
+      "loss": 0.0479,
+      "step": 10708
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5544875648615079,
+      "learning_rate": 2.480597882206948e-07,
+      "loss": 0.0681,
+      "step": 10709
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.394415306368715,
+      "learning_rate": 2.476354819908616e-07,
+      "loss": 0.0702,
+      "step": 10710
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3040860655207578,
+      "learning_rate": 2.472115297448224e-07,
+      "loss": 0.0856,
+      "step": 10711
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4901303532813168,
+      "learning_rate": 2.4678793151415415e-07,
+      "loss": 0.1181,
+      "step": 10712
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4213804271480601,
+      "learning_rate": 2.463646873304115e-07,
+      "loss": 0.094,
+      "step": 10713
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.41597075375835674,
+      "learning_rate": 2.459417972251188e-07,
+      "loss": 0.1252,
+      "step": 10714
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.25987022287689737,
+      "learning_rate": 2.4551926122977575e-07,
+      "loss": 0.0809,
+      "step": 10715
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3392275105352114,
+      "learning_rate": 2.450970793758539e-07,
+      "loss": 0.0739,
+      "step": 10716
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.40045914471456195,
+      "learning_rate": 2.44675251694802e-07,
+      "loss": 0.084,
+      "step": 10717
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.55505291727462,
+      "learning_rate": 2.4425377821803875e-07,
+      "loss": 0.1019,
+      "step": 10718
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.29148165496352124,
+      "learning_rate": 2.4383265897695794e-07,
+      "loss": 0.0486,
+      "step": 10719
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.18579742626577006,
+      "learning_rate": 2.4341189400292784e-07,
+      "loss": 0.0449,
+      "step": 10720
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4755753008009053,
+      "learning_rate": 2.429914833272895e-07,
+      "loss": 0.1071,
+      "step": 10721
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5503858920489393,
+      "learning_rate": 2.425714269813567e-07,
+      "loss": 0.131,
+      "step": 10722
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2840245429736912,
+      "learning_rate": 2.4215172499641784e-07,
+      "loss": 0.0764,
+      "step": 10723
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7959610624823609,
+      "learning_rate": 2.41732377403735e-07,
+      "loss": 0.1311,
+      "step": 10724
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2923292515261836,
+      "learning_rate": 2.413133842345444e-07,
+      "loss": 0.0846,
+      "step": 10725
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5061462059865495,
+      "learning_rate": 2.4089474552005334e-07,
+      "loss": 0.1034,
+      "step": 10726
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.23667007963378325,
+      "learning_rate": 2.404764612914462e-07,
+      "loss": 0.0399,
+      "step": 10727
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.26619665370728324,
+      "learning_rate": 2.4005853157987813e-07,
+      "loss": 0.0789,
+      "step": 10728
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3297421159426025,
+      "learning_rate": 2.3964095641647924e-07,
+      "loss": 0.0824,
+      "step": 10729
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.24051887425020937,
+      "learning_rate": 2.392237358323535e-07,
+      "loss": 0.0532,
+      "step": 10730
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.27031493360775616,
+      "learning_rate": 2.3880686985857714e-07,
+      "loss": 0.0705,
+      "step": 10731
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.22823239887082386,
+      "learning_rate": 2.3839035852620152e-07,
+      "loss": 0.0464,
+      "step": 10732
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3302753035625371,
+      "learning_rate": 2.3797420186624953e-07,
+      "loss": 0.087,
+      "step": 10733
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4609507075350674,
+      "learning_rate": 2.3755839990972086e-07,
+      "loss": 0.126,
+      "step": 10734
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3018838806075609,
+      "learning_rate": 2.371429526875857e-07,
+      "loss": 0.105,
+      "step": 10735
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4825051800044049,
+      "learning_rate": 2.3672786023078876e-07,
+      "loss": 0.0983,
+      "step": 10736
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6981423643813979,
+      "learning_rate": 2.3631312257024918e-07,
+      "loss": 0.1595,
+      "step": 10737
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4292411778929159,
+      "learning_rate": 2.3589873973685839e-07,
+      "loss": 0.1199,
+      "step": 10738
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.43114024003540075,
+      "learning_rate": 2.354847117614828e-07,
+      "loss": 0.155,
+      "step": 10739
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3028976562245556,
+      "learning_rate": 2.3507103867496107e-07,
+      "loss": 0.0807,
+      "step": 10740
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.23623535731175438,
+      "learning_rate": 2.3465772050810632e-07,
+      "loss": 0.0769,
+      "step": 10741
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.20982305609758156,
+      "learning_rate": 2.3424475729170392e-07,
+      "loss": 0.0572,
+      "step": 10742
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3545383002188968,
+      "learning_rate": 2.3383214905651542e-07,
+      "loss": 0.1008,
+      "step": 10743
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5684953917210803,
+      "learning_rate": 2.334198958332734e-07,
+      "loss": 0.1328,
+      "step": 10744
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3151669729825578,
+      "learning_rate": 2.3300799765268445e-07,
+      "loss": 0.0809,
+      "step": 10745
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.45486957500255654,
+      "learning_rate": 2.3259645454542956e-07,
+      "loss": 0.1128,
+      "step": 10746
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3232602887947338,
+      "learning_rate": 2.321852665421631e-07,
+      "loss": 0.0889,
+      "step": 10747
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5640674518744295,
+      "learning_rate": 2.3177443367351338e-07,
+      "loss": 0.1338,
+      "step": 10748
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3821172908457758,
+      "learning_rate": 2.3136395597007976e-07,
+      "loss": 0.0923,
+      "step": 10749
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.43088544534987544,
+      "learning_rate": 2.3095383346243837e-07,
+      "loss": 0.0878,
+      "step": 10750
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.41029279011108094,
+      "learning_rate": 2.30544066181137e-07,
+      "loss": 0.1207,
+      "step": 10751
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.25537402727854897,
+      "learning_rate": 2.3013465415669845e-07,
+      "loss": 0.0665,
+      "step": 10752
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.41650577359191204,
+      "learning_rate": 2.297255974196161e-07,
+      "loss": 0.1124,
+      "step": 10753
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.29977458824375897,
+      "learning_rate": 2.293168960003611e-07,
+      "loss": 0.0611,
+      "step": 10754
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.36985380465025514,
+      "learning_rate": 2.2890854992937528e-07,
+      "loss": 0.1063,
+      "step": 10755
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2866156418507742,
+      "learning_rate": 2.285005592370737e-07,
+      "loss": 0.0785,
+      "step": 10756
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.31616866269159105,
+      "learning_rate": 2.2809292395384652e-07,
+      "loss": 0.1045,
+      "step": 10757
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2785008336608391,
+      "learning_rate": 2.276856441100572e-07,
+      "loss": 0.0791,
+      "step": 10758
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.4713637643662592,
+      "learning_rate": 2.2727871973604155e-07,
+      "loss": 0.0839,
+      "step": 10759
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.33857644626116934,
+      "learning_rate": 2.2687215086210967e-07,
+      "loss": 0.0716,
+      "step": 10760
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.24868341592862014,
+      "learning_rate": 2.2646593751854573e-07,
+      "loss": 0.0595,
+      "step": 10761
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.23295129361557404,
+      "learning_rate": 2.2606007973560716e-07,
+      "loss": 0.0471,
+      "step": 10762
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3220780616468485,
+      "learning_rate": 2.2565457754352372e-07,
+      "loss": 0.0503,
+      "step": 10763
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.29355657181633643,
+      "learning_rate": 2.2524943097250008e-07,
+      "loss": 0.0753,
+      "step": 10764
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3479929942629928,
+      "learning_rate": 2.248446400527138e-07,
+      "loss": 0.0713,
+      "step": 10765
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.45609974581004376,
+      "learning_rate": 2.244402048143163e-07,
+      "loss": 0.1157,
+      "step": 10766
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2980593545857858,
+      "learning_rate": 2.2403612528743125e-07,
+      "loss": 0.0567,
+      "step": 10767
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3035160381075223,
+      "learning_rate": 2.2363240150215848e-07,
+      "loss": 0.0722,
+      "step": 10768
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.21947652143348073,
+      "learning_rate": 2.232290334885684e-07,
+      "loss": 0.0637,
+      "step": 10769
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3706976801572213,
+      "learning_rate": 2.2282602127670638e-07,
+      "loss": 0.0626,
+      "step": 10770
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.348425051174002,
+      "learning_rate": 2.224233648965918e-07,
+      "loss": 0.0765,
+      "step": 10771
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5410431415644702,
+      "learning_rate": 2.220210643782167e-07,
+      "loss": 0.1062,
+      "step": 10772
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3225189081429699,
+      "learning_rate": 2.216191197515466e-07,
+      "loss": 0.0895,
+      "step": 10773
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.36975850408544747,
+      "learning_rate": 2.2121753104651922e-07,
+      "loss": 0.0872,
+      "step": 10774
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.19529394023574156,
+      "learning_rate": 2.2081629829305006e-07,
+      "loss": 0.0553,
+      "step": 10775
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2622552167330584,
+      "learning_rate": 2.2041542152102357e-07,
+      "loss": 0.06,
+      "step": 10776
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3675975920097866,
+      "learning_rate": 2.2001490076029864e-07,
+      "loss": 0.0652,
+      "step": 10777
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.26944115649759387,
+      "learning_rate": 2.1961473604071028e-07,
+      "loss": 0.0776,
+      "step": 10778
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.22129977581100824,
+      "learning_rate": 2.1921492739206463e-07,
+      "loss": 0.0601,
+      "step": 10779
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.40932445802003786,
+      "learning_rate": 2.188154748441401e-07,
+      "loss": 0.1011,
+      "step": 10780
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.23854481823817053,
+      "learning_rate": 2.1841637842669238e-07,
+      "loss": 0.0491,
+      "step": 10781
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2998955767529141,
+      "learning_rate": 2.1801763816944709e-07,
+      "loss": 0.0573,
+      "step": 10782
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.4084412468752217,
+      "learning_rate": 2.176192541021055e-07,
+      "loss": 0.0774,
+      "step": 10783
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.17009352006554865,
+      "learning_rate": 2.1722122625434105e-07,
+      "loss": 0.0378,
+      "step": 10784
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3319436128204301,
+      "learning_rate": 2.1682355465580174e-07,
+      "loss": 0.0702,
+      "step": 10785
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.4237771006846492,
+      "learning_rate": 2.1642623933610773e-07,
+      "loss": 0.1172,
+      "step": 10786
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3927803088198924,
+      "learning_rate": 2.160292803248537e-07,
+      "loss": 0.1032,
+      "step": 10787
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.48859552387621963,
+      "learning_rate": 2.1563267765160823e-07,
+      "loss": 0.1305,
+      "step": 10788
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3673575174905861,
+      "learning_rate": 2.1523643134591154e-07,
+      "loss": 0.0747,
+      "step": 10789
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.4364790867669465,
+      "learning_rate": 2.1484054143727895e-07,
+      "loss": 0.0536,
+      "step": 10790
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.4610076375718733,
+      "learning_rate": 2.1444500795519795e-07,
+      "loss": 0.0972,
+      "step": 10791
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.27598321985357693,
+      "learning_rate": 2.1404983092913112e-07,
+      "loss": 0.0637,
+      "step": 10792
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9370747809669794,
+      "learning_rate": 2.136550103885132e-07,
+      "loss": 0.0915,
+      "step": 10793
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.37028228726295187,
+      "learning_rate": 2.1326054636275185e-07,
+      "loss": 0.0835,
+      "step": 10794
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.30345761105397,
+      "learning_rate": 2.128664388812307e-07,
+      "loss": 0.0674,
+      "step": 10795
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2820681808035894,
+      "learning_rate": 2.1247268797330412e-07,
+      "loss": 0.0685,
+      "step": 10796
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.31159278748779146,
+      "learning_rate": 2.1207929366830082e-07,
+      "loss": 0.075,
+      "step": 10797
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.4162972842982443,
+      "learning_rate": 2.116862559955235e-07,
+      "loss": 0.0844,
+      "step": 10798
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3341701735618537,
+      "learning_rate": 2.1129357498424818e-07,
+      "loss": 0.0517,
+      "step": 10799
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3134727448543742,
+      "learning_rate": 2.1090125066372369e-07,
+      "loss": 0.1079,
+      "step": 10800
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.13005172474605234,
+      "learning_rate": 2.105092830631722e-07,
+      "loss": 0.026,
+      "step": 10801
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5168911581019014,
+      "learning_rate": 2.1011767221179035e-07,
+      "loss": 0.1157,
+      "step": 10802
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.4963179543856095,
+      "learning_rate": 2.097264181387476e-07,
+      "loss": 0.0914,
+      "step": 10803
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3000011557774194,
+      "learning_rate": 2.093355208731862e-07,
+      "loss": 0.0562,
+      "step": 10804
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3423850936825139,
+      "learning_rate": 2.089449804442234e-07,
+      "loss": 0.0583,
+      "step": 10805
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.40032511427242506,
+      "learning_rate": 2.085547968809476e-07,
+      "loss": 0.0917,
+      "step": 10806
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.48629328656186543,
+      "learning_rate": 2.0816497021242388e-07,
+      "loss": 0.1022,
+      "step": 10807
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.23613425211046524,
+      "learning_rate": 2.0777550046768679e-07,
+      "loss": 0.0525,
+      "step": 10808
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7181709009559915,
+      "learning_rate": 2.073863876757476e-07,
+      "loss": 0.0711,
+      "step": 10809
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.45097673588275816,
+      "learning_rate": 2.0699763186558919e-07,
+      "loss": 0.0977,
+      "step": 10810
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.41253262812203184,
+      "learning_rate": 2.0660923306616843e-07,
+      "loss": 0.108,
+      "step": 10811
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2458095715756952,
+      "learning_rate": 2.062211913064155e-07,
+      "loss": 0.0632,
+      "step": 10812
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2858059811956463,
+      "learning_rate": 2.058335066152345e-07,
+      "loss": 0.0649,
+      "step": 10813
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.25240986836099916,
+      "learning_rate": 2.054461790215012e-07,
+      "loss": 0.0657,
+      "step": 10814
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.44864381854731467,
+      "learning_rate": 2.0505920855406757e-07,
+      "loss": 0.0885,
+      "step": 10815
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2893213554504105,
+      "learning_rate": 2.0467259524175663e-07,
+      "loss": 0.0802,
+      "step": 10816
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.472797234700328,
+      "learning_rate": 2.0428633911336537e-07,
+      "loss": 0.0916,
+      "step": 10817
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2815168299084547,
+      "learning_rate": 2.0390044019766409e-07,
+      "loss": 0.1039,
+      "step": 10818
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.27420039753303493,
+      "learning_rate": 2.0351489852339813e-07,
+      "loss": 0.0689,
+      "step": 10819
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2934587813782035,
+      "learning_rate": 2.0312971411928451e-07,
+      "loss": 0.0941,
+      "step": 10820
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5198588389521508,
+      "learning_rate": 2.027448870140125e-07,
+      "loss": 0.0921,
+      "step": 10821
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3194354551861137,
+      "learning_rate": 2.023604172362481e-07,
+      "loss": 0.062,
+      "step": 10822
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.21742643432787762,
+      "learning_rate": 2.0197630481462838e-07,
+      "loss": 0.0492,
+      "step": 10823
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.40320347123352607,
+      "learning_rate": 2.0159254977776376e-07,
+      "loss": 0.0793,
+      "step": 10824
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.32151464874824387,
+      "learning_rate": 2.012091521542381e-07,
+      "loss": 0.0571,
+      "step": 10825
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3188341861624577,
+      "learning_rate": 2.008261119726107e-07,
+      "loss": 0.0684,
+      "step": 10826
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.37037750728114366,
+      "learning_rate": 2.0044342926141158e-07,
+      "loss": 0.0611,
+      "step": 10827
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3821049797498922,
+      "learning_rate": 2.0006110404914457e-07,
+      "loss": 0.0871,
+      "step": 10828
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5153336178514163,
+      "learning_rate": 1.9967913636428914e-07,
+      "loss": 0.0888,
+      "step": 10829
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2831283792782594,
+      "learning_rate": 1.9929752623529587e-07,
+      "loss": 0.0741,
+      "step": 10830
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.21697811813416326,
+      "learning_rate": 1.9891627369058807e-07,
+      "loss": 0.0551,
+      "step": 10831
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5860499456732063,
+      "learning_rate": 1.9853537875856477e-07,
+      "loss": 0.1087,
+      "step": 10832
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.47063105174915953,
+      "learning_rate": 1.9815484146759768e-07,
+      "loss": 0.1516,
+      "step": 10833
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.4221928876832875,
+      "learning_rate": 1.9777466184603022e-07,
+      "loss": 0.0834,
+      "step": 10834
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.35425686659273553,
+      "learning_rate": 1.973948399221809e-07,
+      "loss": 0.0977,
+      "step": 10835
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.4648591204857386,
+      "learning_rate": 1.9701537572434147e-07,
+      "loss": 0.1097,
+      "step": 10836
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3855660883995771,
+      "learning_rate": 1.966362692807766e-07,
+      "loss": 0.076,
+      "step": 10837
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6796185621988937,
+      "learning_rate": 1.9625752061972315e-07,
+      "loss": 0.0729,
+      "step": 10838
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.45776135025254255,
+      "learning_rate": 1.9587912976939406e-07,
+      "loss": 0.0896,
+      "step": 10839
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.27343725983218053,
+      "learning_rate": 1.9550109675797347e-07,
+      "loss": 0.0921,
+      "step": 10840
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.4005520350515181,
+      "learning_rate": 1.9512342161361998e-07,
+      "loss": 0.1207,
+      "step": 10841
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5427669953809398,
+      "learning_rate": 1.947461043644633e-07,
+      "loss": 0.1341,
+      "step": 10842
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.37338738931436405,
+      "learning_rate": 1.9436914503861036e-07,
+      "loss": 0.118,
+      "step": 10843
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.37260938137062105,
+      "learning_rate": 1.9399254366413822e-07,
+      "loss": 0.0916,
+      "step": 10844
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3745274183965353,
+      "learning_rate": 1.9361630026909772e-07,
+      "loss": 0.0684,
+      "step": 10845
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.28433820458532916,
+      "learning_rate": 1.9324041488151534e-07,
+      "loss": 0.0743,
+      "step": 10846
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5438397861938997,
+      "learning_rate": 1.9286488752938814e-07,
+      "loss": 0.1242,
+      "step": 10847
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.22433509642694524,
+      "learning_rate": 1.9248971824068762e-07,
+      "loss": 0.0602,
+      "step": 10848
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3110333744521675,
+      "learning_rate": 1.9211490704335867e-07,
+      "loss": 0.0878,
+      "step": 10849
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2621688801504687,
+      "learning_rate": 1.9174045396531948e-07,
+      "loss": 0.0789,
+      "step": 10850
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.23631606543764658,
+      "learning_rate": 1.9136635903446167e-07,
+      "loss": 0.0553,
+      "step": 10851
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3146764824473073,
+      "learning_rate": 1.90992622278649e-07,
+      "loss": 0.0696,
+      "step": 10852
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7276494102161084,
+      "learning_rate": 1.9061924372572094e-07,
+      "loss": 0.1321,
+      "step": 10853
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.22965072117380494,
+      "learning_rate": 1.9024622340348852e-07,
+      "loss": 0.0567,
+      "step": 10854
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2956052290240375,
+      "learning_rate": 1.8987356133973567e-07,
+      "loss": 0.085,
+      "step": 10855
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.34987882429469974,
+      "learning_rate": 1.8950125756222183e-07,
+      "loss": 0.0886,
+      "step": 10856
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.4881286148236791,
+      "learning_rate": 1.8912931209867758e-07,
+      "loss": 0.1018,
+      "step": 10857
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3576147675942974,
+      "learning_rate": 1.887577249768069e-07,
+      "loss": 0.1271,
+      "step": 10858
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.19334740333300582,
+      "learning_rate": 1.8838649622428819e-07,
+      "loss": 0.0714,
+      "step": 10859
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.28771729860249684,
+      "learning_rate": 1.8801562586877377e-07,
+      "loss": 0.0801,
+      "step": 10860
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2783994197897828,
+      "learning_rate": 1.876451139378871e-07,
+      "loss": 0.0815,
+      "step": 10861
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.41085767117799715,
+      "learning_rate": 1.8727496045922611e-07,
+      "loss": 0.0931,
+      "step": 10862
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.33912183552615693,
+      "learning_rate": 1.8690516546036263e-07,
+      "loss": 0.0986,
+      "step": 10863
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.21884426210098282,
+      "learning_rate": 1.8653572896884075e-07,
+      "loss": 0.0373,
+      "step": 10864
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3255546789654701,
+      "learning_rate": 1.8616665101217846e-07,
+      "loss": 0.0326,
+      "step": 10865
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.24705694909537113,
+      "learning_rate": 1.8579793161786595e-07,
+      "loss": 0.0592,
+      "step": 10866
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3190315028443304,
+      "learning_rate": 1.854295708133691e-07,
+      "loss": 0.0861,
+      "step": 10867
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.24384461393340837,
+      "learning_rate": 1.8506156862612423e-07,
+      "loss": 0.0749,
+      "step": 10868
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.25555627543443904,
+      "learning_rate": 1.8469392508354277e-07,
+      "loss": 0.0733,
+      "step": 10869
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.33043674400772677,
+      "learning_rate": 1.843266402130095e-07,
+      "loss": 0.0912,
+      "step": 10870
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.32432216897747956,
+      "learning_rate": 1.839597140418814e-07,
+      "loss": 0.098,
+      "step": 10871
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.22766680346013396,
+      "learning_rate": 1.8359314659748883e-07,
+      "loss": 0.0566,
+      "step": 10872
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3079250214782278,
+      "learning_rate": 1.832269379071372e-07,
+      "loss": 0.0591,
+      "step": 10873
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.28622843857989033,
+      "learning_rate": 1.8286108799810187e-07,
+      "loss": 0.0914,
+      "step": 10874
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.29928928214107514,
+      "learning_rate": 1.8249559689763497e-07,
+      "loss": 0.0795,
+      "step": 10875
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.49324924849899376,
+      "learning_rate": 1.821304646329608e-07,
+      "loss": 0.1077,
+      "step": 10876
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.35200481842786346,
+      "learning_rate": 1.8176569123127542e-07,
+      "loss": 0.1003,
+      "step": 10877
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5150574270795837,
+      "learning_rate": 1.8140127671974872e-07,
+      "loss": 0.1158,
+      "step": 10878
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.30613341361403507,
+      "learning_rate": 1.8103722112552625e-07,
+      "loss": 0.0648,
+      "step": 10879
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.15308659037169078,
+      "learning_rate": 1.8067352447572405e-07,
+      "loss": 0.0259,
+      "step": 10880
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3529978620391495,
+      "learning_rate": 1.8031018679743217e-07,
+      "loss": 0.0828,
+      "step": 10881
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2841708934665366,
+      "learning_rate": 1.7994720811771506e-07,
+      "loss": 0.0648,
+      "step": 10882
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.49609183772861,
+      "learning_rate": 1.7958458846360772e-07,
+      "loss": 0.1132,
+      "step": 10883
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.34161258445454934,
+      "learning_rate": 1.7922232786212134e-07,
+      "loss": 0.1031,
+      "step": 10884
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2929430792918475,
+      "learning_rate": 1.788604263402399e-07,
+      "loss": 0.0831,
+      "step": 10885
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3834876021791396,
+      "learning_rate": 1.7849888392491799e-07,
+      "loss": 0.069,
+      "step": 10886
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2735056011925727,
+      "learning_rate": 1.7813770064308732e-07,
+      "loss": 0.0767,
+      "step": 10887
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.44478265137258893,
+      "learning_rate": 1.7777687652165031e-07,
+      "loss": 0.0942,
+      "step": 10888
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.21625502462686888,
+      "learning_rate": 1.7741641158748213e-07,
+      "loss": 0.0587,
+      "step": 10889
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6513067873716356,
+      "learning_rate": 1.7705630586743404e-07,
+      "loss": 0.1086,
+      "step": 10890
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.38419730527094703,
+      "learning_rate": 1.7669655938832797e-07,
+      "loss": 0.07,
+      "step": 10891
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7610889784368245,
+      "learning_rate": 1.7633717217695968e-07,
+      "loss": 0.1368,
+      "step": 10892
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.23693384629111947,
+      "learning_rate": 1.7597814426009884e-07,
+      "loss": 0.0701,
+      "step": 10893
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.21713260005515417,
+      "learning_rate": 1.7561947566448746e-07,
+      "loss": 0.0478,
+      "step": 10894
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3439506667074959,
+      "learning_rate": 1.7526116641684243e-07,
+      "loss": 0.0944,
+      "step": 10895
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.24885671747089583,
+      "learning_rate": 1.749032165438508e-07,
+      "loss": 0.0731,
+      "step": 10896
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3039888176124646,
+      "learning_rate": 1.7454562607217674e-07,
+      "loss": 0.0893,
+      "step": 10897
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5286856861986966,
+      "learning_rate": 1.741883950284551e-07,
+      "loss": 0.1275,
+      "step": 10898
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3756186840212498,
+      "learning_rate": 1.7383152343929345e-07,
+      "loss": 0.0739,
+      "step": 10899
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3466034373804129,
+      "learning_rate": 1.7347501133127442e-07,
+      "loss": 0.1077,
+      "step": 10900
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2897371083908611,
+      "learning_rate": 1.7311885873095345e-07,
+      "loss": 0.0989,
+      "step": 10901
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.48578090650974554,
+      "learning_rate": 1.7276306566485879e-07,
+      "loss": 0.1299,
+      "step": 10902
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.30265813612148157,
+      "learning_rate": 1.724076321594914e-07,
+      "loss": 0.0895,
+      "step": 10903
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3773675116387481,
+      "learning_rate": 1.7205255824132627e-07,
+      "loss": 0.0991,
+      "step": 10904
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.31704541276376935,
+      "learning_rate": 1.7169784393681166e-07,
+      "loss": 0.0779,
+      "step": 10905
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5931366756271921,
+      "learning_rate": 1.7134348927236866e-07,
+      "loss": 0.0664,
+      "step": 10906
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.39926128924057525,
+      "learning_rate": 1.7098949427439115e-07,
+      "loss": 0.0888,
+      "step": 10907
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5026792589542943,
+      "learning_rate": 1.7063585896924806e-07,
+      "loss": 0.1328,
+      "step": 10908
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3368691878771733,
+      "learning_rate": 1.7028258338327885e-07,
+      "loss": 0.0676,
+      "step": 10909
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2613155750065307,
+      "learning_rate": 1.699296675427975e-07,
+      "loss": 0.0582,
+      "step": 10910
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3402108387927932,
+      "learning_rate": 1.695771114740924e-07,
+      "loss": 0.0939,
+      "step": 10911
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3250831335260662,
+      "learning_rate": 1.692249152034231e-07,
+      "loss": 0.0903,
+      "step": 10912
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.23631057901512023,
+      "learning_rate": 1.6887307875702307e-07,
+      "loss": 0.0547,
+      "step": 10913
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2548952061065161,
+      "learning_rate": 1.6852160216110026e-07,
+      "loss": 0.0596,
+      "step": 10914
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4531824297082557,
+      "learning_rate": 1.6817048544183368e-07,
+      "loss": 0.1163,
+      "step": 10915
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.23602436320047168,
+      "learning_rate": 1.678197286253769e-07,
+      "loss": 0.0575,
+      "step": 10916
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2604961325603366,
+      "learning_rate": 1.6746933173785563e-07,
+      "loss": 0.07,
+      "step": 10917
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.45600108595951705,
+      "learning_rate": 1.671192948053707e-07,
+      "loss": 0.0855,
+      "step": 10918
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3172426653068614,
+      "learning_rate": 1.667696178539946e-07,
+      "loss": 0.0852,
+      "step": 10919
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.18480010702860972,
+      "learning_rate": 1.6642030090977202e-07,
+      "loss": 0.0553,
+      "step": 10920
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.38001014959891893,
+      "learning_rate": 1.6607134399872382e-07,
+      "loss": 0.0744,
+      "step": 10921
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2746768944561804,
+      "learning_rate": 1.65722747146842e-07,
+      "loss": 0.0209,
+      "step": 10922
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4323663419430902,
+      "learning_rate": 1.6537451038009077e-07,
+      "loss": 0.0805,
+      "step": 10923
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.26041174242435644,
+      "learning_rate": 1.6502663372441053e-07,
+      "loss": 0.0862,
+      "step": 10924
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.351208112802025,
+      "learning_rate": 1.6467911720571216e-07,
+      "loss": 0.0596,
+      "step": 10925
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5413161115406405,
+      "learning_rate": 1.6433196084988113e-07,
+      "loss": 0.1199,
+      "step": 10926
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.36451183211082616,
+      "learning_rate": 1.6398516468277503e-07,
+      "loss": 0.0823,
+      "step": 10927
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.42996920408962125,
+      "learning_rate": 1.63638728730226e-07,
+      "loss": 0.0993,
+      "step": 10928
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4802627540909113,
+      "learning_rate": 1.6329265301803843e-07,
+      "loss": 0.1061,
+      "step": 10929
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3489318677176047,
+      "learning_rate": 1.6294693757198943e-07,
+      "loss": 0.0878,
+      "step": 10930
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4166750667584952,
+      "learning_rate": 1.626015824178312e-07,
+      "loss": 0.0963,
+      "step": 10931
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3907724352636322,
+      "learning_rate": 1.6225658758128703e-07,
+      "loss": 0.0518,
+      "step": 10932
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3015459053800912,
+      "learning_rate": 1.6191195308805417e-07,
+      "loss": 0.0548,
+      "step": 10933
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4361947951910891,
+      "learning_rate": 1.615676789638021e-07,
+      "loss": 0.0825,
+      "step": 10934
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.34419642519696336,
+      "learning_rate": 1.6122376523417638e-07,
+      "loss": 0.057,
+      "step": 10935
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3325474144844211,
+      "learning_rate": 1.6088021192479207e-07,
+      "loss": 0.0831,
+      "step": 10936
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3027950083915135,
+      "learning_rate": 1.605370190612393e-07,
+      "loss": 0.0621,
+      "step": 10937
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.40322321290470015,
+      "learning_rate": 1.601941866690815e-07,
+      "loss": 0.0801,
+      "step": 10938
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3970684844148039,
+      "learning_rate": 1.5985171477385485e-07,
+      "loss": 0.079,
+      "step": 10939
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.19992423798987713,
+      "learning_rate": 1.5950960340106848e-07,
+      "loss": 0.0509,
+      "step": 10940
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.42758933245322533,
+      "learning_rate": 1.591678525762047e-07,
+      "loss": 0.089,
+      "step": 10941
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.24037967554726927,
+      "learning_rate": 1.5882646232471877e-07,
+      "loss": 0.0636,
+      "step": 10942
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2921442944739254,
+      "learning_rate": 1.5848543267203976e-07,
+      "loss": 0.0542,
+      "step": 10943
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4419292576801984,
+      "learning_rate": 1.5814476364357013e-07,
+      "loss": 0.0971,
+      "step": 10944
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3598501085061918,
+      "learning_rate": 1.5780445526468457e-07,
+      "loss": 0.1075,
+      "step": 10945
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.21631671745008493,
+      "learning_rate": 1.5746450756073062e-07,
+      "loss": 0.0462,
+      "step": 10946
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.44433941711797487,
+      "learning_rate": 1.571249205570302e-07,
+      "loss": 0.0699,
+      "step": 10947
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3694008390036305,
+      "learning_rate": 1.5678569427887756e-07,
+      "loss": 0.1017,
+      "step": 10948
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.1973451981976213,
+      "learning_rate": 1.5644682875154026e-07,
+      "loss": 0.0655,
+      "step": 10949
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3170268448730254,
+      "learning_rate": 1.5610832400025921e-07,
+      "loss": 0.0804,
+      "step": 10950
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.21664474621761837,
+      "learning_rate": 1.5577018005024702e-07,
+      "loss": 0.0411,
+      "step": 10951
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5123724062690334,
+      "learning_rate": 1.554323969266919e-07,
+      "loss": 0.0941,
+      "step": 10952
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.1882500177831335,
+      "learning_rate": 1.550949746547542e-07,
+      "loss": 0.0492,
+      "step": 10953
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.40682578384981505,
+      "learning_rate": 1.5475791325956556e-07,
+      "loss": 0.1132,
+      "step": 10954
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.23282010426795893,
+      "learning_rate": 1.5442121276623369e-07,
+      "loss": 0.0563,
+      "step": 10955
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3986976654197232,
+      "learning_rate": 1.5408487319983734e-07,
+      "loss": 0.0396,
+      "step": 10956
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6821326980939498,
+      "learning_rate": 1.5374889458542875e-07,
+      "loss": 0.0988,
+      "step": 10957
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3568034030439121,
+      "learning_rate": 1.5341327694803455e-07,
+      "loss": 0.0681,
+      "step": 10958
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2673461419668013,
+      "learning_rate": 1.5307802031265305e-07,
+      "loss": 0.0826,
+      "step": 10959
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.39532692543599574,
+      "learning_rate": 1.52743124704256e-07,
+      "loss": 0.0718,
+      "step": 10960
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.49744765993420337,
+      "learning_rate": 1.5240859014778787e-07,
+      "loss": 0.0946,
+      "step": 10961
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3844361127497604,
+      "learning_rate": 1.5207441666816758e-07,
+      "loss": 0.0828,
+      "step": 10962
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5035057548365465,
+      "learning_rate": 1.5174060429028582e-07,
+      "loss": 0.1349,
+      "step": 10963
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.22898304472490458,
+      "learning_rate": 1.514071530390071e-07,
+      "loss": 0.0386,
+      "step": 10964
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3564599031160037,
+      "learning_rate": 1.5107406293916883e-07,
+      "loss": 0.0859,
+      "step": 10965
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.25951859024984947,
+      "learning_rate": 1.5074133401558165e-07,
+      "loss": 0.0818,
+      "step": 10966
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.24153081994501283,
+      "learning_rate": 1.504089662930286e-07,
+      "loss": 0.0415,
+      "step": 10967
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.28175278736023496,
+      "learning_rate": 1.5007695979626646e-07,
+      "loss": 0.058,
+      "step": 10968
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.20701268775280623,
+      "learning_rate": 1.4974531455002605e-07,
+      "loss": 0.0746,
+      "step": 10969
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2905471592784245,
+      "learning_rate": 1.4941403057900927e-07,
+      "loss": 0.0929,
+      "step": 10970
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2807491074182197,
+      "learning_rate": 1.4908310790789137e-07,
+      "loss": 0.0835,
+      "step": 10971
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.21969335684062477,
+      "learning_rate": 1.487525465613232e-07,
+      "loss": 0.0639,
+      "step": 10972
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2971679042824664,
+      "learning_rate": 1.4842234656392562e-07,
+      "loss": 0.0492,
+      "step": 10973
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4987830549264698,
+      "learning_rate": 1.4809250794029452e-07,
+      "loss": 0.1133,
+      "step": 10974
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.33337451637349785,
+      "learning_rate": 1.4776303071499687e-07,
+      "loss": 0.0795,
+      "step": 10975
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3033378143128062,
+      "learning_rate": 1.4743391491257585e-07,
+      "loss": 0.0868,
+      "step": 10976
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.48357695957854385,
+      "learning_rate": 1.4710516055754521e-07,
+      "loss": 0.1008,
+      "step": 10977
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.34691315527683,
+      "learning_rate": 1.4677676767439198e-07,
+      "loss": 0.0867,
+      "step": 10978
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.349506733591542,
+      "learning_rate": 1.4644873628757773e-07,
+      "loss": 0.1008,
+      "step": 10979
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3993853506533474,
+      "learning_rate": 1.4612106642153568e-07,
+      "loss": 0.0863,
+      "step": 10980
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5637088142549584,
+      "learning_rate": 1.4579375810067186e-07,
+      "loss": 0.1152,
+      "step": 10981
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2800794203106138,
+      "learning_rate": 1.454668113493668e-07,
+      "loss": 0.0784,
+      "step": 10982
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.317480010413139,
+      "learning_rate": 1.4514022619197431e-07,
+      "loss": 0.0692,
+      "step": 10983
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.23131663785230341,
+      "learning_rate": 1.4481400265281888e-07,
+      "loss": 0.0393,
+      "step": 10984
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3063102832241636,
+      "learning_rate": 1.4448814075619988e-07,
+      "loss": 0.0878,
+      "step": 10985
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.330661163671372,
+      "learning_rate": 1.4416264052638961e-07,
+      "loss": 0.054,
+      "step": 10986
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.29342077034986697,
+      "learning_rate": 1.4383750198763368e-07,
+      "loss": 0.0865,
+      "step": 10987
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.28997534567421346,
+      "learning_rate": 1.4351272516414882e-07,
+      "loss": 0.0555,
+      "step": 10988
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5302516164253191,
+      "learning_rate": 1.4318831008012845e-07,
+      "loss": 0.1007,
+      "step": 10989
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.35094217572194647,
+      "learning_rate": 1.4286425675973549e-07,
+      "loss": 0.0897,
+      "step": 10990
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.287059338660618,
+      "learning_rate": 1.4254056522710779e-07,
+      "loss": 0.0387,
+      "step": 10991
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2870486821576297,
+      "learning_rate": 1.4221723550635446e-07,
+      "loss": 0.0533,
+      "step": 10992
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3076179063433011,
+      "learning_rate": 1.4189426762156122e-07,
+      "loss": 0.0733,
+      "step": 10993
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.24929750374699525,
+      "learning_rate": 1.4157166159678326e-07,
+      "loss": 0.0821,
+      "step": 10994
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.22563166378918179,
+      "learning_rate": 1.4124941745605024e-07,
+      "loss": 0.0478,
+      "step": 10995
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2634247334143495,
+      "learning_rate": 1.4092753522336466e-07,
+      "loss": 0.0496,
+      "step": 10996
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.22570587237133935,
+      "learning_rate": 1.4060601492270288e-07,
+      "loss": 0.075,
+      "step": 10997
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2643235702530205,
+      "learning_rate": 1.4028485657801295e-07,
+      "loss": 0.0638,
+      "step": 10998
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.38302164938698036,
+      "learning_rate": 1.3996406021321686e-07,
+      "loss": 0.0823,
+      "step": 10999
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.32411950697647246,
+      "learning_rate": 1.3964362585220993e-07,
+      "loss": 0.084,
+      "step": 11000
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3690642630777387,
+      "learning_rate": 1.393235535188586e-07,
+      "loss": 0.1092,
+      "step": 11001
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5611291920937465,
+      "learning_rate": 1.39003843237005e-07,
+      "loss": 0.1298,
+      "step": 11002
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3971075828971253,
+      "learning_rate": 1.386844950304622e-07,
+      "loss": 0.0881,
+      "step": 11003
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2640944537587143,
+      "learning_rate": 1.3836550892301792e-07,
+      "loss": 0.0681,
+      "step": 11004
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.31481012095030914,
+      "learning_rate": 1.3804688493843087e-07,
+      "loss": 0.0739,
+      "step": 11005
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2593427515773532,
+      "learning_rate": 1.3772862310043543e-07,
+      "loss": 0.0861,
+      "step": 11006
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5013116804945846,
+      "learning_rate": 1.374107234327371e-07,
+      "loss": 0.0921,
+      "step": 11007
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2734727757925456,
+      "learning_rate": 1.3709318595901465e-07,
+      "loss": 0.0786,
+      "step": 11008
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.32302142223359825,
+      "learning_rate": 1.3677601070292034e-07,
+      "loss": 0.0958,
+      "step": 11009
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3339957079727807,
+      "learning_rate": 1.3645919768807858e-07,
+      "loss": 0.0908,
+      "step": 11010
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.21727381421509548,
+      "learning_rate": 1.361427469380888e-07,
+      "loss": 0.0403,
+      "step": 11011
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2710627617893132,
+      "learning_rate": 1.3582665847652053e-07,
+      "loss": 0.0837,
+      "step": 11012
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7516050556880568,
+      "learning_rate": 1.3551093232691937e-07,
+      "loss": 0.132,
+      "step": 11013
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3416181980584646,
+      "learning_rate": 1.351955685128009e-07,
+      "loss": 0.0609,
+      "step": 11014
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3079639260409698,
+      "learning_rate": 1.348805670576564e-07,
+      "loss": 0.1092,
+      "step": 11015
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2008957145760628,
+      "learning_rate": 1.3456592798494928e-07,
+      "loss": 0.0601,
+      "step": 11016
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3215262886148315,
+      "learning_rate": 1.342516513181147e-07,
+      "loss": 0.0611,
+      "step": 11017
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.4316006367952478,
+      "learning_rate": 1.3393773708056225e-07,
+      "loss": 0.0684,
+      "step": 11018
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.26739681515823316,
+      "learning_rate": 1.3362418529567322e-07,
+      "loss": 0.0508,
+      "step": 11019
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2067938739321848,
+      "learning_rate": 1.3331099598680453e-07,
+      "loss": 0.038,
+      "step": 11020
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5479743881268239,
+      "learning_rate": 1.3299816917728303e-07,
+      "loss": 0.1112,
+      "step": 11021
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5689304550789344,
+      "learning_rate": 1.326857048904101e-07,
+      "loss": 0.0743,
+      "step": 11022
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.4624647981447354,
+      "learning_rate": 1.3237360314946045e-07,
+      "loss": 0.076,
+      "step": 11023
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.28509041814383523,
+      "learning_rate": 1.3206186397768107e-07,
+      "loss": 0.0699,
+      "step": 11024
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.29341525083605907,
+      "learning_rate": 1.3175048739829167e-07,
+      "loss": 0.072,
+      "step": 11025
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2571658055766883,
+      "learning_rate": 1.3143947343448482e-07,
+      "loss": 0.0754,
+      "step": 11026
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.4036928123104843,
+      "learning_rate": 1.311288221094287e-07,
+      "loss": 0.0789,
+      "step": 11027
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3502379375605657,
+      "learning_rate": 1.3081853344626082e-07,
+      "loss": 0.0873,
+      "step": 11028
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.366677074693949,
+      "learning_rate": 1.305086074680928e-07,
+      "loss": 0.0783,
+      "step": 11029
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.27978024354036973,
+      "learning_rate": 1.3019904419801165e-07,
+      "loss": 0.0502,
+      "step": 11030
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5302028471515052,
+      "learning_rate": 1.2988984365907397e-07,
+      "loss": 0.1365,
+      "step": 11031
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.38606569544698016,
+      "learning_rate": 1.2958100587431076e-07,
+      "loss": 0.1038,
+      "step": 11032
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3095819248433764,
+      "learning_rate": 1.2927253086672698e-07,
+      "loss": 0.0688,
+      "step": 11033
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.27863117005278804,
+      "learning_rate": 1.2896441865929976e-07,
+      "loss": 0.0763,
+      "step": 11034
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.26107616828004504,
+      "learning_rate": 1.28656669274978e-07,
+      "loss": 0.0634,
+      "step": 11035
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.20922024806940337,
+      "learning_rate": 1.2834928273668446e-07,
+      "loss": 0.0467,
+      "step": 11036
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.31224538630757276,
+      "learning_rate": 1.2804225906731693e-07,
+      "loss": 0.0769,
+      "step": 11037
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0982170682710088,
+      "learning_rate": 1.2773559828974268e-07,
+      "loss": 0.0771,
+      "step": 11038
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.42984613220516227,
+      "learning_rate": 1.2742930042680402e-07,
+      "loss": 0.1025,
+      "step": 11039
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3211660310350537,
+      "learning_rate": 1.27123365501316e-07,
+      "loss": 0.0648,
+      "step": 11040
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3725324455960209,
+      "learning_rate": 1.268177935360665e-07,
+      "loss": 0.1124,
+      "step": 11041
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.47762734935196544,
+      "learning_rate": 1.265125845538162e-07,
+      "loss": 0.0854,
+      "step": 11042
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.4949293571046767,
+      "learning_rate": 1.2620773857729807e-07,
+      "loss": 0.1227,
+      "step": 11043
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.26003897704203677,
+      "learning_rate": 1.2590325562922002e-07,
+      "loss": 0.0723,
+      "step": 11044
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3492841026661677,
+      "learning_rate": 1.2559913573226168e-07,
+      "loss": 0.107,
+      "step": 11045
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2920559640824417,
+      "learning_rate": 1.2529537890907438e-07,
+      "loss": 0.0473,
+      "step": 11046
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.38668183035908504,
+      "learning_rate": 1.249919851822845e-07,
+      "loss": 0.0994,
+      "step": 11047
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.36097518629495084,
+      "learning_rate": 1.2468895457449115e-07,
+      "loss": 0.0787,
+      "step": 11048
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.30465758635246637,
+      "learning_rate": 1.2438628710826462e-07,
+      "loss": 0.1228,
+      "step": 11049
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2193426357381622,
+      "learning_rate": 1.2408398280615075e-07,
+      "loss": 0.0447,
+      "step": 11050
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3357814261508535,
+      "learning_rate": 1.2378204169066599e-07,
+      "loss": 0.0939,
+      "step": 11051
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.33715502988057505,
+      "learning_rate": 1.2348046378430068e-07,
+      "loss": 0.0776,
+      "step": 11052
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.24794686955078973,
+      "learning_rate": 1.2317924910951794e-07,
+      "loss": 0.061,
+      "step": 11053
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.31391790070466,
+      "learning_rate": 1.2287839768875431e-07,
+      "loss": 0.07,
+      "step": 11054
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.24467535575192373,
+      "learning_rate": 1.2257790954441962e-07,
+      "loss": 0.0254,
+      "step": 11055
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3228074477683021,
+      "learning_rate": 1.2227778469889429e-07,
+      "loss": 0.0867,
+      "step": 11056
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2411432887187595,
+      "learning_rate": 1.2197802317453544e-07,
+      "loss": 0.047,
+      "step": 11057
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.4206553135367485,
+      "learning_rate": 1.2167862499366968e-07,
+      "loss": 0.1286,
+      "step": 11058
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3173055861147901,
+      "learning_rate": 1.2137959017859802e-07,
+      "loss": 0.0784,
+      "step": 11059
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.23820894682227708,
+      "learning_rate": 1.2108091875159432e-07,
+      "loss": 0.0719,
+      "step": 11060
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3714655031189375,
+      "learning_rate": 1.2078261073490581e-07,
+      "loss": 0.1096,
+      "step": 11061
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3060146658426092,
+      "learning_rate": 1.204846661507525e-07,
+      "loss": 0.0915,
+      "step": 11062
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.33058244785948837,
+      "learning_rate": 1.2018708502132602e-07,
+      "loss": 0.0601,
+      "step": 11063
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.29436704459557533,
+      "learning_rate": 1.198898673687926e-07,
+      "loss": 0.096,
+      "step": 11064
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2034673746554507,
+      "learning_rate": 1.195930132152906e-07,
+      "loss": 0.0683,
+      "step": 11065
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2856781654113205,
+      "learning_rate": 1.1929652258293122e-07,
+      "loss": 0.0908,
+      "step": 11066
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.30265539699690486,
+      "learning_rate": 1.1900039549379904e-07,
+      "loss": 0.0406,
+      "step": 11067
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.4239218909472435,
+      "learning_rate": 1.1870463196995196e-07,
+      "loss": 0.1285,
+      "step": 11068
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.27239438551441,
+      "learning_rate": 1.1840923203341903e-07,
+      "loss": 0.0899,
+      "step": 11069
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.47260266589926764,
+      "learning_rate": 1.1811419570620375e-07,
+      "loss": 0.1129,
+      "step": 11070
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.49296411168890064,
+      "learning_rate": 1.1781952301028243e-07,
+      "loss": 0.1371,
+      "step": 11071
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.23522108228345823,
+      "learning_rate": 1.1752521396760419e-07,
+      "loss": 0.0509,
+      "step": 11072
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2938075118739258,
+      "learning_rate": 1.1723126860008981e-07,
+      "loss": 0.0762,
+      "step": 11073
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.18859819434479183,
+      "learning_rate": 1.1693768692963569e-07,
+      "loss": 0.0694,
+      "step": 11074
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2331799941321318,
+      "learning_rate": 1.1664446897810821e-07,
+      "loss": 0.0692,
+      "step": 11075
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2775195088717485,
+      "learning_rate": 1.1635161476734824e-07,
+      "loss": 0.062,
+      "step": 11076
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.40283451705409695,
+      "learning_rate": 1.1605912431916943e-07,
+      "loss": 0.0725,
+      "step": 11077
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.22180221832919114,
+      "learning_rate": 1.157669976553577e-07,
+      "loss": 0.0564,
+      "step": 11078
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.27655848838424457,
+      "learning_rate": 1.1547523479767287e-07,
+      "loss": 0.0726,
+      "step": 11079
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2589735677437674,
+      "learning_rate": 1.1518383576784753e-07,
+      "loss": 0.0589,
+      "step": 11080
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3955122149407037,
+      "learning_rate": 1.1489280058758601e-07,
+      "loss": 0.084,
+      "step": 11081
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3190993815374948,
+      "learning_rate": 1.1460212927856595e-07,
+      "loss": 0.092,
+      "step": 11082
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5942341938638771,
+      "learning_rate": 1.143118218624395e-07,
+      "loss": 0.0883,
+      "step": 11083
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.1862497302499822,
+      "learning_rate": 1.140218783608299e-07,
+      "loss": 0.04,
+      "step": 11084
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3063210335201512,
+      "learning_rate": 1.1373229879533376e-07,
+      "loss": 0.0891,
+      "step": 11085
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.34530531220014465,
+      "learning_rate": 1.1344308318752051e-07,
+      "loss": 0.1037,
+      "step": 11086
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.24713037225827278,
+      "learning_rate": 1.1315423155893179e-07,
+      "loss": 0.0705,
+      "step": 11087
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3231677939429711,
+      "learning_rate": 1.1286574393108485e-07,
+      "loss": 0.0571,
+      "step": 11088
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.31207249153400085,
+      "learning_rate": 1.1257762032546693e-07,
+      "loss": 0.0702,
+      "step": 11089
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.4573930505270725,
+      "learning_rate": 1.1228986076353865e-07,
+      "loss": 0.0937,
+      "step": 11090
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2137385492851112,
+      "learning_rate": 1.1200246526673509e-07,
+      "loss": 0.0467,
+      "step": 11091
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.22265368858365467,
+      "learning_rate": 1.1171543385646244e-07,
+      "loss": 0.0538,
+      "step": 11092
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.40744981240597195,
+      "learning_rate": 1.1142876655410084e-07,
+      "loss": 0.1155,
+      "step": 11093
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2914035617508047,
+      "learning_rate": 1.1114246338100209e-07,
+      "loss": 0.0613,
+      "step": 11094
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.299713112638648,
+      "learning_rate": 1.1085652435849303e-07,
+      "loss": 0.083,
+      "step": 11095
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.4469237316739102,
+      "learning_rate": 1.1057094950787162e-07,
+      "loss": 0.0747,
+      "step": 11096
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3487351181843528,
+      "learning_rate": 1.1028573885040806e-07,
+      "loss": 0.0882,
+      "step": 11097
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.24219512939207158,
+      "learning_rate": 1.1000089240734757e-07,
+      "loss": 0.056,
+      "step": 11098
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.33748241318513894,
+      "learning_rate": 1.0971641019990764e-07,
+      "loss": 0.0827,
+      "step": 11099
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6583702746094607,
+      "learning_rate": 1.0943229224927687e-07,
+      "loss": 0.0906,
+      "step": 11100
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2484965385955238,
+      "learning_rate": 1.091485385766189e-07,
+      "loss": 0.0522,
+      "step": 11101
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.19693323631136342,
+      "learning_rate": 1.0886514920306902e-07,
+      "loss": 0.0489,
+      "step": 11102
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.25318500687297296,
+      "learning_rate": 1.085821241497359e-07,
+      "loss": 0.0547,
+      "step": 11103
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.43407562696556895,
+      "learning_rate": 1.0829946343770048e-07,
+      "loss": 0.1305,
+      "step": 11104
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2887839597129628,
+      "learning_rate": 1.0801716708801757e-07,
+      "loss": 0.0414,
+      "step": 11105
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.31810485943793565,
+      "learning_rate": 1.0773523512171368e-07,
+      "loss": 0.0755,
+      "step": 11106
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5098641948806104,
+      "learning_rate": 1.074536675597887e-07,
+      "loss": 0.1176,
+      "step": 11107
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2641917895628593,
+      "learning_rate": 1.0717246442321638e-07,
+      "loss": 0.0686,
+      "step": 11108
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.23773646542488674,
+      "learning_rate": 1.068916257329422e-07,
+      "loss": 0.0671,
+      "step": 11109
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.45564730259135444,
+      "learning_rate": 1.0661115150988332e-07,
+      "loss": 0.0955,
+      "step": 11110
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6361904933325756,
+      "learning_rate": 1.0633104177493192e-07,
+      "loss": 0.0989,
+      "step": 11111
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.40220513493448534,
+      "learning_rate": 1.0605129654895296e-07,
+      "loss": 0.0838,
+      "step": 11112
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.28176553003860905,
+      "learning_rate": 1.0577191585278257e-07,
+      "loss": 0.072,
+      "step": 11113
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.34817836981770933,
+      "learning_rate": 1.0549289970723021e-07,
+      "loss": 0.1194,
+      "step": 11114
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.4596741557188952,
+      "learning_rate": 1.0521424813307979e-07,
+      "loss": 0.098,
+      "step": 11115
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.30019521486968,
+      "learning_rate": 1.0493596115108695e-07,
+      "loss": 0.0831,
+      "step": 11116
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3730464457498851,
+      "learning_rate": 1.0465803878197899e-07,
+      "loss": 0.0846,
+      "step": 11117
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3825698974745048,
+      "learning_rate": 1.043804810464577e-07,
+      "loss": 0.0843,
+      "step": 11118
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.19465036026403382,
+      "learning_rate": 1.0410328796519764e-07,
+      "loss": 0.0418,
+      "step": 11119
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.4297390873607323,
+      "learning_rate": 1.0382645955884508e-07,
+      "loss": 0.1392,
+      "step": 11120
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.19914969879815567,
+      "learning_rate": 1.0354999584802017e-07,
+      "loss": 0.0543,
+      "step": 11121
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3368781128553592,
+      "learning_rate": 1.0327389685331535e-07,
+      "loss": 0.0742,
+      "step": 11122
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2561567565555721,
+      "learning_rate": 1.0299816259529638e-07,
+      "loss": 0.0449,
+      "step": 11123
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.442824555581664,
+      "learning_rate": 1.0272279309450129e-07,
+      "loss": 0.0722,
+      "step": 11124
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.39412310736001593,
+      "learning_rate": 1.0244778837144198e-07,
+      "loss": 0.1188,
+      "step": 11125
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.22899019583356575,
+      "learning_rate": 1.0217314844660097e-07,
+      "loss": 0.045,
+      "step": 11126
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.24268415690934625,
+      "learning_rate": 1.0189887334043635e-07,
+      "loss": 0.0507,
+      "step": 11127
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.28323668343850633,
+      "learning_rate": 1.0162496307337677e-07,
+      "loss": 0.0374,
+      "step": 11128
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.35385262776531995,
+      "learning_rate": 1.0135141766582535e-07,
+      "loss": 0.0329,
+      "step": 11129
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3649155454418222,
+      "learning_rate": 1.010782371381569e-07,
+      "loss": 0.0601,
+      "step": 11130
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.31108798852298297,
+      "learning_rate": 1.0080542151071959e-07,
+      "loss": 0.0771,
+      "step": 11131
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.734070394559944,
+      "learning_rate": 1.0053297080383496e-07,
+      "loss": 0.12,
+      "step": 11132
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.228704914015767,
+      "learning_rate": 1.0026088503779564e-07,
+      "loss": 0.0588,
+      "step": 11133
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5435423084885356,
+      "learning_rate": 9.998916423286876e-08,
+      "loss": 0.0978,
+      "step": 11134
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.4121767810379705,
+      "learning_rate": 9.971780840929368e-08,
+      "loss": 0.0804,
+      "step": 11135
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3759780590336709,
+      "learning_rate": 9.944681758728258e-08,
+      "loss": 0.0851,
+      "step": 11136
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.33390015919531246,
+      "learning_rate": 9.917619178702043e-08,
+      "loss": 0.0793,
+      "step": 11137
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3245391746862259,
+      "learning_rate": 9.890593102866442e-08,
+      "loss": 0.0757,
+      "step": 11138
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.4999889183245336,
+      "learning_rate": 9.863603533234622e-08,
+      "loss": 0.0925,
+      "step": 11139
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.21210101211848256,
+      "learning_rate": 9.83665047181681e-08,
+      "loss": 0.0462,
+      "step": 11140
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.17663036225687423,
+      "learning_rate": 9.809733920620679e-08,
+      "loss": 0.0495,
+      "step": 11141
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.4829498091272957,
+      "learning_rate": 9.782853881651233e-08,
+      "loss": 0.1027,
+      "step": 11142
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2360634002781989,
+      "learning_rate": 9.756010356910484e-08,
+      "loss": 0.0523,
+      "step": 11143
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5424902744133441,
+      "learning_rate": 9.729203348397997e-08,
+      "loss": 0.068,
+      "step": 11144
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.43894418538831637,
+      "learning_rate": 9.702432858110455e-08,
+      "loss": 0.087,
+      "step": 11145
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2020666164691161,
+      "learning_rate": 9.675698888041873e-08,
+      "loss": 0.0399,
+      "step": 11146
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.36028601050765807,
+      "learning_rate": 9.649001440183658e-08,
+      "loss": 0.0859,
+      "step": 11147
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.25490210576013106,
+      "learning_rate": 9.622340516524332e-08,
+      "loss": 0.0413,
+      "step": 11148
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5328712387866651,
+      "learning_rate": 9.595716119049692e-08,
+      "loss": 0.126,
+      "step": 11149
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5973336135376869,
+      "learning_rate": 9.569128249742932e-08,
+      "loss": 0.0597,
+      "step": 11150
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.46205782605054385,
+      "learning_rate": 9.542576910584467e-08,
+      "loss": 0.1418,
+      "step": 11151
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3655348174753182,
+      "learning_rate": 9.516062103552049e-08,
+      "loss": 0.0957,
+      "step": 11152
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.4592514359092967,
+      "learning_rate": 9.489583830620597e-08,
+      "loss": 0.1065,
+      "step": 11153
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2877827933482134,
+      "learning_rate": 9.463142093762367e-08,
+      "loss": 0.0663,
+      "step": 11154
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.23285224799903795,
+      "learning_rate": 9.436736894946841e-08,
+      "loss": 0.0711,
+      "step": 11155
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.33547084570254815,
+      "learning_rate": 9.410368236140943e-08,
+      "loss": 0.0701,
+      "step": 11156
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.4975086414040959,
+      "learning_rate": 9.384036119308771e-08,
+      "loss": 0.0871,
+      "step": 11157
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2675973146701914,
+      "learning_rate": 9.357740546411531e-08,
+      "loss": 0.0891,
+      "step": 11158
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.28660399788452495,
+      "learning_rate": 9.331481519408048e-08,
+      "loss": 0.0457,
+      "step": 11159
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3596893520389731,
+      "learning_rate": 9.305259040254144e-08,
+      "loss": 0.1055,
+      "step": 11160
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.37220464402171766,
+      "learning_rate": 9.279073110903092e-08,
+      "loss": 0.0772,
+      "step": 11161
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2898081359962696,
+      "learning_rate": 9.252923733305274e-08,
+      "loss": 0.0642,
+      "step": 11162
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5619680933823189,
+      "learning_rate": 9.226810909408579e-08,
+      "loss": 0.0923,
+      "step": 11163
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3438638219043938,
+      "learning_rate": 9.200734641157949e-08,
+      "loss": 0.0596,
+      "step": 11164
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5385374795559705,
+      "learning_rate": 9.174694930495664e-08,
+      "loss": 0.1107,
+      "step": 11165
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.29259661100403284,
+      "learning_rate": 9.148691779361452e-08,
+      "loss": 0.0738,
+      "step": 11166
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.32241658110500787,
+      "learning_rate": 9.122725189692039e-08,
+      "loss": 0.0692,
+      "step": 11167
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.4442942848518488,
+      "learning_rate": 9.096795163421657e-08,
+      "loss": 0.0688,
+      "step": 11168
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.22061675103132072,
+      "learning_rate": 9.07090170248165e-08,
+      "loss": 0.0494,
+      "step": 11169
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.24935385543925348,
+      "learning_rate": 9.045044808800806e-08,
+      "loss": 0.0621,
+      "step": 11170
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.26149146069001444,
+      "learning_rate": 9.019224484305033e-08,
+      "loss": 0.0683,
+      "step": 11171
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.20905267479499953,
+      "learning_rate": 8.993440730917569e-08,
+      "loss": 0.0552,
+      "step": 11172
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.44374962044494315,
+      "learning_rate": 8.967693550559042e-08,
+      "loss": 0.1094,
+      "step": 11173
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.32352189648419505,
+      "learning_rate": 8.941982945147087e-08,
+      "loss": 0.0693,
+      "step": 11174
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.37343560174673135,
+      "learning_rate": 8.916308916596894e-08,
+      "loss": 0.0829,
+      "step": 11175
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3347283913269959,
+      "learning_rate": 8.890671466820821e-08,
+      "loss": 0.0877,
+      "step": 11176
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.406050802050187,
+      "learning_rate": 8.865070597728453e-08,
+      "loss": 0.104,
+      "step": 11177
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.24734928450503685,
+      "learning_rate": 8.839506311226654e-08,
+      "loss": 0.075,
+      "step": 11178
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.307723288515105,
+      "learning_rate": 8.813978609219676e-08,
+      "loss": 0.0794,
+      "step": 11179
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3121892422367239,
+      "learning_rate": 8.788487493608944e-08,
+      "loss": 0.0887,
+      "step": 11180
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.19868817136489092,
+      "learning_rate": 8.763032966293161e-08,
+      "loss": 0.0469,
+      "step": 11181
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.33222524979361934,
+      "learning_rate": 8.737615029168367e-08,
+      "loss": 0.0777,
+      "step": 11182
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.27398542395398534,
+      "learning_rate": 8.712233684127824e-08,
+      "loss": 0.067,
+      "step": 11183
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1366121371041498,
+      "learning_rate": 8.686888933062076e-08,
+      "loss": 0.1125,
+      "step": 11184
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.4938426717019866,
+      "learning_rate": 8.661580777858947e-08,
+      "loss": 0.1099,
+      "step": 11185
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.32501706599396374,
+      "learning_rate": 8.636309220403538e-08,
+      "loss": 0.0395,
+      "step": 11186
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.59213436980137,
+      "learning_rate": 8.611074262578234e-08,
+      "loss": 0.0715,
+      "step": 11187
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.19225049871954866,
+      "learning_rate": 8.585875906262697e-08,
+      "loss": 0.0356,
+      "step": 11188
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.28075366090648485,
+      "learning_rate": 8.56071415333376e-08,
+      "loss": 0.0796,
+      "step": 11189
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.34416319446156557,
+      "learning_rate": 8.535589005665701e-08,
+      "loss": 0.0676,
+      "step": 11190
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3300993629931398,
+      "learning_rate": 8.510500465129967e-08,
+      "loss": 0.0997,
+      "step": 11191
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.23165686311412675,
+      "learning_rate": 8.485448533595287e-08,
+      "loss": 0.0604,
+      "step": 11192
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.38879980953372406,
+      "learning_rate": 8.460433212927721e-08,
+      "loss": 0.0691,
+      "step": 11193
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.23818055724998383,
+      "learning_rate": 8.435454504990503e-08,
+      "loss": 0.0707,
+      "step": 11194
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2671763339942266,
+      "learning_rate": 8.4105124116442e-08,
+      "loss": 0.0623,
+      "step": 11195
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.20741074287060438,
+      "learning_rate": 8.385606934746604e-08,
+      "loss": 0.045,
+      "step": 11196
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.431272790347049,
+      "learning_rate": 8.360738076152953e-08,
+      "loss": 0.1092,
+      "step": 11197
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.47105150349570835,
+      "learning_rate": 8.335905837715485e-08,
+      "loss": 0.1013,
+      "step": 11198
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.28014076928953896,
+      "learning_rate": 8.311110221283947e-08,
+      "loss": 0.0531,
+      "step": 11199
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3032661527429069,
+      "learning_rate": 8.286351228705192e-08,
+      "loss": 0.0559,
+      "step": 11200
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.46947631512391585,
+      "learning_rate": 8.261628861823467e-08,
+      "loss": 0.085,
+      "step": 11201
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.28421430748229026,
+      "learning_rate": 8.23694312248019e-08,
+      "loss": 0.0747,
+      "step": 11202
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.31799200737196076,
+      "learning_rate": 8.21229401251411e-08,
+      "loss": 0.0966,
+      "step": 11203
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.280505804045926,
+      "learning_rate": 8.18768153376126e-08,
+      "loss": 0.0655,
+      "step": 11204
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3987247531982659,
+      "learning_rate": 8.163105688054896e-08,
+      "loss": 0.1045,
+      "step": 11205
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2809256498676631,
+      "learning_rate": 8.138566477225607e-08,
+      "loss": 0.0603,
+      "step": 11206
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5603739508641362,
+      "learning_rate": 8.114063903101154e-08,
+      "loss": 0.135,
+      "step": 11207
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2675628912681819,
+      "learning_rate": 8.089597967506746e-08,
+      "loss": 0.0525,
+      "step": 11208
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2214586840481978,
+      "learning_rate": 8.065168672264589e-08,
+      "loss": 0.0637,
+      "step": 11209
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.30433944747903424,
+      "learning_rate": 8.040776019194396e-08,
+      "loss": 0.08,
+      "step": 11210
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.32926647978052526,
+      "learning_rate": 8.016420010113158e-08,
+      "loss": 0.0888,
+      "step": 11211
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.23041305092221398,
+      "learning_rate": 7.992100646834921e-08,
+      "loss": 0.0391,
+      "step": 11212
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.35414899545237477,
+      "learning_rate": 7.96781793117124e-08,
+      "loss": 0.0924,
+      "step": 11213
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2385904184812834,
+      "learning_rate": 7.943571864930667e-08,
+      "loss": 0.1019,
+      "step": 11214
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.33139354274541266,
+      "learning_rate": 7.919362449919421e-08,
+      "loss": 0.0793,
+      "step": 11215
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.32304354652793965,
+      "learning_rate": 7.895189687940563e-08,
+      "loss": 0.0705,
+      "step": 11216
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.36539358690478235,
+      "learning_rate": 7.871053580794763e-08,
+      "loss": 0.0745,
+      "step": 11217
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.38962341800656736,
+      "learning_rate": 7.846954130279693e-08,
+      "loss": 0.0829,
+      "step": 11218
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2877837003591167,
+      "learning_rate": 7.82289133819053e-08,
+      "loss": 0.0673,
+      "step": 11219
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.26208692568085346,
+      "learning_rate": 7.798865206319561e-08,
+      "loss": 0.0384,
+      "step": 11220
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.19061227930757443,
+      "learning_rate": 7.774875736456356e-08,
+      "loss": 0.0625,
+      "step": 11221
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.21371070667596545,
+      "learning_rate": 7.750922930387871e-08,
+      "loss": 0.0607,
+      "step": 11222
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3625241741603625,
+      "learning_rate": 7.727006789898184e-08,
+      "loss": 0.0844,
+      "step": 11223
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2377337032398751,
+      "learning_rate": 7.703127316768754e-08,
+      "loss": 0.0977,
+      "step": 11224
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3076311041441613,
+      "learning_rate": 7.679284512778218e-08,
+      "loss": 0.0869,
+      "step": 11225
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.24889487445787453,
+      "learning_rate": 7.655478379702597e-08,
+      "loss": 0.0778,
+      "step": 11226
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3041811454513362,
+      "learning_rate": 7.631708919314973e-08,
+      "loss": 0.0739,
+      "step": 11227
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3837462729933248,
+      "learning_rate": 7.607976133385986e-08,
+      "loss": 0.055,
+      "step": 11228
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.22703436802581847,
+      "learning_rate": 7.584280023683333e-08,
+      "loss": 0.053,
+      "step": 11229
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.317863921746909,
+      "learning_rate": 7.560620591972046e-08,
+      "loss": 0.0701,
+      "step": 11230
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3093262347744054,
+      "learning_rate": 7.536997840014382e-08,
+      "loss": 0.0931,
+      "step": 11231
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.19468723014563755,
+      "learning_rate": 7.513411769569933e-08,
+      "loss": 0.0563,
+      "step": 11232
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.457678307666195,
+      "learning_rate": 7.489862382395519e-08,
+      "loss": 0.0929,
+      "step": 11233
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.4859982354400338,
+      "learning_rate": 7.466349680245289e-08,
+      "loss": 0.0986,
+      "step": 11234
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.23526808004245472,
+      "learning_rate": 7.442873664870509e-08,
+      "loss": 0.0597,
+      "step": 11235
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.49011815495816713,
+      "learning_rate": 7.419434338019893e-08,
+      "loss": 0.0911,
+      "step": 11236
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3729100628997059,
+      "learning_rate": 7.39603170143921e-08,
+      "loss": 0.0903,
+      "step": 11237
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3063423895087612,
+      "learning_rate": 7.372665756871789e-08,
+      "loss": 0.0758,
+      "step": 11238
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.24011249765651202,
+      "learning_rate": 7.349336506058014e-08,
+      "loss": 0.076,
+      "step": 11239
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5580741005454749,
+      "learning_rate": 7.326043950735495e-08,
+      "loss": 0.1224,
+      "step": 11240
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.22135717131903207,
+      "learning_rate": 7.30278809263929e-08,
+      "loss": 0.0752,
+      "step": 11241
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.29234339820697935,
+      "learning_rate": 7.279568933501624e-08,
+      "loss": 0.0711,
+      "step": 11242
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.35048190751767727,
+      "learning_rate": 7.256386475051947e-08,
+      "loss": 0.0721,
+      "step": 11243
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.23415601153607596,
+      "learning_rate": 7.233240719016988e-08,
+      "loss": 0.0501,
+      "step": 11244
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.17600387466280934,
+      "learning_rate": 7.210131667120924e-08,
+      "loss": 0.0259,
+      "step": 11245
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.404367456852045,
+      "learning_rate": 7.187059321084933e-08,
+      "loss": 0.0674,
+      "step": 11246
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6044465104869069,
+      "learning_rate": 7.164023682627641e-08,
+      "loss": 0.0912,
+      "step": 11247
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.34610343308176733,
+      "learning_rate": 7.141024753464843e-08,
+      "loss": 0.1046,
+      "step": 11248
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.32783883795532953,
+      "learning_rate": 7.118062535309611e-08,
+      "loss": 0.082,
+      "step": 11249
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3069930541123221,
+      "learning_rate": 7.095137029872301e-08,
+      "loss": 0.0841,
+      "step": 11250
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.27487476985542186,
+      "learning_rate": 7.072248238860602e-08,
+      "loss": 0.068,
+      "step": 11251
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.45129739842020855,
+      "learning_rate": 7.049396163979427e-08,
+      "loss": 0.1189,
+      "step": 11252
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.23670679181427776,
+      "learning_rate": 7.026580806930805e-08,
+      "loss": 0.0647,
+      "step": 11253
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5537412171512616,
+      "learning_rate": 7.003802169414209e-08,
+      "loss": 0.078,
+      "step": 11254
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3229884078768127,
+      "learning_rate": 6.981060253126392e-08,
+      "loss": 0.0823,
+      "step": 11255
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.37962925504819794,
+      "learning_rate": 6.95835505976128e-08,
+      "loss": 0.0776,
+      "step": 11256
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2829364961804394,
+      "learning_rate": 6.935686591010015e-08,
+      "loss": 0.0734,
+      "step": 11257
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.29742728719668166,
+      "learning_rate": 6.91305484856114e-08,
+      "loss": 0.0599,
+      "step": 11258
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.33796634688926985,
+      "learning_rate": 6.89045983410036e-08,
+      "loss": 0.0757,
+      "step": 11259
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5165733335515028,
+      "learning_rate": 6.867901549310774e-08,
+      "loss": 0.1179,
+      "step": 11260
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.4098000336506173,
+      "learning_rate": 6.845379995872481e-08,
+      "loss": 0.0948,
+      "step": 11261
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2680233413513996,
+      "learning_rate": 6.822895175463195e-08,
+      "loss": 0.0588,
+      "step": 11262
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.44541560692359705,
+      "learning_rate": 6.800447089757689e-08,
+      "loss": 0.1417,
+      "step": 11263
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.356844719759819,
+      "learning_rate": 6.778035740427902e-08,
+      "loss": 0.0904,
+      "step": 11264
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.29956693348692304,
+      "learning_rate": 6.755661129143276e-08,
+      "loss": 0.0387,
+      "step": 11265
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.32630871221460434,
+      "learning_rate": 6.733323257570368e-08,
+      "loss": 0.0822,
+      "step": 11266
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.4803437241753069,
+      "learning_rate": 6.711022127373012e-08,
+      "loss": 0.1288,
+      "step": 11267
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3278616547359571,
+      "learning_rate": 6.688757740212382e-08,
+      "loss": 0.0807,
+      "step": 11268
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.509827095141917,
+      "learning_rate": 6.666530097746814e-08,
+      "loss": 0.1139,
+      "step": 11269
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3528096830949959,
+      "learning_rate": 6.644339201631933e-08,
+      "loss": 0.0961,
+      "step": 11270
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2263115088115778,
+      "learning_rate": 6.622185053520691e-08,
+      "loss": 0.0644,
+      "step": 11271
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3074063761545856,
+      "learning_rate": 6.600067655063269e-08,
+      "loss": 0.047,
+      "step": 11272
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.27142098272127957,
+      "learning_rate": 6.577987007907071e-08,
+      "loss": 0.066,
+      "step": 11273
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.32247256335687213,
+      "learning_rate": 6.555943113696783e-08,
+      "loss": 0.0846,
+      "step": 11274
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2766507074880423,
+      "learning_rate": 6.533935974074369e-08,
+      "loss": 0.0677,
+      "step": 11275
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3706051400807488,
+      "learning_rate": 6.511965590679126e-08,
+      "loss": 0.0788,
+      "step": 11276
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5334665315574001,
+      "learning_rate": 6.490031965147415e-08,
+      "loss": 0.0927,
+      "step": 11277
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.37372903024697623,
+      "learning_rate": 6.468135099112982e-08,
+      "loss": 0.0964,
+      "step": 11278
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.39637452293583175,
+      "learning_rate": 6.446274994206969e-08,
+      "loss": 0.1148,
+      "step": 11279
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.39408032747648475,
+      "learning_rate": 6.424451652057517e-08,
+      "loss": 0.056,
+      "step": 11280
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2306670636104225,
+      "learning_rate": 6.402665074290215e-08,
+      "loss": 0.058,
+      "step": 11281
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.41757942126120007,
+      "learning_rate": 6.380915262527765e-08,
+      "loss": 0.0845,
+      "step": 11282
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.30454904877306804,
+      "learning_rate": 6.35920221839037e-08,
+      "loss": 0.065,
+      "step": 11283
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5498314056471072,
+      "learning_rate": 6.337525943495237e-08,
+      "loss": 0.1256,
+      "step": 11284
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.30493502280316315,
+      "learning_rate": 6.315886439456964e-08,
+      "loss": 0.0449,
+      "step": 11285
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.4580618055995727,
+      "learning_rate": 6.294283707887372e-08,
+      "loss": 0.0887,
+      "step": 11286
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.28530437728878005,
+      "learning_rate": 6.272717750395618e-08,
+      "loss": 0.0658,
+      "step": 11287
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2645329507799146,
+      "learning_rate": 6.251188568587973e-08,
+      "loss": 0.081,
+      "step": 11288
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.31904548399563337,
+      "learning_rate": 6.229696164068156e-08,
+      "loss": 0.0714,
+      "step": 11289
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2053897971750405,
+      "learning_rate": 6.208240538436993e-08,
+      "loss": 0.0504,
+      "step": 11290
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3165104364030358,
+      "learning_rate": 6.186821693292599e-08,
+      "loss": 0.0787,
+      "step": 11291
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.28628400124775477,
+      "learning_rate": 6.165439630230363e-08,
+      "loss": 0.0723,
+      "step": 11292
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3514415429696445,
+      "learning_rate": 6.144094350843066e-08,
+      "loss": 0.0562,
+      "step": 11293
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6287995790262101,
+      "learning_rate": 6.122785856720492e-08,
+      "loss": 0.1061,
+      "step": 11294
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.34477923905358565,
+      "learning_rate": 6.101514149449872e-08,
+      "loss": 0.0685,
+      "step": 11295
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.21466988875106274,
+      "learning_rate": 6.08027923061566e-08,
+      "loss": 0.0689,
+      "step": 11296
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3399869100395212,
+      "learning_rate": 6.059081101799535e-08,
+      "loss": 0.0799,
+      "step": 11297
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5183571381734173,
+      "learning_rate": 6.037919764580457e-08,
+      "loss": 0.076,
+      "step": 11298
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3056311888900178,
+      "learning_rate": 6.016795220534721e-08,
+      "loss": 0.0847,
+      "step": 11299
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3469973640215041,
+      "learning_rate": 5.995707471235679e-08,
+      "loss": 0.0855,
+      "step": 11300
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.4321331394795447,
+      "learning_rate": 5.97465651825413e-08,
+      "loss": 0.0739,
+      "step": 11301
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.24505976539533014,
+      "learning_rate": 5.953642363158096e-08,
+      "loss": 0.0715,
+      "step": 11302
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.4520016443839452,
+      "learning_rate": 5.9326650075128276e-08,
+      "loss": 0.0945,
+      "step": 11303
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.47271434561429265,
+      "learning_rate": 5.9117244528808516e-08,
+      "loss": 0.0989,
+      "step": 11304
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3890941574697573,
+      "learning_rate": 5.890820700821809e-08,
+      "loss": 0.0917,
+      "step": 11305
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3616456320561048,
+      "learning_rate": 5.869953752892954e-08,
+      "loss": 0.088,
+      "step": 11306
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.4044091054774401,
+      "learning_rate": 5.8491236106483774e-08,
+      "loss": 0.0935,
+      "step": 11307
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3332779043928037,
+      "learning_rate": 5.828330275639782e-08,
+      "loss": 0.1093,
+      "step": 11308
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6236353017227955,
+      "learning_rate": 5.807573749415873e-08,
+      "loss": 0.1011,
+      "step": 11309
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.320432391036484,
+      "learning_rate": 5.786854033522748e-08,
+      "loss": 0.071,
+      "step": 11310
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.34481205906138357,
+      "learning_rate": 5.7661711295037836e-08,
+      "loss": 0.0689,
+      "step": 11311
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3350922360027979,
+      "learning_rate": 5.745525038899469e-08,
+      "loss": 0.0888,
+      "step": 11312
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.4635520634090538,
+      "learning_rate": 5.724915763247685e-08,
+      "loss": 0.0512,
+      "step": 11313
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2639679656239182,
+      "learning_rate": 5.704343304083593e-08,
+      "loss": 0.0736,
+      "step": 11314
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.16673479432900848,
+      "learning_rate": 5.683807662939467e-08,
+      "loss": 0.0317,
+      "step": 11315
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3974567446005665,
+      "learning_rate": 5.663308841344972e-08,
+      "loss": 0.092,
+      "step": 11316
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3610695959370713,
+      "learning_rate": 5.6428468408269435e-08,
+      "loss": 0.0677,
+      "step": 11317
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3203274888251374,
+      "learning_rate": 5.622421662909494e-08,
+      "loss": 0.0641,
+      "step": 11318
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.17055499379425346,
+      "learning_rate": 5.6020333091140743e-08,
+      "loss": 0.0543,
+      "step": 11319
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3333436329644468,
+      "learning_rate": 5.5816817809593026e-08,
+      "loss": 0.0799,
+      "step": 11320
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.35037645775671616,
+      "learning_rate": 5.5613670799610774e-08,
+      "loss": 0.1119,
+      "step": 11321
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5187449920728862,
+      "learning_rate": 5.541089207632522e-08,
+      "loss": 0.1084,
+      "step": 11322
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.32895272099749806,
+      "learning_rate": 5.520848165484094e-08,
+      "loss": 0.0872,
+      "step": 11323
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.42981996460968186,
+      "learning_rate": 5.500643955023477e-08,
+      "loss": 0.0711,
+      "step": 11324
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.17831005122805985,
+      "learning_rate": 5.480476577755522e-08,
+      "loss": 0.0551,
+      "step": 11325
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3045963024840234,
+      "learning_rate": 5.4603460351824734e-08,
+      "loss": 0.082,
+      "step": 11326
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.35660415002327656,
+      "learning_rate": 5.440252328803741e-08,
+      "loss": 0.0682,
+      "step": 11327
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.34131128869374466,
+      "learning_rate": 5.420195460116073e-08,
+      "loss": 0.0869,
+      "step": 11328
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.26880446199797353,
+      "learning_rate": 5.400175430613386e-08,
+      "loss": 0.0818,
+      "step": 11329
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3434558816849806,
+      "learning_rate": 5.3801922417868745e-08,
+      "loss": 0.0965,
+      "step": 11330
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.36512128709644526,
+      "learning_rate": 5.360245895125016e-08,
+      "loss": 0.0956,
+      "step": 11331
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.304488699632571,
+      "learning_rate": 5.34033639211351e-08,
+      "loss": 0.0824,
+      "step": 11332
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8430855127807839,
+      "learning_rate": 5.320463734235393e-08,
+      "loss": 0.1699,
+      "step": 11333
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3038620563934288,
+      "learning_rate": 5.300627922970869e-08,
+      "loss": 0.0699,
+      "step": 11334
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.30726140059862755,
+      "learning_rate": 5.280828959797424e-08,
+      "loss": 0.0846,
+      "step": 11335
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3455810027705089,
+      "learning_rate": 5.2610668461897105e-08,
+      "loss": 0.0947,
+      "step": 11336
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.24333341421881813,
+      "learning_rate": 5.241341583619886e-08,
+      "loss": 0.0542,
+      "step": 11337
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.30168058041806944,
+      "learning_rate": 5.2216531735571064e-08,
+      "loss": 0.0804,
+      "step": 11338
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2400316864465075,
+      "learning_rate": 5.2020016174678665e-08,
+      "loss": 0.0489,
+      "step": 11339
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.458949354416097,
+      "learning_rate": 5.18238691681594e-08,
+      "loss": 0.0604,
+      "step": 11340
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3867263906489827,
+      "learning_rate": 5.1628090730624355e-08,
+      "loss": 0.0936,
+      "step": 11341
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.224722510843846,
+      "learning_rate": 5.143268087665465e-08,
+      "loss": 0.0641,
+      "step": 11342
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3375220001873892,
+      "learning_rate": 5.123763962080697e-08,
+      "loss": 0.0735,
+      "step": 11343
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3029390357604747,
+      "learning_rate": 5.104296697760858e-08,
+      "loss": 0.0725,
+      "step": 11344
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.1926658989417735,
+      "learning_rate": 5.0848662961560106e-08,
+      "loss": 0.0392,
+      "step": 11345
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3775448902036218,
+      "learning_rate": 5.0654727587133303e-08,
+      "loss": 0.0602,
+      "step": 11346
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4649038124308176,
+      "learning_rate": 5.046116086877495e-08,
+      "loss": 0.1244,
+      "step": 11347
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2687562557865079,
+      "learning_rate": 5.02679628209024e-08,
+      "loss": 0.0804,
+      "step": 11348
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.35274356978120736,
+      "learning_rate": 5.0075133457906934e-08,
+      "loss": 0.0693,
+      "step": 11349
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2150622420007374,
+      "learning_rate": 4.988267279414982e-08,
+      "loss": 0.0545,
+      "step": 11350
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2727031397930467,
+      "learning_rate": 4.96905808439685e-08,
+      "loss": 0.068,
+      "step": 11351
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.33424933316428324,
+      "learning_rate": 4.949885762167039e-08,
+      "loss": 0.0672,
+      "step": 11352
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.44138481099346616,
+      "learning_rate": 4.930750314153632e-08,
+      "loss": 0.0557,
+      "step": 11353
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4013189476395214,
+      "learning_rate": 4.911651741781875e-08,
+      "loss": 0.1293,
+      "step": 11354
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3881659899305356,
+      "learning_rate": 4.8925900464744656e-08,
+      "loss": 0.0968,
+      "step": 11355
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4638157795291874,
+      "learning_rate": 4.8735652296511006e-08,
+      "loss": 0.0827,
+      "step": 11356
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2250924702952408,
+      "learning_rate": 4.854577292728979e-08,
+      "loss": 0.0593,
+      "step": 11357
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.25333487880204586,
+      "learning_rate": 4.8356262371223595e-08,
+      "loss": 0.0561,
+      "step": 11358
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2964428085683523,
+      "learning_rate": 4.816712064242834e-08,
+      "loss": 0.0959,
+      "step": 11359
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.276022621851475,
+      "learning_rate": 4.797834775499222e-08,
+      "loss": 0.0796,
+      "step": 11360
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.28001006366069237,
+      "learning_rate": 4.778994372297674e-08,
+      "loss": 0.0444,
+      "step": 11361
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3495570352991,
+      "learning_rate": 4.760190856041513e-08,
+      "loss": 0.0759,
+      "step": 11362
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0046541657819519,
+      "learning_rate": 4.741424228131286e-08,
+      "loss": 0.122,
+      "step": 11363
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.43400250443587873,
+      "learning_rate": 4.7226944899649296e-08,
+      "loss": 0.118,
+      "step": 11364
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.16205431210960036,
+      "learning_rate": 4.704001642937439e-08,
+      "loss": 0.0412,
+      "step": 11365
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.31436112819749146,
+      "learning_rate": 4.6853456884412004e-08,
+      "loss": 0.0837,
+      "step": 11366
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.34379047261929285,
+      "learning_rate": 4.666726627865881e-08,
+      "loss": 0.0968,
+      "step": 11367
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4771284870334238,
+      "learning_rate": 4.648144462598314e-08,
+      "loss": 0.1117,
+      "step": 11368
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.31492308559775567,
+      "learning_rate": 4.6295991940225603e-08,
+      "loss": 0.0559,
+      "step": 11369
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5790130886573577,
+      "learning_rate": 4.6110908235199595e-08,
+      "loss": 0.0824,
+      "step": 11370
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4433207158290598,
+      "learning_rate": 4.592619352469241e-08,
+      "loss": 0.1202,
+      "step": 11371
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2857450995265564,
+      "learning_rate": 4.574184782246138e-08,
+      "loss": 0.0612,
+      "step": 11372
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.35139243849597784,
+      "learning_rate": 4.555787114223831e-08,
+      "loss": 0.0868,
+      "step": 11373
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3795460743533305,
+      "learning_rate": 4.537426349772722e-08,
+      "loss": 0.1096,
+      "step": 11374
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.30537323141618544,
+      "learning_rate": 4.5191024902603297e-08,
+      "loss": 0.0899,
+      "step": 11375
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.38482976203550645,
+      "learning_rate": 4.5008155370516174e-08,
+      "loss": 0.0924,
+      "step": 11376
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.25073260386258756,
+      "learning_rate": 4.482565491508606e-08,
+      "loss": 0.0843,
+      "step": 11377
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2802971419188257,
+      "learning_rate": 4.464352354990764e-08,
+      "loss": 0.0639,
+      "step": 11378
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.36716505550610323,
+      "learning_rate": 4.4461761288546735e-08,
+      "loss": 0.1027,
+      "step": 11379
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4472378823494747,
+      "learning_rate": 4.4280368144541396e-08,
+      "loss": 0.074,
+      "step": 11380
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.27146838768130216,
+      "learning_rate": 4.409934413140415e-08,
+      "loss": 0.0485,
+      "step": 11381
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.31941422477928805,
+      "learning_rate": 4.391868926261755e-08,
+      "loss": 0.0596,
+      "step": 11382
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3278798804123493,
+      "learning_rate": 4.373840355163805e-08,
+      "loss": 0.0718,
+      "step": 11383
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4530233184171188,
+      "learning_rate": 4.3558487011895464e-08,
+      "loss": 0.0799,
+      "step": 11384
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3788073125037581,
+      "learning_rate": 4.337893965678963e-08,
+      "loss": 0.0844,
+      "step": 11385
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3639434449392199,
+      "learning_rate": 4.319976149969485e-08,
+      "loss": 0.0634,
+      "step": 11386
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2553009224275852,
+      "learning_rate": 4.302095255395711e-08,
+      "loss": 0.0629,
+      "step": 11387
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.305499144133457,
+      "learning_rate": 4.2842512832895754e-08,
+      "loss": 0.0676,
+      "step": 11388
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5192241469953546,
+      "learning_rate": 4.2664442349801273e-08,
+      "loss": 0.1343,
+      "step": 11389
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3865900443191973,
+      "learning_rate": 4.2486741117938044e-08,
+      "loss": 0.112,
+      "step": 11390
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5474079939428419,
+      "learning_rate": 4.2309409150541604e-08,
+      "loss": 0.1175,
+      "step": 11391
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4026515787344727,
+      "learning_rate": 4.213244646082137e-08,
+      "loss": 0.0919,
+      "step": 11392
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3829483578009171,
+      "learning_rate": 4.1955853061958486e-08,
+      "loss": 0.0822,
+      "step": 11393
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2243559674568583,
+      "learning_rate": 4.1779628967105745e-08,
+      "loss": 0.0734,
+      "step": 11394
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.27013529964705557,
+      "learning_rate": 4.160377418939043e-08,
+      "loss": 0.0815,
+      "step": 11395
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.31698877448499074,
+      "learning_rate": 4.142828874191096e-08,
+      "loss": 0.106,
+      "step": 11396
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2828775329304314,
+      "learning_rate": 4.125317263773798e-08,
+      "loss": 0.0856,
+      "step": 11397
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.29835151369007096,
+      "learning_rate": 4.107842588991606e-08,
+      "loss": 0.0633,
+      "step": 11398
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2713615608164801,
+      "learning_rate": 4.0904048511460347e-08,
+      "loss": 0.0452,
+      "step": 11399
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.27501725023227286,
+      "learning_rate": 4.073004051535989e-08,
+      "loss": 0.0628,
+      "step": 11400
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.31856142878808086,
+      "learning_rate": 4.055640191457599e-08,
+      "loss": 0.0656,
+      "step": 11401
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.23178081114532081,
+      "learning_rate": 4.038313272204275e-08,
+      "loss": 0.0752,
+      "step": 11402
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.42608663349695775,
+      "learning_rate": 4.02102329506654e-08,
+      "loss": 0.0916,
+      "step": 11403
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.347003711742509,
+      "learning_rate": 4.003770261332252e-08,
+      "loss": 0.058,
+      "step": 11404
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.24440182950855546,
+      "learning_rate": 3.98655417228655e-08,
+      "loss": 0.0659,
+      "step": 11405
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.19158227399747982,
+      "learning_rate": 3.969375029211797e-08,
+      "loss": 0.0433,
+      "step": 11406
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3341533937365476,
+      "learning_rate": 3.952232833387582e-08,
+      "loss": 0.055,
+      "step": 11407
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.23059763784526194,
+      "learning_rate": 3.9351275860907146e-08,
+      "loss": 0.0693,
+      "step": 11408
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.19838686076757472,
+      "learning_rate": 3.918059288595399e-08,
+      "loss": 0.0423,
+      "step": 11409
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.30479441979708793,
+      "learning_rate": 3.901027942172897e-08,
+      "loss": 0.0753,
+      "step": 11410
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4155106453694752,
+      "learning_rate": 3.884033548091804e-08,
+      "loss": 0.1015,
+      "step": 11411
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2870839132161629,
+      "learning_rate": 3.8670761076180505e-08,
+      "loss": 0.0972,
+      "step": 11412
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3375010673678632,
+      "learning_rate": 3.850155622014573e-08,
+      "loss": 0.0793,
+      "step": 11413
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3222396983060046,
+      "learning_rate": 3.8332720925418064e-08,
+      "loss": 0.0869,
+      "step": 11414
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3717009578882101,
+      "learning_rate": 3.8164255204573566e-08,
+      "loss": 0.1002,
+      "step": 11415
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.29335721506650886,
+      "learning_rate": 3.799615907016052e-08,
+      "loss": 0.0956,
+      "step": 11416
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2965948335629752,
+      "learning_rate": 3.782843253469892e-08,
+      "loss": 0.0914,
+      "step": 11417
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.28385170346303873,
+      "learning_rate": 3.7661075610682104e-08,
+      "loss": 0.0508,
+      "step": 11418
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3172264753490098,
+      "learning_rate": 3.749408831057622e-08,
+      "loss": 0.0765,
+      "step": 11419
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.1819224411021191,
+      "learning_rate": 3.732747064681963e-08,
+      "loss": 0.0372,
+      "step": 11420
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.45227763654834774,
+      "learning_rate": 3.716122263182298e-08,
+      "loss": 0.0827,
+      "step": 11421
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.19256002111206616,
+      "learning_rate": 3.6995344277968584e-08,
+      "loss": 0.0413,
+      "step": 11422
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.34501944133766654,
+      "learning_rate": 3.6829835597612664e-08,
+      "loss": 0.078,
+      "step": 11423
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4492894490879793,
+      "learning_rate": 3.666469660308369e-08,
+      "loss": 0.1071,
+      "step": 11424
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.23785379162235312,
+      "learning_rate": 3.6499927306681284e-08,
+      "loss": 0.0412,
+      "step": 11425
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2543995415106295,
+      "learning_rate": 3.6335527720678945e-08,
+      "loss": 0.0832,
+      "step": 11426
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3442318256900812,
+      "learning_rate": 3.617149785732188e-08,
+      "loss": 0.0397,
+      "step": 11427
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.33131528303423136,
+      "learning_rate": 3.60078377288281e-08,
+      "loss": 0.0866,
+      "step": 11428
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5195593079643612,
+      "learning_rate": 3.584454734738785e-08,
+      "loss": 0.1303,
+      "step": 11429
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.21342066166493776,
+      "learning_rate": 3.568162672516418e-08,
+      "loss": 0.0652,
+      "step": 11430
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7086892704107729,
+      "learning_rate": 3.5519075874292376e-08,
+      "loss": 0.1394,
+      "step": 11431
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.21496705789083373,
+      "learning_rate": 3.535689480687998e-08,
+      "loss": 0.0478,
+      "step": 11432
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5863225671315513,
+      "learning_rate": 3.519508353500733e-08,
+      "loss": 0.1033,
+      "step": 11433
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3741328718146961,
+      "learning_rate": 3.5033642070726457e-08,
+      "loss": 0.0692,
+      "step": 11434
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.27093058300583583,
+      "learning_rate": 3.487257042606329e-08,
+      "loss": 0.0721,
+      "step": 11435
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.24102486483142388,
+      "learning_rate": 3.4711868613015457e-08,
+      "loss": 0.0723,
+      "step": 11436
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.361043161509741,
+      "learning_rate": 3.455153664355226e-08,
+      "loss": 0.1138,
+      "step": 11437
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.34571186902891393,
+      "learning_rate": 3.439157452961639e-08,
+      "loss": 0.0953,
+      "step": 11438
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5790109348841226,
+      "learning_rate": 3.4231982283123296e-08,
+      "loss": 0.1199,
+      "step": 11439
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.24737178493504416,
+      "learning_rate": 3.4072759915959594e-08,
+      "loss": 0.0827,
+      "step": 11440
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4066289355514357,
+      "learning_rate": 3.391390743998524e-08,
+      "loss": 0.1047,
+      "step": 11441
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.28347114644906707,
+      "learning_rate": 3.3755424867032424e-08,
+      "loss": 0.0897,
+      "step": 11442
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.25358990304960355,
+      "learning_rate": 3.359731220890672e-08,
+      "loss": 0.0686,
+      "step": 11443
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.36827734514629457,
+      "learning_rate": 3.34395694773848e-08,
+      "loss": 0.0589,
+      "step": 11444
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.38108222768607725,
+      "learning_rate": 3.328219668421506e-08,
+      "loss": 0.0852,
+      "step": 11445
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.44714109758132564,
+      "learning_rate": 3.312519384112145e-08,
+      "loss": 0.0923,
+      "step": 11446
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3102352687356454,
+      "learning_rate": 3.296856095979739e-08,
+      "loss": 0.0717,
+      "step": 11447
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.30853595692004665,
+      "learning_rate": 3.2812298051909665e-08,
+      "loss": 0.0974,
+      "step": 11448
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.38497342614820634,
+      "learning_rate": 3.26564051290984e-08,
+      "loss": 0.0693,
+      "step": 11449
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4639917283424007,
+      "learning_rate": 3.250088220297487e-08,
+      "loss": 0.1143,
+      "step": 11450
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2761422054016099,
+      "learning_rate": 3.23457292851237e-08,
+      "loss": 0.0863,
+      "step": 11451
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2550914587048357,
+      "learning_rate": 3.219094638710119e-08,
+      "loss": 0.0821,
+      "step": 11452
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.540475332838519,
+      "learning_rate": 3.203653352043645e-08,
+      "loss": 0.1101,
+      "step": 11453
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.35515013916666804,
+      "learning_rate": 3.1882490696631406e-08,
+      "loss": 0.1226,
+      "step": 11454
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.18411805147902935,
+      "learning_rate": 3.172881792715965e-08,
+      "loss": 0.0503,
+      "step": 11455
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.23449252729901343,
+      "learning_rate": 3.157551522346758e-08,
+      "loss": 0.0553,
+      "step": 11456
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3115322918366016,
+      "learning_rate": 3.1422582596974946e-08,
+      "loss": 0.0678,
+      "step": 11457
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.21309368512204982,
+      "learning_rate": 3.127002005907209e-08,
+      "loss": 0.0481,
+      "step": 11458
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.33521428753774857,
+      "learning_rate": 3.1117827621123254e-08,
+      "loss": 0.087,
+      "step": 11459
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.38819730925766033,
+      "learning_rate": 3.0966005294464915e-08,
+      "loss": 0.0878,
+      "step": 11460
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.27438590140715324,
+      "learning_rate": 3.081455309040471e-08,
+      "loss": 0.0713,
+      "step": 11461
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3621791055278678,
+      "learning_rate": 3.0663471020224714e-08,
+      "loss": 0.0885,
+      "step": 11462
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3121781877821618,
+      "learning_rate": 3.0512759095177594e-08,
+      "loss": 0.0808,
+      "step": 11463
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3802480447638035,
+      "learning_rate": 3.036241732648937e-08,
+      "loss": 0.0765,
+      "step": 11464
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3494434942111629,
+      "learning_rate": 3.021244572535886e-08,
+      "loss": 0.103,
+      "step": 11465
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6263959216132077,
+      "learning_rate": 3.0062844302956586e-08,
+      "loss": 0.117,
+      "step": 11466
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.30290312226107696,
+      "learning_rate": 2.991361307042584e-08,
+      "loss": 0.0887,
+      "step": 11467
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.30156162515867974,
+      "learning_rate": 2.9764752038881627e-08,
+      "loss": 0.0832,
+      "step": 11468
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.24069645419334346,
+      "learning_rate": 2.9616261219412856e-08,
+      "loss": 0.053,
+      "step": 11469
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.31810159605174715,
+      "learning_rate": 2.9468140623079012e-08,
+      "loss": 0.0588,
+      "step": 11470
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3137005872257044,
+      "learning_rate": 2.9320390260914046e-08,
+      "loss": 0.0806,
+      "step": 11471
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.47357518913960495,
+      "learning_rate": 2.9173010143922487e-08,
+      "loss": 0.0736,
+      "step": 11472
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.326841670977742,
+      "learning_rate": 2.9026000283082776e-08,
+      "loss": 0.1028,
+      "step": 11473
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4108330679511492,
+      "learning_rate": 2.8879360689344492e-08,
+      "loss": 0.1119,
+      "step": 11474
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.31509106420968125,
+      "learning_rate": 2.8733091373630007e-08,
+      "loss": 0.0957,
+      "step": 11475
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.500468889654519,
+      "learning_rate": 2.8587192346835046e-08,
+      "loss": 0.1113,
+      "step": 11476
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.2600530867703393,
+      "learning_rate": 2.8441663619826477e-08,
+      "loss": 0.0855,
+      "step": 11477
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.2400533743940749,
+      "learning_rate": 2.829650520344396e-08,
+      "loss": 0.0587,
+      "step": 11478
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.24730895228165162,
+      "learning_rate": 2.815171710850051e-08,
+      "loss": 0.0615,
+      "step": 11479
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.26653627030677,
+      "learning_rate": 2.8007299345779727e-08,
+      "loss": 0.0572,
+      "step": 11480
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.34223765757323427,
+      "learning_rate": 2.7863251926040223e-08,
+      "loss": 0.0688,
+      "step": 11481
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.40197556147643376,
+      "learning_rate": 2.7719574860009534e-08,
+      "loss": 0.1034,
+      "step": 11482
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.38240479099749586,
+      "learning_rate": 2.7576268158391317e-08,
+      "loss": 0.0974,
+      "step": 11483
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.388620167106343,
+      "learning_rate": 2.74333318318587e-08,
+      "loss": 0.0825,
+      "step": 11484
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.30764787184791,
+      "learning_rate": 2.729076589105928e-08,
+      "loss": 0.0642,
+      "step": 11485
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.35943874571087586,
+      "learning_rate": 2.7148570346611225e-08,
+      "loss": 0.0954,
+      "step": 11486
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.2663779601904052,
+      "learning_rate": 2.7006745209107176e-08,
+      "loss": 0.0757,
+      "step": 11487
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3553796898844576,
+      "learning_rate": 2.686529048911035e-08,
+      "loss": 0.0931,
+      "step": 11488
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.46927607908348656,
+      "learning_rate": 2.6724206197157875e-08,
+      "loss": 0.1248,
+      "step": 11489
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.2730521614156334,
+      "learning_rate": 2.6583492343757454e-08,
+      "loss": 0.0692,
+      "step": 11490
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.31381966069078626,
+      "learning_rate": 2.6443148939390705e-08,
+      "loss": 0.064,
+      "step": 11491
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.32772199054595486,
+      "learning_rate": 2.630317599451149e-08,
+      "loss": 0.0901,
+      "step": 11492
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.37816602273875527,
+      "learning_rate": 2.6163573519545904e-08,
+      "loss": 0.0784,
+      "step": 11493
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.331847945439843,
+      "learning_rate": 2.602434152489175e-08,
+      "loss": 0.063,
+      "step": 11494
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4007527061843292,
+      "learning_rate": 2.5885480020920173e-08,
+      "loss": 0.0905,
+      "step": 11495
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.21585089844164831,
+      "learning_rate": 2.5746989017974567e-08,
+      "loss": 0.0585,
+      "step": 11496
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3515074925231003,
+      "learning_rate": 2.5608868526370568e-08,
+      "loss": 0.0999,
+      "step": 11497
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.47047882031664207,
+      "learning_rate": 2.5471118556396056e-08,
+      "loss": 0.1227,
+      "step": 11498
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3077140393327332,
+      "learning_rate": 2.5333739118310607e-08,
+      "loss": 0.0815,
+      "step": 11499
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.24090473496678147,
+      "learning_rate": 2.519673022234881e-08,
+      "loss": 0.0897,
+      "step": 11500
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.42061866033185114,
+      "learning_rate": 2.5060091878714167e-08,
+      "loss": 0.097,
+      "step": 11501
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.18005878418477725,
+      "learning_rate": 2.4923824097585202e-08,
+      "loss": 0.054,
+      "step": 11502
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.32607927631017386,
+      "learning_rate": 2.4787926889112134e-08,
+      "loss": 0.1094,
+      "step": 11503
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4369335721510544,
+      "learning_rate": 2.4652400263416864e-08,
+      "loss": 0.0954,
+      "step": 11504
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.26396388373818613,
+      "learning_rate": 2.451724423059465e-08,
+      "loss": 0.0669,
+      "step": 11505
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.25641996677247253,
+      "learning_rate": 2.4382458800711882e-08,
+      "loss": 0.0593,
+      "step": 11506
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.28026813077011703,
+      "learning_rate": 2.4248043983808867e-08,
+      "loss": 0.0827,
+      "step": 11507
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3502519443175744,
+      "learning_rate": 2.411399978989759e-08,
+      "loss": 0.0812,
+      "step": 11508
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.20350464725794706,
+      "learning_rate": 2.398032622896229e-08,
+      "loss": 0.0604,
+      "step": 11509
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.36021756932694515,
+      "learning_rate": 2.384702331095945e-08,
+      "loss": 0.0676,
+      "step": 11510
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.2006316583053726,
+      "learning_rate": 2.37140910458189e-08,
+      "loss": 0.0712,
+      "step": 11511
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3640464624018539,
+      "learning_rate": 2.3581529443441608e-08,
+      "loss": 0.1099,
+      "step": 11512
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3037238310506292,
+      "learning_rate": 2.3449338513701904e-08,
+      "loss": 0.0764,
+      "step": 11513
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3206207545015379,
+      "learning_rate": 2.3317518266445793e-08,
+      "loss": 0.0869,
+      "step": 11514
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.2973364896136456,
+      "learning_rate": 2.318606871149265e-08,
+      "loss": 0.0867,
+      "step": 11515
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4563819497232023,
+      "learning_rate": 2.305498985863297e-08,
+      "loss": 0.0819,
+      "step": 11516
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.2841969646588255,
+      "learning_rate": 2.2924281717630613e-08,
+      "loss": 0.0935,
+      "step": 11517
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.2192323421880098,
+      "learning_rate": 2.2793944298221128e-08,
+      "loss": 0.0569,
+      "step": 11518
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4873349525226456,
+      "learning_rate": 2.2663977610112852e-08,
+      "loss": 0.1165,
+      "step": 11519
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7466334148146463,
+      "learning_rate": 2.2534381662987492e-08,
+      "loss": 0.0984,
+      "step": 11520
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.269561793362833,
+      "learning_rate": 2.240515646649677e-08,
+      "loss": 0.0598,
+      "step": 11521
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4321745938994596,
+      "learning_rate": 2.2276302030266318e-08,
+      "loss": 0.0918,
+      "step": 11522
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.35306318159092687,
+      "learning_rate": 2.2147818363894568e-08,
+      "loss": 0.102,
+      "step": 11523
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.39154363362699474,
+      "learning_rate": 2.2019705476951646e-08,
+      "loss": 0.0719,
+      "step": 11524
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.28021996191133947,
+      "learning_rate": 2.1891963378979918e-08,
+      "loss": 0.0903,
+      "step": 11525
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.28381826649616543,
+      "learning_rate": 2.1764592079493995e-08,
+      "loss": 0.0727,
+      "step": 11526
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.28098312596080793,
+      "learning_rate": 2.1637591587981844e-08,
+      "loss": 0.0684,
+      "step": 11527
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6029841444074632,
+      "learning_rate": 2.1510961913903115e-08,
+      "loss": 0.1255,
+      "step": 11528
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3274631943879163,
+      "learning_rate": 2.138470306668916e-08,
+      "loss": 0.0431,
+      "step": 11529
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.22485210333720665,
+      "learning_rate": 2.1258815055745785e-08,
+      "loss": 0.0429,
+      "step": 11530
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4381206931078247,
+      "learning_rate": 2.1133297890448822e-08,
+      "loss": 0.0869,
+      "step": 11531
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4297777043068957,
+      "learning_rate": 2.1008151580148017e-08,
+      "loss": 0.0621,
+      "step": 11532
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3256433004477139,
+      "learning_rate": 2.08833761341648e-08,
+      "loss": 0.0672,
+      "step": 11533
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.33207108673669844,
+      "learning_rate": 2.0758971561793406e-08,
+      "loss": 0.0503,
+      "step": 11534
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.41434776562921183,
+      "learning_rate": 2.0634937872299754e-08,
+      "loss": 0.1095,
+      "step": 11535
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.26255068569039547,
+      "learning_rate": 2.051127507492312e-08,
+      "loss": 0.0994,
+      "step": 11536
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.49876291100700343,
+      "learning_rate": 2.038798317887447e-08,
+      "loss": 0.1305,
+      "step": 11537
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.35195645737122633,
+      "learning_rate": 2.0265062193337014e-08,
+      "loss": 0.1106,
+      "step": 11538
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3061923993169574,
+      "learning_rate": 2.0142512127466763e-08,
+      "loss": 0.0568,
+      "step": 11539
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3383181090125839,
+      "learning_rate": 2.0020332990391967e-08,
+      "loss": 0.0699,
+      "step": 11540
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.25599434500929386,
+      "learning_rate": 1.989852479121368e-08,
+      "loss": 0.0566,
+      "step": 11541
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3821385913076168,
+      "learning_rate": 1.9777087539004093e-08,
+      "loss": 0.0709,
+      "step": 11542
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.34559006744891935,
+      "learning_rate": 1.9656021242808742e-08,
+      "loss": 0.0656,
+      "step": 11543
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.45413643937277454,
+      "learning_rate": 1.9535325911645974e-08,
+      "loss": 0.0793,
+      "step": 11544
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4554363825208631,
+      "learning_rate": 1.941500155450582e-08,
+      "loss": 0.0855,
+      "step": 11545
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.23241215282092142,
+      "learning_rate": 1.9295048180349994e-08,
+      "loss": 0.0727,
+      "step": 11546
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.2577645239870664,
+      "learning_rate": 1.917546579811358e-08,
+      "loss": 0.0705,
+      "step": 11547
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7350815724816152,
+      "learning_rate": 1.9056254416704446e-08,
+      "loss": 0.1312,
+      "step": 11548
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.46447700153543364,
+      "learning_rate": 1.8937414045001046e-08,
+      "loss": 0.0982,
+      "step": 11549
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.35007205211580567,
+      "learning_rate": 1.88189446918563e-08,
+      "loss": 0.0655,
+      "step": 11550
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.28472921696865283,
+      "learning_rate": 1.8700846366093705e-08,
+      "loss": 0.0789,
+      "step": 11551
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.49510503638709313,
+      "learning_rate": 1.858311907651067e-08,
+      "loss": 0.0845,
+      "step": 11552
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.1784477893714195,
+      "learning_rate": 1.8465762831875734e-08,
+      "loss": 0.0423,
+      "step": 11553
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5431164617101403,
+      "learning_rate": 1.8348777640930236e-08,
+      "loss": 0.0945,
+      "step": 11554
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3925571091483308,
+      "learning_rate": 1.8232163512388325e-08,
+      "loss": 0.0862,
+      "step": 11555
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.34210563674857997,
+      "learning_rate": 1.8115920454935264e-08,
+      "loss": 0.0991,
+      "step": 11556
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3945014467416165,
+      "learning_rate": 1.80000484772308e-08,
+      "loss": 0.0825,
+      "step": 11557
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.30706276657426446,
+      "learning_rate": 1.7884547587904144e-08,
+      "loss": 0.0952,
+      "step": 11558
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.18054078498119652,
+      "learning_rate": 1.7769417795560074e-08,
+      "loss": 0.0359,
+      "step": 11559
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.2818013613223306,
+      "learning_rate": 1.76546591087734e-08,
+      "loss": 0.0817,
+      "step": 11560
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.47933782292267946,
+      "learning_rate": 1.7540271536091726e-08,
+      "loss": 0.0763,
+      "step": 11561
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.258962066596929,
+      "learning_rate": 1.742625508603546e-08,
+      "loss": 0.0416,
+      "step": 11562
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3765095489717722,
+      "learning_rate": 1.731260976709781e-08,
+      "loss": 0.1272,
+      "step": 11563
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4749963916557255,
+      "learning_rate": 1.7199335587743114e-08,
+      "loss": 0.1197,
+      "step": 11564
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8406285668901154,
+      "learning_rate": 1.7086432556408515e-08,
+      "loss": 0.1386,
+      "step": 11565
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.18078421383041954,
+      "learning_rate": 1.6973900681504506e-08,
+      "loss": 0.0535,
+      "step": 11566
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.23152975325523695,
+      "learning_rate": 1.686173997141216e-08,
+      "loss": 0.081,
+      "step": 11567
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.29758257104991653,
+      "learning_rate": 1.6749950434487018e-08,
+      "loss": 0.0864,
+      "step": 11568
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.24545763016054287,
+      "learning_rate": 1.6638532079054638e-08,
+      "loss": 0.0693,
+      "step": 11569
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4482030379949024,
+      "learning_rate": 1.6527484913414494e-08,
+      "loss": 0.0847,
+      "step": 11570
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.2631957732348104,
+      "learning_rate": 1.6416808945838304e-08,
+      "loss": 0.0623,
+      "step": 11571
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.4276227964281277,
+      "learning_rate": 1.6306504184570027e-08,
+      "loss": 0.1005,
+      "step": 11572
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.35335301594794294,
+      "learning_rate": 1.6196570637825317e-08,
+      "loss": 0.0743,
+      "step": 11573
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.38034496661137235,
+      "learning_rate": 1.608700831379262e-08,
+      "loss": 0.0789,
+      "step": 11574
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2171940583455382,
+      "learning_rate": 1.5977817220633184e-08,
+      "loss": 0.0623,
+      "step": 11575
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.3147256449362975,
+      "learning_rate": 1.5868997366479955e-08,
+      "loss": 0.0642,
+      "step": 11576
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.29207929128927024,
+      "learning_rate": 1.576054875943811e-08,
+      "loss": 0.0753,
+      "step": 11577
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.3026095848318076,
+      "learning_rate": 1.5652471407586743e-08,
+      "loss": 0.0911,
+      "step": 11578
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.47134602121213226,
+      "learning_rate": 1.5544765318974975e-08,
+      "loss": 0.1266,
+      "step": 11579
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.42788347066781407,
+      "learning_rate": 1.5437430501625272e-08,
+      "loss": 0.0958,
+      "step": 11580
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2111782848657347,
+      "learning_rate": 1.5330466963532908e-08,
+      "loss": 0.0579,
+      "step": 11581
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.3683916541020962,
+      "learning_rate": 1.5223874712665954e-08,
+      "loss": 0.0729,
+      "step": 11582
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.49798363479012253,
+      "learning_rate": 1.511765375696306e-08,
+      "loss": 0.1201,
+      "step": 11583
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2527166619435074,
+      "learning_rate": 1.5011804104335668e-08,
+      "loss": 0.0647,
+      "step": 11584
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.32205308461034304,
+      "learning_rate": 1.4906325762669705e-08,
+      "loss": 0.0806,
+      "step": 11585
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.327638273724475,
+      "learning_rate": 1.4801218739820544e-08,
+      "loss": 0.0901,
+      "step": 11586
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.35203439740372067,
+      "learning_rate": 1.4696483043617482e-08,
+      "loss": 0.1009,
+      "step": 11587
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.3212084621534735,
+      "learning_rate": 1.4592118681862056e-08,
+      "loss": 0.0749,
+      "step": 11588
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.35593589279390087,
+      "learning_rate": 1.448812566232749e-08,
+      "loss": 0.0905,
+      "step": 11589
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.37372612001663696,
+      "learning_rate": 1.4384503992760368e-08,
+      "loss": 0.0914,
+      "step": 11590
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2711063095822574,
+      "learning_rate": 1.428125368087896e-08,
+      "loss": 0.0627,
+      "step": 11591
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.19430711217725277,
+      "learning_rate": 1.4178374734373224e-08,
+      "loss": 0.0535,
+      "step": 11592
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.38588194515337526,
+      "learning_rate": 1.4075867160907031e-08,
+      "loss": 0.0882,
+      "step": 11593
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.3224665385214907,
+      "learning_rate": 1.3973730968115384e-08,
+      "loss": 0.1002,
+      "step": 11594
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.24794471194302883,
+      "learning_rate": 1.3871966163605533e-08,
+      "loss": 0.0716,
+      "step": 11595
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.36005254407483,
+      "learning_rate": 1.3770572754958633e-08,
+      "loss": 0.1072,
+      "step": 11596
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5281959621898545,
+      "learning_rate": 1.3669550749726424e-08,
+      "loss": 0.1104,
+      "step": 11597
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5774251795584968,
+      "learning_rate": 1.3568900155432885e-08,
+      "loss": 0.1243,
+      "step": 11598
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.19457642851823678,
+      "learning_rate": 1.3468620979576464e-08,
+      "loss": 0.0393,
+      "step": 11599
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2944445371860034,
+      "learning_rate": 1.336871322962563e-08,
+      "loss": 0.097,
+      "step": 11600
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.4060100705979313,
+      "learning_rate": 1.3269176913022208e-08,
+      "loss": 0.0825,
+      "step": 11601
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.27558439411468527,
+      "learning_rate": 1.3170012037180268e-08,
+      "loss": 0.0604,
+      "step": 11602
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.21685231105899788,
+      "learning_rate": 1.307121860948668e-08,
+      "loss": 0.0479,
+      "step": 11603
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.20056050413425275,
+      "learning_rate": 1.2972796637299444e-08,
+      "loss": 0.0596,
+      "step": 11604
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5038161893165299,
+      "learning_rate": 1.2874746127949921e-08,
+      "loss": 0.0762,
+      "step": 11605
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2628471631309903,
+      "learning_rate": 1.2777067088741158e-08,
+      "loss": 0.0674,
+      "step": 11606
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5642598774161195,
+      "learning_rate": 1.2679759526949554e-08,
+      "loss": 0.1118,
+      "step": 11607
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.3937694414584248,
+      "learning_rate": 1.2582823449822646e-08,
+      "loss": 0.1063,
+      "step": 11608
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2729646177002026,
+      "learning_rate": 1.248625886458077e-08,
+      "loss": 0.0713,
+      "step": 11609
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.29737997021788937,
+      "learning_rate": 1.239006577841706e-08,
+      "loss": 0.0784,
+      "step": 11610
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.4031927010267624,
+      "learning_rate": 1.2294244198495786e-08,
+      "loss": 0.1046,
+      "step": 11611
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.3003093190190323,
+      "learning_rate": 1.2198794131955128e-08,
+      "loss": 0.0638,
+      "step": 11612
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.35325147498237114,
+      "learning_rate": 1.2103715585904396e-08,
+      "loss": 0.0951,
+      "step": 11613
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2567778764752918,
+      "learning_rate": 1.200900856742515e-08,
+      "loss": 0.0967,
+      "step": 11614
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.28619974689598177,
+      "learning_rate": 1.1914673083572303e-08,
+      "loss": 0.0656,
+      "step": 11615
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5024672333439948,
+      "learning_rate": 1.1820709141372454e-08,
+      "loss": 0.0991,
+      "step": 11616
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.30415771250022955,
+      "learning_rate": 1.1727116747825008e-08,
+      "loss": 0.0819,
+      "step": 11617
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.36811349359618833,
+      "learning_rate": 1.1633895909899939e-08,
+      "loss": 0.0876,
+      "step": 11618
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.34453175114356055,
+      "learning_rate": 1.1541046634541696e-08,
+      "loss": 0.0877,
+      "step": 11619
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2682487894434015,
+      "learning_rate": 1.144856892866697e-08,
+      "loss": 0.0735,
+      "step": 11620
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.15773572265678204,
+      "learning_rate": 1.1356462799162471e-08,
+      "loss": 0.0153,
+      "step": 11621
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6186343269358531,
+      "learning_rate": 1.1264728252889933e-08,
+      "loss": 0.1171,
+      "step": 11622
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.46354534830920635,
+      "learning_rate": 1.1173365296682226e-08,
+      "loss": 0.1268,
+      "step": 11623
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2404725928518511,
+      "learning_rate": 1.1082373937343904e-08,
+      "loss": 0.0875,
+      "step": 11624
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2748765723076189,
+      "learning_rate": 1.0991754181653436e-08,
+      "loss": 0.0581,
+      "step": 11625
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.39829795422306863,
+      "learning_rate": 1.0901506036359866e-08,
+      "loss": 0.0793,
+      "step": 11626
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.34222473327070596,
+      "learning_rate": 1.081162950818615e-08,
+      "loss": 0.0676,
+      "step": 11627
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.31160889584153484,
+      "learning_rate": 1.0722124603825823e-08,
+      "loss": 0.0733,
+      "step": 11628
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2841515650662639,
+      "learning_rate": 1.0632991329946885e-08,
+      "loss": 0.0994,
+      "step": 11629
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.23118007751351166,
+      "learning_rate": 1.0544229693187913e-08,
+      "loss": 0.0631,
+      "step": 11630
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.31827814344621286,
+      "learning_rate": 1.0455839700160286e-08,
+      "loss": 0.0852,
+      "step": 11631
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.36395846163622325,
+      "learning_rate": 1.0367821357447627e-08,
+      "loss": 0.0587,
+      "step": 11632
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.3400252682964885,
+      "learning_rate": 1.0280174671606912e-08,
+      "loss": 0.0561,
+      "step": 11633
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.24890837008972586,
+      "learning_rate": 1.01928996491657e-08,
+      "loss": 0.0562,
+      "step": 11634
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2876885800084201,
+      "learning_rate": 1.0105996296625454e-08,
+      "loss": 0.0786,
+      "step": 11635
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6265329043142973,
+      "learning_rate": 1.0019464620458775e-08,
+      "loss": 0.0925,
+      "step": 11636
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.27305983999333133,
+      "learning_rate": 9.933304627111063e-09,
+      "loss": 0.0636,
+      "step": 11637
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.251064137937191,
+      "learning_rate": 9.847516322999962e-09,
+      "loss": 0.0414,
+      "step": 11638
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.16489430912878564,
+      "learning_rate": 9.762099714515915e-09,
+      "loss": 0.0423,
+      "step": 11639
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.18434499767430468,
+      "learning_rate": 9.677054808021058e-09,
+      "loss": 0.0457,
+      "step": 11640
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.18950452226970704,
+      "learning_rate": 9.592381609849766e-09,
+      "loss": 0.0338,
+      "step": 11641
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.34940353908481575,
+      "learning_rate": 9.50808012630866e-09,
+      "loss": 0.076,
+      "step": 11642
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.3468091242355309,
+      "learning_rate": 9.424150363678275e-09,
+      "loss": 0.0459,
+      "step": 11643
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.1991970139491377,
+      "learning_rate": 9.340592328209163e-09,
+      "loss": 0.0549,
+      "step": 11644
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2209420340354997,
+      "learning_rate": 9.257406026125238e-09,
+      "loss": 0.0609,
+      "step": 11645
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5451416592577246,
+      "learning_rate": 9.174591463623206e-09,
+      "loss": 0.1665,
+      "step": 11646
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2784633723637343,
+      "learning_rate": 9.092148646871469e-09,
+      "loss": 0.0745,
+      "step": 11647
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.24837036139916127,
+      "learning_rate": 9.010077582010667e-09,
+      "loss": 0.0634,
+      "step": 11648
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5513257189644777,
+      "learning_rate": 8.928378275153693e-09,
+      "loss": 0.1166,
+      "step": 11649
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.35458166894293935,
+      "learning_rate": 8.847050732386231e-09,
+      "loss": 0.0941,
+      "step": 11650
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2470265884225025,
+      "learning_rate": 8.766094959766214e-09,
+      "loss": 0.0478,
+      "step": 11651
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.318524790578844,
+      "learning_rate": 8.685510963323262e-09,
+      "loss": 0.0732,
+      "step": 11652
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5335469697968668,
+      "learning_rate": 8.605298749060353e-09,
+      "loss": 0.0997,
+      "step": 11653
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.292504353987712,
+      "learning_rate": 8.525458322951596e-09,
+      "loss": 0.0774,
+      "step": 11654
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2267313409990472,
+      "learning_rate": 8.445989690944457e-09,
+      "loss": 0.0782,
+      "step": 11655
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.3102383269599712,
+      "learning_rate": 8.366892858957532e-09,
+      "loss": 0.0371,
+      "step": 11656
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.3813447834389275,
+      "learning_rate": 8.288167832882777e-09,
+      "loss": 0.0839,
+      "step": 11657
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2293364294351581,
+      "learning_rate": 8.209814618584944e-09,
+      "loss": 0.0546,
+      "step": 11658
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.3212925121479801,
+      "learning_rate": 8.13183322189881e-09,
+      "loss": 0.1134,
+      "step": 11659
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.30201757307799637,
+      "learning_rate": 8.054223648633618e-09,
+      "loss": 0.0671,
+      "step": 11660
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5008833382816787,
+      "learning_rate": 7.976985904569746e-09,
+      "loss": 0.1078,
+      "step": 11661
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.35644120764923065,
+      "learning_rate": 7.900119995460919e-09,
+      "loss": 0.0479,
+      "step": 11662
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.21798623519603036,
+      "learning_rate": 7.823625927032563e-09,
+      "loss": 0.0435,
+      "step": 11663
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.3693477655023786,
+      "learning_rate": 7.74750370498234e-09,
+      "loss": 0.0303,
+      "step": 11664
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5515599762152675,
+      "learning_rate": 7.671753334979604e-09,
+      "loss": 0.1245,
+      "step": 11665
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.30871027878409035,
+      "learning_rate": 7.596374822667063e-09,
+      "loss": 0.0724,
+      "step": 11666
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.23277483085736977,
+      "learning_rate": 7.52136817365967e-09,
+      "loss": 0.039,
+      "step": 11667
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2306364768487206,
+      "learning_rate": 7.44673339354407e-09,
+      "loss": 0.0667,
+      "step": 11668
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.4225360651366093,
+      "learning_rate": 7.372470487879146e-09,
+      "loss": 0.0941,
+      "step": 11669
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.308623947888744,
+      "learning_rate": 7.298579462197142e-09,
+      "loss": 0.0985,
+      "step": 11670
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2777964899887318,
+      "learning_rate": 7.225060322001986e-09,
+      "loss": 0.0667,
+      "step": 11671
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.38665690749011766,
+      "learning_rate": 7.151913072768746e-09,
+      "loss": 0.1126,
+      "step": 11672
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.22935740037088323,
+      "learning_rate": 7.079137719946949e-09,
+      "loss": 0.0576,
+      "step": 11673
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.3293188296095383,
+      "learning_rate": 7.006734268956706e-09,
+      "loss": 0.0566,
+      "step": 11674
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.3715346070788178,
+      "learning_rate": 6.934702725190923e-09,
+      "loss": 0.0859,
+      "step": 11675
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.24745405478479526,
+      "learning_rate": 6.863043094015864e-09,
+      "loss": 0.0759,
+      "step": 11676
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.31282545410458285,
+      "learning_rate": 6.791755380768372e-09,
+      "loss": 0.0794,
+      "step": 11677
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.3696731280795269,
+      "learning_rate": 6.7208395907580884e-09,
+      "loss": 0.1033,
+      "step": 11678
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.34498518477033374,
+      "learning_rate": 6.650295729268008e-09,
+      "loss": 0.0848,
+      "step": 11679
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.3424817770552819,
+      "learning_rate": 6.580123801552263e-09,
+      "loss": 0.087,
+      "step": 11680
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2998104658487225,
+      "learning_rate": 6.510323812837782e-09,
+      "loss": 0.0931,
+      "step": 11681
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.34499255623758235,
+      "learning_rate": 6.440895768323741e-09,
+      "loss": 0.1026,
+      "step": 11682
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.1898615298303383,
+      "learning_rate": 6.371839673181001e-09,
+      "loss": 0.0419,
+      "step": 11683
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.20100455650589155,
+      "learning_rate": 6.303155532553784e-09,
+      "loss": 0.0592,
+      "step": 11684
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.21830741780046703,
+      "learning_rate": 6.234843351557995e-09,
+      "loss": 0.0632,
+      "step": 11685
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.34962770773632357,
+      "learning_rate": 6.166903135282343e-09,
+      "loss": 0.0989,
+      "step": 11686
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.23244997523901723,
+      "learning_rate": 6.0993348887866674e-09,
+      "loss": 0.0328,
+      "step": 11687
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.22874942758022632,
+      "learning_rate": 6.032138617104166e-09,
+      "loss": 0.0426,
+      "step": 11688
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.38510228914862726,
+      "learning_rate": 5.965314325239724e-09,
+      "loss": 0.0899,
+      "step": 11689
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.4449176584090638,
+      "learning_rate": 5.898862018171026e-09,
+      "loss": 0.0872,
+      "step": 11690
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.32093066418897115,
+      "learning_rate": 5.832781700848556e-09,
+      "loss": 0.074,
+      "step": 11691
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3708209093807455,
+      "learning_rate": 5.767073378193377e-09,
+      "loss": 0.0618,
+      "step": 11692
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.1847916385801939,
+      "learning_rate": 5.701737055099909e-09,
+      "loss": 0.0422,
+      "step": 11693
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5758226941015413,
+      "learning_rate": 5.636772736434814e-09,
+      "loss": 0.133,
+      "step": 11694
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.22094610434679887,
+      "learning_rate": 5.5721804270375545e-09,
+      "loss": 0.0645,
+      "step": 11695
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.31302511008902073,
+      "learning_rate": 5.507960131719281e-09,
+      "loss": 0.0707,
+      "step": 11696
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.34221648898517515,
+      "learning_rate": 5.444111855262835e-09,
+      "loss": 0.1116,
+      "step": 11697
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.29108362659481507,
+      "learning_rate": 5.380635602424411e-09,
+      "loss": 0.0928,
+      "step": 11698
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.32518192527661716,
+      "learning_rate": 5.3175313779318945e-09,
+      "loss": 0.0614,
+      "step": 11699
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2087886676579728,
+      "learning_rate": 5.25479918648597e-09,
+      "loss": 0.0568,
+      "step": 11700
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.23743955714051537,
+      "learning_rate": 5.192439032759011e-09,
+      "loss": 0.0628,
+      "step": 11701
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.21791651637671183,
+      "learning_rate": 5.1304509213967455e-09,
+      "loss": 0.0509,
+      "step": 11702
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.30839713114936845,
+      "learning_rate": 5.068834857014926e-09,
+      "loss": 0.0658,
+      "step": 11703
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.38117575824190525,
+      "learning_rate": 5.007590844204324e-09,
+      "loss": 0.0741,
+      "step": 11704
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2693265829443724,
+      "learning_rate": 4.946718887526292e-09,
+      "loss": 0.0798,
+      "step": 11705
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.25886199458226505,
+      "learning_rate": 4.886218991514979e-09,
+      "loss": 0.1014,
+      "step": 11706
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.20190110017395557,
+      "learning_rate": 4.826091160676782e-09,
+      "loss": 0.0454,
+      "step": 11707
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2908591937595396,
+      "learning_rate": 4.766335399490341e-09,
+      "loss": 0.0552,
+      "step": 11708
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2826969543259274,
+      "learning_rate": 4.706951712406538e-09,
+      "loss": 0.0992,
+      "step": 11709
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2424049771811536,
+      "learning_rate": 4.647940103848503e-09,
+      "loss": 0.0948,
+      "step": 11710
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.32801346753312116,
+      "learning_rate": 4.589300578212163e-09,
+      "loss": 0.1164,
+      "step": 11711
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.19767721035560054,
+      "learning_rate": 4.5310331398651374e-09,
+      "loss": 0.0562,
+      "step": 11712
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3023838447955846,
+      "learning_rate": 4.473137793147842e-09,
+      "loss": 0.0786,
+      "step": 11713
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.1813769833934584,
+      "learning_rate": 4.4156145423718264e-09,
+      "loss": 0.0527,
+      "step": 11714
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3001139513292004,
+      "learning_rate": 4.358463391822554e-09,
+      "loss": 0.105,
+      "step": 11715
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.23605412898680925,
+      "learning_rate": 4.301684345756619e-09,
+      "loss": 0.0521,
+      "step": 11716
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.26957254624627247,
+      "learning_rate": 4.245277408403414e-09,
+      "loss": 0.062,
+      "step": 11717
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2960311829191653,
+      "learning_rate": 4.189242583964581e-09,
+      "loss": 0.0905,
+      "step": 11718
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3222039325305927,
+      "learning_rate": 4.133579876613447e-09,
+      "loss": 0.0981,
+      "step": 11719
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.274492102600491,
+      "learning_rate": 4.07828929049614e-09,
+      "loss": 0.0784,
+      "step": 11720
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3507789854021571,
+      "learning_rate": 4.0233708297315874e-09,
+      "loss": 0.1123,
+      "step": 11721
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3582079842573663,
+      "learning_rate": 3.968824498409851e-09,
+      "loss": 0.0684,
+      "step": 11722
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.38003107181484863,
+      "learning_rate": 3.914650300594347e-09,
+      "loss": 0.089,
+      "step": 11723
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3635450475906826,
+      "learning_rate": 3.8608482403196255e-09,
+      "loss": 0.1099,
+      "step": 11724
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3152907723099724,
+      "learning_rate": 3.807418321594147e-09,
+      "loss": 0.1083,
+      "step": 11725
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.524929835984492,
+      "learning_rate": 3.754360548396951e-09,
+      "loss": 0.1209,
+      "step": 11726
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.36260212920868823,
+      "learning_rate": 3.7016749246798766e-09,
+      "loss": 0.0767,
+      "step": 11727
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.24489584199606804,
+      "learning_rate": 3.6493614543681166e-09,
+      "loss": 0.0578,
+      "step": 11728
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.31041872957734284,
+      "learning_rate": 3.597420141357444e-09,
+      "loss": 0.0589,
+      "step": 11729
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3437706940663031,
+      "learning_rate": 3.545850989517541e-09,
+      "loss": 0.0859,
+      "step": 11730
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.18754634991027802,
+      "learning_rate": 3.494654002688669e-09,
+      "loss": 0.0524,
+      "step": 11731
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2837435599278237,
+      "learning_rate": 3.4438291846849994e-09,
+      "loss": 0.0509,
+      "step": 11732
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3745836738960185,
+      "learning_rate": 3.393376539291837e-09,
+      "loss": 0.104,
+      "step": 11733
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.35939347803926025,
+      "learning_rate": 3.343296070267843e-09,
+      "loss": 0.112,
+      "step": 11734
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3880796222436005,
+      "learning_rate": 3.2935877813422557e-09,
+      "loss": 0.0778,
+      "step": 11735
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.25449122423019377,
+      "learning_rate": 3.244251676218779e-09,
+      "loss": 0.0825,
+      "step": 11736
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.1876866328454259,
+      "learning_rate": 3.195287758571142e-09,
+      "loss": 0.0514,
+      "step": 11737
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.4659302365464322,
+      "learning_rate": 3.146696032047536e-09,
+      "loss": 0.125,
+      "step": 11738
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.21965715903888344,
+      "learning_rate": 3.098476500266734e-09,
+      "loss": 0.0747,
+      "step": 11739
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.4616749118902757,
+      "learning_rate": 3.050629166820307e-09,
+      "loss": 0.0897,
+      "step": 11740
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2924198614917721,
+      "learning_rate": 3.003154035272071e-09,
+      "loss": 0.0832,
+      "step": 11741
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3195176685653182,
+      "learning_rate": 2.956051109159197e-09,
+      "loss": 0.0768,
+      "step": 11742
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2905713077334311,
+      "learning_rate": 2.9093203919894342e-09,
+      "loss": 0.0753,
+      "step": 11743
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.22611534215757356,
+      "learning_rate": 2.862961887243332e-09,
+      "loss": 0.0444,
+      "step": 11744
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.29157363493723704,
+      "learning_rate": 2.8169755983747938e-09,
+      "loss": 0.0723,
+      "step": 11745
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2623712901175991,
+      "learning_rate": 2.771361528808303e-09,
+      "loss": 0.0667,
+      "step": 11746
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2898198342386482,
+      "learning_rate": 2.726119681942252e-09,
+      "loss": 0.0897,
+      "step": 11747
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3878889869242105,
+      "learning_rate": 2.6812500611456128e-09,
+      "loss": 0.1351,
+      "step": 11748
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.4742886126537848,
+      "learning_rate": 2.6367526697612667e-09,
+      "loss": 0.0726,
+      "step": 11749
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.31844921512361263,
+      "learning_rate": 2.5926275111032296e-09,
+      "loss": 0.0744,
+      "step": 11750
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.29980898060181776,
+      "learning_rate": 2.5488745884583166e-09,
+      "loss": 0.0838,
+      "step": 11751
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.44363154521735015,
+      "learning_rate": 2.5054939050855877e-09,
+      "loss": 0.0739,
+      "step": 11752
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.30603154280587114,
+      "learning_rate": 2.4624854642163467e-09,
+      "loss": 0.0709,
+      "step": 11753
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6724538549797336,
+      "learning_rate": 2.419849269053587e-09,
+      "loss": 0.1416,
+      "step": 11754
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3049014330178256,
+      "learning_rate": 2.3775853227736566e-09,
+      "loss": 0.0457,
+      "step": 11755
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5560289687569162,
+      "learning_rate": 2.3356936285251486e-09,
+      "loss": 0.133,
+      "step": 11756
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.304402795354917,
+      "learning_rate": 2.294174189427234e-09,
+      "loss": 0.0966,
+      "step": 11757
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3847321658310377,
+      "learning_rate": 2.2530270085724394e-09,
+      "loss": 0.0769,
+      "step": 11758
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.34332344273527193,
+      "learning_rate": 2.2122520890272005e-09,
+      "loss": 0.0968,
+      "step": 11759
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.1429218124893918,
+      "learning_rate": 2.171849433826867e-09,
+      "loss": 0.0359,
+      "step": 11760
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.29721986374660564,
+      "learning_rate": 2.131819045982364e-09,
+      "loss": 0.0736,
+      "step": 11761
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.26007282749663907,
+      "learning_rate": 2.0921609284746402e-09,
+      "loss": 0.0563,
+      "step": 11762
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8124114389026214,
+      "learning_rate": 2.052875084257444e-09,
+      "loss": 0.0538,
+      "step": 11763
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.22827876270234074,
+      "learning_rate": 2.0139615162573234e-09,
+      "loss": 0.0584,
+      "step": 11764
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.34518652103434205,
+      "learning_rate": 1.9754202273730706e-09,
+      "loss": 0.0854,
+      "step": 11765
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3758834220443019,
+      "learning_rate": 1.937251220474612e-09,
+      "loss": 0.056,
+      "step": 11766
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.34780667092615575,
+      "learning_rate": 1.8994544984057838e-09,
+      "loss": 0.0922,
+      "step": 11767
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.35528637347309605,
+      "learning_rate": 1.862030063982112e-09,
+      "loss": 0.0942,
+      "step": 11768
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3799984134402679,
+      "learning_rate": 1.824977919990256e-09,
+      "loss": 0.082,
+      "step": 11769
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5781437266226297,
+      "learning_rate": 1.788298069190786e-09,
+      "loss": 0.1434,
+      "step": 11770
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3113911224875263,
+      "learning_rate": 1.7519905143154048e-09,
+      "loss": 0.0884,
+      "step": 11771
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.37165810657128423,
+      "learning_rate": 1.7160552580686163e-09,
+      "loss": 0.0581,
+      "step": 11772
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3071167657486759,
+      "learning_rate": 1.680492303127723e-09,
+      "loss": 0.0736,
+      "step": 11773
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.4172984737921329,
+      "learning_rate": 1.6453016521406074e-09,
+      "loss": 0.0389,
+      "step": 11774
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.32295976498516704,
+      "learning_rate": 1.6104833077290605e-09,
+      "loss": 0.0685,
+      "step": 11775
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.42797967652606334,
+      "learning_rate": 1.5760372724871186e-09,
+      "loss": 0.1185,
+      "step": 11776
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.4633207679690751,
+      "learning_rate": 1.541963548979397e-09,
+      "loss": 0.0915,
+      "step": 11777
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.4072940743861392,
+      "learning_rate": 1.5082621397444208e-09,
+      "loss": 0.1355,
+      "step": 11778
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.23769197711983342,
+      "learning_rate": 1.474933047292404e-09,
+      "loss": 0.0584,
+      "step": 11779
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2066007336165977,
+      "learning_rate": 1.4419762741058052e-09,
+      "loss": 0.0583,
+      "step": 11780
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.23935670955594482,
+      "learning_rate": 1.4093918226398829e-09,
+      "loss": 0.0618,
+      "step": 11781
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.24753221993947386,
+      "learning_rate": 1.3771796953210292e-09,
+      "loss": 0.0673,
+      "step": 11782
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5860998978953742,
+      "learning_rate": 1.3453398945495467e-09,
+      "loss": 0.1269,
+      "step": 11783
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.603867184496569,
+      "learning_rate": 1.313872422695761e-09,
+      "loss": 0.0883,
+      "step": 11784
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.32797923400167506,
+      "learning_rate": 1.2827772821044638e-09,
+      "loss": 0.0982,
+      "step": 11785
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8177253745068733,
+      "learning_rate": 1.252054475092135e-09,
+      "loss": 0.1328,
+      "step": 11786
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3881613445963822,
+      "learning_rate": 1.221704003945834e-09,
+      "loss": 0.0895,
+      "step": 11787
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.4614403955053557,
+      "learning_rate": 1.1917258709276402e-09,
+      "loss": 0.1132,
+      "step": 11788
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2756829686052264,
+      "learning_rate": 1.1621200782696573e-09,
+      "loss": 0.0956,
+      "step": 11789
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5739813669314355,
+      "learning_rate": 1.1328866281773432e-09,
+      "loss": 0.1109,
+      "step": 11790
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2540557745636962,
+      "learning_rate": 1.1040255228284002e-09,
+      "loss": 0.0532,
+      "step": 11791
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.373205882408839,
+      "learning_rate": 1.0755367643722203e-09,
+      "loss": 0.1051,
+      "step": 11792
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.4720354757525241,
+      "learning_rate": 1.0474203549309946e-09,
+      "loss": 0.1056,
+      "step": 11793
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3671025661759038,
+      "learning_rate": 1.0196762965991592e-09,
+      "loss": 0.0867,
+      "step": 11794
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3108969214941454,
+      "learning_rate": 9.923045914428387e-10,
+      "loss": 0.0546,
+      "step": 11795
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.42579980270451384,
+      "learning_rate": 9.653052415009578e-10,
+      "loss": 0.1094,
+      "step": 11796
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2067672153837607,
+      "learning_rate": 9.386782487846857e-10,
+      "loss": 0.0644,
+      "step": 11797
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.15849490249565845,
+      "learning_rate": 9.124236152774358e-10,
+      "loss": 0.0112,
+      "step": 11798
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3676063998633194,
+      "learning_rate": 8.865413429348658e-10,
+      "loss": 0.0774,
+      "step": 11799
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5118447288748603,
+      "learning_rate": 8.610314336843229e-10,
+      "loss": 0.055,
+      "step": 11800
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.32184649405325855,
+      "learning_rate": 8.358938894265089e-10,
+      "loss": 0.091,
+      "step": 11801
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.48191383943045984,
+      "learning_rate": 8.1112871203326e-10,
+      "loss": 0.096,
+      "step": 11802
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.4378390971315926,
+      "learning_rate": 7.867359033497668e-10,
+      "loss": 0.0959,
+      "step": 11803
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3224403064762407,
+      "learning_rate": 7.627154651929092e-10,
+      "loss": 0.0776,
+      "step": 11804
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2113921458945713,
+      "learning_rate": 7.390673993518116e-10,
+      "loss": 0.0926,
+      "step": 11805
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3689104176921736,
+      "learning_rate": 7.157917075878429e-10,
+      "loss": 0.1036,
+      "step": 11806
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3229002645865447,
+      "learning_rate": 6.928883916346163e-10,
+      "loss": 0.0708,
+      "step": 11807
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.4798771990897857,
+      "learning_rate": 6.703574531979895e-10,
+      "loss": 0.1192,
+      "step": 11808
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6060549837871293,
+      "learning_rate": 6.481988939566197e-10,
+      "loss": 0.0802,
+      "step": 11809
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.3112226618217803,
+      "learning_rate": 6.264127155602983e-10,
+      "loss": 0.0845,
+      "step": 11810
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.25791594707481535,
+      "learning_rate": 6.049989196327266e-10,
+      "loss": 0.0778,
+      "step": 11811
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.2072148950472449,
+      "learning_rate": 5.839575077681847e-10,
+      "loss": 0.0613,
+      "step": 11812
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.40416325552104143,
+      "learning_rate": 5.632884815343076e-10,
+      "loss": 0.0906,
+      "step": 11813
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.3188732296165041,
+      "learning_rate": 5.429918424709745e-10,
+      "loss": 0.0951,
+      "step": 11814
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.49423694435251964,
+      "learning_rate": 5.230675920891992e-10,
+      "loss": 0.1411,
+      "step": 11815
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.27927466824910835,
+      "learning_rate": 5.035157318733497e-10,
+      "loss": 0.0631,
+      "step": 11816
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.20511027595949646,
+      "learning_rate": 4.843362632800386e-10,
+      "loss": 0.0389,
+      "step": 11817
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.26438975230668127,
+      "learning_rate": 4.6552918773812296e-10,
+      "loss": 0.0703,
+      "step": 11818
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.40528210505853113,
+      "learning_rate": 4.47094506647594e-10,
+      "loss": 0.1062,
+      "step": 11819
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.4291638868690063,
+      "learning_rate": 4.290322213817977e-10,
+      "loss": 0.1282,
+      "step": 11820
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.35866197170922853,
+      "learning_rate": 4.1134233328687934e-10,
+      "loss": 0.07,
+      "step": 11821
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.35621801850970936,
+      "learning_rate": 3.940248436795635e-10,
+      "loss": 0.0929,
+      "step": 11822
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.2747052331328941,
+      "learning_rate": 3.7707975384992933e-10,
+      "loss": 0.0839,
+      "step": 11823
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.2665092032953628,
+      "learning_rate": 3.6050706506085556e-10,
+      "loss": 0.0534,
+      "step": 11824
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.1927948336476631,
+      "learning_rate": 3.4430677854579984e-10,
+      "loss": 0.0607,
+      "step": 11825
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.2877933525528284,
+      "learning_rate": 3.2847889551268497e-10,
+      "loss": 0.0791,
+      "step": 11826
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.2910236027602921,
+      "learning_rate": 3.1302341713890236e-10,
+      "loss": 0.0702,
+      "step": 11827
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.31989387698803146,
+      "learning_rate": 2.979403445768636e-10,
+      "loss": 0.0767,
+      "step": 11828
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.2795830480951221,
+      "learning_rate": 2.8322967894955924e-10,
+      "loss": 0.0655,
+      "step": 11829
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.2895280812639416,
+      "learning_rate": 2.6889142135277935e-10,
+      "loss": 0.0721,
+      "step": 11830
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.2765020444213322,
+      "learning_rate": 2.5492557285455855e-10,
+      "loss": 0.0433,
+      "step": 11831
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.22811937424412942,
+      "learning_rate": 2.413321344951758e-10,
+      "loss": 0.0468,
+      "step": 11832
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.36940869497529055,
+      "learning_rate": 2.2811110728715446e-10,
+      "loss": 0.1222,
+      "step": 11833
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.23832616470487086,
+      "learning_rate": 2.152624922152624e-10,
+      "loss": 0.0697,
+      "step": 11834
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.34869234414938416,
+      "learning_rate": 2.0278629023651187e-10,
+      "loss": 0.1378,
+      "step": 11835
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.23666897967073502,
+      "learning_rate": 1.906825022801595e-10,
+      "loss": 0.0478,
+      "step": 11836
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.2768783721044072,
+      "learning_rate": 1.789511292477064e-10,
+      "loss": 0.0889,
+      "step": 11837
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.47324168891344626,
+      "learning_rate": 1.6759217201289812e-10,
+      "loss": 0.1094,
+      "step": 11838
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.20960050846261782,
+      "learning_rate": 1.566056314222797e-10,
+      "loss": 0.0588,
+      "step": 11839
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.2168419461505182,
+      "learning_rate": 1.4599150829408548e-10,
+      "loss": 0.0493,
+      "step": 11840
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.3725211561619002,
+      "learning_rate": 1.357498034187943e-10,
+      "loss": 0.114,
+      "step": 11841
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.45960117822304525,
+      "learning_rate": 1.2588051755912934e-10,
+      "loss": 0.1351,
+      "step": 11842
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.39346999596671195,
+      "learning_rate": 1.1638365145005825e-10,
+      "loss": 0.0908,
+      "step": 11843
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.23300106828992764,
+      "learning_rate": 1.0725920579934823e-10,
+      "loss": 0.0574,
+      "step": 11844
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.26171702441622763,
+      "learning_rate": 9.850718128645576e-11,
+      "loss": 0.0534,
+      "step": 11845
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.2676382409966388,
+      "learning_rate": 9.012757856308173e-11,
+      "loss": 0.0615,
+      "step": 11846
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.21658409419396413,
+      "learning_rate": 8.212039825428175e-11,
+      "loss": 0.0564,
+      "step": 11847
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.25757406244161735,
+      "learning_rate": 7.448564095513533e-11,
+      "loss": 0.0522,
+      "step": 11848
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.3447296366322997,
+      "learning_rate": 6.7223307235742e-11,
+      "loss": 0.0615,
+      "step": 11849
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.25626614718999186,
+      "learning_rate": 6.033339763567014e-11,
+      "loss": 0.077,
+      "step": 11850
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5193133874999188,
+      "learning_rate": 5.381591266895303e-11,
+      "loss": 0.102,
+      "step": 11851
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.29202820410701,
+      "learning_rate": 4.767085282075812e-11,
+      "loss": 0.0912,
+      "step": 11852
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.3404326817516068,
+      "learning_rate": 4.1898218549607515e-11,
+      "loss": 0.0488,
+      "step": 11853
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.262533264660476,
+      "learning_rate": 3.649801028404731e-11,
+      "loss": 0.0782,
+      "step": 11854
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5546942607194442,
+      "learning_rate": 3.147022842708847e-11,
+      "loss": 0.1401,
+      "step": 11855
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.3351340884470931,
+      "learning_rate": 2.6814873353431248e-11,
+      "loss": 0.0975,
+      "step": 11856
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.3395039146296503,
+      "learning_rate": 2.2531945409465238e-11,
+      "loss": 0.0513,
+      "step": 11857
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.45189046368524854,
+      "learning_rate": 1.862144491437956e-11,
+      "loss": 0.0922,
+      "step": 11858
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.15549157182987108,
+      "learning_rate": 1.5083372159607756e-11,
+      "loss": 0.0314,
+      "step": 11859
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5875523599919215,
+      "learning_rate": 1.1917727408272684e-11,
+      "loss": 0.12,
+      "step": 11860
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.2973782432299746,
+      "learning_rate": 9.124510896851845e-12,
+      "loss": 0.0644,
+      "step": 11861
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.3297972389410682,
+      "learning_rate": 6.703722832401838e-12,
+      "loss": 0.1083,
+      "step": 11862
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.8574829781047126,
+      "learning_rate": 4.655363395889012e-12,
+      "loss": 0.1404,
+      "step": 11863
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.4144200820695359,
+      "learning_rate": 2.9794327399690347e-12,
+      "loss": 0.0433,
+      "step": 11864
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6252006663130905,
+      "learning_rate": 1.6759309889868847e-12,
+      "loss": 0.0852,
+      "step": 11865
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5665002798454156,
+      "learning_rate": 7.448582406421878e-13,
+      "loss": 0.0929,
+      "step": 11866
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.24785827729567578,
+      "learning_rate": 1.8621456376877177e-13,
+      "loss": 0.0672,
+      "step": 11867
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.35324169525443694,
+      "learning_rate": 0.0,
+      "loss": 0.0631,
+      "step": 11868
+    },
+    {
+      "epoch": 1.0,
+      "step": 11868,
+      "total_flos": 452060205940736.0,
+      "train_loss": 0.08946360924744487,
+      "train_runtime": 32513.5487,
+      "train_samples_per_second": 0.73,
+      "train_steps_per_second": 0.365
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 11868,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 50000,
+  "total_flos": 452060205940736.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}
diff --git a/llavaqwen2-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split06_all_mm_tune_olora256_512_llm/README.md b/llavaqwen2-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split06_all_mm_tune_olora256_512_llm/README.md
new file mode 100644
index 0000000000000000000000000000000000000000..d4576fe074287232d3836bf69c21d3f2593290d9
--- /dev/null
+++ b/llavaqwen2-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split06_all_mm_tune_olora256_512_llm/README.md
@@ -0,0 +1,9 @@
+---
+library_name: peft
+---
+## Training procedure
+
+### Framework versions
+
+
+- PEFT 0.4.0
diff --git a/llavaqwen2-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split06_all_mm_tune_olora256_512_llm/adapter_config.json b/llavaqwen2-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split06_all_mm_tune_olora256_512_llm/adapter_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..522aa87957d3a02332bdcf639384a9346e2c385e
--- /dev/null
+++ b/llavaqwen2-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split06_all_mm_tune_olora256_512_llm/adapter_config.json
@@ -0,0 +1,26 @@
+{
+  "auto_mapping": null,
+  "base_model_name_or_path": "lmms-lab/LLaVA-Video-7B-Qwen2",
+  "bias": "none",
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": "olora",
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "lora_alpha": 512,
+  "lora_dropout": 0.05,
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "r": 256,
+  "revision": null,
+  "target_modules": [
+    "up_proj",
+    "v_proj",
+    "down_proj",
+    "k_proj",
+    "o_proj",
+    "q_proj",
+    "gate_proj"
+  ],
+  "task_type": "CAUSAL_LM"
+}
\ No newline at end of file
diff --git a/llavaqwen2-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split06_all_mm_tune_olora256_512_llm/adapter_model.bin b/llavaqwen2-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split06_all_mm_tune_olora256_512_llm/adapter_model.bin
new file mode 100644
index 0000000000000000000000000000000000000000..44715c42e785279c52bced63ae9d7c31e64040b4
--- /dev/null
+++ b/llavaqwen2-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split06_all_mm_tune_olora256_512_llm/adapter_model.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:acbd28cec4b78cb2d066189c430142cb98fff08ae2e35307dba74e28fe87f02d
+size 1384057050
diff --git a/llavaqwen2-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split06_all_mm_tune_olora256_512_llm/config.json b/llavaqwen2-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split06_all_mm_tune_olora256_512_llm/config.json
new file mode 100644
index 0000000000000000000000000000000000000000..1a67f02d0063c3de7740207b9ab2a3eb7be1cbe3
--- /dev/null
+++ b/llavaqwen2-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split06_all_mm_tune_olora256_512_llm/config.json
@@ -0,0 +1,221 @@
+{
+  "_name_or_path": "lmms-lab/LLaVA-Video-7B-Qwen2",
+  "add_faster_video": false,
+  "add_time_instruction": true,
+  "architectures": [
+    "LlavaQwenForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 151643,
+  "eos_token_id": 151645,
+  "faster_token_stride": 10,
+  "force_sample": true,
+  "hidden_act": "silu",
+  "hidden_size": 3584,
+  "ignore_index": -100,
+  "image_aspect_ratio": "anyres_max_9",
+  "image_crop_resolution": null,
+  "image_grid_pinpoints": [
+    [
+      384,
+      384
+    ],
+    [
+      384,
+      768
+    ],
+    [
+      384,
+      1152
+    ],
+    [
+      384,
+      1536
+    ],
+    [
+      384,
+      1920
+    ],
+    [
+      384,
+      2304
+    ],
+    [
+      768,
+      384
+    ],
+    [
+      768,
+      768
+    ],
+    [
+      768,
+      1152
+    ],
+    [
+      768,
+      1536
+    ],
+    [
+      768,
+      1920
+    ],
+    [
+      768,
+      2304
+    ],
+    [
+      1152,
+      384
+    ],
+    [
+      1152,
+      768
+    ],
+    [
+      1152,
+      1152
+    ],
+    [
+      1152,
+      1536
+    ],
+    [
+      1152,
+      1920
+    ],
+    [
+      1152,
+      2304
+    ],
+    [
+      1536,
+      384
+    ],
+    [
+      1536,
+      768
+    ],
+    [
+      1536,
+      1152
+    ],
+    [
+      1536,
+      1536
+    ],
+    [
+      1536,
+      1920
+    ],
+    [
+      1536,
+      2304
+    ],
+    [
+      1920,
+      384
+    ],
+    [
+      1920,
+      768
+    ],
+    [
+      1920,
+      1152
+    ],
+    [
+      1920,
+      1536
+    ],
+    [
+      1920,
+      1920
+    ],
+    [
+      1920,
+      2304
+    ],
+    [
+      2304,
+      384
+    ],
+    [
+      2304,
+      768
+    ],
+    [
+      2304,
+      1152
+    ],
+    [
+      2304,
+      1536
+    ],
+    [
+      2304,
+      1920
+    ],
+    [
+      2304,
+      2304
+    ]
+  ],
+  "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": "grid",
+  "mm_patch_merge_type": "spatial_unpad",
+  "mm_projector_lr": 2e-05,
+  "mm_projector_type": "mlp2x_gelu",
+  "mm_resampler_type": null,
+  "mm_spatial_pool_mode": "bilinear",
+  "mm_spatial_pool_stride": 2,
+  "mm_tunable_parts": "mm_vision_tower,mm_mlp_adapter,mm_language_model",
+  "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",
+  "num_attention_heads": 28,
+  "num_hidden_layers": 28,
+  "num_key_value_heads": 4,
+  "pos_skipping_range": 4096,
+  "projector_hidden_act": "gelu",
+  "rms_norm_eps": 1e-06,
+  "rope_scaling": null,
+  "rope_theta": 1000000.0,
+  "sliding_window": 131072,
+  "text_config": {
+    "model_type": "llama"
+  },
+  "tie_word_embeddings": false,
+  "tokenizer_model_max_length": 32768,
+  "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_config": {
+    "hidden_size": 1024,
+    "image_size": 336,
+    "intermediate_size": 4096,
+    "model_type": "clip_vision_model",
+    "num_attention_heads": 16,
+    "num_hidden_layers": 24,
+    "patch_size": 14,
+    "projection_dim": 768,
+    "vocab_size": 32000
+  },
+  "vision_feature_layer": -2,
+  "vision_feature_select_strategy": "default",
+  "vision_tower_pretrained": null
+}
diff --git a/llavaqwen2-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split06_all_mm_tune_olora256_512_llm/generation_config.json b/llavaqwen2-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split06_all_mm_tune_olora256_512_llm/generation_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..19a297221acb87418d4388a3decef2282c6d7316
--- /dev/null
+++ b/llavaqwen2-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split06_all_mm_tune_olora256_512_llm/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/llavaqwen2-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split06_all_mm_tune_olora256_512_llm/non_lora_trainables.bin b/llavaqwen2-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split06_all_mm_tune_olora256_512_llm/non_lora_trainables.bin
new file mode 100644
index 0000000000000000000000000000000000000000..76e56a68946b28450ca5e8d8b7eb2d0005823ce8
--- /dev/null
+++ b/llavaqwen2-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split06_all_mm_tune_olora256_512_llm/non_lora_trainables.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:55e845b34e8e2d9fb8f5b43868586b3bbc65e76f0957e62ba1f6478d5254b4e1
+size 33964208
diff --git a/llavaqwen2-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split06_all_mm_tune_olora256_512_llm/trainer_state.json b/llavaqwen2-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split06_all_mm_tune_olora256_512_llm/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..b545131aa0b283015d4d31c8e9aedc97d05183c9
--- /dev/null
+++ b/llavaqwen2-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split06_all_mm_tune_olora256_512_llm/trainer_state.json
@@ -0,0 +1,24530 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 3500,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.6559280289916924,
+      "learning_rate": 9.523809523809525e-08,
+      "loss": 0.7224,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.0096688648254406,
+      "learning_rate": 1.904761904761905e-07,
+      "loss": 0.4464,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.9032770462485882,
+      "learning_rate": 2.8571428571428575e-07,
+      "loss": 0.5203,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.9667735520114558,
+      "learning_rate": 3.80952380952381e-07,
+      "loss": 0.511,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.9814548453165757,
+      "learning_rate": 4.7619047619047623e-07,
+      "loss": 0.4734,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.1984625427959634,
+      "learning_rate": 5.714285714285715e-07,
+      "loss": 0.4914,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.0612080516412417,
+      "learning_rate": 6.666666666666667e-07,
+      "loss": 0.4076,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.792849022681951,
+      "learning_rate": 7.61904761904762e-07,
+      "loss": 0.5003,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.175499560707213,
+      "learning_rate": 8.571428571428572e-07,
+      "loss": 0.5635,
+      "step": 9
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.0311328561733566,
+      "learning_rate": 9.523809523809525e-07,
+      "loss": 0.7983,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.8360302365975791,
+      "learning_rate": 1.0476190476190478e-06,
+      "loss": 0.5042,
+      "step": 11
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.383169128978254,
+      "learning_rate": 1.142857142857143e-06,
+      "loss": 0.3251,
+      "step": 12
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.9034153840049772,
+      "learning_rate": 1.2380952380952382e-06,
+      "loss": 0.465,
+      "step": 13
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.775896938987523,
+      "learning_rate": 1.3333333333333334e-06,
+      "loss": 1.2175,
+      "step": 14
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.30536972073011,
+      "learning_rate": 1.4285714285714286e-06,
+      "loss": 0.6381,
+      "step": 15
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.722675415522589,
+      "learning_rate": 1.523809523809524e-06,
+      "loss": 0.4503,
+      "step": 16
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.614486736580859,
+      "learning_rate": 1.6190476190476193e-06,
+      "loss": 0.4714,
+      "step": 17
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5196186528148072,
+      "learning_rate": 1.7142857142857145e-06,
+      "loss": 0.4606,
+      "step": 18
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.497045818500941,
+      "learning_rate": 1.8095238095238097e-06,
+      "loss": 0.4179,
+      "step": 19
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0436549709147536,
+      "learning_rate": 1.904761904761905e-06,
+      "loss": 0.6045,
+      "step": 20
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6123040371644173,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 0.4135,
+      "step": 21
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2013571425957115,
+      "learning_rate": 2.0952380952380955e-06,
+      "loss": 0.5183,
+      "step": 22
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6030687602677367,
+      "learning_rate": 2.1904761904761908e-06,
+      "loss": 0.4749,
+      "step": 23
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9166891584638006,
+      "learning_rate": 2.285714285714286e-06,
+      "loss": 0.5479,
+      "step": 24
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2055675952706235,
+      "learning_rate": 2.380952380952381e-06,
+      "loss": 0.3448,
+      "step": 25
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7046011655338569,
+      "learning_rate": 2.4761904761904764e-06,
+      "loss": 0.4165,
+      "step": 26
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3200060583775266,
+      "learning_rate": 2.571428571428571e-06,
+      "loss": 0.3484,
+      "step": 27
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.156488808318543,
+      "learning_rate": 2.666666666666667e-06,
+      "loss": 0.593,
+      "step": 28
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3724071169263563,
+      "learning_rate": 2.7619047619047625e-06,
+      "loss": 0.3157,
+      "step": 29
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0618373948373527,
+      "learning_rate": 2.8571428571428573e-06,
+      "loss": 0.2136,
+      "step": 30
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2814610028366067,
+      "learning_rate": 2.9523809523809525e-06,
+      "loss": 0.3498,
+      "step": 31
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5198689941228432,
+      "learning_rate": 3.047619047619048e-06,
+      "loss": 0.3199,
+      "step": 32
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0259596723635873,
+      "learning_rate": 3.142857142857143e-06,
+      "loss": 0.2961,
+      "step": 33
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.313082934999788,
+      "learning_rate": 3.2380952380952385e-06,
+      "loss": 0.2747,
+      "step": 34
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.340821897956387,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 0.2578,
+      "step": 35
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2018022664265944,
+      "learning_rate": 3.428571428571429e-06,
+      "loss": 0.2718,
+      "step": 36
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3186914769581646,
+      "learning_rate": 3.523809523809524e-06,
+      "loss": 0.2489,
+      "step": 37
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2992416650203018,
+      "learning_rate": 3.6190476190476194e-06,
+      "loss": 0.2993,
+      "step": 38
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.772125365448492,
+      "learning_rate": 3.7142857142857146e-06,
+      "loss": 0.3999,
+      "step": 39
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8770991535129182,
+      "learning_rate": 3.80952380952381e-06,
+      "loss": 0.1851,
+      "step": 40
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8745028120708158,
+      "learning_rate": 3.9047619047619055e-06,
+      "loss": 0.1878,
+      "step": 41
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.321115347273228,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 0.1735,
+      "step": 42
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0741926723658475,
+      "learning_rate": 4.095238095238096e-06,
+      "loss": 0.2032,
+      "step": 43
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.1082442933568981,
+      "learning_rate": 4.190476190476191e-06,
+      "loss": 0.2134,
+      "step": 44
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0796947842685276,
+      "learning_rate": 4.2857142857142855e-06,
+      "loss": 0.2507,
+      "step": 45
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.4517323089964154,
+      "learning_rate": 4.3809523809523815e-06,
+      "loss": 0.3937,
+      "step": 46
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5717801997670124,
+      "learning_rate": 4.476190476190477e-06,
+      "loss": 0.1819,
+      "step": 47
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1780618394564493,
+      "learning_rate": 4.571428571428572e-06,
+      "loss": 0.2757,
+      "step": 48
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9538752405202381,
+      "learning_rate": 4.666666666666667e-06,
+      "loss": 0.1732,
+      "step": 49
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5594028468303063,
+      "learning_rate": 4.761904761904762e-06,
+      "loss": 0.1659,
+      "step": 50
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9013709697820569,
+      "learning_rate": 4.857142857142858e-06,
+      "loss": 0.1457,
+      "step": 51
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.316396604855784,
+      "learning_rate": 4.952380952380953e-06,
+      "loss": 0.2838,
+      "step": 52
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0144195208183198,
+      "learning_rate": 5.047619047619048e-06,
+      "loss": 0.2073,
+      "step": 53
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7753323685166975,
+      "learning_rate": 5.142857142857142e-06,
+      "loss": 0.1442,
+      "step": 54
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8248824555139371,
+      "learning_rate": 5.2380952380952384e-06,
+      "loss": 0.2562,
+      "step": 55
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7063235284172091,
+      "learning_rate": 5.333333333333334e-06,
+      "loss": 0.1598,
+      "step": 56
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9732468260011179,
+      "learning_rate": 5.428571428571429e-06,
+      "loss": 0.1233,
+      "step": 57
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.165700586222653,
+      "learning_rate": 5.523809523809525e-06,
+      "loss": 0.2323,
+      "step": 58
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3168009929121847,
+      "learning_rate": 5.619047619047619e-06,
+      "loss": 0.2518,
+      "step": 59
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.233880179354882,
+      "learning_rate": 5.7142857142857145e-06,
+      "loss": 0.1262,
+      "step": 60
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0209711977795521,
+      "learning_rate": 5.8095238095238106e-06,
+      "loss": 0.1771,
+      "step": 61
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2692656195439345,
+      "learning_rate": 5.904761904761905e-06,
+      "loss": 0.1833,
+      "step": 62
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2779359909698755,
+      "learning_rate": 6e-06,
+      "loss": 0.1395,
+      "step": 63
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9967245397604485,
+      "learning_rate": 6.095238095238096e-06,
+      "loss": 0.1365,
+      "step": 64
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.094172667670302,
+      "learning_rate": 6.1904761904761914e-06,
+      "loss": 0.1587,
+      "step": 65
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.837919991285461,
+      "learning_rate": 6.285714285714286e-06,
+      "loss": 0.3666,
+      "step": 66
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5700070650907721,
+      "learning_rate": 6.380952380952381e-06,
+      "loss": 0.1186,
+      "step": 67
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3198936509025527,
+      "learning_rate": 6.476190476190477e-06,
+      "loss": 0.1711,
+      "step": 68
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8045994016978977,
+      "learning_rate": 6.571428571428572e-06,
+      "loss": 0.209,
+      "step": 69
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6758420381151559,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 0.1143,
+      "step": 70
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8025693699018324,
+      "learning_rate": 6.761904761904763e-06,
+      "loss": 0.1159,
+      "step": 71
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8422640734211968,
+      "learning_rate": 6.857142857142858e-06,
+      "loss": 0.085,
+      "step": 72
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6451656843101513,
+      "learning_rate": 6.952380952380952e-06,
+      "loss": 0.0848,
+      "step": 73
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5481845662341531,
+      "learning_rate": 7.047619047619048e-06,
+      "loss": 0.1838,
+      "step": 74
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6814230718344307,
+      "learning_rate": 7.1428571428571436e-06,
+      "loss": 0.0924,
+      "step": 75
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6534419780740991,
+      "learning_rate": 7.238095238095239e-06,
+      "loss": 0.0892,
+      "step": 76
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6311335037710308,
+      "learning_rate": 7.333333333333333e-06,
+      "loss": 0.138,
+      "step": 77
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7107753773840428,
+      "learning_rate": 7.428571428571429e-06,
+      "loss": 0.1209,
+      "step": 78
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.18785073387853,
+      "learning_rate": 7.523809523809524e-06,
+      "loss": 0.1696,
+      "step": 79
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8769210319622615,
+      "learning_rate": 7.61904761904762e-06,
+      "loss": 0.1147,
+      "step": 80
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.9972481202089707,
+      "learning_rate": 7.714285714285716e-06,
+      "loss": 0.2136,
+      "step": 81
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.871714731096346,
+      "learning_rate": 7.809523809523811e-06,
+      "loss": 0.1671,
+      "step": 82
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9040197324423482,
+      "learning_rate": 7.904761904761904e-06,
+      "loss": 0.1314,
+      "step": 83
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8353400083573649,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 0.1492,
+      "step": 84
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1965167723105985,
+      "learning_rate": 8.095238095238097e-06,
+      "loss": 0.1899,
+      "step": 85
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6970684913235916,
+      "learning_rate": 8.190476190476192e-06,
+      "loss": 0.1167,
+      "step": 86
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0004972692904666,
+      "learning_rate": 8.285714285714287e-06,
+      "loss": 0.1797,
+      "step": 87
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.682149243805965,
+      "learning_rate": 8.380952380952382e-06,
+      "loss": 0.14,
+      "step": 88
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6844206391813945,
+      "learning_rate": 8.476190476190477e-06,
+      "loss": 0.0611,
+      "step": 89
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9019633116200761,
+      "learning_rate": 8.571428571428571e-06,
+      "loss": 0.1573,
+      "step": 90
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0176943533714722,
+      "learning_rate": 8.666666666666668e-06,
+      "loss": 0.1504,
+      "step": 91
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7358111333939853,
+      "learning_rate": 8.761904761904763e-06,
+      "loss": 0.1435,
+      "step": 92
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5235847912021016,
+      "learning_rate": 8.857142857142858e-06,
+      "loss": 0.0699,
+      "step": 93
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0174063008813663,
+      "learning_rate": 8.952380952380953e-06,
+      "loss": 0.1368,
+      "step": 94
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7967069556065175,
+      "learning_rate": 9.047619047619049e-06,
+      "loss": 0.1081,
+      "step": 95
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3654155000995285,
+      "learning_rate": 9.142857142857144e-06,
+      "loss": 0.1395,
+      "step": 96
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6357033696427046,
+      "learning_rate": 9.238095238095239e-06,
+      "loss": 0.1652,
+      "step": 97
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8980953495205466,
+      "learning_rate": 9.333333333333334e-06,
+      "loss": 0.1171,
+      "step": 98
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8913864148939407,
+      "learning_rate": 9.42857142857143e-06,
+      "loss": 0.1599,
+      "step": 99
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.544543553668204,
+      "learning_rate": 9.523809523809525e-06,
+      "loss": 0.1868,
+      "step": 100
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8894016880965862,
+      "learning_rate": 9.61904761904762e-06,
+      "loss": 0.086,
+      "step": 101
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9990694939814734,
+      "learning_rate": 9.714285714285715e-06,
+      "loss": 0.1195,
+      "step": 102
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8996897056834422,
+      "learning_rate": 9.80952380952381e-06,
+      "loss": 0.0993,
+      "step": 103
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7239166665180842,
+      "learning_rate": 9.904761904761906e-06,
+      "loss": 0.0803,
+      "step": 104
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5666278173832199,
+      "learning_rate": 1e-05,
+      "loss": 0.0864,
+      "step": 105
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6233138913957039,
+      "learning_rate": 9.999997859278526e-06,
+      "loss": 0.1164,
+      "step": 106
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.883625227676478,
+      "learning_rate": 9.999991437115933e-06,
+      "loss": 0.214,
+      "step": 107
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9319416016251865,
+      "learning_rate": 9.999980733517723e-06,
+      "loss": 0.1669,
+      "step": 108
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6660261953169586,
+      "learning_rate": 9.999965748493058e-06,
+      "loss": 0.0889,
+      "step": 109
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8812531589326963,
+      "learning_rate": 9.999946482054771e-06,
+      "loss": 0.1087,
+      "step": 110
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.013938509272412,
+      "learning_rate": 9.999922934219364e-06,
+      "loss": 0.16,
+      "step": 111
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.779924825342806,
+      "learning_rate": 9.999895105006995e-06,
+      "loss": 0.1171,
+      "step": 112
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5518981538318566,
+      "learning_rate": 9.999862994441496e-06,
+      "loss": 0.1687,
+      "step": 113
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.48999398014184997,
+      "learning_rate": 9.999826602550361e-06,
+      "loss": 0.0823,
+      "step": 114
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0379184831245838,
+      "learning_rate": 9.999785929364756e-06,
+      "loss": 0.0916,
+      "step": 115
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7814552871771392,
+      "learning_rate": 9.999740974919506e-06,
+      "loss": 0.1048,
+      "step": 116
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0459600546375207,
+      "learning_rate": 9.999691739253106e-06,
+      "loss": 0.1067,
+      "step": 117
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8687944034663304,
+      "learning_rate": 9.999638222407715e-06,
+      "loss": 0.0879,
+      "step": 118
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9992034541083337,
+      "learning_rate": 9.99958042442916e-06,
+      "loss": 0.1812,
+      "step": 119
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6549484173810968,
+      "learning_rate": 9.999518345366933e-06,
+      "loss": 0.1085,
+      "step": 120
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9620760826195979,
+      "learning_rate": 9.999451985274188e-06,
+      "loss": 0.1608,
+      "step": 121
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7995800634969509,
+      "learning_rate": 9.999381344207753e-06,
+      "loss": 0.0873,
+      "step": 122
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6696605238491989,
+      "learning_rate": 9.999306422228117e-06,
+      "loss": 0.1115,
+      "step": 123
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7509671990302248,
+      "learning_rate": 9.99922721939943e-06,
+      "loss": 0.1154,
+      "step": 124
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9825079833026981,
+      "learning_rate": 9.999143735789518e-06,
+      "loss": 0.1376,
+      "step": 125
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0799052477502589,
+      "learning_rate": 9.999055971469864e-06,
+      "loss": 0.1994,
+      "step": 126
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6075380890888262,
+      "learning_rate": 9.99896392651562e-06,
+      "loss": 0.1142,
+      "step": 127
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9286626226849513,
+      "learning_rate": 9.998867601005604e-06,
+      "loss": 0.1005,
+      "step": 128
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.487131521035367,
+      "learning_rate": 9.998766995022297e-06,
+      "loss": 0.1936,
+      "step": 129
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6084846497507734,
+      "learning_rate": 9.99866210865185e-06,
+      "loss": 0.1044,
+      "step": 130
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7328283271168189,
+      "learning_rate": 9.99855294198407e-06,
+      "loss": 0.1114,
+      "step": 131
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9095013063381632,
+      "learning_rate": 9.998439495112439e-06,
+      "loss": 0.0944,
+      "step": 132
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8867638564351887,
+      "learning_rate": 9.998321768134101e-06,
+      "loss": 0.1078,
+      "step": 133
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5672820705386512,
+      "learning_rate": 9.998199761149865e-06,
+      "loss": 0.0866,
+      "step": 134
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.725001728131153,
+      "learning_rate": 9.9980734742642e-06,
+      "loss": 0.0989,
+      "step": 135
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.775556653527424,
+      "learning_rate": 9.997942907585247e-06,
+      "loss": 0.1185,
+      "step": 136
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0529065649602682,
+      "learning_rate": 9.99780806122481e-06,
+      "loss": 0.1798,
+      "step": 137
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5803911239780993,
+      "learning_rate": 9.997668935298353e-06,
+      "loss": 0.1033,
+      "step": 138
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9432373674731988,
+      "learning_rate": 9.997525529925008e-06,
+      "loss": 0.1244,
+      "step": 139
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.41233635150763165,
+      "learning_rate": 9.997377845227577e-06,
+      "loss": 0.0681,
+      "step": 140
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7619125890832402,
+      "learning_rate": 9.997225881332513e-06,
+      "loss": 0.1189,
+      "step": 141
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.731063900905247,
+      "learning_rate": 9.997069638369945e-06,
+      "loss": 0.1726,
+      "step": 142
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7350723280008996,
+      "learning_rate": 9.996909116473663e-06,
+      "loss": 0.132,
+      "step": 143
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.885499903881568,
+      "learning_rate": 9.996744315781119e-06,
+      "loss": 0.1003,
+      "step": 144
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6107676956417163,
+      "learning_rate": 9.996575236433428e-06,
+      "loss": 0.0939,
+      "step": 145
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6855064867583867,
+      "learning_rate": 9.996401878575374e-06,
+      "loss": 0.1145,
+      "step": 146
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7232702300952882,
+      "learning_rate": 9.9962242423554e-06,
+      "loss": 0.1199,
+      "step": 147
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6744310151901042,
+      "learning_rate": 9.996042327925613e-06,
+      "loss": 0.0989,
+      "step": 148
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5170328702922494,
+      "learning_rate": 9.995856135441785e-06,
+      "loss": 0.0878,
+      "step": 149
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7400293830583283,
+      "learning_rate": 9.995665665063349e-06,
+      "loss": 0.1073,
+      "step": 150
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.338869327388438,
+      "learning_rate": 9.995470916953405e-06,
+      "loss": 0.112,
+      "step": 151
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9744608681284141,
+      "learning_rate": 9.995271891278712e-06,
+      "loss": 0.1578,
+      "step": 152
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7245415196591839,
+      "learning_rate": 9.995068588209695e-06,
+      "loss": 0.1492,
+      "step": 153
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1864074350226086,
+      "learning_rate": 9.99486100792044e-06,
+      "loss": 0.151,
+      "step": 154
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0998672949312496,
+      "learning_rate": 9.994649150588694e-06,
+      "loss": 0.1328,
+      "step": 155
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7878447133281667,
+      "learning_rate": 9.994433016395868e-06,
+      "loss": 0.1173,
+      "step": 156
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.179693118007151,
+      "learning_rate": 9.994212605527036e-06,
+      "loss": 0.1761,
+      "step": 157
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7837054114786819,
+      "learning_rate": 9.993987918170934e-06,
+      "loss": 0.0911,
+      "step": 158
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9016876537829598,
+      "learning_rate": 9.993758954519957e-06,
+      "loss": 0.1287,
+      "step": 159
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5654627349734971,
+      "learning_rate": 9.993525714770167e-06,
+      "loss": 0.1209,
+      "step": 160
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9325146601897986,
+      "learning_rate": 9.993288199121283e-06,
+      "loss": 0.1456,
+      "step": 161
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8011664414955603,
+      "learning_rate": 9.993046407776687e-06,
+      "loss": 0.0878,
+      "step": 162
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5265780114164411,
+      "learning_rate": 9.992800340943421e-06,
+      "loss": 0.09,
+      "step": 163
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7413159658677737,
+      "learning_rate": 9.992549998832192e-06,
+      "loss": 0.1522,
+      "step": 164
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.64797029238668,
+      "learning_rate": 9.992295381657361e-06,
+      "loss": 0.2691,
+      "step": 165
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7338495948785921,
+      "learning_rate": 9.992036489636959e-06,
+      "loss": 0.1198,
+      "step": 166
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6518044519539281,
+      "learning_rate": 9.991773322992668e-06,
+      "loss": 0.1048,
+      "step": 167
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2989102160862043,
+      "learning_rate": 9.991505881949837e-06,
+      "loss": 0.1561,
+      "step": 168
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5184978052251222,
+      "learning_rate": 9.991234166737472e-06,
+      "loss": 0.0994,
+      "step": 169
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8114606660007323,
+      "learning_rate": 9.990958177588236e-06,
+      "loss": 0.0921,
+      "step": 170
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5405227025965889,
+      "learning_rate": 9.990677914738462e-06,
+      "loss": 0.1154,
+      "step": 171
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6854086725894425,
+      "learning_rate": 9.990393378428133e-06,
+      "loss": 0.0664,
+      "step": 172
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.607785348904251,
+      "learning_rate": 9.990104568900894e-06,
+      "loss": 0.1144,
+      "step": 173
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3499216602173765,
+      "learning_rate": 9.989811486404047e-06,
+      "loss": 0.1298,
+      "step": 174
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6640856742707212,
+      "learning_rate": 9.98951413118856e-06,
+      "loss": 0.1199,
+      "step": 175
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7258902684307403,
+      "learning_rate": 9.989212503509051e-06,
+      "loss": 0.1582,
+      "step": 176
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6374866990595666,
+      "learning_rate": 9.9889066036238e-06,
+      "loss": 0.123,
+      "step": 177
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6473255925101078,
+      "learning_rate": 9.988596431794749e-06,
+      "loss": 0.0846,
+      "step": 178
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7056436245707706,
+      "learning_rate": 9.988281988287494e-06,
+      "loss": 0.1242,
+      "step": 179
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5768846856508759,
+      "learning_rate": 9.987963273371287e-06,
+      "loss": 0.0912,
+      "step": 180
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2463463020099645,
+      "learning_rate": 9.987640287319041e-06,
+      "loss": 0.2249,
+      "step": 181
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7353683654933066,
+      "learning_rate": 9.987313030407325e-06,
+      "loss": 0.1127,
+      "step": 182
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1174838155664126,
+      "learning_rate": 9.986981502916364e-06,
+      "loss": 0.1731,
+      "step": 183
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2635041592029812,
+      "learning_rate": 9.986645705130046e-06,
+      "loss": 0.2071,
+      "step": 184
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9739326159678383,
+      "learning_rate": 9.986305637335907e-06,
+      "loss": 0.1323,
+      "step": 185
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1746961119873447,
+      "learning_rate": 9.985961299825144e-06,
+      "loss": 0.1474,
+      "step": 186
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0224320471521582,
+      "learning_rate": 9.985612692892608e-06,
+      "loss": 0.1555,
+      "step": 187
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7115155142503841,
+      "learning_rate": 9.985259816836809e-06,
+      "loss": 0.1235,
+      "step": 188
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3880477340080042,
+      "learning_rate": 9.984902671959911e-06,
+      "loss": 0.0968,
+      "step": 189
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5229836416130678,
+      "learning_rate": 9.984541258567732e-06,
+      "loss": 0.091,
+      "step": 190
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9859563440891488,
+      "learning_rate": 9.984175576969746e-06,
+      "loss": 0.1875,
+      "step": 191
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8937978532083918,
+      "learning_rate": 9.983805627479081e-06,
+      "loss": 0.1528,
+      "step": 192
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6714263410693669,
+      "learning_rate": 9.983431410412525e-06,
+      "loss": 0.1248,
+      "step": 193
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6051274231913827,
+      "learning_rate": 9.98305292609051e-06,
+      "loss": 0.0963,
+      "step": 194
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2791179950988925,
+      "learning_rate": 9.982670174837131e-06,
+      "loss": 0.1546,
+      "step": 195
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5526581263372871,
+      "learning_rate": 9.982283156980133e-06,
+      "loss": 0.0974,
+      "step": 196
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5469715280915156,
+      "learning_rate": 9.981891872850915e-06,
+      "loss": 0.0827,
+      "step": 197
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.3269280337142058,
+      "learning_rate": 9.98149632278453e-06,
+      "loss": 0.1549,
+      "step": 198
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6792026352415177,
+      "learning_rate": 9.98109650711968e-06,
+      "loss": 0.1237,
+      "step": 199
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5262828220255815,
+      "learning_rate": 9.980692426198728e-06,
+      "loss": 0.0865,
+      "step": 200
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5655283422306094,
+      "learning_rate": 9.98028408036768e-06,
+      "loss": 0.0797,
+      "step": 201
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.49899879221140164,
+      "learning_rate": 9.979871469976197e-06,
+      "loss": 0.0872,
+      "step": 202
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.023023912058545,
+      "learning_rate": 9.979454595377594e-06,
+      "loss": 0.1188,
+      "step": 203
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9585597431837045,
+      "learning_rate": 9.979033456928834e-06,
+      "loss": 0.0951,
+      "step": 204
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7593341762338881,
+      "learning_rate": 9.97860805499054e-06,
+      "loss": 0.1618,
+      "step": 205
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9878764238767545,
+      "learning_rate": 9.97817838992697e-06,
+      "loss": 0.1287,
+      "step": 206
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6226373819190023,
+      "learning_rate": 9.977744462106048e-06,
+      "loss": 0.1225,
+      "step": 207
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9282611042570976,
+      "learning_rate": 9.977306271899336e-06,
+      "loss": 0.1514,
+      "step": 208
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6124797469277562,
+      "learning_rate": 9.976863819682055e-06,
+      "loss": 0.11,
+      "step": 209
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7966969818313082,
+      "learning_rate": 9.97641710583307e-06,
+      "loss": 0.1481,
+      "step": 210
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6708286500114512,
+      "learning_rate": 9.9759661307349e-06,
+      "loss": 0.1071,
+      "step": 211
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3359148785200574,
+      "learning_rate": 9.975510894773705e-06,
+      "loss": 0.139,
+      "step": 212
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1732557367642897,
+      "learning_rate": 9.975051398339302e-06,
+      "loss": 0.1458,
+      "step": 213
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0954552864142109,
+      "learning_rate": 9.97458764182515e-06,
+      "loss": 0.1851,
+      "step": 214
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5668284518872119,
+      "learning_rate": 9.97411962562836e-06,
+      "loss": 0.1049,
+      "step": 215
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.413407712524076,
+      "learning_rate": 9.97364735014969e-06,
+      "loss": 0.0924,
+      "step": 216
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6846820606475784,
+      "learning_rate": 9.973170815793543e-06,
+      "loss": 0.1567,
+      "step": 217
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7104955831606532,
+      "learning_rate": 9.97269002296797e-06,
+      "loss": 0.0961,
+      "step": 218
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9088643628647692,
+      "learning_rate": 9.972204972084667e-06,
+      "loss": 0.1577,
+      "step": 219
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2063379291826812,
+      "learning_rate": 9.971715663558978e-06,
+      "loss": 0.1569,
+      "step": 220
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6950750179869627,
+      "learning_rate": 9.971222097809895e-06,
+      "loss": 0.1179,
+      "step": 221
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6239492884456627,
+      "learning_rate": 9.97072427526005e-06,
+      "loss": 0.0964,
+      "step": 222
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.074717542517687,
+      "learning_rate": 9.970222196335724e-06,
+      "loss": 0.1077,
+      "step": 223
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9295098911783852,
+      "learning_rate": 9.969715861466839e-06,
+      "loss": 0.1238,
+      "step": 224
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7725413999922282,
+      "learning_rate": 9.969205271086969e-06,
+      "loss": 0.157,
+      "step": 225
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0391873748724498,
+      "learning_rate": 9.96869042563332e-06,
+      "loss": 0.1547,
+      "step": 226
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7404030523492874,
+      "learning_rate": 9.968171325546754e-06,
+      "loss": 0.1077,
+      "step": 227
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0163578185462376,
+      "learning_rate": 9.967647971271769e-06,
+      "loss": 0.2065,
+      "step": 228
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4508114700967041,
+      "learning_rate": 9.967120363256503e-06,
+      "loss": 0.0665,
+      "step": 229
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.5555517780970542,
+      "learning_rate": 9.966588501952747e-06,
+      "loss": 0.0767,
+      "step": 230
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7702021933859821,
+      "learning_rate": 9.966052387815923e-06,
+      "loss": 0.1436,
+      "step": 231
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8275984175997171,
+      "learning_rate": 9.965512021305104e-06,
+      "loss": 0.1149,
+      "step": 232
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4995501608117536,
+      "learning_rate": 9.964967402882995e-06,
+      "loss": 0.0996,
+      "step": 233
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5660501473102331,
+      "learning_rate": 9.96441853301595e-06,
+      "loss": 0.1113,
+      "step": 234
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8238187403852497,
+      "learning_rate": 9.963865412173958e-06,
+      "loss": 0.1186,
+      "step": 235
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7875749448428275,
+      "learning_rate": 9.96330804083065e-06,
+      "loss": 0.1162,
+      "step": 236
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0298150673502589,
+      "learning_rate": 9.962746419463298e-06,
+      "loss": 0.1724,
+      "step": 237
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7735969323058663,
+      "learning_rate": 9.962180548552812e-06,
+      "loss": 0.1237,
+      "step": 238
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.930852149800409,
+      "learning_rate": 9.96161042858374e-06,
+      "loss": 0.169,
+      "step": 239
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5392105608782614,
+      "learning_rate": 9.961036060044268e-06,
+      "loss": 0.0713,
+      "step": 240
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7987543249051764,
+      "learning_rate": 9.960457443426227e-06,
+      "loss": 0.1622,
+      "step": 241
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5520057170888215,
+      "learning_rate": 9.959874579225071e-06,
+      "loss": 0.1214,
+      "step": 242
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7714202480896567,
+      "learning_rate": 9.959287467939905e-06,
+      "loss": 0.1012,
+      "step": 243
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7018270463464699,
+      "learning_rate": 9.958696110073467e-06,
+      "loss": 0.1301,
+      "step": 244
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.38352681428144836,
+      "learning_rate": 9.958100506132127e-06,
+      "loss": 0.0897,
+      "step": 245
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0087660315330838,
+      "learning_rate": 9.957500656625894e-06,
+      "loss": 0.0879,
+      "step": 246
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8437066569382713,
+      "learning_rate": 9.956896562068414e-06,
+      "loss": 0.1374,
+      "step": 247
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0657285913367234,
+      "learning_rate": 9.956288222976967e-06,
+      "loss": 0.1117,
+      "step": 248
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7761536227390066,
+      "learning_rate": 9.955675639872465e-06,
+      "loss": 0.102,
+      "step": 249
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6343701991714403,
+      "learning_rate": 9.955058813279454e-06,
+      "loss": 0.1198,
+      "step": 250
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.304116025053766,
+      "learning_rate": 9.954437743726119e-06,
+      "loss": 0.1093,
+      "step": 251
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8200651903998959,
+      "learning_rate": 9.953812431744274e-06,
+      "loss": 0.079,
+      "step": 252
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6296748927291802,
+      "learning_rate": 9.953182877869367e-06,
+      "loss": 0.0982,
+      "step": 253
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5565623273658198,
+      "learning_rate": 9.952549082640477e-06,
+      "loss": 0.0789,
+      "step": 254
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4177577620070906,
+      "learning_rate": 9.951911046600313e-06,
+      "loss": 0.0502,
+      "step": 255
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6088489128825829,
+      "learning_rate": 9.951268770295223e-06,
+      "loss": 0.1344,
+      "step": 256
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0327549751969483,
+      "learning_rate": 9.950622254275177e-06,
+      "loss": 0.1345,
+      "step": 257
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8541531254505703,
+      "learning_rate": 9.949971499093783e-06,
+      "loss": 0.1761,
+      "step": 258
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8841526214312907,
+      "learning_rate": 9.94931650530827e-06,
+      "loss": 0.098,
+      "step": 259
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9100164150700225,
+      "learning_rate": 9.948657273479508e-06,
+      "loss": 0.1796,
+      "step": 260
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7086473035157004,
+      "learning_rate": 9.947993804171984e-06,
+      "loss": 0.1044,
+      "step": 261
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7222672977162455,
+      "learning_rate": 9.947326097953822e-06,
+      "loss": 0.135,
+      "step": 262
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8827591249828824,
+      "learning_rate": 9.94665415539677e-06,
+      "loss": 0.1175,
+      "step": 263
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4623982686609934,
+      "learning_rate": 9.945977977076206e-06,
+      "loss": 0.1072,
+      "step": 264
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4658633145163734,
+      "learning_rate": 9.945297563571135e-06,
+      "loss": 0.0803,
+      "step": 265
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5577025537296104,
+      "learning_rate": 9.944612915464183e-06,
+      "loss": 0.0987,
+      "step": 266
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9151149107638324,
+      "learning_rate": 9.94392403334161e-06,
+      "loss": 0.1404,
+      "step": 267
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6881193923523389,
+      "learning_rate": 9.943230917793297e-06,
+      "loss": 0.085,
+      "step": 268
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1367716264658339,
+      "learning_rate": 9.942533569412751e-06,
+      "loss": 0.1158,
+      "step": 269
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6198641287768313,
+      "learning_rate": 9.941831988797104e-06,
+      "loss": 0.1357,
+      "step": 270
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5584956978186149,
+      "learning_rate": 9.94112617654711e-06,
+      "loss": 0.0635,
+      "step": 271
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7710831743353788,
+      "learning_rate": 9.940416133267147e-06,
+      "loss": 0.1128,
+      "step": 272
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5757367164546061,
+      "learning_rate": 9.93970185956522e-06,
+      "loss": 0.1268,
+      "step": 273
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.78935651150386,
+      "learning_rate": 9.938983356052953e-06,
+      "loss": 0.0948,
+      "step": 274
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1938775083474606,
+      "learning_rate": 9.938260623345591e-06,
+      "loss": 0.1125,
+      "step": 275
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5934987143439762,
+      "learning_rate": 9.937533662062002e-06,
+      "loss": 0.1266,
+      "step": 276
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.634122518529506,
+      "learning_rate": 9.936802472824675e-06,
+      "loss": 0.1645,
+      "step": 277
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5640956349326679,
+      "learning_rate": 9.93606705625972e-06,
+      "loss": 0.1316,
+      "step": 278
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8946303659615805,
+      "learning_rate": 9.935327412996863e-06,
+      "loss": 0.1282,
+      "step": 279
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.40685726080093326,
+      "learning_rate": 9.934583543669454e-06,
+      "loss": 0.067,
+      "step": 280
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6363030971304591,
+      "learning_rate": 9.933835448914459e-06,
+      "loss": 0.1211,
+      "step": 281
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.37747426502340686,
+      "learning_rate": 9.933083129372467e-06,
+      "loss": 0.069,
+      "step": 282
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0026041546645483,
+      "learning_rate": 9.932326585687675e-06,
+      "loss": 0.1103,
+      "step": 283
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6237086866445896,
+      "learning_rate": 9.931565818507904e-06,
+      "loss": 0.1459,
+      "step": 284
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5198324908614264,
+      "learning_rate": 9.930800828484593e-06,
+      "loss": 0.0949,
+      "step": 285
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.52901423899486,
+      "learning_rate": 9.930031616272791e-06,
+      "loss": 0.1416,
+      "step": 286
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.768077322285948,
+      "learning_rate": 9.929258182531167e-06,
+      "loss": 0.1069,
+      "step": 287
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.43670884798857934,
+      "learning_rate": 9.928480527922004e-06,
+      "loss": 0.1021,
+      "step": 288
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9349301943026924,
+      "learning_rate": 9.9276986531112e-06,
+      "loss": 0.1397,
+      "step": 289
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.46386221384249404,
+      "learning_rate": 9.926912558768261e-06,
+      "loss": 0.056,
+      "step": 290
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.794965781068542,
+      "learning_rate": 9.926122245566315e-06,
+      "loss": 0.1069,
+      "step": 291
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6382219405542284,
+      "learning_rate": 9.925327714182098e-06,
+      "loss": 0.1126,
+      "step": 292
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6005837617420782,
+      "learning_rate": 9.924528965295955e-06,
+      "loss": 0.0878,
+      "step": 293
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1359604968686856,
+      "learning_rate": 9.923725999591846e-06,
+      "loss": 0.1481,
+      "step": 294
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4115889109907375,
+      "learning_rate": 9.922918817757346e-06,
+      "loss": 0.0788,
+      "step": 295
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7082289559538197,
+      "learning_rate": 9.92210742048363e-06,
+      "loss": 0.122,
+      "step": 296
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2019615043832632,
+      "learning_rate": 9.921291808465492e-06,
+      "loss": 0.1827,
+      "step": 297
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7559303887743698,
+      "learning_rate": 9.920471982401328e-06,
+      "loss": 0.1124,
+      "step": 298
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.479984057936419,
+      "learning_rate": 9.91964794299315e-06,
+      "loss": 0.102,
+      "step": 299
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.44094109838821893,
+      "learning_rate": 9.918819690946568e-06,
+      "loss": 0.0815,
+      "step": 300
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.39589012538937873,
+      "learning_rate": 9.917987226970811e-06,
+      "loss": 0.1,
+      "step": 301
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5132120943206462,
+      "learning_rate": 9.917150551778702e-06,
+      "loss": 0.0835,
+      "step": 302
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4717486642036433,
+      "learning_rate": 9.916309666086682e-06,
+      "loss": 0.0935,
+      "step": 303
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7547071809281382,
+      "learning_rate": 9.91546457061479e-06,
+      "loss": 0.1458,
+      "step": 304
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3951791292961316,
+      "learning_rate": 9.914615266086668e-06,
+      "loss": 0.0738,
+      "step": 305
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4726660966593668,
+      "learning_rate": 9.91376175322957e-06,
+      "loss": 0.0852,
+      "step": 306
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9099276522464423,
+      "learning_rate": 9.912904032774351e-06,
+      "loss": 0.1884,
+      "step": 307
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6985873768508254,
+      "learning_rate": 9.912042105455462e-06,
+      "loss": 0.1273,
+      "step": 308
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7065357650409732,
+      "learning_rate": 9.911175972010965e-06,
+      "loss": 0.1125,
+      "step": 309
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6304748039856666,
+      "learning_rate": 9.910305633182518e-06,
+      "loss": 0.117,
+      "step": 310
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.045398760806309,
+      "learning_rate": 9.909431089715384e-06,
+      "loss": 0.131,
+      "step": 311
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6531289786555421,
+      "learning_rate": 9.908552342358424e-06,
+      "loss": 0.1659,
+      "step": 312
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4342512824963871,
+      "learning_rate": 9.9076693918641e-06,
+      "loss": 0.1085,
+      "step": 313
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5572430880433067,
+      "learning_rate": 9.906782238988471e-06,
+      "loss": 0.0838,
+      "step": 314
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6234216153610849,
+      "learning_rate": 9.905890884491196e-06,
+      "loss": 0.082,
+      "step": 315
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4573970084251661,
+      "learning_rate": 9.904995329135533e-06,
+      "loss": 0.0803,
+      "step": 316
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5842062537706539,
+      "learning_rate": 9.904095573688334e-06,
+      "loss": 0.106,
+      "step": 317
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6506065102648938,
+      "learning_rate": 9.903191618920052e-06,
+      "loss": 0.1377,
+      "step": 318
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.42497362854593346,
+      "learning_rate": 9.902283465604731e-06,
+      "loss": 0.0594,
+      "step": 319
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8118712222286686,
+      "learning_rate": 9.901371114520014e-06,
+      "loss": 0.095,
+      "step": 320
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5546005108949749,
+      "learning_rate": 9.900454566447133e-06,
+      "loss": 0.1164,
+      "step": 321
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.642648777360291,
+      "learning_rate": 9.899533822170922e-06,
+      "loss": 0.1053,
+      "step": 322
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7432919092539422,
+      "learning_rate": 9.898608882479803e-06,
+      "loss": 0.1503,
+      "step": 323
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7774030160511798,
+      "learning_rate": 9.897679748165789e-06,
+      "loss": 0.1801,
+      "step": 324
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6358064884579139,
+      "learning_rate": 9.89674642002449e-06,
+      "loss": 0.1638,
+      "step": 325
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5941447561513429,
+      "learning_rate": 9.895808898855102e-06,
+      "loss": 0.1336,
+      "step": 326
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0164412114525911,
+      "learning_rate": 9.894867185460416e-06,
+      "loss": 0.1635,
+      "step": 327
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.125245132533889,
+      "learning_rate": 9.893921280646806e-06,
+      "loss": 0.1492,
+      "step": 328
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.919610050412465,
+      "learning_rate": 9.892971185224244e-06,
+      "loss": 0.1193,
+      "step": 329
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9071384652106338,
+      "learning_rate": 9.892016900006284e-06,
+      "loss": 0.1122,
+      "step": 330
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9191595453235826,
+      "learning_rate": 9.891058425810072e-06,
+      "loss": 0.1706,
+      "step": 331
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.40444134298820555,
+      "learning_rate": 9.890095763456335e-06,
+      "loss": 0.102,
+      "step": 332
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5795507667218293,
+      "learning_rate": 9.88912891376939e-06,
+      "loss": 0.1184,
+      "step": 333
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5159918037293236,
+      "learning_rate": 9.888157877577142e-06,
+      "loss": 0.1555,
+      "step": 334
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7181342541230561,
+      "learning_rate": 9.887182655711078e-06,
+      "loss": 0.1283,
+      "step": 335
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5437508320370097,
+      "learning_rate": 9.886203249006265e-06,
+      "loss": 0.0821,
+      "step": 336
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4871363822091073,
+      "learning_rate": 9.885219658301364e-06,
+      "loss": 0.1376,
+      "step": 337
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.49116677149229987,
+      "learning_rate": 9.884231884438607e-06,
+      "loss": 0.1075,
+      "step": 338
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6662276223253167,
+      "learning_rate": 9.883239928263817e-06,
+      "loss": 0.1418,
+      "step": 339
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7085309618070171,
+      "learning_rate": 9.882243790626393e-06,
+      "loss": 0.1491,
+      "step": 340
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4442275206621628,
+      "learning_rate": 9.881243472379318e-06,
+      "loss": 0.1132,
+      "step": 341
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8460443182094871,
+      "learning_rate": 9.880238974379151e-06,
+      "loss": 0.1599,
+      "step": 342
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1208304678283527,
+      "learning_rate": 9.879230297486034e-06,
+      "loss": 0.1617,
+      "step": 343
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6311369962916906,
+      "learning_rate": 9.878217442563684e-06,
+      "loss": 0.073,
+      "step": 344
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.607057662787762,
+      "learning_rate": 9.877200410479399e-06,
+      "loss": 0.0814,
+      "step": 345
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6660247333119897,
+      "learning_rate": 9.87617920210405e-06,
+      "loss": 0.1606,
+      "step": 346
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1808944353739446,
+      "learning_rate": 9.875153818312087e-06,
+      "loss": 0.181,
+      "step": 347
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.544513639908992,
+      "learning_rate": 9.874124259981535e-06,
+      "loss": 0.0945,
+      "step": 348
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3040066447276715,
+      "learning_rate": 9.873090527993991e-06,
+      "loss": 0.093,
+      "step": 349
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4463180589496372,
+      "learning_rate": 9.872052623234632e-06,
+      "loss": 0.1903,
+      "step": 350
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7270762926748952,
+      "learning_rate": 9.871010546592199e-06,
+      "loss": 0.1297,
+      "step": 351
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5589878973559803,
+      "learning_rate": 9.869964298959013e-06,
+      "loss": 0.1001,
+      "step": 352
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5514658148916033,
+      "learning_rate": 9.868913881230964e-06,
+      "loss": 0.108,
+      "step": 353
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.39676003138608923,
+      "learning_rate": 9.867859294307512e-06,
+      "loss": 0.0845,
+      "step": 354
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8529347030868513,
+      "learning_rate": 9.866800539091688e-06,
+      "loss": 0.144,
+      "step": 355
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.43061480486765114,
+      "learning_rate": 9.865737616490092e-06,
+      "loss": 0.0703,
+      "step": 356
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7716491329388591,
+      "learning_rate": 9.864670527412891e-06,
+      "loss": 0.1309,
+      "step": 357
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7008222888423032,
+      "learning_rate": 9.863599272773825e-06,
+      "loss": 0.1516,
+      "step": 358
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5992066599365296,
+      "learning_rate": 9.862523853490193e-06,
+      "loss": 0.0902,
+      "step": 359
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.40837576918803054,
+      "learning_rate": 9.861444270482869e-06,
+      "loss": 0.0707,
+      "step": 360
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3809539796674921,
+      "learning_rate": 9.860360524676282e-06,
+      "loss": 0.0618,
+      "step": 361
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8405820843448246,
+      "learning_rate": 9.859272616998435e-06,
+      "loss": 0.1492,
+      "step": 362
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6296256234106073,
+      "learning_rate": 9.85818054838089e-06,
+      "loss": 0.1309,
+      "step": 363
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.34282425587313214,
+      "learning_rate": 9.857084319758772e-06,
+      "loss": 0.0816,
+      "step": 364
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.43307965248951774,
+      "learning_rate": 9.855983932070771e-06,
+      "loss": 0.0864,
+      "step": 365
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3853600765959287,
+      "learning_rate": 9.854879386259133e-06,
+      "loss": 0.0845,
+      "step": 366
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6166113837784739,
+      "learning_rate": 9.853770683269672e-06,
+      "loss": 0.115,
+      "step": 367
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.523962182786505,
+      "learning_rate": 9.852657824051756e-06,
+      "loss": 0.0883,
+      "step": 368
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.48823667476209875,
+      "learning_rate": 9.851540809558314e-06,
+      "loss": 0.0686,
+      "step": 369
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5456778299346657,
+      "learning_rate": 9.85041964074583e-06,
+      "loss": 0.0864,
+      "step": 370
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3571644499775195,
+      "learning_rate": 9.849294318574353e-06,
+      "loss": 0.1546,
+      "step": 371
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6307858386939773,
+      "learning_rate": 9.84816484400748e-06,
+      "loss": 0.0954,
+      "step": 372
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.189465103768443,
+      "learning_rate": 9.847031218012366e-06,
+      "loss": 0.1403,
+      "step": 373
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6678800569095343,
+      "learning_rate": 9.845893441559726e-06,
+      "loss": 0.1179,
+      "step": 374
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.48783364991932904,
+      "learning_rate": 9.844751515623824e-06,
+      "loss": 0.0581,
+      "step": 375
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7854315167901365,
+      "learning_rate": 9.843605441182476e-06,
+      "loss": 0.1179,
+      "step": 376
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8646577020314478,
+      "learning_rate": 9.842455219217054e-06,
+      "loss": 0.1421,
+      "step": 377
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6472659091717089,
+      "learning_rate": 9.841300850712479e-06,
+      "loss": 0.1106,
+      "step": 378
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4881078709914351,
+      "learning_rate": 9.840142336657225e-06,
+      "loss": 0.0784,
+      "step": 379
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4071832924612178,
+      "learning_rate": 9.838979678043314e-06,
+      "loss": 0.0925,
+      "step": 380
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5635616531793785,
+      "learning_rate": 9.837812875866317e-06,
+      "loss": 0.0882,
+      "step": 381
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8490835049175628,
+      "learning_rate": 9.836641931125352e-06,
+      "loss": 0.1173,
+      "step": 382
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.46353250195994344,
+      "learning_rate": 9.835466844823089e-06,
+      "loss": 0.0914,
+      "step": 383
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.44003404589085987,
+      "learning_rate": 9.834287617965737e-06,
+      "loss": 0.0917,
+      "step": 384
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4382669571034664,
+      "learning_rate": 9.833104251563058e-06,
+      "loss": 0.0922,
+      "step": 385
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6622412790128882,
+      "learning_rate": 9.831916746628352e-06,
+      "loss": 0.124,
+      "step": 386
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8740197276848335,
+      "learning_rate": 9.830725104178467e-06,
+      "loss": 0.1218,
+      "step": 387
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3418488166373136,
+      "learning_rate": 9.829529325233795e-06,
+      "loss": 0.0539,
+      "step": 388
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8820005412720228,
+      "learning_rate": 9.828329410818265e-06,
+      "loss": 0.094,
+      "step": 389
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6348576839659562,
+      "learning_rate": 9.827125361959353e-06,
+      "loss": 0.1595,
+      "step": 390
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7776623517993152,
+      "learning_rate": 9.825917179688069e-06,
+      "loss": 0.1286,
+      "step": 391
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3652804147520181,
+      "learning_rate": 9.824704865038967e-06,
+      "loss": 0.0893,
+      "step": 392
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6098195027818251,
+      "learning_rate": 9.82348841905014e-06,
+      "loss": 0.1259,
+      "step": 393
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6818619351526606,
+      "learning_rate": 9.822267842763214e-06,
+      "loss": 0.1382,
+      "step": 394
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.41464173850333236,
+      "learning_rate": 9.821043137223356e-06,
+      "loss": 0.0875,
+      "step": 395
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7801461228744722,
+      "learning_rate": 9.819814303479268e-06,
+      "loss": 0.1142,
+      "step": 396
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6547149840231421,
+      "learning_rate": 9.818581342583184e-06,
+      "loss": 0.0917,
+      "step": 397
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5861194330529879,
+      "learning_rate": 9.817344255590878e-06,
+      "loss": 0.1201,
+      "step": 398
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.557594547924238,
+      "learning_rate": 9.816103043561648e-06,
+      "loss": 0.0807,
+      "step": 399
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2760885179588771,
+      "learning_rate": 9.814857707558334e-06,
+      "loss": 0.0492,
+      "step": 400
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4693531378375744,
+      "learning_rate": 9.813608248647303e-06,
+      "loss": 0.0882,
+      "step": 401
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5731728776422602,
+      "learning_rate": 9.812354667898452e-06,
+      "loss": 0.0913,
+      "step": 402
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5198319925114208,
+      "learning_rate": 9.811096966385208e-06,
+      "loss": 0.1292,
+      "step": 403
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8579699415577815,
+      "learning_rate": 9.809835145184523e-06,
+      "loss": 0.1655,
+      "step": 404
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7003979193764588,
+      "learning_rate": 9.808569205376885e-06,
+      "loss": 0.1396,
+      "step": 405
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.378968657022864,
+      "learning_rate": 9.807299148046301e-06,
+      "loss": 0.1405,
+      "step": 406
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4222883847606399,
+      "learning_rate": 9.806024974280308e-06,
+      "loss": 0.0845,
+      "step": 407
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5636086148045607,
+      "learning_rate": 9.804746685169964e-06,
+      "loss": 0.1077,
+      "step": 408
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0657574716078748,
+      "learning_rate": 9.803464281809856e-06,
+      "loss": 0.1338,
+      "step": 409
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4145608930789154,
+      "learning_rate": 9.802177765298091e-06,
+      "loss": 0.1138,
+      "step": 410
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.577848034737305,
+      "learning_rate": 9.800887136736297e-06,
+      "loss": 0.0831,
+      "step": 411
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.589083321715794,
+      "learning_rate": 9.799592397229626e-06,
+      "loss": 0.1157,
+      "step": 412
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.37488285060769505,
+      "learning_rate": 9.798293547886748e-06,
+      "loss": 0.0834,
+      "step": 413
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.882238271623321,
+      "learning_rate": 9.796990589819852e-06,
+      "loss": 0.1014,
+      "step": 414
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8510571435671381,
+      "learning_rate": 9.795683524144649e-06,
+      "loss": 0.0881,
+      "step": 415
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0433820395292264,
+      "learning_rate": 9.794372351980361e-06,
+      "loss": 0.159,
+      "step": 416
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5690948101339289,
+      "learning_rate": 9.793057074449732e-06,
+      "loss": 0.083,
+      "step": 417
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5635245868859875,
+      "learning_rate": 9.791737692679017e-06,
+      "loss": 0.0943,
+      "step": 418
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.48896492458869784,
+      "learning_rate": 9.790414207797992e-06,
+      "loss": 0.1345,
+      "step": 419
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5080777969890101,
+      "learning_rate": 9.789086620939936e-06,
+      "loss": 0.0959,
+      "step": 420
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1796292883213124,
+      "learning_rate": 9.78775493324165e-06,
+      "loss": 0.1619,
+      "step": 421
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9168967688087062,
+      "learning_rate": 9.786419145843444e-06,
+      "loss": 0.1372,
+      "step": 422
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8450339717912574,
+      "learning_rate": 9.785079259889134e-06,
+      "loss": 0.1329,
+      "step": 423
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7086652838313081,
+      "learning_rate": 9.783735276526052e-06,
+      "loss": 0.1039,
+      "step": 424
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6658796117357779,
+      "learning_rate": 9.782387196905034e-06,
+      "loss": 0.1575,
+      "step": 425
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6668359446272558,
+      "learning_rate": 9.781035022180428e-06,
+      "loss": 0.1045,
+      "step": 426
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5404750055461947,
+      "learning_rate": 9.779678753510082e-06,
+      "loss": 0.1249,
+      "step": 427
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6652977074482611,
+      "learning_rate": 9.778318392055354e-06,
+      "loss": 0.092,
+      "step": 428
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6220798892890224,
+      "learning_rate": 9.776953938981107e-06,
+      "loss": 0.0951,
+      "step": 429
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7757080092841647,
+      "learning_rate": 9.775585395455708e-06,
+      "loss": 0.1066,
+      "step": 430
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5810251634899445,
+      "learning_rate": 9.774212762651023e-06,
+      "loss": 0.1342,
+      "step": 431
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5469304876227978,
+      "learning_rate": 9.772836041742423e-06,
+      "loss": 0.1351,
+      "step": 432
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5860406866309947,
+      "learning_rate": 9.771455233908778e-06,
+      "loss": 0.1137,
+      "step": 433
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5141147693391175,
+      "learning_rate": 9.770070340332457e-06,
+      "loss": 0.1005,
+      "step": 434
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6755000524610992,
+      "learning_rate": 9.76868136219933e-06,
+      "loss": 0.1028,
+      "step": 435
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.35347845941056094,
+      "learning_rate": 9.767288300698762e-06,
+      "loss": 0.0929,
+      "step": 436
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3531724115204999,
+      "learning_rate": 9.76589115702362e-06,
+      "loss": 0.0692,
+      "step": 437
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3476338504585057,
+      "learning_rate": 9.764489932370254e-06,
+      "loss": 0.0632,
+      "step": 438
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5519961970875953,
+      "learning_rate": 9.763084627938524e-06,
+      "loss": 0.1219,
+      "step": 439
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1119189227138462,
+      "learning_rate": 9.761675244931772e-06,
+      "loss": 0.1654,
+      "step": 440
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.670153662815508,
+      "learning_rate": 9.76026178455684e-06,
+      "loss": 0.1381,
+      "step": 441
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.42626841703523144,
+      "learning_rate": 9.758844248024053e-06,
+      "loss": 0.1204,
+      "step": 442
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5487977644272218,
+      "learning_rate": 9.757422636547236e-06,
+      "loss": 0.1061,
+      "step": 443
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.31267067712952507,
+      "learning_rate": 9.755996951343696e-06,
+      "loss": 0.0741,
+      "step": 444
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6992283598924897,
+      "learning_rate": 9.754567193634232e-06,
+      "loss": 0.0992,
+      "step": 445
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6011903547464283,
+      "learning_rate": 9.75313336464313e-06,
+      "loss": 0.1503,
+      "step": 446
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3818976630825426,
+      "learning_rate": 9.751695465598161e-06,
+      "loss": 0.0697,
+      "step": 447
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5769177820484979,
+      "learning_rate": 9.75025349773058e-06,
+      "loss": 0.1103,
+      "step": 448
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5619632209692563,
+      "learning_rate": 9.748807462275129e-06,
+      "loss": 0.1048,
+      "step": 449
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.106285406278163,
+      "learning_rate": 9.747357360470033e-06,
+      "loss": 0.1484,
+      "step": 450
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4834893229443317,
+      "learning_rate": 9.745903193556994e-06,
+      "loss": 0.0645,
+      "step": 451
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9196394433430203,
+      "learning_rate": 9.7444449627812e-06,
+      "loss": 0.0938,
+      "step": 452
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7994197717698222,
+      "learning_rate": 9.742982669391321e-06,
+      "loss": 0.1251,
+      "step": 453
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5959381420740328,
+      "learning_rate": 9.741516314639496e-06,
+      "loss": 0.1066,
+      "step": 454
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.57828463794402,
+      "learning_rate": 9.740045899781353e-06,
+      "loss": 0.1151,
+      "step": 455
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5054311221075344,
+      "learning_rate": 9.73857142607599e-06,
+      "loss": 0.1233,
+      "step": 456
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6722451213544615,
+      "learning_rate": 9.737092894785978e-06,
+      "loss": 0.1213,
+      "step": 457
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6038351017174377,
+      "learning_rate": 9.735610307177374e-06,
+      "loss": 0.0783,
+      "step": 458
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5161675228685113,
+      "learning_rate": 9.734123664519695e-06,
+      "loss": 0.1081,
+      "step": 459
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3431757659984759,
+      "learning_rate": 9.732632968085937e-06,
+      "loss": 0.0634,
+      "step": 460
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.49936599615693683,
+      "learning_rate": 9.731138219152566e-06,
+      "loss": 0.0998,
+      "step": 461
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6865581145913818,
+      "learning_rate": 9.729639418999524e-06,
+      "loss": 0.1101,
+      "step": 462
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9076153368211043,
+      "learning_rate": 9.728136568910209e-06,
+      "loss": 0.1028,
+      "step": 463
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8907933608275562,
+      "learning_rate": 9.7266296701715e-06,
+      "loss": 0.161,
+      "step": 464
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1085238664126051,
+      "learning_rate": 9.725118724073732e-06,
+      "loss": 0.1504,
+      "step": 465
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6157213503360681,
+      "learning_rate": 9.723603731910715e-06,
+      "loss": 0.1169,
+      "step": 466
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5561371323209214,
+      "learning_rate": 9.72208469497972e-06,
+      "loss": 0.1235,
+      "step": 467
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4442295890258957,
+      "learning_rate": 9.720561614581475e-06,
+      "loss": 0.1019,
+      "step": 468
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3757130394438029,
+      "learning_rate": 9.719034492020183e-06,
+      "loss": 0.0695,
+      "step": 469
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7063966445456047,
+      "learning_rate": 9.717503328603499e-06,
+      "loss": 0.1342,
+      "step": 470
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9155691047755337,
+      "learning_rate": 9.71596812564254e-06,
+      "loss": 0.1061,
+      "step": 471
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.9842753198583956,
+      "learning_rate": 9.714428884451886e-06,
+      "loss": 0.1689,
+      "step": 472
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6728142908542221,
+      "learning_rate": 9.712885606349567e-06,
+      "loss": 0.0818,
+      "step": 473
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1648373090904836,
+      "learning_rate": 9.711338292657078e-06,
+      "loss": 0.1476,
+      "step": 474
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5377512468558412,
+      "learning_rate": 9.709786944699364e-06,
+      "loss": 0.1165,
+      "step": 475
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5056295002520085,
+      "learning_rate": 9.708231563804828e-06,
+      "loss": 0.1046,
+      "step": 476
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5148043840776015,
+      "learning_rate": 9.706672151305324e-06,
+      "loss": 0.0667,
+      "step": 477
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6616997554763696,
+      "learning_rate": 9.70510870853616e-06,
+      "loss": 0.0847,
+      "step": 478
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5431856851371992,
+      "learning_rate": 9.703541236836094e-06,
+      "loss": 0.1055,
+      "step": 479
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5847676253243631,
+      "learning_rate": 9.701969737547332e-06,
+      "loss": 0.1157,
+      "step": 480
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.46457743196055556,
+      "learning_rate": 9.700394212015533e-06,
+      "loss": 0.0875,
+      "step": 481
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6501521201959152,
+      "learning_rate": 9.698814661589804e-06,
+      "loss": 0.1197,
+      "step": 482
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5799873171236264,
+      "learning_rate": 9.697231087622691e-06,
+      "loss": 0.0995,
+      "step": 483
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7622329310948606,
+      "learning_rate": 9.695643491470192e-06,
+      "loss": 0.1341,
+      "step": 484
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3683783284062726,
+      "learning_rate": 9.694051874491748e-06,
+      "loss": 0.0901,
+      "step": 485
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.43007029560479687,
+      "learning_rate": 9.692456238050245e-06,
+      "loss": 0.0731,
+      "step": 486
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4154136935256243,
+      "learning_rate": 9.690856583512002e-06,
+      "loss": 0.0768,
+      "step": 487
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9367202073860206,
+      "learning_rate": 9.689252912246792e-06,
+      "loss": 0.1284,
+      "step": 488
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4594907018923105,
+      "learning_rate": 9.687645225627817e-06,
+      "loss": 0.0815,
+      "step": 489
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6653529192373276,
+      "learning_rate": 9.68603352503172e-06,
+      "loss": 0.1173,
+      "step": 490
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4147604844498556,
+      "learning_rate": 9.68441781183858e-06,
+      "loss": 0.1052,
+      "step": 491
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4719536721308244,
+      "learning_rate": 9.68279808743192e-06,
+      "loss": 0.09,
+      "step": 492
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3180783346206244,
+      "learning_rate": 9.681174353198687e-06,
+      "loss": 0.1377,
+      "step": 493
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.30528484487601437,
+      "learning_rate": 9.679546610529267e-06,
+      "loss": 0.0573,
+      "step": 494
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5632376819736666,
+      "learning_rate": 9.677914860817476e-06,
+      "loss": 0.0851,
+      "step": 495
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5931974641813189,
+      "learning_rate": 9.676279105460567e-06,
+      "loss": 0.0908,
+      "step": 496
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6363861396511807,
+      "learning_rate": 9.674639345859213e-06,
+      "loss": 0.1158,
+      "step": 497
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4182628237754652,
+      "learning_rate": 9.672995583417526e-06,
+      "loss": 0.0883,
+      "step": 498
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5090871258925077,
+      "learning_rate": 9.671347819543039e-06,
+      "loss": 0.1205,
+      "step": 499
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1133692096561043,
+      "learning_rate": 9.669696055646713e-06,
+      "loss": 0.0881,
+      "step": 500
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4833512750349395,
+      "learning_rate": 9.668040293142937e-06,
+      "loss": 0.1377,
+      "step": 501
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3457146551502528,
+      "learning_rate": 9.666380533449517e-06,
+      "loss": 0.0589,
+      "step": 502
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.50304511972244,
+      "learning_rate": 9.664716777987691e-06,
+      "loss": 0.1298,
+      "step": 503
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4533229168514692,
+      "learning_rate": 9.663049028182112e-06,
+      "loss": 0.0664,
+      "step": 504
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5655907790886503,
+      "learning_rate": 9.661377285460856e-06,
+      "loss": 0.1235,
+      "step": 505
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.716552282403697,
+      "learning_rate": 9.659701551255415e-06,
+      "loss": 0.1111,
+      "step": 506
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7380319278342574,
+      "learning_rate": 9.658021827000702e-06,
+      "loss": 0.1298,
+      "step": 507
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.371408881428638,
+      "learning_rate": 9.656338114135045e-06,
+      "loss": 0.0718,
+      "step": 508
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.587187527009652,
+      "learning_rate": 9.65465041410019e-06,
+      "loss": 0.1381,
+      "step": 509
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6578247223161997,
+      "learning_rate": 9.652958728341296e-06,
+      "loss": 0.1205,
+      "step": 510
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.39107218775046754,
+      "learning_rate": 9.651263058306932e-06,
+      "loss": 0.1052,
+      "step": 511
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6450822857601923,
+      "learning_rate": 9.64956340544908e-06,
+      "loss": 0.0798,
+      "step": 512
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8606836095404234,
+      "learning_rate": 9.647859771223137e-06,
+      "loss": 0.1416,
+      "step": 513
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5190929098082754,
+      "learning_rate": 9.646152157087903e-06,
+      "loss": 0.1125,
+      "step": 514
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4424401320689641,
+      "learning_rate": 9.644440564505589e-06,
+      "loss": 0.1151,
+      "step": 515
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1476742980713732,
+      "learning_rate": 9.64272499494181e-06,
+      "loss": 0.109,
+      "step": 516
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5946358724686177,
+      "learning_rate": 9.641005449865593e-06,
+      "loss": 0.1347,
+      "step": 517
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.76152838679737,
+      "learning_rate": 9.639281930749363e-06,
+      "loss": 0.1563,
+      "step": 518
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.47081786131225006,
+      "learning_rate": 9.637554439068949e-06,
+      "loss": 0.0578,
+      "step": 519
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1990414757673342,
+      "learning_rate": 9.635822976303582e-06,
+      "loss": 0.1554,
+      "step": 520
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.448209401779092,
+      "learning_rate": 9.634087543935894e-06,
+      "loss": 0.1159,
+      "step": 521
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4996623490869529,
+      "learning_rate": 9.632348143451918e-06,
+      "loss": 0.106,
+      "step": 522
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7512792764643562,
+      "learning_rate": 9.630604776341082e-06,
+      "loss": 0.0803,
+      "step": 523
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3930266590077223,
+      "learning_rate": 9.628857444096209e-06,
+      "loss": 0.074,
+      "step": 524
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3841164058938006,
+      "learning_rate": 9.627106148213521e-06,
+      "loss": 0.0724,
+      "step": 525
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0102148139482265,
+      "learning_rate": 9.625350890192634e-06,
+      "loss": 0.178,
+      "step": 526
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8634464485184761,
+      "learning_rate": 9.623591671536554e-06,
+      "loss": 0.1394,
+      "step": 527
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.33868117647467416,
+      "learning_rate": 9.621828493751681e-06,
+      "loss": 0.0795,
+      "step": 528
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5374296491118014,
+      "learning_rate": 9.620061358347805e-06,
+      "loss": 0.1129,
+      "step": 529
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5393537120310781,
+      "learning_rate": 9.6182902668381e-06,
+      "loss": 0.1291,
+      "step": 530
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6608501322157044,
+      "learning_rate": 9.616515220739134e-06,
+      "loss": 0.1484,
+      "step": 531
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8839234605871841,
+      "learning_rate": 9.61473622157086e-06,
+      "loss": 0.1588,
+      "step": 532
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.48316007077989875,
+      "learning_rate": 9.61295327085661e-06,
+      "loss": 0.1078,
+      "step": 533
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8887117248606631,
+      "learning_rate": 9.611166370123109e-06,
+      "loss": 0.1719,
+      "step": 534
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4890278731456642,
+      "learning_rate": 9.60937552090046e-06,
+      "loss": 0.1185,
+      "step": 535
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2707751582206246,
+      "learning_rate": 9.607580724722141e-06,
+      "loss": 0.1571,
+      "step": 536
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.42484621527305594,
+      "learning_rate": 9.605781983125023e-06,
+      "loss": 0.0993,
+      "step": 537
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7346021626296555,
+      "learning_rate": 9.603979297649344e-06,
+      "loss": 0.1441,
+      "step": 538
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.44585308353686126,
+      "learning_rate": 9.602172669838721e-06,
+      "loss": 0.061,
+      "step": 539
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.34525815165431484,
+      "learning_rate": 9.600362101240153e-06,
+      "loss": 0.11,
+      "step": 540
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6344262740370257,
+      "learning_rate": 9.598547593404007e-06,
+      "loss": 0.1526,
+      "step": 541
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.393301102748828,
+      "learning_rate": 9.596729147884024e-06,
+      "loss": 0.0947,
+      "step": 542
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.776761271158952,
+      "learning_rate": 9.594906766237322e-06,
+      "loss": 0.102,
+      "step": 543
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3314025179857403,
+      "learning_rate": 9.593080450024382e-06,
+      "loss": 0.0585,
+      "step": 544
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5065702669137127,
+      "learning_rate": 9.591250200809061e-06,
+      "loss": 0.1066,
+      "step": 545
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7828868868785803,
+      "learning_rate": 9.589416020158577e-06,
+      "loss": 0.1324,
+      "step": 546
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5464686564794162,
+      "learning_rate": 9.58757790964352e-06,
+      "loss": 0.0948,
+      "step": 547
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.564572501433098,
+      "learning_rate": 9.585735870837842e-06,
+      "loss": 0.0813,
+      "step": 548
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.45973238827855767,
+      "learning_rate": 9.583889905318863e-06,
+      "loss": 0.0839,
+      "step": 549
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6380586075479202,
+      "learning_rate": 9.582040014667258e-06,
+      "loss": 0.0813,
+      "step": 550
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3444860266036195,
+      "learning_rate": 9.58018620046707e-06,
+      "loss": 0.0741,
+      "step": 551
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8103692222613961,
+      "learning_rate": 9.578328464305697e-06,
+      "loss": 0.1618,
+      "step": 552
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.30131060469646403,
+      "learning_rate": 9.5764668077739e-06,
+      "loss": 0.0595,
+      "step": 553
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7401579131128008,
+      "learning_rate": 9.57460123246579e-06,
+      "loss": 0.1484,
+      "step": 554
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5854614695945147,
+      "learning_rate": 9.57273173997884e-06,
+      "loss": 0.0956,
+      "step": 555
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4661820227358909,
+      "learning_rate": 9.570858331913877e-06,
+      "loss": 0.067,
+      "step": 556
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5485571108713299,
+      "learning_rate": 9.568981009875078e-06,
+      "loss": 0.1233,
+      "step": 557
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4455101604075415,
+      "learning_rate": 9.56709977546997e-06,
+      "loss": 0.1008,
+      "step": 558
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.46579763917353784,
+      "learning_rate": 9.565214630309437e-06,
+      "loss": 0.1065,
+      "step": 559
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8301614026074754,
+      "learning_rate": 9.563325576007702e-06,
+      "loss": 0.1845,
+      "step": 560
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4677538331059423,
+      "learning_rate": 9.561432614182344e-06,
+      "loss": 0.1058,
+      "step": 561
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7161536191110569,
+      "learning_rate": 9.559535746454285e-06,
+      "loss": 0.1176,
+      "step": 562
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.48154650097884616,
+      "learning_rate": 9.55763497444779e-06,
+      "loss": 0.08,
+      "step": 563
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.411701023144594,
+      "learning_rate": 9.555730299790467e-06,
+      "loss": 0.0787,
+      "step": 564
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7955992073989979,
+      "learning_rate": 9.553821724113268e-06,
+      "loss": 0.0849,
+      "step": 565
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4804114124676437,
+      "learning_rate": 9.551909249050489e-06,
+      "loss": 0.1282,
+      "step": 566
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5893379062898435,
+      "learning_rate": 9.549992876239753e-06,
+      "loss": 0.1583,
+      "step": 567
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.781153614929025,
+      "learning_rate": 9.548072607322035e-06,
+      "loss": 0.1309,
+      "step": 568
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.092893497688046,
+      "learning_rate": 9.546148443941634e-06,
+      "loss": 0.1531,
+      "step": 569
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.931755470555799,
+      "learning_rate": 9.544220387746193e-06,
+      "loss": 0.1008,
+      "step": 570
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.37262268518549607,
+      "learning_rate": 9.54228844038668e-06,
+      "loss": 0.0809,
+      "step": 571
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8229771839333507,
+      "learning_rate": 9.540352603517407e-06,
+      "loss": 0.1297,
+      "step": 572
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.42212543135085095,
+      "learning_rate": 9.538412878796001e-06,
+      "loss": 0.0814,
+      "step": 573
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7491251188140415,
+      "learning_rate": 9.536469267883432e-06,
+      "loss": 0.1239,
+      "step": 574
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5777544621930815,
+      "learning_rate": 9.534521772443989e-06,
+      "loss": 0.1275,
+      "step": 575
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4013350836379305,
+      "learning_rate": 9.532570394145292e-06,
+      "loss": 0.0945,
+      "step": 576
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7159831233187581,
+      "learning_rate": 9.530615134658278e-06,
+      "loss": 0.0986,
+      "step": 577
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.40874424417924476,
+      "learning_rate": 9.528655995657222e-06,
+      "loss": 0.063,
+      "step": 578
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.48580392186182875,
+      "learning_rate": 9.526692978819706e-06,
+      "loss": 0.102,
+      "step": 579
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5600308979814005,
+      "learning_rate": 9.524726085826645e-06,
+      "loss": 0.0907,
+      "step": 580
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6358393813816943,
+      "learning_rate": 9.52275531836226e-06,
+      "loss": 0.1216,
+      "step": 581
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5206580267327036,
+      "learning_rate": 9.5207806781141e-06,
+      "loss": 0.0985,
+      "step": 582
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3799897525179568,
+      "learning_rate": 9.518802166773028e-06,
+      "loss": 0.0885,
+      "step": 583
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2813068295543023,
+      "learning_rate": 9.51681978603322e-06,
+      "loss": 0.0508,
+      "step": 584
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3522678471071529,
+      "learning_rate": 9.514833537592167e-06,
+      "loss": 0.07,
+      "step": 585
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.45470245267937753,
+      "learning_rate": 9.512843423150666e-06,
+      "loss": 0.0813,
+      "step": 586
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6978287764047405,
+      "learning_rate": 9.510849444412835e-06,
+      "loss": 0.1432,
+      "step": 587
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5919384658602628,
+      "learning_rate": 9.508851603086094e-06,
+      "loss": 0.0984,
+      "step": 588
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4508118904856478,
+      "learning_rate": 9.506849900881169e-06,
+      "loss": 0.1059,
+      "step": 589
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7844312756182817,
+      "learning_rate": 9.504844339512096e-06,
+      "loss": 0.1245,
+      "step": 590
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.35067890977456656,
+      "learning_rate": 9.502834920696216e-06,
+      "loss": 0.0813,
+      "step": 591
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4916953871743061,
+      "learning_rate": 9.50082164615417e-06,
+      "loss": 0.073,
+      "step": 592
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3728444034470237,
+      "learning_rate": 9.498804517609903e-06,
+      "loss": 0.0875,
+      "step": 593
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.35651214288355376,
+      "learning_rate": 9.496783536790658e-06,
+      "loss": 0.0765,
+      "step": 594
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.31375965590627275,
+      "learning_rate": 9.494758705426978e-06,
+      "loss": 0.0528,
+      "step": 595
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5356005361879115,
+      "learning_rate": 9.492730025252704e-06,
+      "loss": 0.0594,
+      "step": 596
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5648101065684547,
+      "learning_rate": 9.490697498004968e-06,
+      "loss": 0.1364,
+      "step": 597
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7114698023455776,
+      "learning_rate": 9.488661125424205e-06,
+      "loss": 0.1575,
+      "step": 598
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6531081938971721,
+      "learning_rate": 9.486620909254136e-06,
+      "loss": 0.1356,
+      "step": 599
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7131636474188195,
+      "learning_rate": 9.484576851241774e-06,
+      "loss": 0.1292,
+      "step": 600
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.35752307596870525,
+      "learning_rate": 9.482528953137422e-06,
+      "loss": 0.0472,
+      "step": 601
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5298837090546342,
+      "learning_rate": 9.480477216694674e-06,
+      "loss": 0.0987,
+      "step": 602
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5747005138242258,
+      "learning_rate": 9.478421643670406e-06,
+      "loss": 0.1086,
+      "step": 603
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.45450892628952655,
+      "learning_rate": 9.476362235824783e-06,
+      "loss": 0.0609,
+      "step": 604
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.45746596718396365,
+      "learning_rate": 9.474298994921252e-06,
+      "loss": 0.1131,
+      "step": 605
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.67000800703229,
+      "learning_rate": 9.472231922726544e-06,
+      "loss": 0.1091,
+      "step": 606
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.47934571718907415,
+      "learning_rate": 9.470161021010667e-06,
+      "loss": 0.1048,
+      "step": 607
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4194734317041921,
+      "learning_rate": 9.468086291546913e-06,
+      "loss": 0.1048,
+      "step": 608
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6524051051139869,
+      "learning_rate": 9.466007736111846e-06,
+      "loss": 0.1262,
+      "step": 609
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6188530913371063,
+      "learning_rate": 9.463925356485313e-06,
+      "loss": 0.139,
+      "step": 610
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5035072039530757,
+      "learning_rate": 9.46183915445043e-06,
+      "loss": 0.1044,
+      "step": 611
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.41982057323405875,
+      "learning_rate": 9.459749131793588e-06,
+      "loss": 0.1046,
+      "step": 612
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4670070943276553,
+      "learning_rate": 9.457655290304449e-06,
+      "loss": 0.0803,
+      "step": 613
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5615915375373741,
+      "learning_rate": 9.455557631775946e-06,
+      "loss": 0.1195,
+      "step": 614
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4503642746310023,
+      "learning_rate": 9.45345615800428e-06,
+      "loss": 0.0769,
+      "step": 615
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.36719871467858317,
+      "learning_rate": 9.451350870788922e-06,
+      "loss": 0.0987,
+      "step": 616
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3667326749266243,
+      "learning_rate": 9.4492417719326e-06,
+      "loss": 0.0801,
+      "step": 617
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3488566704855922,
+      "learning_rate": 9.447128863241315e-06,
+      "loss": 0.1018,
+      "step": 618
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6368301632708415,
+      "learning_rate": 9.445012146524326e-06,
+      "loss": 0.1721,
+      "step": 619
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.47525937384417927,
+      "learning_rate": 9.442891623594153e-06,
+      "loss": 0.1081,
+      "step": 620
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.33093487651641074,
+      "learning_rate": 9.440767296266576e-06,
+      "loss": 0.0706,
+      "step": 621
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.36351252074473267,
+      "learning_rate": 9.43863916636063e-06,
+      "loss": 0.0473,
+      "step": 622
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.530672538411266,
+      "learning_rate": 9.436507235698613e-06,
+      "loss": 0.1207,
+      "step": 623
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7421951703191766,
+      "learning_rate": 9.434371506106068e-06,
+      "loss": 0.1246,
+      "step": 624
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5299383321675147,
+      "learning_rate": 9.432231979411799e-06,
+      "loss": 0.0752,
+      "step": 625
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5208153217297306,
+      "learning_rate": 9.430088657447856e-06,
+      "loss": 0.1453,
+      "step": 626
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4653774048923359,
+      "learning_rate": 9.427941542049543e-06,
+      "loss": 0.1156,
+      "step": 627
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.43697657357216074,
+      "learning_rate": 9.42579063505541e-06,
+      "loss": 0.0883,
+      "step": 628
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6115594859303495,
+      "learning_rate": 9.423635938307256e-06,
+      "loss": 0.1428,
+      "step": 629
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.41136842035064647,
+      "learning_rate": 9.421477453650118e-06,
+      "loss": 0.0984,
+      "step": 630
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4929570243812377,
+      "learning_rate": 9.419315182932287e-06,
+      "loss": 0.0927,
+      "step": 631
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5358411779767442,
+      "learning_rate": 9.417149128005288e-06,
+      "loss": 0.0679,
+      "step": 632
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5663438385482606,
+      "learning_rate": 9.414979290723891e-06,
+      "loss": 0.1458,
+      "step": 633
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3177486118184364,
+      "learning_rate": 9.412805672946102e-06,
+      "loss": 0.055,
+      "step": 634
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5447180871590981,
+      "learning_rate": 9.410628276533163e-06,
+      "loss": 0.098,
+      "step": 635
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.48325418160280365,
+      "learning_rate": 9.408447103349556e-06,
+      "loss": 0.0826,
+      "step": 636
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5790689085859455,
+      "learning_rate": 9.406262155262995e-06,
+      "loss": 0.0737,
+      "step": 637
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4039519034914288,
+      "learning_rate": 9.404073434144424e-06,
+      "loss": 0.0809,
+      "step": 638
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4461995600952947,
+      "learning_rate": 9.401880941868024e-06,
+      "loss": 0.1006,
+      "step": 639
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.45319261228159363,
+      "learning_rate": 9.399684680311197e-06,
+      "loss": 0.1137,
+      "step": 640
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7312794169976998,
+      "learning_rate": 9.397484651354578e-06,
+      "loss": 0.0816,
+      "step": 641
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.42358036696294604,
+      "learning_rate": 9.395280856882026e-06,
+      "loss": 0.1044,
+      "step": 642
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7109159286986856,
+      "learning_rate": 9.393073298780627e-06,
+      "loss": 0.104,
+      "step": 643
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2283321580262665,
+      "learning_rate": 9.390861978940687e-06,
+      "loss": 0.058,
+      "step": 644
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5407706797610025,
+      "learning_rate": 9.388646899255733e-06,
+      "loss": 0.0865,
+      "step": 645
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5305289253000061,
+      "learning_rate": 9.386428061622513e-06,
+      "loss": 0.1291,
+      "step": 646
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4608935788913227,
+      "learning_rate": 9.384205467940993e-06,
+      "loss": 0.1044,
+      "step": 647
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4187809752103644,
+      "learning_rate": 9.381979120114354e-06,
+      "loss": 0.0986,
+      "step": 648
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.40910352465299,
+      "learning_rate": 9.379749020048992e-06,
+      "loss": 0.1548,
+      "step": 649
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4904574292595726,
+      "learning_rate": 9.377515169654518e-06,
+      "loss": 0.0779,
+      "step": 650
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.27826549346693685,
+      "learning_rate": 9.37527757084375e-06,
+      "loss": 0.0335,
+      "step": 651
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8832023529321562,
+      "learning_rate": 9.373036225532719e-06,
+      "loss": 0.0866,
+      "step": 652
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8272570531872325,
+      "learning_rate": 9.370791135640665e-06,
+      "loss": 0.1265,
+      "step": 653
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.683414814615958,
+      "learning_rate": 9.368542303090032e-06,
+      "loss": 0.1107,
+      "step": 654
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4537591785209198,
+      "learning_rate": 9.366289729806468e-06,
+      "loss": 0.0909,
+      "step": 655
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3611325743188263,
+      "learning_rate": 9.36403341771883e-06,
+      "loss": 0.0776,
+      "step": 656
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7715225557983743,
+      "learning_rate": 9.361773368759167e-06,
+      "loss": 0.1438,
+      "step": 657
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8229300546981989,
+      "learning_rate": 9.359509584862735e-06,
+      "loss": 0.0597,
+      "step": 658
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.39645132593923904,
+      "learning_rate": 9.35724206796799e-06,
+      "loss": 0.0917,
+      "step": 659
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7286126262325562,
+      "learning_rate": 9.354970820016576e-06,
+      "loss": 0.0756,
+      "step": 660
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5107501028216207,
+      "learning_rate": 9.35269584295334e-06,
+      "loss": 0.0908,
+      "step": 661
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5590486582264846,
+      "learning_rate": 9.350417138726315e-06,
+      "loss": 0.1172,
+      "step": 662
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0684748553819996,
+      "learning_rate": 9.348134709286732e-06,
+      "loss": 0.1464,
+      "step": 663
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.308743027902976,
+      "learning_rate": 9.345848556589011e-06,
+      "loss": 0.0602,
+      "step": 664
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.556289940922841,
+      "learning_rate": 9.343558682590757e-06,
+      "loss": 0.1205,
+      "step": 665
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4636388866468853,
+      "learning_rate": 9.34126508925276e-06,
+      "loss": 0.1141,
+      "step": 666
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2961531246516484,
+      "learning_rate": 9.338967778539003e-06,
+      "loss": 0.0587,
+      "step": 667
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6908136801369276,
+      "learning_rate": 9.336666752416643e-06,
+      "loss": 0.1248,
+      "step": 668
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7189385091006617,
+      "learning_rate": 9.334362012856025e-06,
+      "loss": 0.0899,
+      "step": 669
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.34565430257336976,
+      "learning_rate": 9.332053561830669e-06,
+      "loss": 0.0779,
+      "step": 670
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4313753350170771,
+      "learning_rate": 9.329741401317275e-06,
+      "loss": 0.0497,
+      "step": 671
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5036708096897136,
+      "learning_rate": 9.327425533295725e-06,
+      "loss": 0.078,
+      "step": 672
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6683806773452055,
+      "learning_rate": 9.325105959749062e-06,
+      "loss": 0.1216,
+      "step": 673
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6964085263385951,
+      "learning_rate": 9.322782682663516e-06,
+      "loss": 0.1461,
+      "step": 674
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.306454261021034,
+      "learning_rate": 9.320455704028482e-06,
+      "loss": 0.0757,
+      "step": 675
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7014155264403084,
+      "learning_rate": 9.318125025836524e-06,
+      "loss": 0.1415,
+      "step": 676
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3516675920524297,
+      "learning_rate": 9.315790650083376e-06,
+      "loss": 0.0648,
+      "step": 677
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.20928109827686087,
+      "learning_rate": 9.313452578767937e-06,
+      "loss": 0.0372,
+      "step": 678
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5061408737680055,
+      "learning_rate": 9.31111081389227e-06,
+      "loss": 0.0905,
+      "step": 679
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5234568909313362,
+      "learning_rate": 9.308765357461604e-06,
+      "loss": 0.0942,
+      "step": 680
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5961880907562055,
+      "learning_rate": 9.306416211484323e-06,
+      "loss": 0.0715,
+      "step": 681
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4436182651534206,
+      "learning_rate": 9.304063377971978e-06,
+      "loss": 0.1086,
+      "step": 682
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8996073654251279,
+      "learning_rate": 9.30170685893927e-06,
+      "loss": 0.1333,
+      "step": 683
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.46052297409844706,
+      "learning_rate": 9.299346656404061e-06,
+      "loss": 0.0797,
+      "step": 684
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6021166938690489,
+      "learning_rate": 9.296982772387366e-06,
+      "loss": 0.1053,
+      "step": 685
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.2674255269406825,
+      "learning_rate": 9.29461520891335e-06,
+      "loss": 0.0663,
+      "step": 686
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3045359359469913,
+      "learning_rate": 9.292243968009332e-06,
+      "loss": 0.0763,
+      "step": 687
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5834508891312872,
+      "learning_rate": 9.289869051705777e-06,
+      "loss": 0.1361,
+      "step": 688
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6611334537227191,
+      "learning_rate": 9.287490462036301e-06,
+      "loss": 0.1085,
+      "step": 689
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5636213905563565,
+      "learning_rate": 9.285108201037663e-06,
+      "loss": 0.0558,
+      "step": 690
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7057820765338262,
+      "learning_rate": 9.282722270749764e-06,
+      "loss": 0.107,
+      "step": 691
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8620729301203439,
+      "learning_rate": 9.280332673215651e-06,
+      "loss": 0.1444,
+      "step": 692
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6183703632004979,
+      "learning_rate": 9.277939410481507e-06,
+      "loss": 0.099,
+      "step": 693
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7057134143581547,
+      "learning_rate": 9.275542484596658e-06,
+      "loss": 0.1041,
+      "step": 694
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.2628334930423759,
+      "learning_rate": 9.27314189761356e-06,
+      "loss": 0.0642,
+      "step": 695
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6088153193058466,
+      "learning_rate": 9.270737651587813e-06,
+      "loss": 0.1295,
+      "step": 696
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.744151412836345,
+      "learning_rate": 9.268329748578144e-06,
+      "loss": 0.1621,
+      "step": 697
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5618375618090972,
+      "learning_rate": 9.265918190646413e-06,
+      "loss": 0.1311,
+      "step": 698
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6350495934711705,
+      "learning_rate": 9.263502979857608e-06,
+      "loss": 0.1249,
+      "step": 699
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7429219334181577,
+      "learning_rate": 9.261084118279846e-06,
+      "loss": 0.1236,
+      "step": 700
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6834825143828912,
+      "learning_rate": 9.258661607984374e-06,
+      "loss": 0.1622,
+      "step": 701
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4665496520433196,
+      "learning_rate": 9.256235451045558e-06,
+      "loss": 0.0687,
+      "step": 702
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.43041703878012194,
+      "learning_rate": 9.253805649540888e-06,
+      "loss": 0.0915,
+      "step": 703
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4323839035861413,
+      "learning_rate": 9.251372205550975e-06,
+      "loss": 0.0915,
+      "step": 704
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5734425170958729,
+      "learning_rate": 9.248935121159552e-06,
+      "loss": 0.1654,
+      "step": 705
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.37270001001552855,
+      "learning_rate": 9.246494398453462e-06,
+      "loss": 0.0889,
+      "step": 706
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4087403745286263,
+      "learning_rate": 9.244050039522673e-06,
+      "loss": 0.1161,
+      "step": 707
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.37248115377091445,
+      "learning_rate": 9.241602046460259e-06,
+      "loss": 0.0912,
+      "step": 708
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3926233305358984,
+      "learning_rate": 9.239150421362409e-06,
+      "loss": 0.0778,
+      "step": 709
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.43737638808273627,
+      "learning_rate": 9.23669516632842e-06,
+      "loss": 0.0643,
+      "step": 710
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.684772974774738,
+      "learning_rate": 9.2342362834607e-06,
+      "loss": 0.1126,
+      "step": 711
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5215146223231149,
+      "learning_rate": 9.231773774864764e-06,
+      "loss": 0.1385,
+      "step": 712
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5365368655609294,
+      "learning_rate": 9.229307642649227e-06,
+      "loss": 0.0933,
+      "step": 713
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4725361435553599,
+      "learning_rate": 9.226837888925813e-06,
+      "loss": 0.0765,
+      "step": 714
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3904654870756912,
+      "learning_rate": 9.224364515809344e-06,
+      "loss": 0.103,
+      "step": 715
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1742846534420257,
+      "learning_rate": 9.221887525417737e-06,
+      "loss": 0.1431,
+      "step": 716
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7353839708273526,
+      "learning_rate": 9.219406919872014e-06,
+      "loss": 0.1401,
+      "step": 717
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3523709264724253,
+      "learning_rate": 9.216922701296289e-06,
+      "loss": 0.0586,
+      "step": 718
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3411766976346362,
+      "learning_rate": 9.214434871817769e-06,
+      "loss": 0.0769,
+      "step": 719
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2009156501089908,
+      "learning_rate": 9.211943433566755e-06,
+      "loss": 0.1282,
+      "step": 720
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5582439527709884,
+      "learning_rate": 9.209448388676636e-06,
+      "loss": 0.1446,
+      "step": 721
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4252995547914209,
+      "learning_rate": 9.20694973928389e-06,
+      "loss": 0.0921,
+      "step": 722
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.37333649499305366,
+      "learning_rate": 9.204447487528085e-06,
+      "loss": 0.0922,
+      "step": 723
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3364452080789254,
+      "learning_rate": 9.20194163555187e-06,
+      "loss": 0.0688,
+      "step": 724
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.841864169024913,
+      "learning_rate": 9.199432185500972e-06,
+      "loss": 0.1456,
+      "step": 725
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6276528959560425,
+      "learning_rate": 9.196919139524213e-06,
+      "loss": 0.0969,
+      "step": 726
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6504575615211694,
+      "learning_rate": 9.194402499773478e-06,
+      "loss": 0.1332,
+      "step": 727
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.189817192486516,
+      "learning_rate": 9.191882268403743e-06,
+      "loss": 0.1409,
+      "step": 728
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5374540414308913,
+      "learning_rate": 9.189358447573047e-06,
+      "loss": 0.1292,
+      "step": 729
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3552774265718138,
+      "learning_rate": 9.186831039442514e-06,
+      "loss": 0.0677,
+      "step": 730
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6988232672750744,
+      "learning_rate": 9.184300046176333e-06,
+      "loss": 0.1037,
+      "step": 731
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5337454702787745,
+      "learning_rate": 9.181765469941765e-06,
+      "loss": 0.1033,
+      "step": 732
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5276214220760512,
+      "learning_rate": 9.17922731290914e-06,
+      "loss": 0.1027,
+      "step": 733
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.35890532134151637,
+      "learning_rate": 9.176685577251851e-06,
+      "loss": 0.0808,
+      "step": 734
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4825929864132039,
+      "learning_rate": 9.174140265146356e-06,
+      "loss": 0.1101,
+      "step": 735
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7291551338337156,
+      "learning_rate": 9.17159137877218e-06,
+      "loss": 0.17,
+      "step": 736
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.750892141102736,
+      "learning_rate": 9.169038920311904e-06,
+      "loss": 0.1176,
+      "step": 737
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6081493782398585,
+      "learning_rate": 9.166482891951167e-06,
+      "loss": 0.1273,
+      "step": 738
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7329853233032924,
+      "learning_rate": 9.163923295878671e-06,
+      "loss": 0.1187,
+      "step": 739
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4699516253473333,
+      "learning_rate": 9.161360134286166e-06,
+      "loss": 0.0968,
+      "step": 740
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3577997147083906,
+      "learning_rate": 9.158793409368457e-06,
+      "loss": 0.0828,
+      "step": 741
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5160848593239924,
+      "learning_rate": 9.156223123323405e-06,
+      "loss": 0.1055,
+      "step": 742
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5661886219563604,
+      "learning_rate": 9.153649278351912e-06,
+      "loss": 0.1159,
+      "step": 743
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5506370893870068,
+      "learning_rate": 9.151071876657938e-06,
+      "loss": 0.0952,
+      "step": 744
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.49667306715332216,
+      "learning_rate": 9.148490920448476e-06,
+      "loss": 0.1312,
+      "step": 745
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6028460999658366,
+      "learning_rate": 9.145906411933576e-06,
+      "loss": 0.1018,
+      "step": 746
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7536199812633009,
+      "learning_rate": 9.143318353326316e-06,
+      "loss": 0.1419,
+      "step": 747
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3783696079417029,
+      "learning_rate": 9.140726746842827e-06,
+      "loss": 0.0511,
+      "step": 748
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4995011041289576,
+      "learning_rate": 9.13813159470227e-06,
+      "loss": 0.1112,
+      "step": 749
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5643956899537229,
+      "learning_rate": 9.135532899126844e-06,
+      "loss": 0.1254,
+      "step": 750
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.42022233300605943,
+      "learning_rate": 9.132930662341783e-06,
+      "loss": 0.0708,
+      "step": 751
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5380086554905054,
+      "learning_rate": 9.130324886575351e-06,
+      "loss": 0.123,
+      "step": 752
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4400408465782521,
+      "learning_rate": 9.127715574058847e-06,
+      "loss": 0.0716,
+      "step": 753
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.37534544197208,
+      "learning_rate": 9.125102727026592e-06,
+      "loss": 0.1099,
+      "step": 754
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6991202727796607,
+      "learning_rate": 9.122486347715937e-06,
+      "loss": 0.1367,
+      "step": 755
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.41246221387968546,
+      "learning_rate": 9.119866438367263e-06,
+      "loss": 0.0825,
+      "step": 756
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.2983856613021006,
+      "learning_rate": 9.117243001223963e-06,
+      "loss": 0.0793,
+      "step": 757
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4286379751579008,
+      "learning_rate": 9.11461603853246e-06,
+      "loss": 0.0718,
+      "step": 758
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.30316378745638267,
+      "learning_rate": 9.111985552542188e-06,
+      "loss": 0.0703,
+      "step": 759
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.36592788858390773,
+      "learning_rate": 9.109351545505607e-06,
+      "loss": 0.0647,
+      "step": 760
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5676181326297737,
+      "learning_rate": 9.106714019678185e-06,
+      "loss": 0.0779,
+      "step": 761
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4835641579608436,
+      "learning_rate": 9.104072977318403e-06,
+      "loss": 0.0641,
+      "step": 762
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5508660668429857,
+      "learning_rate": 9.101428420687759e-06,
+      "loss": 0.0954,
+      "step": 763
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.36934478548767896,
+      "learning_rate": 9.098780352050756e-06,
+      "loss": 0.083,
+      "step": 764
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8650730665456191,
+      "learning_rate": 9.096128773674902e-06,
+      "loss": 0.1289,
+      "step": 765
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.48078309658654167,
+      "learning_rate": 9.093473687830718e-06,
+      "loss": 0.1137,
+      "step": 766
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.556726383146375,
+      "learning_rate": 9.090815096791719e-06,
+      "loss": 0.1399,
+      "step": 767
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5062090329432174,
+      "learning_rate": 9.08815300283443e-06,
+      "loss": 0.1002,
+      "step": 768
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3857594535941769,
+      "learning_rate": 9.08548740823837e-06,
+      "loss": 0.0979,
+      "step": 769
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.42728429643783955,
+      "learning_rate": 9.082818315286054e-06,
+      "loss": 0.0873,
+      "step": 770
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4544981939951406,
+      "learning_rate": 9.080145726263003e-06,
+      "loss": 0.0838,
+      "step": 771
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6013625857886737,
+      "learning_rate": 9.077469643457719e-06,
+      "loss": 0.1116,
+      "step": 772
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6236641276359599,
+      "learning_rate": 9.074790069161703e-06,
+      "loss": 0.1359,
+      "step": 773
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5043621989493956,
+      "learning_rate": 9.072107005669444e-06,
+      "loss": 0.1426,
+      "step": 774
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4141078190743088,
+      "learning_rate": 9.069420455278418e-06,
+      "loss": 0.1124,
+      "step": 775
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3135133375118733,
+      "learning_rate": 9.066730420289088e-06,
+      "loss": 0.0566,
+      "step": 776
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.34710548595849217,
+      "learning_rate": 9.0640369030049e-06,
+      "loss": 0.0884,
+      "step": 777
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5349865562937722,
+      "learning_rate": 9.061339905732282e-06,
+      "loss": 0.1306,
+      "step": 778
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5997926521709435,
+      "learning_rate": 9.058639430780642e-06,
+      "loss": 0.1752,
+      "step": 779
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3936916454698406,
+      "learning_rate": 9.055935480462366e-06,
+      "loss": 0.0914,
+      "step": 780
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3228969309279008,
+      "learning_rate": 9.053228057092817e-06,
+      "loss": 0.0814,
+      "step": 781
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.24771035108866218,
+      "learning_rate": 9.050517162990329e-06,
+      "loss": 0.0531,
+      "step": 782
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.47770744193018433,
+      "learning_rate": 9.04780280047621e-06,
+      "loss": 0.1322,
+      "step": 783
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5372195573790007,
+      "learning_rate": 9.045084971874738e-06,
+      "loss": 0.0865,
+      "step": 784
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8220141399078558,
+      "learning_rate": 9.042363679513158e-06,
+      "loss": 0.1439,
+      "step": 785
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5064666522255065,
+      "learning_rate": 9.039638925721683e-06,
+      "loss": 0.1297,
+      "step": 786
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6144017014891907,
+      "learning_rate": 9.036910712833488e-06,
+      "loss": 0.1163,
+      "step": 787
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4727152509436765,
+      "learning_rate": 9.03417904318471e-06,
+      "loss": 0.1106,
+      "step": 788
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5433509786648136,
+      "learning_rate": 9.031443919114447e-06,
+      "loss": 0.1118,
+      "step": 789
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6010335333394724,
+      "learning_rate": 9.028705342964752e-06,
+      "loss": 0.139,
+      "step": 790
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.32402803259529556,
+      "learning_rate": 9.025963317080641e-06,
+      "loss": 0.0852,
+      "step": 791
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5616724571525001,
+      "learning_rate": 9.023217843810078e-06,
+      "loss": 0.1097,
+      "step": 792
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4277656347530322,
+      "learning_rate": 9.02046892550398e-06,
+      "loss": 0.0922,
+      "step": 793
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5865243436306229,
+      "learning_rate": 9.017716564516213e-06,
+      "loss": 0.0731,
+      "step": 794
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.389559883434402,
+      "learning_rate": 9.014960763203592e-06,
+      "loss": 0.0986,
+      "step": 795
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5123520630529207,
+      "learning_rate": 9.012201523925883e-06,
+      "loss": 0.083,
+      "step": 796
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6005092454859025,
+      "learning_rate": 9.009438849045787e-06,
+      "loss": 0.1422,
+      "step": 797
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7752913470234684,
+      "learning_rate": 9.006672740928952e-06,
+      "loss": 0.1155,
+      "step": 798
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.30730440396061093,
+      "learning_rate": 9.003903201943963e-06,
+      "loss": 0.0668,
+      "step": 799
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3973499918699411,
+      "learning_rate": 9.001130234462348e-06,
+      "loss": 0.1166,
+      "step": 800
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.43049515189178245,
+      "learning_rate": 8.998353840858565e-06,
+      "loss": 0.0933,
+      "step": 801
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6827994479548175,
+      "learning_rate": 8.995574023510007e-06,
+      "loss": 0.119,
+      "step": 802
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.48279321530860336,
+      "learning_rate": 8.992790784797004e-06,
+      "loss": 0.0967,
+      "step": 803
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5756268706713004,
+      "learning_rate": 8.99000412710281e-06,
+      "loss": 0.094,
+      "step": 804
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5182683702974817,
+      "learning_rate": 8.987214052813605e-06,
+      "loss": 0.1155,
+      "step": 805
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.42356209279728074,
+      "learning_rate": 8.984420564318501e-06,
+      "loss": 0.0938,
+      "step": 806
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.069788771610959,
+      "learning_rate": 8.98162366400953e-06,
+      "loss": 0.1668,
+      "step": 807
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4550416977253712,
+      "learning_rate": 8.978823354281645e-06,
+      "loss": 0.1007,
+      "step": 808
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7524921906573384,
+      "learning_rate": 8.97601963753272e-06,
+      "loss": 0.119,
+      "step": 809
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.32381408525164634,
+      "learning_rate": 8.973212516163545e-06,
+      "loss": 0.0755,
+      "step": 810
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7182312122919299,
+      "learning_rate": 8.970401992577828e-06,
+      "loss": 0.091,
+      "step": 811
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.505126853830724,
+      "learning_rate": 8.967588069182184e-06,
+      "loss": 0.0636,
+      "step": 812
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6576481777177007,
+      "learning_rate": 8.964770748386149e-06,
+      "loss": 0.1209,
+      "step": 813
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.692326398992742,
+      "learning_rate": 8.961950032602157e-06,
+      "loss": 0.2793,
+      "step": 814
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6519662848792224,
+      "learning_rate": 8.959125924245559e-06,
+      "loss": 0.1369,
+      "step": 815
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6731876604424843,
+      "learning_rate": 8.956298425734606e-06,
+      "loss": 0.113,
+      "step": 816
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.631265442700046,
+      "learning_rate": 8.95346753949045e-06,
+      "loss": 0.1409,
+      "step": 817
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4359598787752724,
+      "learning_rate": 8.95063326793715e-06,
+      "loss": 0.0561,
+      "step": 818
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5325726032748053,
+      "learning_rate": 8.947795613501658e-06,
+      "loss": 0.1352,
+      "step": 819
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.36788253328153797,
+      "learning_rate": 8.944954578613826e-06,
+      "loss": 0.1112,
+      "step": 820
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5167892458757593,
+      "learning_rate": 8.942110165706402e-06,
+      "loss": 0.0984,
+      "step": 821
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5279357960312524,
+      "learning_rate": 8.939262377215019e-06,
+      "loss": 0.1245,
+      "step": 822
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3902659964321454,
+      "learning_rate": 8.936411215578211e-06,
+      "loss": 0.097,
+      "step": 823
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4741580517118749,
+      "learning_rate": 8.933556683237392e-06,
+      "loss": 0.0967,
+      "step": 824
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5879684966050115,
+      "learning_rate": 8.930698782636868e-06,
+      "loss": 0.125,
+      "step": 825
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.295025135184848,
+      "learning_rate": 8.927837516223824e-06,
+      "loss": 0.0856,
+      "step": 826
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8211083633805236,
+      "learning_rate": 8.924972886448331e-06,
+      "loss": 0.1354,
+      "step": 827
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4820327564594707,
+      "learning_rate": 8.922104895763339e-06,
+      "loss": 0.1181,
+      "step": 828
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.36098422965878196,
+      "learning_rate": 8.919233546624677e-06,
+      "loss": 0.0807,
+      "step": 829
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3319217140421602,
+      "learning_rate": 8.916358841491046e-06,
+      "loss": 0.0769,
+      "step": 830
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.40274034029099054,
+      "learning_rate": 8.913480782824025e-06,
+      "loss": 0.1067,
+      "step": 831
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4767912541059932,
+      "learning_rate": 8.91059937308806e-06,
+      "loss": 0.1207,
+      "step": 832
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4908848791522766,
+      "learning_rate": 8.907714614750473e-06,
+      "loss": 0.0925,
+      "step": 833
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5120088266126959,
+      "learning_rate": 8.904826510281448e-06,
+      "loss": 0.1213,
+      "step": 834
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5826862448715896,
+      "learning_rate": 8.901935062154035e-06,
+      "loss": 0.1064,
+      "step": 835
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.45695283672301107,
+      "learning_rate": 8.899040272844149e-06,
+      "loss": 0.0915,
+      "step": 836
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.38119870235578857,
+      "learning_rate": 8.896142144830565e-06,
+      "loss": 0.0889,
+      "step": 837
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5054610878270641,
+      "learning_rate": 8.893240680594916e-06,
+      "loss": 0.1025,
+      "step": 838
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.44381009273730515,
+      "learning_rate": 8.890335882621697e-06,
+      "loss": 0.0896,
+      "step": 839
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.34372743762996427,
+      "learning_rate": 8.887427753398249e-06,
+      "loss": 0.0924,
+      "step": 840
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.588425547513867,
+      "learning_rate": 8.88451629541477e-06,
+      "loss": 0.1044,
+      "step": 841
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.45866727011355307,
+      "learning_rate": 8.881601511164306e-06,
+      "loss": 0.0641,
+      "step": 842
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5856834534904065,
+      "learning_rate": 8.87868340314276e-06,
+      "loss": 0.091,
+      "step": 843
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5321123695622559,
+      "learning_rate": 8.87576197384887e-06,
+      "loss": 0.0782,
+      "step": 844
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.379052329623936,
+      "learning_rate": 8.872837225784227e-06,
+      "loss": 0.0861,
+      "step": 845
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5868727341308448,
+      "learning_rate": 8.869909161453254e-06,
+      "loss": 0.1258,
+      "step": 846
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3688506033382923,
+      "learning_rate": 8.866977783363219e-06,
+      "loss": 0.0621,
+      "step": 847
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4016791149564159,
+      "learning_rate": 8.864043094024233e-06,
+      "loss": 0.0849,
+      "step": 848
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.43053240748250243,
+      "learning_rate": 8.861105095949234e-06,
+      "loss": 0.0921,
+      "step": 849
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.629742603869501,
+      "learning_rate": 8.858163791653994e-06,
+      "loss": 0.1201,
+      "step": 850
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.43952511190365295,
+      "learning_rate": 8.855219183657122e-06,
+      "loss": 0.079,
+      "step": 851
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5248966692778887,
+      "learning_rate": 8.85227127448005e-06,
+      "loss": 0.1004,
+      "step": 852
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4758828808847174,
+      "learning_rate": 8.84932006664704e-06,
+      "loss": 0.117,
+      "step": 853
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.36892765193767757,
+      "learning_rate": 8.846365562685178e-06,
+      "loss": 0.0968,
+      "step": 854
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5009491205345755,
+      "learning_rate": 8.84340776512437e-06,
+      "loss": 0.1131,
+      "step": 855
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9089159097110372,
+      "learning_rate": 8.840446676497344e-06,
+      "loss": 0.1165,
+      "step": 856
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4161456090989337,
+      "learning_rate": 8.837482299339652e-06,
+      "loss": 0.1069,
+      "step": 857
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0317820640217963,
+      "learning_rate": 8.83451463618965e-06,
+      "loss": 0.0809,
+      "step": 858
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5800774784418494,
+      "learning_rate": 8.831543689588517e-06,
+      "loss": 0.1071,
+      "step": 859
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4007033123605079,
+      "learning_rate": 8.82856946208024e-06,
+      "loss": 0.0969,
+      "step": 860
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3946324231265691,
+      "learning_rate": 8.825591956211614e-06,
+      "loss": 0.084,
+      "step": 861
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5111307803372946,
+      "learning_rate": 8.822611174532248e-06,
+      "loss": 0.1191,
+      "step": 862
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6711085048466602,
+      "learning_rate": 8.819627119594549e-06,
+      "loss": 0.11,
+      "step": 863
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6188807241146488,
+      "learning_rate": 8.816639793953727e-06,
+      "loss": 0.0849,
+      "step": 864
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6968074134324962,
+      "learning_rate": 8.8136492001678e-06,
+      "loss": 0.1275,
+      "step": 865
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5287857662766221,
+      "learning_rate": 8.810655340797574e-06,
+      "loss": 0.0807,
+      "step": 866
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3558440872750076,
+      "learning_rate": 8.807658218406658e-06,
+      "loss": 0.0755,
+      "step": 867
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3331007464909322,
+      "learning_rate": 8.804657835561456e-06,
+      "loss": 0.1036,
+      "step": 868
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4497950761853012,
+      "learning_rate": 8.801654194831159e-06,
+      "loss": 0.0757,
+      "step": 869
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5087848866630862,
+      "learning_rate": 8.798647298787754e-06,
+      "loss": 0.0893,
+      "step": 870
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.45894665768788273,
+      "learning_rate": 8.795637150006007e-06,
+      "loss": 0.1201,
+      "step": 871
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.354858690530361,
+      "learning_rate": 8.792623751063476e-06,
+      "loss": 0.0733,
+      "step": 872
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6719833618670081,
+      "learning_rate": 8.789607104540502e-06,
+      "loss": 0.16,
+      "step": 873
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.44781858187738477,
+      "learning_rate": 8.786587213020202e-06,
+      "loss": 0.0757,
+      "step": 874
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.45570009334354245,
+      "learning_rate": 8.783564079088478e-06,
+      "loss": 0.0983,
+      "step": 875
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5766552732083019,
+      "learning_rate": 8.780537705334e-06,
+      "loss": 0.0833,
+      "step": 876
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3320289562900106,
+      "learning_rate": 8.777508094348222e-06,
+      "loss": 0.0785,
+      "step": 877
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6086385199833315,
+      "learning_rate": 8.774475248725366e-06,
+      "loss": 0.1006,
+      "step": 878
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.48025884615280756,
+      "learning_rate": 8.771439171062417e-06,
+      "loss": 0.1355,
+      "step": 879
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4580674740736033,
+      "learning_rate": 8.76839986395914e-06,
+      "loss": 0.1165,
+      "step": 880
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3298155358437963,
+      "learning_rate": 8.765357330018056e-06,
+      "loss": 0.0939,
+      "step": 881
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3922558669302778,
+      "learning_rate": 8.762311571844453e-06,
+      "loss": 0.1046,
+      "step": 882
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7558459488933367,
+      "learning_rate": 8.759262592046378e-06,
+      "loss": 0.0952,
+      "step": 883
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4258296794792277,
+      "learning_rate": 8.756210393234636e-06,
+      "loss": 0.0799,
+      "step": 884
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3555262486853882,
+      "learning_rate": 8.753154978022795e-06,
+      "loss": 0.0826,
+      "step": 885
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7015025200179268,
+      "learning_rate": 8.75009634902717e-06,
+      "loss": 0.108,
+      "step": 886
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7530615840579308,
+      "learning_rate": 8.747034508866828e-06,
+      "loss": 0.21,
+      "step": 887
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5315933652038956,
+      "learning_rate": 8.74396946016359e-06,
+      "loss": 0.0863,
+      "step": 888
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0945944899723976,
+      "learning_rate": 8.74090120554202e-06,
+      "loss": 0.0902,
+      "step": 889
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5683446077431457,
+      "learning_rate": 8.737829747629432e-06,
+      "loss": 0.0809,
+      "step": 890
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.29614996842243607,
+      "learning_rate": 8.73475508905588e-06,
+      "loss": 0.0735,
+      "step": 891
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.46585249998082157,
+      "learning_rate": 8.731677232454159e-06,
+      "loss": 0.0861,
+      "step": 892
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.46383857614209745,
+      "learning_rate": 8.728596180459799e-06,
+      "loss": 0.0833,
+      "step": 893
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.387209190084108,
+      "learning_rate": 8.725511935711074e-06,
+      "loss": 0.0739,
+      "step": 894
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5959494360000862,
+      "learning_rate": 8.722424500848988e-06,
+      "loss": 0.1125,
+      "step": 895
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4229318299158801,
+      "learning_rate": 8.719333878517274e-06,
+      "loss": 0.117,
+      "step": 896
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4931293812890095,
+      "learning_rate": 8.716240071362394e-06,
+      "loss": 0.1337,
+      "step": 897
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.34618356814511986,
+      "learning_rate": 8.713143082033546e-06,
+      "loss": 0.0839,
+      "step": 898
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7007740491097011,
+      "learning_rate": 8.710042913182642e-06,
+      "loss": 0.1235,
+      "step": 899
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.39627229141713144,
+      "learning_rate": 8.706939567464322e-06,
+      "loss": 0.0882,
+      "step": 900
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.44060741286425664,
+      "learning_rate": 8.703833047535946e-06,
+      "loss": 0.1002,
+      "step": 901
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3488303699913442,
+      "learning_rate": 8.700723356057593e-06,
+      "loss": 0.0698,
+      "step": 902
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5795997255716634,
+      "learning_rate": 8.697610495692055e-06,
+      "loss": 0.1081,
+      "step": 903
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6542499486577853,
+      "learning_rate": 8.694494469104839e-06,
+      "loss": 0.0741,
+      "step": 904
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4638422976582731,
+      "learning_rate": 8.691375278964161e-06,
+      "loss": 0.0892,
+      "step": 905
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.263318870009058,
+      "learning_rate": 8.688252927940951e-06,
+      "loss": 0.0715,
+      "step": 906
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3974683063077847,
+      "learning_rate": 8.685127418708843e-06,
+      "loss": 0.0891,
+      "step": 907
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.49445402902436825,
+      "learning_rate": 8.68199875394417e-06,
+      "loss": 0.077,
+      "step": 908
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4449436932278595,
+      "learning_rate": 8.678866936325978e-06,
+      "loss": 0.0768,
+      "step": 909
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4215841710611692,
+      "learning_rate": 8.675731968536004e-06,
+      "loss": 0.0832,
+      "step": 910
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.407905652057663,
+      "learning_rate": 8.672593853258683e-06,
+      "loss": 0.0913,
+      "step": 911
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.574925733644576,
+      "learning_rate": 8.66945259318115e-06,
+      "loss": 0.093,
+      "step": 912
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6442804473503869,
+      "learning_rate": 8.66630819099323e-06,
+      "loss": 0.0873,
+      "step": 913
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7524676718124926,
+      "learning_rate": 8.663160649387437e-06,
+      "loss": 0.1657,
+      "step": 914
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.34435275238634266,
+      "learning_rate": 8.660009971058977e-06,
+      "loss": 0.0915,
+      "step": 915
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.35403476880377843,
+      "learning_rate": 8.656856158705739e-06,
+      "loss": 0.0709,
+      "step": 916
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.382013326900908,
+      "learning_rate": 8.653699215028298e-06,
+      "loss": 0.0887,
+      "step": 917
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2969627627438944,
+      "learning_rate": 8.650539142729906e-06,
+      "loss": 0.0568,
+      "step": 918
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.23465454383079634,
+      "learning_rate": 8.647375944516498e-06,
+      "loss": 0.0311,
+      "step": 919
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5186825503251482,
+      "learning_rate": 8.644209623096686e-06,
+      "loss": 0.1222,
+      "step": 920
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4625313033137186,
+      "learning_rate": 8.641040181181755e-06,
+      "loss": 0.0929,
+      "step": 921
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5628437161261859,
+      "learning_rate": 8.637867621485658e-06,
+      "loss": 0.104,
+      "step": 922
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.39877521396518395,
+      "learning_rate": 8.634691946725026e-06,
+      "loss": 0.083,
+      "step": 923
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4711420645294162,
+      "learning_rate": 8.63151315961915e-06,
+      "loss": 0.0914,
+      "step": 924
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.42262596072091024,
+      "learning_rate": 8.628331262889992e-06,
+      "loss": 0.0926,
+      "step": 925
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2583481971720481,
+      "learning_rate": 8.625146259262171e-06,
+      "loss": 0.0497,
+      "step": 926
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.42162082330689993,
+      "learning_rate": 8.621958151462972e-06,
+      "loss": 0.1033,
+      "step": 927
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.39189691045759195,
+      "learning_rate": 8.618766942222334e-06,
+      "loss": 0.0963,
+      "step": 928
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5795328923387679,
+      "learning_rate": 8.615572634272853e-06,
+      "loss": 0.0996,
+      "step": 929
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5871640470001664,
+      "learning_rate": 8.612375230349779e-06,
+      "loss": 0.0997,
+      "step": 930
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3740854127694093,
+      "learning_rate": 8.609174733191012e-06,
+      "loss": 0.1078,
+      "step": 931
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3675860631682876,
+      "learning_rate": 8.6059711455371e-06,
+      "loss": 0.0919,
+      "step": 932
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8034782523140173,
+      "learning_rate": 8.602764470131241e-06,
+      "loss": 0.115,
+      "step": 933
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5684175011157698,
+      "learning_rate": 8.599554709719273e-06,
+      "loss": 0.0553,
+      "step": 934
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4621484777944867,
+      "learning_rate": 8.596341867049677e-06,
+      "loss": 0.0608,
+      "step": 935
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7909439828218208,
+      "learning_rate": 8.593125944873575e-06,
+      "loss": 0.1247,
+      "step": 936
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4000895031285387,
+      "learning_rate": 8.589906945944722e-06,
+      "loss": 0.0717,
+      "step": 937
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5060977475727799,
+      "learning_rate": 8.586684873019513e-06,
+      "loss": 0.119,
+      "step": 938
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.449585847206327,
+      "learning_rate": 8.583459728856972e-06,
+      "loss": 0.0956,
+      "step": 939
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6216744445303553,
+      "learning_rate": 8.58023151621875e-06,
+      "loss": 0.1282,
+      "step": 940
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.36326462305814344,
+      "learning_rate": 8.577000237869131e-06,
+      "loss": 0.0674,
+      "step": 941
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4122286497124454,
+      "learning_rate": 8.573765896575022e-06,
+      "loss": 0.072,
+      "step": 942
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.36702351763564395,
+      "learning_rate": 8.570528495105952e-06,
+      "loss": 0.0956,
+      "step": 943
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3055341362571203,
+      "learning_rate": 8.567288036234071e-06,
+      "loss": 0.0782,
+      "step": 944
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5808103442156552,
+      "learning_rate": 8.564044522734147e-06,
+      "loss": 0.087,
+      "step": 945
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3718400606813192,
+      "learning_rate": 8.560797957383564e-06,
+      "loss": 0.0889,
+      "step": 946
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.228691714152912,
+      "learning_rate": 8.557548342962318e-06,
+      "loss": 0.1751,
+      "step": 947
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.30425561951883184,
+      "learning_rate": 8.554295682253016e-06,
+      "loss": 0.0844,
+      "step": 948
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5269394360838388,
+      "learning_rate": 8.551039978040877e-06,
+      "loss": 0.1292,
+      "step": 949
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3307915158146362,
+      "learning_rate": 8.54778123311372e-06,
+      "loss": 0.0721,
+      "step": 950
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6568645637523918,
+      "learning_rate": 8.544519450261976e-06,
+      "loss": 0.0465,
+      "step": 951
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3854610769587849,
+      "learning_rate": 8.541254632278667e-06,
+      "loss": 0.1248,
+      "step": 952
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.44736870573530385,
+      "learning_rate": 8.537986781959423e-06,
+      "loss": 0.1002,
+      "step": 953
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4998194788485879,
+      "learning_rate": 8.534715902102463e-06,
+      "loss": 0.0828,
+      "step": 954
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5464549611667345,
+      "learning_rate": 8.531441995508609e-06,
+      "loss": 0.1342,
+      "step": 955
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7232068112164314,
+      "learning_rate": 8.528165064981266e-06,
+      "loss": 0.0999,
+      "step": 956
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.29674152937097625,
+      "learning_rate": 8.524885113326435e-06,
+      "loss": 0.0752,
+      "step": 957
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4164296072770713,
+      "learning_rate": 8.5216021433527e-06,
+      "loss": 0.0838,
+      "step": 958
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.43165949526934533,
+      "learning_rate": 8.518316157871232e-06,
+      "loss": 0.1017,
+      "step": 959
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.659632022701993,
+      "learning_rate": 8.515027159695781e-06,
+      "loss": 0.1027,
+      "step": 960
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4494197565037416,
+      "learning_rate": 8.511735151642678e-06,
+      "loss": 0.1264,
+      "step": 961
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6056482137264181,
+      "learning_rate": 8.508440136530833e-06,
+      "loss": 0.1275,
+      "step": 962
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6873658026634043,
+      "learning_rate": 8.505142117181732e-06,
+      "loss": 0.0999,
+      "step": 963
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6179202466779846,
+      "learning_rate": 8.501841096419428e-06,
+      "loss": 0.0961,
+      "step": 964
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.31766812424786744,
+      "learning_rate": 8.498537077070548e-06,
+      "loss": 0.0805,
+      "step": 965
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.47004606745251126,
+      "learning_rate": 8.495230061964289e-06,
+      "loss": 0.0889,
+      "step": 966
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.38399650573413024,
+      "learning_rate": 8.491920053932406e-06,
+      "loss": 0.0606,
+      "step": 967
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5825205963634896,
+      "learning_rate": 8.488607055809223e-06,
+      "loss": 0.0706,
+      "step": 968
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.45989814660217077,
+      "learning_rate": 8.485291070431625e-06,
+      "loss": 0.0948,
+      "step": 969
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6746008813092472,
+      "learning_rate": 8.481972100639049e-06,
+      "loss": 0.1472,
+      "step": 970
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3041188595130515,
+      "learning_rate": 8.478650149273493e-06,
+      "loss": 0.0709,
+      "step": 971
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4066723941499203,
+      "learning_rate": 8.475325219179502e-06,
+      "loss": 0.0808,
+      "step": 972
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4273194677897243,
+      "learning_rate": 8.471997313204183e-06,
+      "loss": 0.0861,
+      "step": 973
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3656756351013927,
+      "learning_rate": 8.468666434197177e-06,
+      "loss": 0.0818,
+      "step": 974
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7699240643770021,
+      "learning_rate": 8.465332585010682e-06,
+      "loss": 0.0904,
+      "step": 975
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.49151957054172846,
+      "learning_rate": 8.461995768499433e-06,
+      "loss": 0.071,
+      "step": 976
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.586020044524259,
+      "learning_rate": 8.458655987520708e-06,
+      "loss": 0.1369,
+      "step": 977
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.36776011299451683,
+      "learning_rate": 8.455313244934324e-06,
+      "loss": 0.0748,
+      "step": 978
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7619473882290455,
+      "learning_rate": 8.451967543602635e-06,
+      "loss": 0.1337,
+      "step": 979
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5303405477984192,
+      "learning_rate": 8.448618886390523e-06,
+      "loss": 0.0969,
+      "step": 980
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.31130010552918846,
+      "learning_rate": 8.445267276165406e-06,
+      "loss": 0.0563,
+      "step": 981
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7283887396492364,
+      "learning_rate": 8.441912715797231e-06,
+      "loss": 0.1205,
+      "step": 982
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5454777059466462,
+      "learning_rate": 8.43855520815847e-06,
+      "loss": 0.141,
+      "step": 983
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.49972607460894874,
+      "learning_rate": 8.435194756124118e-06,
+      "loss": 0.1073,
+      "step": 984
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.43957090825164086,
+      "learning_rate": 8.431831362571692e-06,
+      "loss": 0.127,
+      "step": 985
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3446840899666005,
+      "learning_rate": 8.428465030381227e-06,
+      "loss": 0.0793,
+      "step": 986
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5881703597812649,
+      "learning_rate": 8.425095762435274e-06,
+      "loss": 0.0844,
+      "step": 987
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5494944307253974,
+      "learning_rate": 8.4217235616189e-06,
+      "loss": 0.0808,
+      "step": 988
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7162758642553052,
+      "learning_rate": 8.418348430819681e-06,
+      "loss": 0.1203,
+      "step": 989
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.39597458879945174,
+      "learning_rate": 8.414970372927705e-06,
+      "loss": 0.0744,
+      "step": 990
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.38118073928129964,
+      "learning_rate": 8.411589390835561e-06,
+      "loss": 0.0963,
+      "step": 991
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.48131735200581544,
+      "learning_rate": 8.408205487438351e-06,
+      "loss": 0.105,
+      "step": 992
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6537539054016474,
+      "learning_rate": 8.404818665633666e-06,
+      "loss": 0.0946,
+      "step": 993
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6881627111782112,
+      "learning_rate": 8.401428928321607e-06,
+      "loss": 0.108,
+      "step": 994
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6937707564738863,
+      "learning_rate": 8.398036278404768e-06,
+      "loss": 0.074,
+      "step": 995
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.42510209933061105,
+      "learning_rate": 8.394640718788234e-06,
+      "loss": 0.076,
+      "step": 996
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.39062853225418204,
+      "learning_rate": 8.391242252379586e-06,
+      "loss": 0.0945,
+      "step": 997
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.47000992816828807,
+      "learning_rate": 8.387840882088889e-06,
+      "loss": 0.0949,
+      "step": 998
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3630042614344615,
+      "learning_rate": 8.384436610828702e-06,
+      "loss": 0.0894,
+      "step": 999
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4473125165778053,
+      "learning_rate": 8.38102944151406e-06,
+      "loss": 0.0729,
+      "step": 1000
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2925190783459522,
+      "learning_rate": 8.377619377062483e-06,
+      "loss": 0.0498,
+      "step": 1001
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.39742597520065365,
+      "learning_rate": 8.374206420393973e-06,
+      "loss": 0.0623,
+      "step": 1002
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4478276878530916,
+      "learning_rate": 8.370790574431005e-06,
+      "loss": 0.1058,
+      "step": 1003
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.42328523998064277,
+      "learning_rate": 8.367371842098528e-06,
+      "loss": 0.1258,
+      "step": 1004
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5080160093202588,
+      "learning_rate": 8.363950226323963e-06,
+      "loss": 0.1033,
+      "step": 1005
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.33864954751696186,
+      "learning_rate": 8.360525730037201e-06,
+      "loss": 0.0551,
+      "step": 1006
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.36966845092822054,
+      "learning_rate": 8.357098356170603e-06,
+      "loss": 0.1034,
+      "step": 1007
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8101625761239845,
+      "learning_rate": 8.353668107658984e-06,
+      "loss": 0.1378,
+      "step": 1008
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.36101811088061997,
+      "learning_rate": 8.35023498743963e-06,
+      "loss": 0.0811,
+      "step": 1009
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.44657333797588306,
+      "learning_rate": 8.346798998452283e-06,
+      "loss": 0.087,
+      "step": 1010
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.48977663941340943,
+      "learning_rate": 8.343360143639138e-06,
+      "loss": 0.0931,
+      "step": 1011
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5098982142975803,
+      "learning_rate": 8.339918425944851e-06,
+      "loss": 0.0953,
+      "step": 1012
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3262565161128692,
+      "learning_rate": 8.336473848316524e-06,
+      "loss": 0.0631,
+      "step": 1013
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.35421779112164864,
+      "learning_rate": 8.33302641370371e-06,
+      "loss": 0.0977,
+      "step": 1014
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2869545712405884,
+      "learning_rate": 8.329576125058406e-06,
+      "loss": 0.0639,
+      "step": 1015
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.20287072416268473,
+      "learning_rate": 8.326122985335057e-06,
+      "loss": 0.0685,
+      "step": 1016
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7583154745559284,
+      "learning_rate": 8.322666997490547e-06,
+      "loss": 0.0875,
+      "step": 1017
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4885199266295902,
+      "learning_rate": 8.319208164484197e-06,
+      "loss": 0.1041,
+      "step": 1018
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4043930830511361,
+      "learning_rate": 8.315746489277769e-06,
+      "loss": 0.1157,
+      "step": 1019
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.648302343309106,
+      "learning_rate": 8.312281974835452e-06,
+      "loss": 0.1407,
+      "step": 1020
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5627686264207775,
+      "learning_rate": 8.308814624123875e-06,
+      "loss": 0.1281,
+      "step": 1021
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7904679232456217,
+      "learning_rate": 8.305344440112089e-06,
+      "loss": 0.1918,
+      "step": 1022
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1811965424431194,
+      "learning_rate": 8.30187142577157e-06,
+      "loss": 0.1582,
+      "step": 1023
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4097653619564071,
+      "learning_rate": 8.298395584076225e-06,
+      "loss": 0.0797,
+      "step": 1024
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5762426286667174,
+      "learning_rate": 8.294916918002377e-06,
+      "loss": 0.1297,
+      "step": 1025
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6371882149026293,
+      "learning_rate": 8.291435430528762e-06,
+      "loss": 0.1266,
+      "step": 1026
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6461947680286652,
+      "learning_rate": 8.287951124636546e-06,
+      "loss": 0.1466,
+      "step": 1027
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.33244417086603184,
+      "learning_rate": 8.284464003309298e-06,
+      "loss": 0.1034,
+      "step": 1028
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4759563814993332,
+      "learning_rate": 8.280974069532999e-06,
+      "loss": 0.1027,
+      "step": 1029
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.45943737713144167,
+      "learning_rate": 8.277481326296039e-06,
+      "loss": 0.1161,
+      "step": 1030
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.29575566208843385,
+      "learning_rate": 8.273985776589215e-06,
+      "loss": 0.053,
+      "step": 1031
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4466903877659513,
+      "learning_rate": 8.270487423405727e-06,
+      "loss": 0.1189,
+      "step": 1032
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5105463418911009,
+      "learning_rate": 8.266986269741173e-06,
+      "loss": 0.0778,
+      "step": 1033
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.31254158962842304,
+      "learning_rate": 8.263482318593553e-06,
+      "loss": 0.0893,
+      "step": 1034
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8478579221449796,
+      "learning_rate": 8.259975572963257e-06,
+      "loss": 0.123,
+      "step": 1035
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.36709753626424013,
+      "learning_rate": 8.256466035853077e-06,
+      "loss": 0.0497,
+      "step": 1036
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6835638168250348,
+      "learning_rate": 8.252953710268185e-06,
+      "loss": 0.0725,
+      "step": 1037
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4340829342289024,
+      "learning_rate": 8.249438599216149e-06,
+      "loss": 0.0809,
+      "step": 1038
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.36138727045471125,
+      "learning_rate": 8.245920705706913e-06,
+      "loss": 0.0889,
+      "step": 1039
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2993361892304179,
+      "learning_rate": 8.242400032752813e-06,
+      "loss": 0.0526,
+      "step": 1040
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5094376855780399,
+      "learning_rate": 8.238876583368563e-06,
+      "loss": 0.1006,
+      "step": 1041
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5634589100474235,
+      "learning_rate": 8.235350360571249e-06,
+      "loss": 0.1157,
+      "step": 1042
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.47421182884319807,
+      "learning_rate": 8.231821367380335e-06,
+      "loss": 0.1418,
+      "step": 1043
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3645236056670461,
+      "learning_rate": 8.228289606817658e-06,
+      "loss": 0.0614,
+      "step": 1044
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4875507064892114,
+      "learning_rate": 8.224755081907427e-06,
+      "loss": 0.0763,
+      "step": 1045
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7407249078045226,
+      "learning_rate": 8.221217795676213e-06,
+      "loss": 0.1152,
+      "step": 1046
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3649004315172086,
+      "learning_rate": 8.217677751152954e-06,
+      "loss": 0.0814,
+      "step": 1047
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.536336536250725,
+      "learning_rate": 8.21413495136895e-06,
+      "loss": 0.1196,
+      "step": 1048
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2855081985137709,
+      "learning_rate": 8.21058939935786e-06,
+      "loss": 0.029,
+      "step": 1049
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2680957073915973,
+      "learning_rate": 8.207041098155701e-06,
+      "loss": 0.0786,
+      "step": 1050
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.45054152748401677,
+      "learning_rate": 8.20349005080084e-06,
+      "loss": 0.1169,
+      "step": 1051
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6139110124025449,
+      "learning_rate": 8.199936260334e-06,
+      "loss": 0.1165,
+      "step": 1052
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.33867977009829214,
+      "learning_rate": 8.196379729798252e-06,
+      "loss": 0.1048,
+      "step": 1053
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.37942391245149126,
+      "learning_rate": 8.192820462239012e-06,
+      "loss": 0.1012,
+      "step": 1054
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5645099693984074,
+      "learning_rate": 8.189258460704039e-06,
+      "loss": 0.1294,
+      "step": 1055
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4853097526050573,
+      "learning_rate": 8.185693728243435e-06,
+      "loss": 0.1012,
+      "step": 1056
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9300273392049366,
+      "learning_rate": 8.182126267909642e-06,
+      "loss": 0.1265,
+      "step": 1057
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6999893711864746,
+      "learning_rate": 8.178556082757431e-06,
+      "loss": 0.1152,
+      "step": 1058
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9138579044020679,
+      "learning_rate": 8.174983175843915e-06,
+      "loss": 0.1931,
+      "step": 1059
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.41861793606440134,
+      "learning_rate": 8.171407550228532e-06,
+      "loss": 0.1189,
+      "step": 1060
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4241523484325529,
+      "learning_rate": 8.167829208973048e-06,
+      "loss": 0.0863,
+      "step": 1061
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6830500366547438,
+      "learning_rate": 8.164248155141557e-06,
+      "loss": 0.1394,
+      "step": 1062
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4446345716747391,
+      "learning_rate": 8.160664391800475e-06,
+      "loss": 0.1029,
+      "step": 1063
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4953977838410171,
+      "learning_rate": 8.157077922018537e-06,
+      "loss": 0.1023,
+      "step": 1064
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.32633677272271966,
+      "learning_rate": 8.153488748866795e-06,
+      "loss": 0.0624,
+      "step": 1065
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8318169681199975,
+      "learning_rate": 8.14989687541862e-06,
+      "loss": 0.1282,
+      "step": 1066
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3466502879782476,
+      "learning_rate": 8.146302304749689e-06,
+      "loss": 0.0852,
+      "step": 1067
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.28109079631108735,
+      "learning_rate": 8.142705039937994e-06,
+      "loss": 0.0665,
+      "step": 1068
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7215831667805335,
+      "learning_rate": 8.139105084063832e-06,
+      "loss": 0.1144,
+      "step": 1069
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.31182985041749733,
+      "learning_rate": 8.135502440209803e-06,
+      "loss": 0.0874,
+      "step": 1070
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4850093138487995,
+      "learning_rate": 8.13189711146081e-06,
+      "loss": 0.128,
+      "step": 1071
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2797988827907325,
+      "learning_rate": 8.128289100904056e-06,
+      "loss": 0.0691,
+      "step": 1072
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8220920414966311,
+      "learning_rate": 8.124678411629037e-06,
+      "loss": 0.1486,
+      "step": 1073
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0613437851569856,
+      "learning_rate": 8.121065046727547e-06,
+      "loss": 0.1489,
+      "step": 1074
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5075715468063482,
+      "learning_rate": 8.117449009293668e-06,
+      "loss": 0.1168,
+      "step": 1075
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8245510726304975,
+      "learning_rate": 8.113830302423773e-06,
+      "loss": 0.1081,
+      "step": 1076
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.34847096383222875,
+      "learning_rate": 8.110208929216518e-06,
+      "loss": 0.0714,
+      "step": 1077
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.513751607536443,
+      "learning_rate": 8.106584892772844e-06,
+      "loss": 0.1279,
+      "step": 1078
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.47401253237123436,
+      "learning_rate": 8.102958196195972e-06,
+      "loss": 0.1397,
+      "step": 1079
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.40449775116622433,
+      "learning_rate": 8.0993288425914e-06,
+      "loss": 0.0794,
+      "step": 1080
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5951573885001263,
+      "learning_rate": 8.095696835066906e-06,
+      "loss": 0.15,
+      "step": 1081
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5569632760651496,
+      "learning_rate": 8.092062176732531e-06,
+      "loss": 0.1026,
+      "step": 1082
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6521827209880674,
+      "learning_rate": 8.088424870700595e-06,
+      "loss": 0.1117,
+      "step": 1083
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6260366045494986,
+      "learning_rate": 8.084784920085682e-06,
+      "loss": 0.1141,
+      "step": 1084
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.37305490577484035,
+      "learning_rate": 8.081142328004638e-06,
+      "loss": 0.0886,
+      "step": 1085
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.49597637338839584,
+      "learning_rate": 8.077497097576573e-06,
+      "loss": 0.1086,
+      "step": 1086
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8158622365930779,
+      "learning_rate": 8.073849231922859e-06,
+      "loss": 0.1661,
+      "step": 1087
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.458905363150963,
+      "learning_rate": 8.070198734167119e-06,
+      "loss": 0.0989,
+      "step": 1088
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4259627728408611,
+      "learning_rate": 8.066545607435232e-06,
+      "loss": 0.1298,
+      "step": 1089
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.33877904788872976,
+      "learning_rate": 8.062889854855334e-06,
+      "loss": 0.1014,
+      "step": 1090
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.26955903561824496,
+      "learning_rate": 8.059231479557797e-06,
+      "loss": 0.0526,
+      "step": 1091
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4960593268108397,
+      "learning_rate": 8.055570484675252e-06,
+      "loss": 0.1127,
+      "step": 1092
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5296626194371825,
+      "learning_rate": 8.051906873342563e-06,
+      "loss": 0.1129,
+      "step": 1093
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.24575969028576672,
+      "learning_rate": 8.048240648696842e-06,
+      "loss": 0.0442,
+      "step": 1094
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3513992957638845,
+      "learning_rate": 8.044571813877431e-06,
+      "loss": 0.0905,
+      "step": 1095
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3203419179094259,
+      "learning_rate": 8.040900372025916e-06,
+      "loss": 0.0793,
+      "step": 1096
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4700084002956758,
+      "learning_rate": 8.03722632628611e-06,
+      "loss": 0.109,
+      "step": 1097
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2483215033554251,
+      "learning_rate": 8.033549679804052e-06,
+      "loss": 0.0561,
+      "step": 1098
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3597357354623874,
+      "learning_rate": 8.029870435728018e-06,
+      "loss": 0.0996,
+      "step": 1099
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3000531772010142,
+      "learning_rate": 8.0261885972085e-06,
+      "loss": 0.0623,
+      "step": 1100
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.41445398585338367,
+      "learning_rate": 8.022504167398214e-06,
+      "loss": 0.0942,
+      "step": 1101
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5169367410952866,
+      "learning_rate": 8.018817149452096e-06,
+      "loss": 0.0746,
+      "step": 1102
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.44796645479944325,
+      "learning_rate": 8.015127546527299e-06,
+      "loss": 0.0998,
+      "step": 1103
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.44025839130861805,
+      "learning_rate": 8.011435361783184e-06,
+      "loss": 0.1114,
+      "step": 1104
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3937437646224052,
+      "learning_rate": 8.007740598381329e-06,
+      "loss": 0.0594,
+      "step": 1105
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.30715817255303246,
+      "learning_rate": 8.004043259485519e-06,
+      "loss": 0.0619,
+      "step": 1106
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3406652493401725,
+      "learning_rate": 8.000343348261741e-06,
+      "loss": 0.078,
+      "step": 1107
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3548624704379509,
+      "learning_rate": 7.996640867878188e-06,
+      "loss": 0.0767,
+      "step": 1108
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.34396569921394493,
+      "learning_rate": 7.99293582150525e-06,
+      "loss": 0.0535,
+      "step": 1109
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5047146478790318,
+      "learning_rate": 7.989228212315516e-06,
+      "loss": 0.0967,
+      "step": 1110
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.19463895010201257,
+      "learning_rate": 7.985518043483774e-06,
+      "loss": 0.0522,
+      "step": 1111
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.27575738138782074,
+      "learning_rate": 7.981805318186992e-06,
+      "loss": 0.0653,
+      "step": 1112
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.44116358775158315,
+      "learning_rate": 7.978090039604342e-06,
+      "loss": 0.0946,
+      "step": 1113
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.30929616820651784,
+      "learning_rate": 7.974372210917168e-06,
+      "loss": 0.053,
+      "step": 1114
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.47226923600043097,
+      "learning_rate": 7.970651835309009e-06,
+      "loss": 0.0934,
+      "step": 1115
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3661054432573276,
+      "learning_rate": 7.966928915965578e-06,
+      "loss": 0.0551,
+      "step": 1116
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5489439870804758,
+      "learning_rate": 7.963203456074767e-06,
+      "loss": 0.1272,
+      "step": 1117
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3052756891544375,
+      "learning_rate": 7.959475458826647e-06,
+      "loss": 0.0638,
+      "step": 1118
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5124683390357987,
+      "learning_rate": 7.95574492741346e-06,
+      "loss": 0.108,
+      "step": 1119
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4310133133606615,
+      "learning_rate": 7.952011865029614e-06,
+      "loss": 0.0785,
+      "step": 1120
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4294669892547384,
+      "learning_rate": 7.94827627487169e-06,
+      "loss": 0.0714,
+      "step": 1121
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.40917875591040775,
+      "learning_rate": 7.944538160138435e-06,
+      "loss": 0.104,
+      "step": 1122
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.36318268664891995,
+      "learning_rate": 7.940797524030748e-06,
+      "loss": 0.1028,
+      "step": 1123
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.25870197230093644,
+      "learning_rate": 7.937054369751696e-06,
+      "loss": 0.0613,
+      "step": 1124
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5492907488895081,
+      "learning_rate": 7.933308700506497e-06,
+      "loss": 0.1046,
+      "step": 1125
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.33197899047198387,
+      "learning_rate": 7.929560519502528e-06,
+      "loss": 0.0656,
+      "step": 1126
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6347481219081285,
+      "learning_rate": 7.925809829949312e-06,
+      "loss": 0.0981,
+      "step": 1127
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.34912028643554205,
+      "learning_rate": 7.922056635058522e-06,
+      "loss": 0.0651,
+      "step": 1128
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5078039159530806,
+      "learning_rate": 7.918300938043974e-06,
+      "loss": 0.1305,
+      "step": 1129
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5697315782808061,
+      "learning_rate": 7.914542742121632e-06,
+      "loss": 0.1025,
+      "step": 1130
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.32530947380410763,
+      "learning_rate": 7.910782050509596e-06,
+      "loss": 0.0383,
+      "step": 1131
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.32869399919836406,
+      "learning_rate": 7.9070188664281e-06,
+      "loss": 0.0593,
+      "step": 1132
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.33694692372972257,
+      "learning_rate": 7.903253193099516e-06,
+      "loss": 0.0837,
+      "step": 1133
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2904129730568934,
+      "learning_rate": 7.89948503374835e-06,
+      "loss": 0.0819,
+      "step": 1134
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.49514352849483495,
+      "learning_rate": 7.895714391601232e-06,
+      "loss": 0.0783,
+      "step": 1135
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.34327959303358785,
+      "learning_rate": 7.891941269886922e-06,
+      "loss": 0.0928,
+      "step": 1136
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5075477782555474,
+      "learning_rate": 7.888165671836297e-06,
+      "loss": 0.0748,
+      "step": 1137
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3502041423172233,
+      "learning_rate": 7.884387600682362e-06,
+      "loss": 0.0789,
+      "step": 1138
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.39308755449196536,
+      "learning_rate": 7.880607059660235e-06,
+      "loss": 0.0823,
+      "step": 1139
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4244428577656742,
+      "learning_rate": 7.87682405200715e-06,
+      "loss": 0.0993,
+      "step": 1140
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.43510702450394106,
+      "learning_rate": 7.873038580962453e-06,
+      "loss": 0.0891,
+      "step": 1141
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4456295982152556,
+      "learning_rate": 7.869250649767601e-06,
+      "loss": 0.1099,
+      "step": 1142
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.39182860845353573,
+      "learning_rate": 7.865460261666155e-06,
+      "loss": 0.0798,
+      "step": 1143
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5344187460056024,
+      "learning_rate": 7.861667419903783e-06,
+      "loss": 0.0985,
+      "step": 1144
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7891369466523669,
+      "learning_rate": 7.857872127728248e-06,
+      "loss": 0.0819,
+      "step": 1145
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5215596996530159,
+      "learning_rate": 7.854074388389421e-06,
+      "loss": 0.0814,
+      "step": 1146
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.46852218611317137,
+      "learning_rate": 7.850274205139258e-06,
+      "loss": 0.0667,
+      "step": 1147
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5449716291985394,
+      "learning_rate": 7.846471581231814e-06,
+      "loss": 0.0671,
+      "step": 1148
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5031026351604613,
+      "learning_rate": 7.842666519923235e-06,
+      "loss": 0.0975,
+      "step": 1149
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.37185250654919283,
+      "learning_rate": 7.838859024471747e-06,
+      "loss": 0.0683,
+      "step": 1150
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6139521239306095,
+      "learning_rate": 7.835049098137669e-06,
+      "loss": 0.1509,
+      "step": 1151
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.40880878757373446,
+      "learning_rate": 7.831236744183395e-06,
+      "loss": 0.1124,
+      "step": 1152
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2575037307931132,
+      "learning_rate": 7.827421965873403e-06,
+      "loss": 0.1537,
+      "step": 1153
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6546122947184678,
+      "learning_rate": 7.823604766474239e-06,
+      "loss": 0.1225,
+      "step": 1154
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5498410531971489,
+      "learning_rate": 7.819785149254534e-06,
+      "loss": 0.1383,
+      "step": 1155
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6197056198658327,
+      "learning_rate": 7.815963117484977e-06,
+      "loss": 0.0945,
+      "step": 1156
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4924690545451198,
+      "learning_rate": 7.812138674438332e-06,
+      "loss": 0.1279,
+      "step": 1157
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6229791928315794,
+      "learning_rate": 7.808311823389428e-06,
+      "loss": 0.099,
+      "step": 1158
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7343873627959406,
+      "learning_rate": 7.80448256761515e-06,
+      "loss": 0.14,
+      "step": 1159
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.40830450132964397,
+      "learning_rate": 7.80065091039445e-06,
+      "loss": 0.0654,
+      "step": 1160
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5631745495802228,
+      "learning_rate": 7.79681685500833e-06,
+      "loss": 0.0971,
+      "step": 1161
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.32978136538134334,
+      "learning_rate": 7.792980404739849e-06,
+      "loss": 0.0787,
+      "step": 1162
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.39506083561245803,
+      "learning_rate": 7.789141562874114e-06,
+      "loss": 0.0561,
+      "step": 1163
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.2966677165805511,
+      "learning_rate": 7.785300332698282e-06,
+      "loss": 0.0803,
+      "step": 1164
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4966997858510323,
+      "learning_rate": 7.781456717501557e-06,
+      "loss": 0.1164,
+      "step": 1165
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.43113926216677567,
+      "learning_rate": 7.77761072057518e-06,
+      "loss": 0.1117,
+      "step": 1166
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4856468390318449,
+      "learning_rate": 7.773762345212434e-06,
+      "loss": 0.0683,
+      "step": 1167
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3968324268099564,
+      "learning_rate": 7.76991159470864e-06,
+      "loss": 0.0768,
+      "step": 1168
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5842974160225576,
+      "learning_rate": 7.766058472361154e-06,
+      "loss": 0.0605,
+      "step": 1169
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7117350909882223,
+      "learning_rate": 7.762202981469358e-06,
+      "loss": 0.1317,
+      "step": 1170
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.42990855690076935,
+      "learning_rate": 7.758345125334665e-06,
+      "loss": 0.1125,
+      "step": 1171
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6010850452916343,
+      "learning_rate": 7.754484907260513e-06,
+      "loss": 0.1592,
+      "step": 1172
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5278953074527011,
+      "learning_rate": 7.750622330552365e-06,
+      "loss": 0.137,
+      "step": 1173
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2708849122181835,
+      "learning_rate": 7.746757398517696e-06,
+      "loss": 0.0611,
+      "step": 1174
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6684456893451128,
+      "learning_rate": 7.74289011446601e-06,
+      "loss": 0.0861,
+      "step": 1175
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3248118299458125,
+      "learning_rate": 7.739020481708816e-06,
+      "loss": 0.0752,
+      "step": 1176
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2840766394037429,
+      "learning_rate": 7.735148503559633e-06,
+      "loss": 0.0845,
+      "step": 1177
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3729838347067792,
+      "learning_rate": 7.731274183333995e-06,
+      "loss": 0.0722,
+      "step": 1178
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3818306176788432,
+      "learning_rate": 7.727397524349437e-06,
+      "loss": 0.0638,
+      "step": 1179
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.32945247662636773,
+      "learning_rate": 7.7235185299255e-06,
+      "loss": 0.0896,
+      "step": 1180
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.47214454470745265,
+      "learning_rate": 7.719637203383718e-06,
+      "loss": 0.1208,
+      "step": 1181
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.313714152344336,
+      "learning_rate": 7.715753548047632e-06,
+      "loss": 0.0842,
+      "step": 1182
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6362640918581809,
+      "learning_rate": 7.711867567242769e-06,
+      "loss": 0.0806,
+      "step": 1183
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2668214451753407,
+      "learning_rate": 7.707979264296649e-06,
+      "loss": 0.0272,
+      "step": 1184
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.35934940489629713,
+      "learning_rate": 7.704088642538782e-06,
+      "loss": 0.0791,
+      "step": 1185
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.30190362982869606,
+      "learning_rate": 7.700195705300667e-06,
+      "loss": 0.0248,
+      "step": 1186
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7878973347362918,
+      "learning_rate": 7.696300455915775e-06,
+      "loss": 0.1631,
+      "step": 1187
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4697535069263179,
+      "learning_rate": 7.692402897719568e-06,
+      "loss": 0.0834,
+      "step": 1188
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4593108243237563,
+      "learning_rate": 7.68850303404948e-06,
+      "loss": 0.092,
+      "step": 1189
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5735805679363245,
+      "learning_rate": 7.68460086824492e-06,
+      "loss": 0.0949,
+      "step": 1190
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6014239675928695,
+      "learning_rate": 7.680696403647268e-06,
+      "loss": 0.1592,
+      "step": 1191
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.36052541407201133,
+      "learning_rate": 7.676789643599871e-06,
+      "loss": 0.1029,
+      "step": 1192
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4811788369901242,
+      "learning_rate": 7.672880591448043e-06,
+      "loss": 0.0768,
+      "step": 1193
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7774597072896245,
+      "learning_rate": 7.668969250539063e-06,
+      "loss": 0.0671,
+      "step": 1194
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.29186336807886026,
+      "learning_rate": 7.665055624222166e-06,
+      "loss": 0.0852,
+      "step": 1195
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.637159785983851,
+      "learning_rate": 7.661139715848547e-06,
+      "loss": 0.0847,
+      "step": 1196
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7085992935208999,
+      "learning_rate": 7.657221528771352e-06,
+      "loss": 0.1229,
+      "step": 1197
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4036885370956982,
+      "learning_rate": 7.653301066345681e-06,
+      "loss": 0.0963,
+      "step": 1198
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6550127017484116,
+      "learning_rate": 7.649378331928581e-06,
+      "loss": 0.1257,
+      "step": 1199
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.45172260752731824,
+      "learning_rate": 7.645453328879042e-06,
+      "loss": 0.1076,
+      "step": 1200
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.33171439155135335,
+      "learning_rate": 7.641526060558005e-06,
+      "loss": 0.0782,
+      "step": 1201
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3023781409312154,
+      "learning_rate": 7.63759653032834e-06,
+      "loss": 0.077,
+      "step": 1202
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.194190932076639,
+      "learning_rate": 7.633664741554863e-06,
+      "loss": 0.0655,
+      "step": 1203
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5366691119548006,
+      "learning_rate": 7.629730697604314e-06,
+      "loss": 0.1036,
+      "step": 1204
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4074123782121894,
+      "learning_rate": 7.625794401845376e-06,
+      "loss": 0.1139,
+      "step": 1205
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4496057248969518,
+      "learning_rate": 7.621855857648651e-06,
+      "loss": 0.116,
+      "step": 1206
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4215659597010978,
+      "learning_rate": 7.617915068386671e-06,
+      "loss": 0.1036,
+      "step": 1207
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3945657635817372,
+      "learning_rate": 7.613972037433886e-06,
+      "loss": 0.085,
+      "step": 1208
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3006720640134774,
+      "learning_rate": 7.610026768166674e-06,
+      "loss": 0.0695,
+      "step": 1209
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5542776181889553,
+      "learning_rate": 7.606079263963318e-06,
+      "loss": 0.1502,
+      "step": 1210
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6237299007530562,
+      "learning_rate": 7.602129528204023e-06,
+      "loss": 0.0802,
+      "step": 1211
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3318324299310615,
+      "learning_rate": 7.5981775642709056e-06,
+      "loss": 0.0765,
+      "step": 1212
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6140720417596165,
+      "learning_rate": 7.594223375547982e-06,
+      "loss": 0.1221,
+      "step": 1213
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.39126371219320183,
+      "learning_rate": 7.590266965421182e-06,
+      "loss": 0.0855,
+      "step": 1214
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.25712230229700495,
+      "learning_rate": 7.5863083372783365e-06,
+      "loss": 0.0582,
+      "step": 1215
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2968649920922504,
+      "learning_rate": 7.58234749450917e-06,
+      "loss": 0.0624,
+      "step": 1216
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5077027081562454,
+      "learning_rate": 7.5783844405053064e-06,
+      "loss": 0.0922,
+      "step": 1217
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0093736048263418,
+      "learning_rate": 7.574419178660269e-06,
+      "loss": 0.1619,
+      "step": 1218
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5041760420555579,
+      "learning_rate": 7.57045171236946e-06,
+      "loss": 0.0968,
+      "step": 1219
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.32562538777616085,
+      "learning_rate": 7.566482045030179e-06,
+      "loss": 0.1013,
+      "step": 1220
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6546255816024102,
+      "learning_rate": 7.5625101800416055e-06,
+      "loss": 0.1335,
+      "step": 1221
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4318407082961103,
+      "learning_rate": 7.558536120804804e-06,
+      "loss": 0.1222,
+      "step": 1222
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2712329289615602,
+      "learning_rate": 7.554559870722714e-06,
+      "loss": 0.0619,
+      "step": 1223
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.38976509691487654,
+      "learning_rate": 7.550581433200155e-06,
+      "loss": 0.0983,
+      "step": 1224
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.39089260900309103,
+      "learning_rate": 7.546600811643816e-06,
+      "loss": 0.0704,
+      "step": 1225
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8971786888240914,
+      "learning_rate": 7.542618009462258e-06,
+      "loss": 0.1157,
+      "step": 1226
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4520304215432081,
+      "learning_rate": 7.538633030065909e-06,
+      "loss": 0.0899,
+      "step": 1227
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4206029596159851,
+      "learning_rate": 7.534645876867064e-06,
+      "loss": 0.1164,
+      "step": 1228
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.35222251931872256,
+      "learning_rate": 7.530656553279873e-06,
+      "loss": 0.0895,
+      "step": 1229
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5895836229513604,
+      "learning_rate": 7.526665062720351e-06,
+      "loss": 0.1304,
+      "step": 1230
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4194301687103519,
+      "learning_rate": 7.522671408606363e-06,
+      "loss": 0.0969,
+      "step": 1231
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4186167818878908,
+      "learning_rate": 7.5186755943576324e-06,
+      "loss": 0.0957,
+      "step": 1232
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3626156880181813,
+      "learning_rate": 7.51467762339573e-06,
+      "loss": 0.0801,
+      "step": 1233
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.48752646939344874,
+      "learning_rate": 7.510677499144068e-06,
+      "loss": 0.1013,
+      "step": 1234
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4865304274990717,
+      "learning_rate": 7.5066752250279104e-06,
+      "loss": 0.0828,
+      "step": 1235
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4398756534676093,
+      "learning_rate": 7.502670804474359e-06,
+      "loss": 0.0917,
+      "step": 1236
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4890809214016786,
+      "learning_rate": 7.498664240912354e-06,
+      "loss": 0.0881,
+      "step": 1237
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.48589244001927956,
+      "learning_rate": 7.494655537772667e-06,
+      "loss": 0.0783,
+      "step": 1238
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4906795310797478,
+      "learning_rate": 7.490644698487909e-06,
+      "loss": 0.0824,
+      "step": 1239
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.37049834734013626,
+      "learning_rate": 7.486631726492511e-06,
+      "loss": 0.0985,
+      "step": 1240
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3350345268942506,
+      "learning_rate": 7.482616625222741e-06,
+      "loss": 0.0652,
+      "step": 1241
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.43138496833814577,
+      "learning_rate": 7.478599398116678e-06,
+      "loss": 0.0964,
+      "step": 1242
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4699342535775678,
+      "learning_rate": 7.474580048614233e-06,
+      "loss": 0.0704,
+      "step": 1243
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6043115294856498,
+      "learning_rate": 7.470558580157126e-06,
+      "loss": 0.1101,
+      "step": 1244
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3253143486775578,
+      "learning_rate": 7.466534996188897e-06,
+      "loss": 0.0824,
+      "step": 1245
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7001258558864883,
+      "learning_rate": 7.462509300154892e-06,
+      "loss": 0.0983,
+      "step": 1246
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.49755871463826296,
+      "learning_rate": 7.45848149550227e-06,
+      "loss": 0.1227,
+      "step": 1247
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5260467794728861,
+      "learning_rate": 7.454451585679996e-06,
+      "loss": 0.081,
+      "step": 1248
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5080804721923634,
+      "learning_rate": 7.450419574138833e-06,
+      "loss": 0.1053,
+      "step": 1249
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3791342728629578,
+      "learning_rate": 7.446385464331349e-06,
+      "loss": 0.107,
+      "step": 1250
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5399108324070111,
+      "learning_rate": 7.442349259711904e-06,
+      "loss": 0.1181,
+      "step": 1251
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6639879999689687,
+      "learning_rate": 7.438310963736655e-06,
+      "loss": 0.1495,
+      "step": 1252
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1723017539269172,
+      "learning_rate": 7.434270579863549e-06,
+      "loss": 0.0658,
+      "step": 1253
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3034849286931736,
+      "learning_rate": 7.43022811155232e-06,
+      "loss": 0.0832,
+      "step": 1254
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.36893262943136473,
+      "learning_rate": 7.426183562264487e-06,
+      "loss": 0.0971,
+      "step": 1255
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5974855328671015,
+      "learning_rate": 7.422136935463354e-06,
+      "loss": 0.1166,
+      "step": 1256
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3229655949631019,
+      "learning_rate": 7.418088234613997e-06,
+      "loss": 0.0785,
+      "step": 1257
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4571919749930792,
+      "learning_rate": 7.414037463183276e-06,
+      "loss": 0.0725,
+      "step": 1258
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4660452723925712,
+      "learning_rate": 7.4099846246398185e-06,
+      "loss": 0.0995,
+      "step": 1259
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7272549656849545,
+      "learning_rate": 7.405929722454026e-06,
+      "loss": 0.0774,
+      "step": 1260
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5476897323036802,
+      "learning_rate": 7.401872760098063e-06,
+      "loss": 0.1126,
+      "step": 1261
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3810391769235816,
+      "learning_rate": 7.3978137410458595e-06,
+      "loss": 0.11,
+      "step": 1262
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2522541561241933,
+      "learning_rate": 7.39375266877311e-06,
+      "loss": 0.0678,
+      "step": 1263
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6174401914519404,
+      "learning_rate": 7.3896895467572616e-06,
+      "loss": 0.1361,
+      "step": 1264
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5310372812239782,
+      "learning_rate": 7.385624378477521e-06,
+      "loss": 0.1132,
+      "step": 1265
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5532722658518993,
+      "learning_rate": 7.381557167414844e-06,
+      "loss": 0.1235,
+      "step": 1266
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5359325875315766,
+      "learning_rate": 7.3774879170519386e-06,
+      "loss": 0.1386,
+      "step": 1267
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7013688142544489,
+      "learning_rate": 7.373416630873255e-06,
+      "loss": 0.1025,
+      "step": 1268
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.21878971083627835,
+      "learning_rate": 7.369343312364994e-06,
+      "loss": 0.0706,
+      "step": 1269
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5444098139359753,
+      "learning_rate": 7.365267965015086e-06,
+      "loss": 0.1428,
+      "step": 1270
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.40046008322119825,
+      "learning_rate": 7.361190592313209e-06,
+      "loss": 0.0911,
+      "step": 1271
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4753066450054673,
+      "learning_rate": 7.357111197750768e-06,
+      "loss": 0.1104,
+      "step": 1272
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.34302904252531163,
+      "learning_rate": 7.353029784820902e-06,
+      "loss": 0.0894,
+      "step": 1273
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5175561729935336,
+      "learning_rate": 7.348946357018479e-06,
+      "loss": 0.0685,
+      "step": 1274
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.33407349978091094,
+      "learning_rate": 7.344860917840092e-06,
+      "loss": 0.0769,
+      "step": 1275
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.537296939644874,
+      "learning_rate": 7.3407734707840575e-06,
+      "loss": 0.1415,
+      "step": 1276
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4714955294845538,
+      "learning_rate": 7.336684019350405e-06,
+      "loss": 0.0602,
+      "step": 1277
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5231687005390631,
+      "learning_rate": 7.332592567040889e-06,
+      "loss": 0.1012,
+      "step": 1278
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6451651770169565,
+      "learning_rate": 7.328499117358973e-06,
+      "loss": 0.1086,
+      "step": 1279
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4396753615304459,
+      "learning_rate": 7.324403673809831e-06,
+      "loss": 0.0974,
+      "step": 1280
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3351717334860073,
+      "learning_rate": 7.320306239900343e-06,
+      "loss": 0.0444,
+      "step": 1281
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6057477710603322,
+      "learning_rate": 7.316206819139098e-06,
+      "loss": 0.0557,
+      "step": 1282
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4853121311355041,
+      "learning_rate": 7.312105415036379e-06,
+      "loss": 0.1157,
+      "step": 1283
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3956609692436241,
+      "learning_rate": 7.308002031104177e-06,
+      "loss": 0.0959,
+      "step": 1284
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.36579601627698144,
+      "learning_rate": 7.303896670856168e-06,
+      "loss": 0.0818,
+      "step": 1285
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5950403488333442,
+      "learning_rate": 7.299789337807727e-06,
+      "loss": 0.0962,
+      "step": 1286
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7302893406165183,
+      "learning_rate": 7.2956800354759165e-06,
+      "loss": 0.1507,
+      "step": 1287
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.35248968306116857,
+      "learning_rate": 7.291568767379484e-06,
+      "loss": 0.0677,
+      "step": 1288
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.35954962807442997,
+      "learning_rate": 7.287455537038865e-06,
+      "loss": 0.0789,
+      "step": 1289
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2545344927730057,
+      "learning_rate": 7.283340347976167e-06,
+      "loss": 0.0438,
+      "step": 1290
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6622217651425762,
+      "learning_rate": 7.279223203715183e-06,
+      "loss": 0.1154,
+      "step": 1291
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6497852254352842,
+      "learning_rate": 7.275104107781374e-06,
+      "loss": 0.1187,
+      "step": 1292
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2948404321113314,
+      "learning_rate": 7.270983063701878e-06,
+      "loss": 0.0834,
+      "step": 1293
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.323389471633835,
+      "learning_rate": 7.2668600750054955e-06,
+      "loss": 0.0593,
+      "step": 1294
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6256941837415155,
+      "learning_rate": 7.262735145222696e-06,
+      "loss": 0.1153,
+      "step": 1295
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4091099856115394,
+      "learning_rate": 7.258608277885608e-06,
+      "loss": 0.1209,
+      "step": 1296
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4989328325430237,
+      "learning_rate": 7.254479476528023e-06,
+      "loss": 0.1174,
+      "step": 1297
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.437548733354485,
+      "learning_rate": 7.250348744685385e-06,
+      "loss": 0.0895,
+      "step": 1298
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4457904798020956,
+      "learning_rate": 7.246216085894793e-06,
+      "loss": 0.0738,
+      "step": 1299
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4490462670693121,
+      "learning_rate": 7.242081503694996e-06,
+      "loss": 0.1226,
+      "step": 1300
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.394034010362623,
+      "learning_rate": 7.237945001626388e-06,
+      "loss": 0.0882,
+      "step": 1301
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5396818419239576,
+      "learning_rate": 7.233806583231012e-06,
+      "loss": 0.0992,
+      "step": 1302
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.42430139774657966,
+      "learning_rate": 7.229666252052545e-06,
+      "loss": 0.0947,
+      "step": 1303
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6830348155509963,
+      "learning_rate": 7.225524011636308e-06,
+      "loss": 0.144,
+      "step": 1304
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4371075646071382,
+      "learning_rate": 7.221379865529251e-06,
+      "loss": 0.0821,
+      "step": 1305
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.453529637864777,
+      "learning_rate": 7.2172338172799625e-06,
+      "loss": 0.0848,
+      "step": 1306
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5704528490914308,
+      "learning_rate": 7.213085870438653e-06,
+      "loss": 0.0566,
+      "step": 1307
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6960155872822488,
+      "learning_rate": 7.208936028557165e-06,
+      "loss": 0.1474,
+      "step": 1308
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5817205230296579,
+      "learning_rate": 7.204784295188959e-06,
+      "loss": 0.1037,
+      "step": 1309
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.40251750204790904,
+      "learning_rate": 7.200630673889118e-06,
+      "loss": 0.0951,
+      "step": 1310
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.44423514359563776,
+      "learning_rate": 7.1964751682143385e-06,
+      "loss": 0.1336,
+      "step": 1311
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.31689798590143897,
+      "learning_rate": 7.192317781722935e-06,
+      "loss": 0.0477,
+      "step": 1312
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4498326110158957,
+      "learning_rate": 7.18815851797483e-06,
+      "loss": 0.0755,
+      "step": 1313
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4214579949532243,
+      "learning_rate": 7.183997380531551e-06,
+      "loss": 0.0916,
+      "step": 1314
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3629707871148528,
+      "learning_rate": 7.179834372956236e-06,
+      "loss": 0.0457,
+      "step": 1315
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.23890986120364538,
+      "learning_rate": 7.1756694988136165e-06,
+      "loss": 0.0317,
+      "step": 1316
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3178539575728721,
+      "learning_rate": 7.171502761670032e-06,
+      "loss": 0.0839,
+      "step": 1317
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.36087840792972364,
+      "learning_rate": 7.167334165093407e-06,
+      "loss": 0.0576,
+      "step": 1318
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7533321546317311,
+      "learning_rate": 7.163163712653267e-06,
+      "loss": 0.072,
+      "step": 1319
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8250500485970101,
+      "learning_rate": 7.158991407920721e-06,
+      "loss": 0.0744,
+      "step": 1320
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.22566404858687902,
+      "learning_rate": 7.154817254468467e-06,
+      "loss": 0.0532,
+      "step": 1321
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.35811467808863606,
+      "learning_rate": 7.150641255870783e-06,
+      "loss": 0.089,
+      "step": 1322
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4505878258632678,
+      "learning_rate": 7.14646341570353e-06,
+      "loss": 0.0848,
+      "step": 1323
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.34669718798707816,
+      "learning_rate": 7.142283737544146e-06,
+      "loss": 0.0839,
+      "step": 1324
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.36144477138663855,
+      "learning_rate": 7.13810222497164e-06,
+      "loss": 0.0737,
+      "step": 1325
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.34491353044678136,
+      "learning_rate": 7.133918881566594e-06,
+      "loss": 0.0974,
+      "step": 1326
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4263799438922573,
+      "learning_rate": 7.129733710911159e-06,
+      "loss": 0.0922,
+      "step": 1327
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.2239485715232622,
+      "learning_rate": 7.125546716589046e-06,
+      "loss": 0.0212,
+      "step": 1328
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4832278651309879,
+      "learning_rate": 7.1213579021855325e-06,
+      "loss": 0.099,
+      "step": 1329
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5101730154765555,
+      "learning_rate": 7.117167271287453e-06,
+      "loss": 0.1017,
+      "step": 1330
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.40492803572216257,
+      "learning_rate": 7.112974827483193e-06,
+      "loss": 0.1275,
+      "step": 1331
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.43584343014282445,
+      "learning_rate": 7.108780574362699e-06,
+      "loss": 0.1082,
+      "step": 1332
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.43533816162835104,
+      "learning_rate": 7.104584515517459e-06,
+      "loss": 0.0986,
+      "step": 1333
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.2963252094256265,
+      "learning_rate": 7.100386654540512e-06,
+      "loss": 0.0628,
+      "step": 1334
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4259754685969777,
+      "learning_rate": 7.096186995026439e-06,
+      "loss": 0.0915,
+      "step": 1335
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.2668788584851861,
+      "learning_rate": 7.091985540571358e-06,
+      "loss": 0.0635,
+      "step": 1336
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.47725951760375,
+      "learning_rate": 7.0877822947729265e-06,
+      "loss": 0.1054,
+      "step": 1337
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3467052183988419,
+      "learning_rate": 7.083577261230341e-06,
+      "loss": 0.0859,
+      "step": 1338
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.577650273671718,
+      "learning_rate": 7.079370443544318e-06,
+      "loss": 0.1355,
+      "step": 1339
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.45582154635810423,
+      "learning_rate": 7.07516184531711e-06,
+      "loss": 0.101,
+      "step": 1340
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3014023955248303,
+      "learning_rate": 7.07095147015249e-06,
+      "loss": 0.0802,
+      "step": 1341
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5655097720036469,
+      "learning_rate": 7.066739321655757e-06,
+      "loss": 0.0684,
+      "step": 1342
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.41287951668563994,
+      "learning_rate": 7.062525403433723e-06,
+      "loss": 0.0954,
+      "step": 1343
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.28403130016009487,
+      "learning_rate": 7.05830971909472e-06,
+      "loss": 0.084,
+      "step": 1344
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.30073241076499924,
+      "learning_rate": 7.05409227224859e-06,
+      "loss": 0.083,
+      "step": 1345
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3742776309578899,
+      "learning_rate": 7.049873066506684e-06,
+      "loss": 0.0885,
+      "step": 1346
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.444158701035714,
+      "learning_rate": 7.04565210548186e-06,
+      "loss": 0.1088,
+      "step": 1347
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6000393018205739,
+      "learning_rate": 7.041429392788477e-06,
+      "loss": 0.1029,
+      "step": 1348
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4965731557138154,
+      "learning_rate": 7.0372049320424e-06,
+      "loss": 0.0789,
+      "step": 1349
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6645267790695691,
+      "learning_rate": 7.032978726860981e-06,
+      "loss": 0.1253,
+      "step": 1350
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.41647744727531605,
+      "learning_rate": 7.028750780863078e-06,
+      "loss": 0.0782,
+      "step": 1351
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.49520716867368675,
+      "learning_rate": 7.024521097669026e-06,
+      "loss": 0.1119,
+      "step": 1352
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.885596141103793,
+      "learning_rate": 7.020289680900658e-06,
+      "loss": 0.0834,
+      "step": 1353
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.400198089989043,
+      "learning_rate": 7.0160565341812885e-06,
+      "loss": 0.0735,
+      "step": 1354
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.465346235006083,
+      "learning_rate": 7.0118216611357125e-06,
+      "loss": 0.0857,
+      "step": 1355
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6727623425168257,
+      "learning_rate": 7.007585065390203e-06,
+      "loss": 0.1098,
+      "step": 1356
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4291856686417795,
+      "learning_rate": 7.003346750572506e-06,
+      "loss": 0.0712,
+      "step": 1357
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4471592586419728,
+      "learning_rate": 6.999106720311846e-06,
+      "loss": 0.0685,
+      "step": 1358
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2479345132477074,
+      "learning_rate": 6.99486497823891e-06,
+      "loss": 0.0783,
+      "step": 1359
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5948936709525752,
+      "learning_rate": 6.990621527985856e-06,
+      "loss": 0.0794,
+      "step": 1360
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5471409914024203,
+      "learning_rate": 6.9863763731862984e-06,
+      "loss": 0.1365,
+      "step": 1361
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5551858472302786,
+      "learning_rate": 6.9821295174753175e-06,
+      "loss": 0.0861,
+      "step": 1362
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5523454522986879,
+      "learning_rate": 6.9778809644894475e-06,
+      "loss": 0.1092,
+      "step": 1363
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2730055038951132,
+      "learning_rate": 6.973630717866675e-06,
+      "loss": 0.0525,
+      "step": 1364
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5811598336796037,
+      "learning_rate": 6.969378781246436e-06,
+      "loss": 0.0605,
+      "step": 1365
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3005758151587515,
+      "learning_rate": 6.965125158269619e-06,
+      "loss": 0.0455,
+      "step": 1366
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6381799184781491,
+      "learning_rate": 6.960869852578549e-06,
+      "loss": 0.1451,
+      "step": 1367
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5810543533037971,
+      "learning_rate": 6.956612867816999e-06,
+      "loss": 0.1039,
+      "step": 1368
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3466037402685036,
+      "learning_rate": 6.952354207630174e-06,
+      "loss": 0.0969,
+      "step": 1369
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4397121792763084,
+      "learning_rate": 6.948093875664719e-06,
+      "loss": 0.0933,
+      "step": 1370
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9929070614731769,
+      "learning_rate": 6.943831875568703e-06,
+      "loss": 0.1126,
+      "step": 1371
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.35228268609643915,
+      "learning_rate": 6.939568210991633e-06,
+      "loss": 0.0849,
+      "step": 1372
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.31320050463586374,
+      "learning_rate": 6.935302885584434e-06,
+      "loss": 0.0526,
+      "step": 1373
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.31814123779422676,
+      "learning_rate": 6.931035902999454e-06,
+      "loss": 0.05,
+      "step": 1374
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5358566718177803,
+      "learning_rate": 6.926767266890466e-06,
+      "loss": 0.1168,
+      "step": 1375
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3585165835216518,
+      "learning_rate": 6.92249698091265e-06,
+      "loss": 0.117,
+      "step": 1376
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.24361369021060725,
+      "learning_rate": 6.918225048722604e-06,
+      "loss": 0.0565,
+      "step": 1377
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5300060580289199,
+      "learning_rate": 6.9139514739783364e-06,
+      "loss": 0.1047,
+      "step": 1378
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3235397107082942,
+      "learning_rate": 6.9096762603392595e-06,
+      "loss": 0.0678,
+      "step": 1379
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.43862995403368876,
+      "learning_rate": 6.90539941146619e-06,
+      "loss": 0.1168,
+      "step": 1380
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.19116774365664,
+      "learning_rate": 6.901120931021345e-06,
+      "loss": 0.0935,
+      "step": 1381
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4868001872212555,
+      "learning_rate": 6.896840822668337e-06,
+      "loss": 0.0654,
+      "step": 1382
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3875830997642732,
+      "learning_rate": 6.892559090072177e-06,
+      "loss": 0.1,
+      "step": 1383
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6759822116009601,
+      "learning_rate": 6.888275736899262e-06,
+      "loss": 0.1395,
+      "step": 1384
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.48997515323127844,
+      "learning_rate": 6.883990766817378e-06,
+      "loss": 0.0786,
+      "step": 1385
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.38317298569150376,
+      "learning_rate": 6.8797041834956955e-06,
+      "loss": 0.081,
+      "step": 1386
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5123538753261575,
+      "learning_rate": 6.8754159906047706e-06,
+      "loss": 0.0973,
+      "step": 1387
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4060250824613997,
+      "learning_rate": 6.871126191816529e-06,
+      "loss": 0.1088,
+      "step": 1388
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6699260467691455,
+      "learning_rate": 6.866834790804278e-06,
+      "loss": 0.1409,
+      "step": 1389
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.39217404810140766,
+      "learning_rate": 6.862541791242698e-06,
+      "loss": 0.0903,
+      "step": 1390
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3319049584807593,
+      "learning_rate": 6.858247196807833e-06,
+      "loss": 0.0907,
+      "step": 1391
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.48402241987122746,
+      "learning_rate": 6.853951011177095e-06,
+      "loss": 0.0993,
+      "step": 1392
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.43227601551456063,
+      "learning_rate": 6.849653238029261e-06,
+      "loss": 0.1217,
+      "step": 1393
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3561639464417621,
+      "learning_rate": 6.845353881044463e-06,
+      "loss": 0.0921,
+      "step": 1394
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5367946800369504,
+      "learning_rate": 6.84105294390419e-06,
+      "loss": 0.1309,
+      "step": 1395
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3147138255535657,
+      "learning_rate": 6.83675043029129e-06,
+      "loss": 0.0813,
+      "step": 1396
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.42929135778736255,
+      "learning_rate": 6.832446343889952e-06,
+      "loss": 0.1122,
+      "step": 1397
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2808487389242168,
+      "learning_rate": 6.8281406883857194e-06,
+      "loss": 0.0924,
+      "step": 1398
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3213911099478094,
+      "learning_rate": 6.823833467465473e-06,
+      "loss": 0.0806,
+      "step": 1399
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3767345159241612,
+      "learning_rate": 6.819524684817439e-06,
+      "loss": 0.0874,
+      "step": 1400
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.321662199553273,
+      "learning_rate": 6.8152143441311765e-06,
+      "loss": 0.0403,
+      "step": 1401
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3167712405692955,
+      "learning_rate": 6.810902449097584e-06,
+      "loss": 0.1029,
+      "step": 1402
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3093713780061254,
+      "learning_rate": 6.806589003408886e-06,
+      "loss": 0.0899,
+      "step": 1403
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7724714628357934,
+      "learning_rate": 6.802274010758637e-06,
+      "loss": 0.1413,
+      "step": 1404
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3961863310948659,
+      "learning_rate": 6.797957474841717e-06,
+      "loss": 0.0574,
+      "step": 1405
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3871784418101659,
+      "learning_rate": 6.793639399354324e-06,
+      "loss": 0.1117,
+      "step": 1406
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5945824916656249,
+      "learning_rate": 6.78931978799398e-06,
+      "loss": 0.0919,
+      "step": 1407
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.30112716546299095,
+      "learning_rate": 6.784998644459517e-06,
+      "loss": 0.0491,
+      "step": 1408
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.33613988205627376,
+      "learning_rate": 6.780675972451083e-06,
+      "loss": 0.0663,
+      "step": 1409
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3645211692038736,
+      "learning_rate": 6.776351775670129e-06,
+      "loss": 0.0865,
+      "step": 1410
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.253234124055534,
+      "learning_rate": 6.7720260578194185e-06,
+      "loss": 0.0514,
+      "step": 1411
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3508724331773692,
+      "learning_rate": 6.767698822603013e-06,
+      "loss": 0.0804,
+      "step": 1412
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.26034274444884775,
+      "learning_rate": 6.763370073726276e-06,
+      "loss": 0.0605,
+      "step": 1413
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6546642097245746,
+      "learning_rate": 6.7590398148958625e-06,
+      "loss": 0.1923,
+      "step": 1414
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.40170447802361675,
+      "learning_rate": 6.754708049819728e-06,
+      "loss": 0.0923,
+      "step": 1415
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4235066268178241,
+      "learning_rate": 6.750374782207112e-06,
+      "loss": 0.0516,
+      "step": 1416
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3383309060193509,
+      "learning_rate": 6.74604001576854e-06,
+      "loss": 0.0898,
+      "step": 1417
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4183934545627158,
+      "learning_rate": 6.741703754215825e-06,
+      "loss": 0.1187,
+      "step": 1418
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.29694276385261686,
+      "learning_rate": 6.7373660012620575e-06,
+      "loss": 0.0801,
+      "step": 1419
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.35023179019387557,
+      "learning_rate": 6.733026760621607e-06,
+      "loss": 0.079,
+      "step": 1420
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4280954679718074,
+      "learning_rate": 6.728686036010115e-06,
+      "loss": 0.0545,
+      "step": 1421
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.36828927733308037,
+      "learning_rate": 6.724343831144494e-06,
+      "loss": 0.1062,
+      "step": 1422
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.275826926343428,
+      "learning_rate": 6.720000149742925e-06,
+      "loss": 0.0613,
+      "step": 1423
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4212043430155424,
+      "learning_rate": 6.715654995524853e-06,
+      "loss": 0.106,
+      "step": 1424
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2702659912031575,
+      "learning_rate": 6.711308372210983e-06,
+      "loss": 0.0876,
+      "step": 1425
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4948367930619922,
+      "learning_rate": 6.706960283523282e-06,
+      "loss": 0.0878,
+      "step": 1426
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5187744933374802,
+      "learning_rate": 6.702610733184965e-06,
+      "loss": 0.106,
+      "step": 1427
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.48915435202940594,
+      "learning_rate": 6.698259724920503e-06,
+      "loss": 0.1079,
+      "step": 1428
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3784939134253547,
+      "learning_rate": 6.6939072624556155e-06,
+      "loss": 0.0674,
+      "step": 1429
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.28085894067989636,
+      "learning_rate": 6.689553349517268e-06,
+      "loss": 0.0776,
+      "step": 1430
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.89551810622946,
+      "learning_rate": 6.685197989833665e-06,
+      "loss": 0.1587,
+      "step": 1431
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.41601809001145373,
+      "learning_rate": 6.68084118713425e-06,
+      "loss": 0.0905,
+      "step": 1432
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7815321799772386,
+      "learning_rate": 6.676482945149705e-06,
+      "loss": 0.1905,
+      "step": 1433
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7439023887847122,
+      "learning_rate": 6.672123267611942e-06,
+      "loss": 0.1396,
+      "step": 1434
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.28561834571202693,
+      "learning_rate": 6.667762158254104e-06,
+      "loss": 0.0548,
+      "step": 1435
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.37121560968314343,
+      "learning_rate": 6.663399620810559e-06,
+      "loss": 0.0713,
+      "step": 1436
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5599469503795745,
+      "learning_rate": 6.659035659016898e-06,
+      "loss": 0.1428,
+      "step": 1437
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4501706729423448,
+      "learning_rate": 6.654670276609932e-06,
+      "loss": 0.1264,
+      "step": 1438
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5229707122928983,
+      "learning_rate": 6.650303477327686e-06,
+      "loss": 0.1049,
+      "step": 1439
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.859696260542617,
+      "learning_rate": 6.645935264909404e-06,
+      "loss": 0.118,
+      "step": 1440
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5737426947982244,
+      "learning_rate": 6.641565643095534e-06,
+      "loss": 0.1233,
+      "step": 1441
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5524141557794943,
+      "learning_rate": 6.637194615627733e-06,
+      "loss": 0.0829,
+      "step": 1442
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2981411094203284,
+      "learning_rate": 6.632822186248865e-06,
+      "loss": 0.0589,
+      "step": 1443
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.40743741308632836,
+      "learning_rate": 6.628448358702988e-06,
+      "loss": 0.098,
+      "step": 1444
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4791202138734424,
+      "learning_rate": 6.6240731367353624e-06,
+      "loss": 0.1442,
+      "step": 1445
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7185057917214397,
+      "learning_rate": 6.619696524092439e-06,
+      "loss": 0.1549,
+      "step": 1446
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.34993534334995197,
+      "learning_rate": 6.6153185245218645e-06,
+      "loss": 0.0995,
+      "step": 1447
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4304917727215731,
+      "learning_rate": 6.610939141772467e-06,
+      "loss": 0.1462,
+      "step": 1448
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5365579848682441,
+      "learning_rate": 6.6065583795942625e-06,
+      "loss": 0.1064,
+      "step": 1449
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.26302471257781734,
+      "learning_rate": 6.602176241738449e-06,
+      "loss": 0.0799,
+      "step": 1450
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.29495288160584304,
+      "learning_rate": 6.597792731957399e-06,
+      "loss": 0.0733,
+      "step": 1451
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5462077923402763,
+      "learning_rate": 6.593407854004666e-06,
+      "loss": 0.1174,
+      "step": 1452
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.49316568084281137,
+      "learning_rate": 6.589021611634966e-06,
+      "loss": 0.1103,
+      "step": 1453
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5132963762830282,
+      "learning_rate": 6.584634008604191e-06,
+      "loss": 0.0898,
+      "step": 1454
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.26852926328328786,
+      "learning_rate": 6.580245048669395e-06,
+      "loss": 0.0703,
+      "step": 1455
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.195147398121515,
+      "learning_rate": 6.5758547355887944e-06,
+      "loss": 0.0509,
+      "step": 1456
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2298303915335172,
+      "learning_rate": 6.571463073121763e-06,
+      "loss": 0.0582,
+      "step": 1457
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3338043228952181,
+      "learning_rate": 6.567070065028833e-06,
+      "loss": 0.0654,
+      "step": 1458
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.46569377694297515,
+      "learning_rate": 6.562675715071687e-06,
+      "loss": 0.1032,
+      "step": 1459
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.23718009946674504,
+      "learning_rate": 6.558280027013155e-06,
+      "loss": 0.0574,
+      "step": 1460
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.36490650797169855,
+      "learning_rate": 6.5538830046172175e-06,
+      "loss": 0.0872,
+      "step": 1461
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.27201650887410783,
+      "learning_rate": 6.549484651648991e-06,
+      "loss": 0.0726,
+      "step": 1462
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8062505615723329,
+      "learning_rate": 6.545084971874738e-06,
+      "loss": 0.1241,
+      "step": 1463
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.31937336599729654,
+      "learning_rate": 6.540683969061852e-06,
+      "loss": 0.0846,
+      "step": 1464
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3313888405197613,
+      "learning_rate": 6.536281646978863e-06,
+      "loss": 0.1053,
+      "step": 1465
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.599728482274412,
+      "learning_rate": 6.5318780093954274e-06,
+      "loss": 0.1326,
+      "step": 1466
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.41134979401673594,
+      "learning_rate": 6.527473060082332e-06,
+      "loss": 0.0952,
+      "step": 1467
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.46696607191491313,
+      "learning_rate": 6.523066802811483e-06,
+      "loss": 0.0854,
+      "step": 1468
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7528784570553645,
+      "learning_rate": 6.5186592413559095e-06,
+      "loss": 0.149,
+      "step": 1469
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3896412581457436,
+      "learning_rate": 6.514250379489754e-06,
+      "loss": 0.0834,
+      "step": 1470
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3704572654202895,
+      "learning_rate": 6.509840220988277e-06,
+      "loss": 0.1132,
+      "step": 1471
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3044333766039238,
+      "learning_rate": 6.5054287696278455e-06,
+      "loss": 0.0467,
+      "step": 1472
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3783897809573116,
+      "learning_rate": 6.501016029185936e-06,
+      "loss": 0.0907,
+      "step": 1473
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.30012817272031916,
+      "learning_rate": 6.496602003441127e-06,
+      "loss": 0.0789,
+      "step": 1474
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.44966823770999975,
+      "learning_rate": 6.492186696173097e-06,
+      "loss": 0.0966,
+      "step": 1475
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3876318539712018,
+      "learning_rate": 6.4877701111626266e-06,
+      "loss": 0.0542,
+      "step": 1476
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.33770978651482897,
+      "learning_rate": 6.483352252191585e-06,
+      "loss": 0.0642,
+      "step": 1477
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3674837351329926,
+      "learning_rate": 6.478933123042934e-06,
+      "loss": 0.0373,
+      "step": 1478
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.30121807029193237,
+      "learning_rate": 6.474512727500725e-06,
+      "loss": 0.0827,
+      "step": 1479
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5511239489168589,
+      "learning_rate": 6.47009106935009e-06,
+      "loss": 0.1201,
+      "step": 1480
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5329639725909865,
+      "learning_rate": 6.465668152377247e-06,
+      "loss": 0.1094,
+      "step": 1481
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.467998118357438,
+      "learning_rate": 6.461243980369489e-06,
+      "loss": 0.098,
+      "step": 1482
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.27611797369931845,
+      "learning_rate": 6.456818557115182e-06,
+      "loss": 0.0666,
+      "step": 1483
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.504297881148237,
+      "learning_rate": 6.452391886403767e-06,
+      "loss": 0.0926,
+      "step": 1484
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.47484383288801263,
+      "learning_rate": 6.447963972025752e-06,
+      "loss": 0.0678,
+      "step": 1485
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7171332966002333,
+      "learning_rate": 6.443534817772707e-06,
+      "loss": 0.1475,
+      "step": 1486
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5516107285527765,
+      "learning_rate": 6.4391044274372706e-06,
+      "loss": 0.1202,
+      "step": 1487
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7818961227000709,
+      "learning_rate": 6.434672804813131e-06,
+      "loss": 0.1328,
+      "step": 1488
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.38446665166782235,
+      "learning_rate": 6.430239953695039e-06,
+      "loss": 0.0565,
+      "step": 1489
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6605993525909495,
+      "learning_rate": 6.425805877878794e-06,
+      "loss": 0.0956,
+      "step": 1490
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6157783453818049,
+      "learning_rate": 6.421370581161244e-06,
+      "loss": 0.1527,
+      "step": 1491
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.41730228265956526,
+      "learning_rate": 6.4169340673402814e-06,
+      "loss": 0.1015,
+      "step": 1492
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7657718379287314,
+      "learning_rate": 6.412496340214846e-06,
+      "loss": 0.0909,
+      "step": 1493
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4930339127764775,
+      "learning_rate": 6.4080574035849096e-06,
+      "loss": 0.0811,
+      "step": 1494
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6344653860002332,
+      "learning_rate": 6.403617261251485e-06,
+      "loss": 0.0954,
+      "step": 1495
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.550946449107486,
+      "learning_rate": 6.399175917016613e-06,
+      "loss": 0.0728,
+      "step": 1496
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2713422016625722,
+      "learning_rate": 6.394733374683371e-06,
+      "loss": 0.0615,
+      "step": 1497
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9531303100198866,
+      "learning_rate": 6.390289638055851e-06,
+      "loss": 0.0648,
+      "step": 1498
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.40299138120665573,
+      "learning_rate": 6.385844710939179e-06,
+      "loss": 0.0745,
+      "step": 1499
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5637010517456598,
+      "learning_rate": 6.381398597139492e-06,
+      "loss": 0.1182,
+      "step": 1500
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.28354064126598905,
+      "learning_rate": 6.376951300463948e-06,
+      "loss": 0.0579,
+      "step": 1501
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5843975672098469,
+      "learning_rate": 6.372502824720716e-06,
+      "loss": 0.1196,
+      "step": 1502
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5052794192271838,
+      "learning_rate": 6.368053173718978e-06,
+      "loss": 0.1069,
+      "step": 1503
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5564935999984026,
+      "learning_rate": 6.363602351268913e-06,
+      "loss": 0.1397,
+      "step": 1504
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5095795974359075,
+      "learning_rate": 6.3591503611817155e-06,
+      "loss": 0.1064,
+      "step": 1505
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6050161213150755,
+      "learning_rate": 6.35469720726957e-06,
+      "loss": 0.1015,
+      "step": 1506
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2588092761354788,
+      "learning_rate": 6.350242893345664e-06,
+      "loss": 0.0446,
+      "step": 1507
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4729876848485526,
+      "learning_rate": 6.345787423224174e-06,
+      "loss": 0.1086,
+      "step": 1508
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2992442928232585,
+      "learning_rate": 6.341330800720269e-06,
+      "loss": 0.0793,
+      "step": 1509
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4261129994450151,
+      "learning_rate": 6.336873029650104e-06,
+      "loss": 0.089,
+      "step": 1510
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4443368196552463,
+      "learning_rate": 6.332414113830816e-06,
+      "loss": 0.1577,
+      "step": 1511
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.27405575464330806,
+      "learning_rate": 6.3279540570805265e-06,
+      "loss": 0.049,
+      "step": 1512
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2245225957989012,
+      "learning_rate": 6.3234928632183276e-06,
+      "loss": 0.1352,
+      "step": 1513
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5191731693081545,
+      "learning_rate": 6.319030536064295e-06,
+      "loss": 0.1246,
+      "step": 1514
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4021125762969779,
+      "learning_rate": 6.3145670794394595e-06,
+      "loss": 0.093,
+      "step": 1515
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2847701154707854,
+      "learning_rate": 6.310102497165836e-06,
+      "loss": 0.0495,
+      "step": 1516
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2923960049003912,
+      "learning_rate": 6.305636793066392e-06,
+      "loss": 0.081,
+      "step": 1517
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.617457696014792,
+      "learning_rate": 6.3011699709650586e-06,
+      "loss": 0.0788,
+      "step": 1518
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4874484095341976,
+      "learning_rate": 6.296702034686726e-06,
+      "loss": 0.0467,
+      "step": 1519
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.26254787568666776,
+      "learning_rate": 6.292232988057235e-06,
+      "loss": 0.0705,
+      "step": 1520
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.45079075618602954,
+      "learning_rate": 6.2877628349033825e-06,
+      "loss": 0.1232,
+      "step": 1521
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4535865482010611,
+      "learning_rate": 6.283291579052906e-06,
+      "loss": 0.0748,
+      "step": 1522
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3932746346044528,
+      "learning_rate": 6.2788192243344935e-06,
+      "loss": 0.0929,
+      "step": 1523
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5699864981861645,
+      "learning_rate": 6.27434577457777e-06,
+      "loss": 0.0876,
+      "step": 1524
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2960336863436895,
+      "learning_rate": 6.2698712336133e-06,
+      "loss": 0.075,
+      "step": 1525
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2852590643376082,
+      "learning_rate": 6.265395605272581e-06,
+      "loss": 0.0726,
+      "step": 1526
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.30239596225304155,
+      "learning_rate": 6.260918893388045e-06,
+      "loss": 0.0838,
+      "step": 1527
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3464618142262354,
+      "learning_rate": 6.256441101793046e-06,
+      "loss": 0.0872,
+      "step": 1528
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3643088366043018,
+      "learning_rate": 6.251962234321869e-06,
+      "loss": 0.0881,
+      "step": 1529
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.512353931165177,
+      "learning_rate": 6.247482294809712e-06,
+      "loss": 0.141,
+      "step": 1530
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3101410930114144,
+      "learning_rate": 6.243001287092704e-06,
+      "loss": 0.0491,
+      "step": 1531
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3687686011837704,
+      "learning_rate": 6.238519215007874e-06,
+      "loss": 0.1096,
+      "step": 1532
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.34002579100593294,
+      "learning_rate": 6.234036082393171e-06,
+      "loss": 0.0914,
+      "step": 1533
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3508100813297846,
+      "learning_rate": 6.229551893087453e-06,
+      "loss": 0.099,
+      "step": 1534
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4404280287492867,
+      "learning_rate": 6.225066650930476e-06,
+      "loss": 0.1161,
+      "step": 1535
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5343769153846271,
+      "learning_rate": 6.2205803597629054e-06,
+      "loss": 0.103,
+      "step": 1536
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.27792273844699805,
+      "learning_rate": 6.2160930234262976e-06,
+      "loss": 0.0686,
+      "step": 1537
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4600398166711245,
+      "learning_rate": 6.21160464576311e-06,
+      "loss": 0.1058,
+      "step": 1538
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.43877281003570706,
+      "learning_rate": 6.207115230616689e-06,
+      "loss": 0.0945,
+      "step": 1539
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5050427102883335,
+      "learning_rate": 6.202624781831269e-06,
+      "loss": 0.122,
+      "step": 1540
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5784410529318813,
+      "learning_rate": 6.19813330325197e-06,
+      "loss": 0.1267,
+      "step": 1541
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4745251410287598,
+      "learning_rate": 6.193640798724794e-06,
+      "loss": 0.1139,
+      "step": 1542
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8324889040644113,
+      "learning_rate": 6.18914727209662e-06,
+      "loss": 0.1506,
+      "step": 1543
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.570287586282885,
+      "learning_rate": 6.184652727215207e-06,
+      "loss": 0.0856,
+      "step": 1544
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5169236252677503,
+      "learning_rate": 6.18015716792918e-06,
+      "loss": 0.1102,
+      "step": 1545
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.48575848359046936,
+      "learning_rate": 6.175660598088034e-06,
+      "loss": 0.092,
+      "step": 1546
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.33627824765172376,
+      "learning_rate": 6.171163021542134e-06,
+      "loss": 0.0916,
+      "step": 1547
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.39397531807969416,
+      "learning_rate": 6.1666644421427015e-06,
+      "loss": 0.07,
+      "step": 1548
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6584198092141058,
+      "learning_rate": 6.162164863741817e-06,
+      "loss": 0.1271,
+      "step": 1549
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6403512448013448,
+      "learning_rate": 6.157664290192421e-06,
+      "loss": 0.1301,
+      "step": 1550
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3767507345736296,
+      "learning_rate": 6.1531627253483025e-06,
+      "loss": 0.073,
+      "step": 1551
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.34170917949479634,
+      "learning_rate": 6.148660173064098e-06,
+      "loss": 0.088,
+      "step": 1552
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4116452204062582,
+      "learning_rate": 6.144156637195296e-06,
+      "loss": 0.0668,
+      "step": 1553
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.42638117894199135,
+      "learning_rate": 6.139652121598219e-06,
+      "loss": 0.0918,
+      "step": 1554
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3220134925525686,
+      "learning_rate": 6.135146630130033e-06,
+      "loss": 0.0539,
+      "step": 1555
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.44043160920308577,
+      "learning_rate": 6.13064016664874e-06,
+      "loss": 0.108,
+      "step": 1556
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3188262864721422,
+      "learning_rate": 6.1261327350131726e-06,
+      "loss": 0.0688,
+      "step": 1557
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4866909580068489,
+      "learning_rate": 6.1216243390829926e-06,
+      "loss": 0.0987,
+      "step": 1558
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3897816169344711,
+      "learning_rate": 6.11711498271869e-06,
+      "loss": 0.113,
+      "step": 1559
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.33934496018117577,
+      "learning_rate": 6.112604669781572e-06,
+      "loss": 0.0818,
+      "step": 1560
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.47644992139601133,
+      "learning_rate": 6.108093404133772e-06,
+      "loss": 0.0799,
+      "step": 1561
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.29261316731482023,
+      "learning_rate": 6.103581189638232e-06,
+      "loss": 0.0595,
+      "step": 1562
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.41203875162572184,
+      "learning_rate": 6.099068030158712e-06,
+      "loss": 0.073,
+      "step": 1563
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2928798237478887,
+      "learning_rate": 6.094553929559778e-06,
+      "loss": 0.0478,
+      "step": 1564
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.44021163728740104,
+      "learning_rate": 6.090038891706801e-06,
+      "loss": 0.1198,
+      "step": 1565
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.291893015226659,
+      "learning_rate": 6.08552292046596e-06,
+      "loss": 0.0969,
+      "step": 1566
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.24413792006020263,
+      "learning_rate": 6.081006019704227e-06,
+      "loss": 0.0499,
+      "step": 1567
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5508667269722093,
+      "learning_rate": 6.076488193289375e-06,
+      "loss": 0.1126,
+      "step": 1568
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2722830665827193,
+      "learning_rate": 6.071969445089965e-06,
+      "loss": 0.0641,
+      "step": 1569
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4868550402933533,
+      "learning_rate": 6.06744977897535e-06,
+      "loss": 0.1031,
+      "step": 1570
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.607048435151134,
+      "learning_rate": 6.0629291988156676e-06,
+      "loss": 0.1195,
+      "step": 1571
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4025357090256397,
+      "learning_rate": 6.058407708481843e-06,
+      "loss": 0.1115,
+      "step": 1572
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5038586230503376,
+      "learning_rate": 6.053885311845571e-06,
+      "loss": 0.131,
+      "step": 1573
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.34539097482969766,
+      "learning_rate": 6.049362012779333e-06,
+      "loss": 0.0633,
+      "step": 1574
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3174818014038329,
+      "learning_rate": 6.044837815156377e-06,
+      "loss": 0.0796,
+      "step": 1575
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5081576105379931,
+      "learning_rate": 6.04031272285072e-06,
+      "loss": 0.0929,
+      "step": 1576
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.38081862575066916,
+      "learning_rate": 6.035786739737148e-06,
+      "loss": 0.0959,
+      "step": 1577
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4209994258378272,
+      "learning_rate": 6.0312598696912096e-06,
+      "loss": 0.0468,
+      "step": 1578
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4831905612133384,
+      "learning_rate": 6.026732116589211e-06,
+      "loss": 0.1195,
+      "step": 1579
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.554745652322106,
+      "learning_rate": 6.022203484308217e-06,
+      "loss": 0.0795,
+      "step": 1580
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8422194257132133,
+      "learning_rate": 6.0176739767260415e-06,
+      "loss": 0.1266,
+      "step": 1581
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2629273377921984,
+      "learning_rate": 6.013143597721252e-06,
+      "loss": 0.0778,
+      "step": 1582
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.46135694152117934,
+      "learning_rate": 6.008612351173159e-06,
+      "loss": 0.1339,
+      "step": 1583
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5323434048873409,
+      "learning_rate": 6.004080240961818e-06,
+      "loss": 0.0572,
+      "step": 1584
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.35903437337253497,
+      "learning_rate": 5.999547270968024e-06,
+      "loss": 0.0696,
+      "step": 1585
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3355668525678312,
+      "learning_rate": 5.9950134450733045e-06,
+      "loss": 0.0958,
+      "step": 1586
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5217738759488709,
+      "learning_rate": 5.990478767159926e-06,
+      "loss": 0.06,
+      "step": 1587
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.008012013800437,
+      "learning_rate": 5.985943241110881e-06,
+      "loss": 0.0652,
+      "step": 1588
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.22873010594932353,
+      "learning_rate": 5.981406870809889e-06,
+      "loss": 0.0562,
+      "step": 1589
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3172153423836355,
+      "learning_rate": 5.976869660141389e-06,
+      "loss": 0.0641,
+      "step": 1590
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.22765343451518014,
+      "learning_rate": 5.972331612990546e-06,
+      "loss": 0.0622,
+      "step": 1591
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.44709331552672515,
+      "learning_rate": 5.967792733243239e-06,
+      "loss": 0.11,
+      "step": 1592
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7527874138092266,
+      "learning_rate": 5.963253024786053e-06,
+      "loss": 0.1596,
+      "step": 1593
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.461994906572781,
+      "learning_rate": 5.958712491506295e-06,
+      "loss": 0.0871,
+      "step": 1594
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.42772132151415737,
+      "learning_rate": 5.954171137291968e-06,
+      "loss": 0.1114,
+      "step": 1595
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3103664339231245,
+      "learning_rate": 5.949628966031785e-06,
+      "loss": 0.0807,
+      "step": 1596
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.44769070484034534,
+      "learning_rate": 5.945085981615153e-06,
+      "loss": 0.1122,
+      "step": 1597
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.273929006442074,
+      "learning_rate": 5.9405421879321775e-06,
+      "loss": 0.0522,
+      "step": 1598
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6064356448255825,
+      "learning_rate": 5.9359975888736575e-06,
+      "loss": 0.1255,
+      "step": 1599
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5084612825724896,
+      "learning_rate": 5.931452188331084e-06,
+      "loss": 0.0977,
+      "step": 1600
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5017692422046749,
+      "learning_rate": 5.9269059901966276e-06,
+      "loss": 0.1213,
+      "step": 1601
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5983936099829104,
+      "learning_rate": 5.922358998363148e-06,
+      "loss": 0.1326,
+      "step": 1602
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5481813785522884,
+      "learning_rate": 5.9178112167241805e-06,
+      "loss": 0.1221,
+      "step": 1603
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.28442226686611066,
+      "learning_rate": 5.9132626491739434e-06,
+      "loss": 0.0912,
+      "step": 1604
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.521759258678626,
+      "learning_rate": 5.908713299607318e-06,
+      "loss": 0.1018,
+      "step": 1605
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3343345513119317,
+      "learning_rate": 5.904163171919863e-06,
+      "loss": 0.0762,
+      "step": 1606
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4843724584068069,
+      "learning_rate": 5.8996122700077995e-06,
+      "loss": 0.0725,
+      "step": 1607
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5069068934350319,
+      "learning_rate": 5.8950605977680156e-06,
+      "loss": 0.1411,
+      "step": 1608
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8410660748434027,
+      "learning_rate": 5.890508159098054e-06,
+      "loss": 0.1587,
+      "step": 1609
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4906275464104598,
+      "learning_rate": 5.885954957896115e-06,
+      "loss": 0.0849,
+      "step": 1610
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.41330337650584437,
+      "learning_rate": 5.8814009980610556e-06,
+      "loss": 0.0716,
+      "step": 1611
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.34820977706585926,
+      "learning_rate": 5.8768462834923765e-06,
+      "loss": 0.1004,
+      "step": 1612
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.29546894717527333,
+      "learning_rate": 5.87229081809023e-06,
+      "loss": 0.0455,
+      "step": 1613
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4006065237789604,
+      "learning_rate": 5.86773460575541e-06,
+      "loss": 0.0864,
+      "step": 1614
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.371814294402616,
+      "learning_rate": 5.863177650389346e-06,
+      "loss": 0.0568,
+      "step": 1615
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2632650599541928,
+      "learning_rate": 5.85861995589411e-06,
+      "loss": 0.0801,
+      "step": 1616
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3400357185661022,
+      "learning_rate": 5.854061526172402e-06,
+      "loss": 0.0604,
+      "step": 1617
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2766265717611615,
+      "learning_rate": 5.849502365127555e-06,
+      "loss": 0.0538,
+      "step": 1618
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.32857827802430734,
+      "learning_rate": 5.844942476663524e-06,
+      "loss": 0.0723,
+      "step": 1619
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.34501487504396006,
+      "learning_rate": 5.840381864684892e-06,
+      "loss": 0.0568,
+      "step": 1620
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2920261742900055,
+      "learning_rate": 5.8358205330968565e-06,
+      "loss": 0.0889,
+      "step": 1621
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.26203902687466885,
+      "learning_rate": 5.8312584858052366e-06,
+      "loss": 0.0574,
+      "step": 1622
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5179357804745687,
+      "learning_rate": 5.826695726716459e-06,
+      "loss": 0.1241,
+      "step": 1623
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5270969682793168,
+      "learning_rate": 5.822132259737565e-06,
+      "loss": 0.119,
+      "step": 1624
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.43987989675138955,
+      "learning_rate": 5.817568088776195e-06,
+      "loss": 0.1053,
+      "step": 1625
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3553046774849037,
+      "learning_rate": 5.8130032177406e-06,
+      "loss": 0.0751,
+      "step": 1626
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.38753349293150197,
+      "learning_rate": 5.808437650539625e-06,
+      "loss": 0.0901,
+      "step": 1627
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3550301524423936,
+      "learning_rate": 5.803871391082716e-06,
+      "loss": 0.076,
+      "step": 1628
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.47481754395175735,
+      "learning_rate": 5.799304443279905e-06,
+      "loss": 0.1049,
+      "step": 1629
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5041031416435303,
+      "learning_rate": 5.794736811041821e-06,
+      "loss": 0.1303,
+      "step": 1630
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.40276753935917553,
+      "learning_rate": 5.7901684982796716e-06,
+      "loss": 0.1103,
+      "step": 1631
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3539687959012876,
+      "learning_rate": 5.785599508905254e-06,
+      "loss": 0.0908,
+      "step": 1632
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.36296725857750745,
+      "learning_rate": 5.7810298468309404e-06,
+      "loss": 0.0909,
+      "step": 1633
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4679603141707834,
+      "learning_rate": 5.776459515969681e-06,
+      "loss": 0.1416,
+      "step": 1634
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3552108408298559,
+      "learning_rate": 5.771888520234997e-06,
+      "loss": 0.0715,
+      "step": 1635
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.45845237300279773,
+      "learning_rate": 5.767316863540979e-06,
+      "loss": 0.0903,
+      "step": 1636
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3333494974439071,
+      "learning_rate": 5.762744549802288e-06,
+      "loss": 0.0787,
+      "step": 1637
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3663448907549116,
+      "learning_rate": 5.75817158293414e-06,
+      "loss": 0.0858,
+      "step": 1638
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.33256693244323554,
+      "learning_rate": 5.7535979668523175e-06,
+      "loss": 0.0868,
+      "step": 1639
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.30085328869337624,
+      "learning_rate": 5.749023705473154e-06,
+      "loss": 0.0613,
+      "step": 1640
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.31145374124401287,
+      "learning_rate": 5.744448802713537e-06,
+      "loss": 0.0643,
+      "step": 1641
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5108038140962555,
+      "learning_rate": 5.739873262490905e-06,
+      "loss": 0.0952,
+      "step": 1642
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5800985739639948,
+      "learning_rate": 5.7352970887232395e-06,
+      "loss": 0.1517,
+      "step": 1643
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3264374011859333,
+      "learning_rate": 5.730720285329067e-06,
+      "loss": 0.0551,
+      "step": 1644
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2857478479789638,
+      "learning_rate": 5.726142856227453e-06,
+      "loss": 0.0789,
+      "step": 1645
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.42996187472922226,
+      "learning_rate": 5.721564805337994e-06,
+      "loss": 0.0867,
+      "step": 1646
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7146067840887904,
+      "learning_rate": 5.716986136580827e-06,
+      "loss": 0.102,
+      "step": 1647
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4221017557141259,
+      "learning_rate": 5.712406853876611e-06,
+      "loss": 0.0924,
+      "step": 1648
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.44854720995597286,
+      "learning_rate": 5.7078269611465355e-06,
+      "loss": 0.1095,
+      "step": 1649
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5802845790628374,
+      "learning_rate": 5.703246462312307e-06,
+      "loss": 0.1045,
+      "step": 1650
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3941843737136875,
+      "learning_rate": 5.698665361296159e-06,
+      "loss": 0.0834,
+      "step": 1651
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3819219745328019,
+      "learning_rate": 5.694083662020835e-06,
+      "loss": 0.0957,
+      "step": 1652
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.41415081906132956,
+      "learning_rate": 5.689501368409588e-06,
+      "loss": 0.1024,
+      "step": 1653
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4979139055435071,
+      "learning_rate": 5.684918484386188e-06,
+      "loss": 0.086,
+      "step": 1654
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3073305612027108,
+      "learning_rate": 5.680335013874903e-06,
+      "loss": 0.0687,
+      "step": 1655
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3042939310338812,
+      "learning_rate": 5.6757509608005104e-06,
+      "loss": 0.0628,
+      "step": 1656
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4752393760198132,
+      "learning_rate": 5.671166329088278e-06,
+      "loss": 0.1245,
+      "step": 1657
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6190283409663718,
+      "learning_rate": 5.666581122663978e-06,
+      "loss": 0.1435,
+      "step": 1658
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4020750810602729,
+      "learning_rate": 5.661995345453867e-06,
+      "loss": 0.1079,
+      "step": 1659
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4453041951082366,
+      "learning_rate": 5.657409001384695e-06,
+      "loss": 0.0819,
+      "step": 1660
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5326253661866354,
+      "learning_rate": 5.652822094383697e-06,
+      "loss": 0.1189,
+      "step": 1661
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6304797285655436,
+      "learning_rate": 5.6482346283785875e-06,
+      "loss": 0.1007,
+      "step": 1662
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.345498055989681,
+      "learning_rate": 5.643646607297562e-06,
+      "loss": 0.068,
+      "step": 1663
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.30374633281909824,
+      "learning_rate": 5.63905803506929e-06,
+      "loss": 0.0869,
+      "step": 1664
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3167988165275447,
+      "learning_rate": 5.634468915622915e-06,
+      "loss": 0.0779,
+      "step": 1665
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.34570834283436513,
+      "learning_rate": 5.629879252888046e-06,
+      "loss": 0.0842,
+      "step": 1666
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.28167780729170155,
+      "learning_rate": 5.625289050794761e-06,
+      "loss": 0.0797,
+      "step": 1667
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3643783364716649,
+      "learning_rate": 5.6206983132735946e-06,
+      "loss": 0.1014,
+      "step": 1668
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.43282734392081657,
+      "learning_rate": 5.6161070442555465e-06,
+      "loss": 0.0943,
+      "step": 1669
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.49469597540911775,
+      "learning_rate": 5.611515247672063e-06,
+      "loss": 0.128,
+      "step": 1670
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3600198581892039,
+      "learning_rate": 5.606922927455054e-06,
+      "loss": 0.0927,
+      "step": 1671
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4649005442976091,
+      "learning_rate": 5.602330087536865e-06,
+      "loss": 0.1025,
+      "step": 1672
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4889004397083401,
+      "learning_rate": 5.597736731850295e-06,
+      "loss": 0.1147,
+      "step": 1673
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5386122076842417,
+      "learning_rate": 5.593142864328581e-06,
+      "loss": 0.0858,
+      "step": 1674
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5213397170894769,
+      "learning_rate": 5.588548488905402e-06,
+      "loss": 0.105,
+      "step": 1675
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.339827614017928,
+      "learning_rate": 5.583953609514865e-06,
+      "loss": 0.0575,
+      "step": 1676
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4752875943339977,
+      "learning_rate": 5.579358230091516e-06,
+      "loss": 0.0552,
+      "step": 1677
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6416464185421734,
+      "learning_rate": 5.574762354570326e-06,
+      "loss": 0.1486,
+      "step": 1678
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3603957152814472,
+      "learning_rate": 5.570165986886689e-06,
+      "loss": 0.1159,
+      "step": 1679
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.39091485209605414,
+      "learning_rate": 5.5655691309764225e-06,
+      "loss": 0.0692,
+      "step": 1680
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4486385227761031,
+      "learning_rate": 5.560971790775762e-06,
+      "loss": 0.1328,
+      "step": 1681
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3301996352135462,
+      "learning_rate": 5.556373970221358e-06,
+      "loss": 0.0804,
+      "step": 1682
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.48044802766468475,
+      "learning_rate": 5.551775673250271e-06,
+      "loss": 0.1247,
+      "step": 1683
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4630108842458435,
+      "learning_rate": 5.547176903799972e-06,
+      "loss": 0.0786,
+      "step": 1684
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5722662598875865,
+      "learning_rate": 5.542577665808332e-06,
+      "loss": 0.1361,
+      "step": 1685
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4614918847166946,
+      "learning_rate": 5.5379779632136284e-06,
+      "loss": 0.1133,
+      "step": 1686
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.38964247144139563,
+      "learning_rate": 5.533377799954532e-06,
+      "loss": 0.1278,
+      "step": 1687
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.381542864421644,
+      "learning_rate": 5.528777179970114e-06,
+      "loss": 0.1053,
+      "step": 1688
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5333533581918855,
+      "learning_rate": 5.524176107199828e-06,
+      "loss": 0.1012,
+      "step": 1689
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.2590788715383924,
+      "learning_rate": 5.519574585583523e-06,
+      "loss": 0.0649,
+      "step": 1690
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5322182709840814,
+      "learning_rate": 5.514972619061427e-06,
+      "loss": 0.1026,
+      "step": 1691
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.35366858025515724,
+      "learning_rate": 5.510370211574156e-06,
+      "loss": 0.0888,
+      "step": 1692
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3886167683604698,
+      "learning_rate": 5.505767367062695e-06,
+      "loss": 0.1082,
+      "step": 1693
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.32878539158919307,
+      "learning_rate": 5.501164089468406e-06,
+      "loss": 0.0599,
+      "step": 1694
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.449508249233077,
+      "learning_rate": 5.496560382733028e-06,
+      "loss": 0.08,
+      "step": 1695
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3450803689585386,
+      "learning_rate": 5.491956250798658e-06,
+      "loss": 0.101,
+      "step": 1696
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.38664765688421115,
+      "learning_rate": 5.487351697607765e-06,
+      "loss": 0.0458,
+      "step": 1697
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6695114807096503,
+      "learning_rate": 5.482746727103174e-06,
+      "loss": 0.0793,
+      "step": 1698
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.34257299124486956,
+      "learning_rate": 5.4781413432280685e-06,
+      "loss": 0.0911,
+      "step": 1699
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2698521025005036,
+      "learning_rate": 5.473535549925986e-06,
+      "loss": 0.0419,
+      "step": 1700
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4596661061762761,
+      "learning_rate": 5.4689293511408155e-06,
+      "loss": 0.1221,
+      "step": 1701
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6884284811535573,
+      "learning_rate": 5.464322750816791e-06,
+      "loss": 0.1311,
+      "step": 1702
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.70145172750144,
+      "learning_rate": 5.459715752898494e-06,
+      "loss": 0.1311,
+      "step": 1703
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3882456722292968,
+      "learning_rate": 5.455108361330843e-06,
+      "loss": 0.0722,
+      "step": 1704
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7701544381387703,
+      "learning_rate": 5.450500580059095e-06,
+      "loss": 0.162,
+      "step": 1705
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.39029789328923986,
+      "learning_rate": 5.445892413028839e-06,
+      "loss": 0.094,
+      "step": 1706
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2728642985894729,
+      "learning_rate": 5.441283864186e-06,
+      "loss": 0.0599,
+      "step": 1707
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5439590225895382,
+      "learning_rate": 5.43667493747682e-06,
+      "loss": 0.1258,
+      "step": 1708
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3635704725591149,
+      "learning_rate": 5.432065636847876e-06,
+      "loss": 0.0937,
+      "step": 1709
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.44263235253746064,
+      "learning_rate": 5.427455966246057e-06,
+      "loss": 0.0911,
+      "step": 1710
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3852561623062434,
+      "learning_rate": 5.4228459296185696e-06,
+      "loss": 0.0943,
+      "step": 1711
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2710827522782717,
+      "learning_rate": 5.418235530912939e-06,
+      "loss": 0.1595,
+      "step": 1712
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.38866509341925787,
+      "learning_rate": 5.413624774076993e-06,
+      "loss": 0.0485,
+      "step": 1713
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3768872146849069,
+      "learning_rate": 5.409013663058874e-06,
+      "loss": 0.0941,
+      "step": 1714
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4843679193799688,
+      "learning_rate": 5.404402201807022e-06,
+      "loss": 0.0787,
+      "step": 1715
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6257603145756986,
+      "learning_rate": 5.399790394270179e-06,
+      "loss": 0.1034,
+      "step": 1716
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3771225744028005,
+      "learning_rate": 5.395178244397382e-06,
+      "loss": 0.1121,
+      "step": 1717
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.36688068268341084,
+      "learning_rate": 5.390565756137964e-06,
+      "loss": 0.0905,
+      "step": 1718
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.33871418359254224,
+      "learning_rate": 5.385952933441545e-06,
+      "loss": 0.0691,
+      "step": 1719
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.35771477481335084,
+      "learning_rate": 5.381339780258034e-06,
+      "loss": 0.0669,
+      "step": 1720
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7362815592084997,
+      "learning_rate": 5.376726300537618e-06,
+      "loss": 0.128,
+      "step": 1721
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2955159047101439,
+      "learning_rate": 5.372112498230771e-06,
+      "loss": 0.055,
+      "step": 1722
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3805913987843583,
+      "learning_rate": 5.367498377288236e-06,
+      "loss": 0.0606,
+      "step": 1723
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5074061802339812,
+      "learning_rate": 5.362883941661034e-06,
+      "loss": 0.1399,
+      "step": 1724
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4624573222358884,
+      "learning_rate": 5.358269195300454e-06,
+      "loss": 0.0929,
+      "step": 1725
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4293300336910408,
+      "learning_rate": 5.353654142158049e-06,
+      "loss": 0.0977,
+      "step": 1726
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.39845874593890546,
+      "learning_rate": 5.349038786185639e-06,
+      "loss": 0.092,
+      "step": 1727
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3505025908314534,
+      "learning_rate": 5.3444231313352965e-06,
+      "loss": 0.0897,
+      "step": 1728
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.42745382667252335,
+      "learning_rate": 5.339807181559359e-06,
+      "loss": 0.0938,
+      "step": 1729
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.22585096868673774,
+      "learning_rate": 5.335190940810407e-06,
+      "loss": 0.0339,
+      "step": 1730
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8104057816148835,
+      "learning_rate": 5.330574413041278e-06,
+      "loss": 0.1438,
+      "step": 1731
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5442941487882357,
+      "learning_rate": 5.325957602205051e-06,
+      "loss": 0.1154,
+      "step": 1732
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3295448531274127,
+      "learning_rate": 5.321340512255047e-06,
+      "loss": 0.0724,
+      "step": 1733
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4071347177236589,
+      "learning_rate": 5.3167231471448296e-06,
+      "loss": 0.0483,
+      "step": 1734
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3818177023038044,
+      "learning_rate": 5.312105510828196e-06,
+      "loss": 0.0599,
+      "step": 1735
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.40413321433612515,
+      "learning_rate": 5.307487607259175e-06,
+      "loss": 0.096,
+      "step": 1736
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.27123886566658834,
+      "learning_rate": 5.302869440392022e-06,
+      "loss": 0.0602,
+      "step": 1737
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5408898387276906,
+      "learning_rate": 5.2982510141812245e-06,
+      "loss": 0.1086,
+      "step": 1738
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5563925108749879,
+      "learning_rate": 5.293632332581487e-06,
+      "loss": 0.1442,
+      "step": 1739
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4135835036813737,
+      "learning_rate": 5.289013399547732e-06,
+      "loss": 0.1028,
+      "step": 1740
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4467060010997873,
+      "learning_rate": 5.2843942190351e-06,
+      "loss": 0.0994,
+      "step": 1741
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8541430244209335,
+      "learning_rate": 5.2797747949989454e-06,
+      "loss": 0.1531,
+      "step": 1742
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6015915655857912,
+      "learning_rate": 5.275155131394825e-06,
+      "loss": 0.1221,
+      "step": 1743
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6835209558694298,
+      "learning_rate": 5.270535232178505e-06,
+      "loss": 0.1235,
+      "step": 1744
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4752525613133181,
+      "learning_rate": 5.265915101305952e-06,
+      "loss": 0.0737,
+      "step": 1745
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.41208071341670405,
+      "learning_rate": 5.261294742733333e-06,
+      "loss": 0.0614,
+      "step": 1746
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.31722740389800774,
+      "learning_rate": 5.256674160417006e-06,
+      "loss": 0.0607,
+      "step": 1747
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4338570013808057,
+      "learning_rate": 5.252053358313525e-06,
+      "loss": 0.1098,
+      "step": 1748
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.47930442265747286,
+      "learning_rate": 5.247432340379628e-06,
+      "loss": 0.1117,
+      "step": 1749
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2795485943598554,
+      "learning_rate": 5.242811110572243e-06,
+      "loss": 0.0771,
+      "step": 1750
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8587182145827098,
+      "learning_rate": 5.238189672848472e-06,
+      "loss": 0.174,
+      "step": 1751
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3850768937056276,
+      "learning_rate": 5.233568031165603e-06,
+      "loss": 0.0831,
+      "step": 1752
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5811047288629524,
+      "learning_rate": 5.228946189481094e-06,
+      "loss": 0.0985,
+      "step": 1753
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4938779692662693,
+      "learning_rate": 5.224324151752575e-06,
+      "loss": 0.103,
+      "step": 1754
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.40324228456050887,
+      "learning_rate": 5.219701921937845e-06,
+      "loss": 0.08,
+      "step": 1755
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4794459631357845,
+      "learning_rate": 5.215079503994866e-06,
+      "loss": 0.0994,
+      "step": 1756
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.42398369004618675,
+      "learning_rate": 5.210456901881761e-06,
+      "loss": 0.1033,
+      "step": 1757
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.39932128986874454,
+      "learning_rate": 5.2058341195568115e-06,
+      "loss": 0.084,
+      "step": 1758
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.33495593223377956,
+      "learning_rate": 5.201211160978457e-06,
+      "loss": 0.0842,
+      "step": 1759
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5776028108630641,
+      "learning_rate": 5.196588030105278e-06,
+      "loss": 0.0797,
+      "step": 1760
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4644351553118459,
+      "learning_rate": 5.191964730896013e-06,
+      "loss": 0.0793,
+      "step": 1761
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.42614134097942846,
+      "learning_rate": 5.187341267309539e-06,
+      "loss": 0.108,
+      "step": 1762
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4491794654933965,
+      "learning_rate": 5.182717643304876e-06,
+      "loss": 0.079,
+      "step": 1763
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.29289051004587,
+      "learning_rate": 5.1780938628411795e-06,
+      "loss": 0.1037,
+      "step": 1764
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.21499893698290443,
+      "learning_rate": 5.173469929877741e-06,
+      "loss": 0.0443,
+      "step": 1765
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.35423226480291703,
+      "learning_rate": 5.168845848373979e-06,
+      "loss": 0.1047,
+      "step": 1766
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.33833942312962095,
+      "learning_rate": 5.164221622289445e-06,
+      "loss": 0.0491,
+      "step": 1767
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.588866523484762,
+      "learning_rate": 5.159597255583808e-06,
+      "loss": 0.0909,
+      "step": 1768
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.1910207915071211,
+      "learning_rate": 5.154972752216865e-06,
+      "loss": 0.0629,
+      "step": 1769
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3875092536142003,
+      "learning_rate": 5.1503481161485206e-06,
+      "loss": 0.0757,
+      "step": 1770
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.47769256631299484,
+      "learning_rate": 5.145723351338799e-06,
+      "loss": 0.117,
+      "step": 1771
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3002554362509087,
+      "learning_rate": 5.141098461747837e-06,
+      "loss": 0.0825,
+      "step": 1772
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4122361039490995,
+      "learning_rate": 5.136473451335869e-06,
+      "loss": 0.1097,
+      "step": 1773
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5074687911950858,
+      "learning_rate": 5.131848324063243e-06,
+      "loss": 0.1234,
+      "step": 1774
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.26160101336100056,
+      "learning_rate": 5.127223083890402e-06,
+      "loss": 0.0723,
+      "step": 1775
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.41088471896219075,
+      "learning_rate": 5.122597734777884e-06,
+      "loss": 0.0918,
+      "step": 1776
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.41377911072083023,
+      "learning_rate": 5.1179722806863264e-06,
+      "loss": 0.1046,
+      "step": 1777
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.40122840114603303,
+      "learning_rate": 5.11334672557645e-06,
+      "loss": 0.0643,
+      "step": 1778
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5176300532922882,
+      "learning_rate": 5.108721073409067e-06,
+      "loss": 0.0717,
+      "step": 1779
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.44071189944171935,
+      "learning_rate": 5.104095328145069e-06,
+      "loss": 0.1066,
+      "step": 1780
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.45067724330772735,
+      "learning_rate": 5.099469493745429e-06,
+      "loss": 0.0912,
+      "step": 1781
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.304169288511332,
+      "learning_rate": 5.094843574171195e-06,
+      "loss": 0.0633,
+      "step": 1782
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4923736256734507,
+      "learning_rate": 5.0902175733834926e-06,
+      "loss": 0.1282,
+      "step": 1783
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3045856973069122,
+      "learning_rate": 5.08559149534351e-06,
+      "loss": 0.0645,
+      "step": 1784
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3108558213473051,
+      "learning_rate": 5.080965344012509e-06,
+      "loss": 0.0559,
+      "step": 1785
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5671978774407673,
+      "learning_rate": 5.076339123351805e-06,
+      "loss": 0.0905,
+      "step": 1786
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4856432131264074,
+      "learning_rate": 5.071712837322782e-06,
+      "loss": 0.1296,
+      "step": 1787
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3389170179946792,
+      "learning_rate": 5.067086489886873e-06,
+      "loss": 0.096,
+      "step": 1788
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5632118394665078,
+      "learning_rate": 5.06246008500557e-06,
+      "loss": 0.0964,
+      "step": 1789
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.648512793509673,
+      "learning_rate": 5.0578336266404085e-06,
+      "loss": 0.122,
+      "step": 1790
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3258354727516121,
+      "learning_rate": 5.053207118752973e-06,
+      "loss": 0.068,
+      "step": 1791
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7180310667839394,
+      "learning_rate": 5.048580565304887e-06,
+      "loss": 0.1101,
+      "step": 1792
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2764883777437533,
+      "learning_rate": 5.043953970257819e-06,
+      "loss": 0.0713,
+      "step": 1793
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.508040561230719,
+      "learning_rate": 5.0393273375734664e-06,
+      "loss": 0.1132,
+      "step": 1794
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3303481225347844,
+      "learning_rate": 5.034700671213565e-06,
+      "loss": 0.0791,
+      "step": 1795
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.46792985989080804,
+      "learning_rate": 5.0300739751398744e-06,
+      "loss": 0.0796,
+      "step": 1796
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3827573561801135,
+      "learning_rate": 5.025447253314181e-06,
+      "loss": 0.1074,
+      "step": 1797
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.36133266664408525,
+      "learning_rate": 5.020820509698296e-06,
+      "loss": 0.0908,
+      "step": 1798
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.35464439574957596,
+      "learning_rate": 5.016193748254045e-06,
+      "loss": 0.0874,
+      "step": 1799
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4123590202063127,
+      "learning_rate": 5.0115669729432725e-06,
+      "loss": 0.0671,
+      "step": 1800
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.44071860293528237,
+      "learning_rate": 5.006940187727832e-06,
+      "loss": 0.0894,
+      "step": 1801
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.35522717609795534,
+      "learning_rate": 5.00231339656959e-06,
+      "loss": 0.1204,
+      "step": 1802
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.38112553943593,
+      "learning_rate": 4.9976866034304116e-06,
+      "loss": 0.0683,
+      "step": 1803
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4659657997607652,
+      "learning_rate": 4.9930598122721695e-06,
+      "loss": 0.0628,
+      "step": 1804
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.2967690693715893,
+      "learning_rate": 4.988433027056729e-06,
+      "loss": 0.0427,
+      "step": 1805
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.44824372545590935,
+      "learning_rate": 4.983806251745958e-06,
+      "loss": 0.0956,
+      "step": 1806
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.37800412975486114,
+      "learning_rate": 4.979179490301706e-06,
+      "loss": 0.1326,
+      "step": 1807
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7885437464260369,
+      "learning_rate": 4.9745527466858215e-06,
+      "loss": 0.1018,
+      "step": 1808
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4051763334362205,
+      "learning_rate": 4.969926024860127e-06,
+      "loss": 0.0654,
+      "step": 1809
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3781882626945609,
+      "learning_rate": 4.965299328786437e-06,
+      "loss": 0.0695,
+      "step": 1810
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3830674250490139,
+      "learning_rate": 4.960672662426535e-06,
+      "loss": 0.0909,
+      "step": 1811
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5041786500328712,
+      "learning_rate": 4.956046029742183e-06,
+      "loss": 0.1102,
+      "step": 1812
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5880617457405704,
+      "learning_rate": 4.951419434695115e-06,
+      "loss": 0.1123,
+      "step": 1813
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.28522135886218086,
+      "learning_rate": 4.946792881247028e-06,
+      "loss": 0.0734,
+      "step": 1814
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3294796218255551,
+      "learning_rate": 4.942166373359593e-06,
+      "loss": 0.0801,
+      "step": 1815
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.29668438106948947,
+      "learning_rate": 4.9375399149944305e-06,
+      "loss": 0.0702,
+      "step": 1816
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.27321131758204753,
+      "learning_rate": 4.932913510113128e-06,
+      "loss": 0.0679,
+      "step": 1817
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3614485415983663,
+      "learning_rate": 4.9282871626772195e-06,
+      "loss": 0.0752,
+      "step": 1818
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3685190017228016,
+      "learning_rate": 4.923660876648197e-06,
+      "loss": 0.0686,
+      "step": 1819
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3290672961024479,
+      "learning_rate": 4.919034655987493e-06,
+      "loss": 0.0784,
+      "step": 1820
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5367577533454958,
+      "learning_rate": 4.914408504656491e-06,
+      "loss": 0.1332,
+      "step": 1821
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3384153629731904,
+      "learning_rate": 4.909782426616508e-06,
+      "loss": 0.0936,
+      "step": 1822
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.29261661358632907,
+      "learning_rate": 4.9051564258288055e-06,
+      "loss": 0.0543,
+      "step": 1823
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6011233546440757,
+      "learning_rate": 4.900530506254573e-06,
+      "loss": 0.0882,
+      "step": 1824
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4096237989727531,
+      "learning_rate": 4.895904671854933e-06,
+      "loss": 0.1072,
+      "step": 1825
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4846911623910697,
+      "learning_rate": 4.8912789265909335e-06,
+      "loss": 0.1136,
+      "step": 1826
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.42455388147756434,
+      "learning_rate": 4.886653274423551e-06,
+      "loss": 0.1202,
+      "step": 1827
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4470496073472172,
+      "learning_rate": 4.882027719313675e-06,
+      "loss": 0.115,
+      "step": 1828
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4737130846179845,
+      "learning_rate": 4.877402265222117e-06,
+      "loss": 0.1264,
+      "step": 1829
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5387389491188319,
+      "learning_rate": 4.872776916109601e-06,
+      "loss": 0.1138,
+      "step": 1830
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.40017293609374827,
+      "learning_rate": 4.8681516759367595e-06,
+      "loss": 0.0819,
+      "step": 1831
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6302318750178969,
+      "learning_rate": 4.863526548664133e-06,
+      "loss": 0.0975,
+      "step": 1832
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4792126507643318,
+      "learning_rate": 4.858901538252166e-06,
+      "loss": 0.0995,
+      "step": 1833
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.762190300077111,
+      "learning_rate": 4.8542766486612035e-06,
+      "loss": 0.0864,
+      "step": 1834
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4152887567328987,
+      "learning_rate": 4.849651883851482e-06,
+      "loss": 0.0966,
+      "step": 1835
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.24764893449554634,
+      "learning_rate": 4.845027247783138e-06,
+      "loss": 0.0567,
+      "step": 1836
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.39746673426372264,
+      "learning_rate": 4.840402744416193e-06,
+      "loss": 0.1,
+      "step": 1837
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3066734894909942,
+      "learning_rate": 4.8357783777105575e-06,
+      "loss": 0.0674,
+      "step": 1838
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.40823531386248535,
+      "learning_rate": 4.831154151626022e-06,
+      "loss": 0.0943,
+      "step": 1839
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.29000721163358345,
+      "learning_rate": 4.826530070122262e-06,
+      "loss": 0.0657,
+      "step": 1840
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.45437564693786975,
+      "learning_rate": 4.821906137158822e-06,
+      "loss": 0.0965,
+      "step": 1841
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4105245817105171,
+      "learning_rate": 4.817282356695126e-06,
+      "loss": 0.0596,
+      "step": 1842
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3657179553981144,
+      "learning_rate": 4.812658732690463e-06,
+      "loss": 0.0913,
+      "step": 1843
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6054162389353315,
+      "learning_rate": 4.808035269103989e-06,
+      "loss": 0.1201,
+      "step": 1844
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5274034016500073,
+      "learning_rate": 4.8034119698947244e-06,
+      "loss": 0.1235,
+      "step": 1845
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7871638385148911,
+      "learning_rate": 4.798788839021546e-06,
+      "loss": 0.1209,
+      "step": 1846
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3949687348906425,
+      "learning_rate": 4.79416588044319e-06,
+      "loss": 0.0842,
+      "step": 1847
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.33777294776552913,
+      "learning_rate": 4.7895430981182415e-06,
+      "loss": 0.0735,
+      "step": 1848
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3567016111842002,
+      "learning_rate": 4.784920496005137e-06,
+      "loss": 0.0698,
+      "step": 1849
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5219398402468832,
+      "learning_rate": 4.780298078062157e-06,
+      "loss": 0.0613,
+      "step": 1850
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4723601537455463,
+      "learning_rate": 4.775675848247427e-06,
+      "loss": 0.109,
+      "step": 1851
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4621040919355528,
+      "learning_rate": 4.771053810518908e-06,
+      "loss": 0.1276,
+      "step": 1852
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5518875067551766,
+      "learning_rate": 4.766431968834399e-06,
+      "loss": 0.1421,
+      "step": 1853
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.45749371003943096,
+      "learning_rate": 4.76181032715153e-06,
+      "loss": 0.0731,
+      "step": 1854
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.38560233895242374,
+      "learning_rate": 4.757188889427761e-06,
+      "loss": 0.0788,
+      "step": 1855
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3570216875704459,
+      "learning_rate": 4.7525676596203726e-06,
+      "loss": 0.0885,
+      "step": 1856
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5116624767912511,
+      "learning_rate": 4.747946641686475e-06,
+      "loss": 0.1175,
+      "step": 1857
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4595799526423975,
+      "learning_rate": 4.743325839582995e-06,
+      "loss": 0.0946,
+      "step": 1858
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4031297689490686,
+      "learning_rate": 4.738705257266667e-06,
+      "loss": 0.078,
+      "step": 1859
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.412017964505888,
+      "learning_rate": 4.734084898694049e-06,
+      "loss": 0.1005,
+      "step": 1860
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2721539235413136,
+      "learning_rate": 4.729464767821496e-06,
+      "loss": 0.0865,
+      "step": 1861
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6004550191016197,
+      "learning_rate": 4.724844868605176e-06,
+      "loss": 0.1226,
+      "step": 1862
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5098773103913625,
+      "learning_rate": 4.720225205001056e-06,
+      "loss": 0.0415,
+      "step": 1863
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.40815745605351283,
+      "learning_rate": 4.7156057809649e-06,
+      "loss": 0.0859,
+      "step": 1864
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3686636187580829,
+      "learning_rate": 4.710986600452269e-06,
+      "loss": 0.0814,
+      "step": 1865
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5084807939262025,
+      "learning_rate": 4.706367667418514e-06,
+      "loss": 0.109,
+      "step": 1866
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.23827947558998486,
+      "learning_rate": 4.701748985818776e-06,
+      "loss": 0.064,
+      "step": 1867
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.454437908383136,
+      "learning_rate": 4.697130559607978e-06,
+      "loss": 0.1086,
+      "step": 1868
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2937112277707689,
+      "learning_rate": 4.6925123927408265e-06,
+      "loss": 0.0724,
+      "step": 1869
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5127832515625701,
+      "learning_rate": 4.687894489171804e-06,
+      "loss": 0.0572,
+      "step": 1870
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7056047133510236,
+      "learning_rate": 4.68327685285517e-06,
+      "loss": 0.1115,
+      "step": 1871
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.39573545656865444,
+      "learning_rate": 4.678659487744953e-06,
+      "loss": 0.1019,
+      "step": 1872
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4083774884476607,
+      "learning_rate": 4.67404239779495e-06,
+      "loss": 0.1011,
+      "step": 1873
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.49253317487897935,
+      "learning_rate": 4.669425586958723e-06,
+      "loss": 0.1171,
+      "step": 1874
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3729798762255827,
+      "learning_rate": 4.664809059189594e-06,
+      "loss": 0.0734,
+      "step": 1875
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3317724517590702,
+      "learning_rate": 4.660192818440642e-06,
+      "loss": 0.0642,
+      "step": 1876
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6723221845107754,
+      "learning_rate": 4.6555768686647035e-06,
+      "loss": 0.0904,
+      "step": 1877
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.47709826160936375,
+      "learning_rate": 4.650961213814362e-06,
+      "loss": 0.1241,
+      "step": 1878
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4713509969286293,
+      "learning_rate": 4.6463458578419504e-06,
+      "loss": 0.103,
+      "step": 1879
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.39502541556809767,
+      "learning_rate": 4.641730804699547e-06,
+      "loss": 0.0943,
+      "step": 1880
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.26264286988087127,
+      "learning_rate": 4.637116058338966e-06,
+      "loss": 0.0679,
+      "step": 1881
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.332228478121911,
+      "learning_rate": 4.6325016227117655e-06,
+      "loss": 0.104,
+      "step": 1882
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3450264960357861,
+      "learning_rate": 4.627887501769231e-06,
+      "loss": 0.0698,
+      "step": 1883
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3570095299234063,
+      "learning_rate": 4.623273699462384e-06,
+      "loss": 0.0925,
+      "step": 1884
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.025992789630151,
+      "learning_rate": 4.618660219741968e-06,
+      "loss": 0.1786,
+      "step": 1885
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3313978386930275,
+      "learning_rate": 4.614047066558457e-06,
+      "loss": 0.0977,
+      "step": 1886
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.35813875580660093,
+      "learning_rate": 4.609434243862037e-06,
+      "loss": 0.1059,
+      "step": 1887
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4691471990597548,
+      "learning_rate": 4.60482175560262e-06,
+      "loss": 0.1158,
+      "step": 1888
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6836332521112931,
+      "learning_rate": 4.600209605729823e-06,
+      "loss": 0.1469,
+      "step": 1889
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.2599298811639539,
+      "learning_rate": 4.59559779819298e-06,
+      "loss": 0.0797,
+      "step": 1890
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.48795608016248054,
+      "learning_rate": 4.5909863369411275e-06,
+      "loss": 0.1116,
+      "step": 1891
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.46588600797664814,
+      "learning_rate": 4.5863752259230085e-06,
+      "loss": 0.115,
+      "step": 1892
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3217361498668818,
+      "learning_rate": 4.581764469087064e-06,
+      "loss": 0.1014,
+      "step": 1893
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.42944486698052714,
+      "learning_rate": 4.577154070381432e-06,
+      "loss": 0.0852,
+      "step": 1894
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4183007935702926,
+      "learning_rate": 4.572544033753945e-06,
+      "loss": 0.0915,
+      "step": 1895
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4803824728872886,
+      "learning_rate": 4.567934363152126e-06,
+      "loss": 0.081,
+      "step": 1896
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.31338340306069945,
+      "learning_rate": 4.5633250625231806e-06,
+      "loss": 0.0825,
+      "step": 1897
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3241077163131851,
+      "learning_rate": 4.558716135814002e-06,
+      "loss": 0.0945,
+      "step": 1898
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5323237759812478,
+      "learning_rate": 4.554107586971162e-06,
+      "loss": 0.1013,
+      "step": 1899
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3229637312120239,
+      "learning_rate": 4.549499419940906e-06,
+      "loss": 0.0801,
+      "step": 1900
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.40184114865363024,
+      "learning_rate": 4.544891638669159e-06,
+      "loss": 0.1009,
+      "step": 1901
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.29438681164914965,
+      "learning_rate": 4.540284247101507e-06,
+      "loss": 0.0422,
+      "step": 1902
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.38210881048051887,
+      "learning_rate": 4.53567724918321e-06,
+      "loss": 0.0932,
+      "step": 1903
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4300572949739541,
+      "learning_rate": 4.531070648859186e-06,
+      "loss": 0.0986,
+      "step": 1904
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.27597139234993856,
+      "learning_rate": 4.526464450074016e-06,
+      "loss": 0.0832,
+      "step": 1905
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.789166463160726,
+      "learning_rate": 4.521858656771933e-06,
+      "loss": 0.1544,
+      "step": 1906
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.462613706943806,
+      "learning_rate": 4.517253272896827e-06,
+      "loss": 0.1217,
+      "step": 1907
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.34773668379669725,
+      "learning_rate": 4.5126483023922354e-06,
+      "loss": 0.0915,
+      "step": 1908
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.42662219449198857,
+      "learning_rate": 4.508043749201343e-06,
+      "loss": 0.0931,
+      "step": 1909
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4031899824903054,
+      "learning_rate": 4.503439617266974e-06,
+      "loss": 0.0724,
+      "step": 1910
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5801786850476682,
+      "learning_rate": 4.498835910531595e-06,
+      "loss": 0.1087,
+      "step": 1911
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2857207517449495,
+      "learning_rate": 4.494232632937308e-06,
+      "loss": 0.0618,
+      "step": 1912
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.25686815839937577,
+      "learning_rate": 4.489629788425847e-06,
+      "loss": 0.0689,
+      "step": 1913
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.45778970543379055,
+      "learning_rate": 4.485027380938574e-06,
+      "loss": 0.0666,
+      "step": 1914
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.31223254043849147,
+      "learning_rate": 4.480425414416479e-06,
+      "loss": 0.0681,
+      "step": 1915
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2713060954370277,
+      "learning_rate": 4.475823892800174e-06,
+      "loss": 0.0851,
+      "step": 1916
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2373650242836078,
+      "learning_rate": 4.471222820029888e-06,
+      "loss": 0.0594,
+      "step": 1917
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.29806082273090767,
+      "learning_rate": 4.4666222000454685e-06,
+      "loss": 0.0695,
+      "step": 1918
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3278947720919759,
+      "learning_rate": 4.462022036786372e-06,
+      "loss": 0.0612,
+      "step": 1919
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.39896879891390513,
+      "learning_rate": 4.45742233419167e-06,
+      "loss": 0.0901,
+      "step": 1920
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.42255087490743387,
+      "learning_rate": 4.452823096200029e-06,
+      "loss": 0.059,
+      "step": 1921
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4469970144861933,
+      "learning_rate": 4.4482243267497304e-06,
+      "loss": 0.1199,
+      "step": 1922
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.49217727261777366,
+      "learning_rate": 4.443626029778643e-06,
+      "loss": 0.1083,
+      "step": 1923
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.28235135435374265,
+      "learning_rate": 4.43902820922424e-06,
+      "loss": 0.0699,
+      "step": 1924
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4553360312689322,
+      "learning_rate": 4.434430869023579e-06,
+      "loss": 0.0723,
+      "step": 1925
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3516303902725191,
+      "learning_rate": 4.4298340131133135e-06,
+      "loss": 0.0555,
+      "step": 1926
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4504339813681332,
+      "learning_rate": 4.425237645429675e-06,
+      "loss": 0.1104,
+      "step": 1927
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.33091913082621083,
+      "learning_rate": 4.420641769908485e-06,
+      "loss": 0.0665,
+      "step": 1928
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.26775536702424624,
+      "learning_rate": 4.416046390485136e-06,
+      "loss": 0.0492,
+      "step": 1929
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.44832676079468037,
+      "learning_rate": 4.4114515110946e-06,
+      "loss": 0.1076,
+      "step": 1930
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4178039232742665,
+      "learning_rate": 4.40685713567142e-06,
+      "loss": 0.0842,
+      "step": 1931
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2860223061574364,
+      "learning_rate": 4.402263268149707e-06,
+      "loss": 0.0599,
+      "step": 1932
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4050827275327297,
+      "learning_rate": 4.397669912463137e-06,
+      "loss": 0.1322,
+      "step": 1933
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5299214420573273,
+      "learning_rate": 4.393077072544948e-06,
+      "loss": 0.1088,
+      "step": 1934
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2887170502060338,
+      "learning_rate": 4.3884847523279374e-06,
+      "loss": 0.0837,
+      "step": 1935
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6404760499001632,
+      "learning_rate": 4.383892955744456e-06,
+      "loss": 0.1495,
+      "step": 1936
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3196506146253498,
+      "learning_rate": 4.379301686726407e-06,
+      "loss": 0.1011,
+      "step": 1937
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3140363793235934,
+      "learning_rate": 4.374710949205241e-06,
+      "loss": 0.0865,
+      "step": 1938
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.702703285669765,
+      "learning_rate": 4.370120747111956e-06,
+      "loss": 0.0791,
+      "step": 1939
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.39221731152640016,
+      "learning_rate": 4.365531084377087e-06,
+      "loss": 0.0864,
+      "step": 1940
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4947895718262526,
+      "learning_rate": 4.360941964930712e-06,
+      "loss": 0.1428,
+      "step": 1941
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7603743640074522,
+      "learning_rate": 4.35635339270244e-06,
+      "loss": 0.1211,
+      "step": 1942
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3384779650531658,
+      "learning_rate": 4.351765371621415e-06,
+      "loss": 0.0828,
+      "step": 1943
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.26062740442709526,
+      "learning_rate": 4.347177905616306e-06,
+      "loss": 0.0456,
+      "step": 1944
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.33304292863241364,
+      "learning_rate": 4.342590998615308e-06,
+      "loss": 0.1021,
+      "step": 1945
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5634434685367286,
+      "learning_rate": 4.338004654546136e-06,
+      "loss": 0.089,
+      "step": 1946
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.2939575586372781,
+      "learning_rate": 4.333418877336024e-06,
+      "loss": 0.0561,
+      "step": 1947
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5901392544645243,
+      "learning_rate": 4.3288336709117246e-06,
+      "loss": 0.1279,
+      "step": 1948
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4660849817812266,
+      "learning_rate": 4.324249039199492e-06,
+      "loss": 0.0747,
+      "step": 1949
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3023997208506098,
+      "learning_rate": 4.319664986125099e-06,
+      "loss": 0.0599,
+      "step": 1950
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4394114231502662,
+      "learning_rate": 4.315081515613815e-06,
+      "loss": 0.0753,
+      "step": 1951
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.30320345750113625,
+      "learning_rate": 4.3104986315904144e-06,
+      "loss": 0.0692,
+      "step": 1952
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4215626390647296,
+      "learning_rate": 4.3059163379791676e-06,
+      "loss": 0.0632,
+      "step": 1953
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3450477954279993,
+      "learning_rate": 4.301334638703843e-06,
+      "loss": 0.0853,
+      "step": 1954
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3439957224189273,
+      "learning_rate": 4.296753537687694e-06,
+      "loss": 0.0473,
+      "step": 1955
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5498819847692374,
+      "learning_rate": 4.292173038853468e-06,
+      "loss": 0.0957,
+      "step": 1956
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.43969107091704507,
+      "learning_rate": 4.287593146123391e-06,
+      "loss": 0.1049,
+      "step": 1957
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.42322994487068916,
+      "learning_rate": 4.283013863419176e-06,
+      "loss": 0.1158,
+      "step": 1958
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7882983420945847,
+      "learning_rate": 4.278435194662007e-06,
+      "loss": 0.1018,
+      "step": 1959
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.2879325362757321,
+      "learning_rate": 4.27385714377255e-06,
+      "loss": 0.0454,
+      "step": 1960
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4926947096035045,
+      "learning_rate": 4.269279714670934e-06,
+      "loss": 0.097,
+      "step": 1961
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.40498556832141896,
+      "learning_rate": 4.2647029112767605e-06,
+      "loss": 0.0935,
+      "step": 1962
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.44690871261554077,
+      "learning_rate": 4.260126737509096e-06,
+      "loss": 0.0926,
+      "step": 1963
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4218838023368967,
+      "learning_rate": 4.2555511972864634e-06,
+      "loss": 0.0915,
+      "step": 1964
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.49503167389875924,
+      "learning_rate": 4.250976294526847e-06,
+      "loss": 0.1332,
+      "step": 1965
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.42508535712984674,
+      "learning_rate": 4.246402033147684e-06,
+      "loss": 0.0763,
+      "step": 1966
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.42110059103624764,
+      "learning_rate": 4.24182841706586e-06,
+      "loss": 0.0633,
+      "step": 1967
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5772184031699668,
+      "learning_rate": 4.237255450197714e-06,
+      "loss": 0.1092,
+      "step": 1968
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.2812932305462848,
+      "learning_rate": 4.232683136459021e-06,
+      "loss": 0.0877,
+      "step": 1969
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9032745882328743,
+      "learning_rate": 4.228111479765004e-06,
+      "loss": 0.1463,
+      "step": 1970
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.42769977338437704,
+      "learning_rate": 4.22354048403032e-06,
+      "loss": 0.1148,
+      "step": 1971
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.445114007386137,
+      "learning_rate": 4.2189701531690595e-06,
+      "loss": 0.0952,
+      "step": 1972
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5369118283803712,
+      "learning_rate": 4.214400491094746e-06,
+      "loss": 0.0951,
+      "step": 1973
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5754994017134322,
+      "learning_rate": 4.209831501720328e-06,
+      "loss": 0.1224,
+      "step": 1974
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4266059198848772,
+      "learning_rate": 4.20526318895818e-06,
+      "loss": 0.0713,
+      "step": 1975
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.2870313247515032,
+      "learning_rate": 4.200695556720095e-06,
+      "loss": 0.0745,
+      "step": 1976
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4307908012439384,
+      "learning_rate": 4.196128608917284e-06,
+      "loss": 0.1243,
+      "step": 1977
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.39299932618763994,
+      "learning_rate": 4.191562349460375e-06,
+      "loss": 0.0588,
+      "step": 1978
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5861435780886223,
+      "learning_rate": 4.1869967822594e-06,
+      "loss": 0.0869,
+      "step": 1979
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.30422620184804433,
+      "learning_rate": 4.182431911223806e-06,
+      "loss": 0.0542,
+      "step": 1980
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.311026370712942,
+      "learning_rate": 4.177867740262437e-06,
+      "loss": 0.0431,
+      "step": 1981
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3465167976558529,
+      "learning_rate": 4.173304273283541e-06,
+      "loss": 0.1026,
+      "step": 1982
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3469748721584691,
+      "learning_rate": 4.168741514194764e-06,
+      "loss": 0.0888,
+      "step": 1983
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5800733373315916,
+      "learning_rate": 4.1641794669031435e-06,
+      "loss": 0.1066,
+      "step": 1984
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3243784108523238,
+      "learning_rate": 4.159618135315109e-06,
+      "loss": 0.0686,
+      "step": 1985
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7434421684735129,
+      "learning_rate": 4.155057523336477e-06,
+      "loss": 0.1197,
+      "step": 1986
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7143013220770097,
+      "learning_rate": 4.1504976348724465e-06,
+      "loss": 0.1333,
+      "step": 1987
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.458633239134645,
+      "learning_rate": 4.145938473827598e-06,
+      "loss": 0.0704,
+      "step": 1988
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.32373332074690114,
+      "learning_rate": 4.141380044105891e-06,
+      "loss": 0.081,
+      "step": 1989
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3113298579819848,
+      "learning_rate": 4.1368223496106544e-06,
+      "loss": 0.0648,
+      "step": 1990
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.46622683878776333,
+      "learning_rate": 4.1322653942445925e-06,
+      "loss": 0.1127,
+      "step": 1991
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5308035018252141,
+      "learning_rate": 4.127709181909771e-06,
+      "loss": 0.1024,
+      "step": 1992
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5468851125410406,
+      "learning_rate": 4.123153716507625e-06,
+      "loss": 0.1032,
+      "step": 1993
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.33844503605315174,
+      "learning_rate": 4.118599001938947e-06,
+      "loss": 0.0941,
+      "step": 1994
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3068677222704898,
+      "learning_rate": 4.1140450421038865e-06,
+      "loss": 0.0779,
+      "step": 1995
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.48738135575860847,
+      "learning_rate": 4.109491840901948e-06,
+      "loss": 0.1086,
+      "step": 1996
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.28863622022466223,
+      "learning_rate": 4.104939402231986e-06,
+      "loss": 0.0662,
+      "step": 1997
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.368527455061649,
+      "learning_rate": 4.100387729992201e-06,
+      "loss": 0.0493,
+      "step": 1998
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.37516101208183883,
+      "learning_rate": 4.095836828080138e-06,
+      "loss": 0.0636,
+      "step": 1999
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.2591543806192558,
+      "learning_rate": 4.091286700392683e-06,
+      "loss": 0.0429,
+      "step": 2000
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4028847043574642,
+      "learning_rate": 4.086737350826058e-06,
+      "loss": 0.0783,
+      "step": 2001
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4184077498856964,
+      "learning_rate": 4.08218878327582e-06,
+      "loss": 0.0786,
+      "step": 2002
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.32607010234508654,
+      "learning_rate": 4.077641001636854e-06,
+      "loss": 0.0705,
+      "step": 2003
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.29771386911145054,
+      "learning_rate": 4.073094009803374e-06,
+      "loss": 0.0918,
+      "step": 2004
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4315479249247474,
+      "learning_rate": 4.068547811668918e-06,
+      "loss": 0.0978,
+      "step": 2005
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5330447499926427,
+      "learning_rate": 4.064002411126343e-06,
+      "loss": 0.1061,
+      "step": 2006
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.258613621124329,
+      "learning_rate": 4.059457812067823e-06,
+      "loss": 0.0641,
+      "step": 2007
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3220392034549073,
+      "learning_rate": 4.054914018384849e-06,
+      "loss": 0.1006,
+      "step": 2008
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.36690079395341285,
+      "learning_rate": 4.050371033968216e-06,
+      "loss": 0.0729,
+      "step": 2009
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4473416973250645,
+      "learning_rate": 4.0458288627080325e-06,
+      "loss": 0.0851,
+      "step": 2010
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3053859056768928,
+      "learning_rate": 4.041287508493706e-06,
+      "loss": 0.0828,
+      "step": 2011
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.40630274324795607,
+      "learning_rate": 4.0367469752139475e-06,
+      "loss": 0.0763,
+      "step": 2012
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3701731626037116,
+      "learning_rate": 4.032207266756764e-06,
+      "loss": 0.0885,
+      "step": 2013
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4265760298185957,
+      "learning_rate": 4.027668387009455e-06,
+      "loss": 0.0938,
+      "step": 2014
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.28060969979017275,
+      "learning_rate": 4.0231303398586124e-06,
+      "loss": 0.0652,
+      "step": 2015
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.29222129070474345,
+      "learning_rate": 4.018593129190113e-06,
+      "loss": 0.0643,
+      "step": 2016
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.27138955197208636,
+      "learning_rate": 4.014056758889121e-06,
+      "loss": 0.0716,
+      "step": 2017
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6951548257280417,
+      "learning_rate": 4.009521232840075e-06,
+      "loss": 0.121,
+      "step": 2018
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7563475757578632,
+      "learning_rate": 4.004986554926697e-06,
+      "loss": 0.1071,
+      "step": 2019
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5375498548744064,
+      "learning_rate": 4.000452729031978e-06,
+      "loss": 0.0747,
+      "step": 2020
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.22798813992235856,
+      "learning_rate": 3.995919759038184e-06,
+      "loss": 0.0466,
+      "step": 2021
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3595766444515047,
+      "learning_rate": 3.991387648826842e-06,
+      "loss": 0.0774,
+      "step": 2022
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2998093674841255,
+      "learning_rate": 3.98685640227875e-06,
+      "loss": 0.0473,
+      "step": 2023
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3985626427420547,
+      "learning_rate": 3.982326023273959e-06,
+      "loss": 0.1184,
+      "step": 2024
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6113417200642814,
+      "learning_rate": 3.977796515691785e-06,
+      "loss": 0.1277,
+      "step": 2025
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5718721177873624,
+      "learning_rate": 3.97326788341079e-06,
+      "loss": 0.1053,
+      "step": 2026
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5847265187586651,
+      "learning_rate": 3.968740130308792e-06,
+      "loss": 0.1068,
+      "step": 2027
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6206156584815474,
+      "learning_rate": 3.964213260262853e-06,
+      "loss": 0.1301,
+      "step": 2028
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.501407405321876,
+      "learning_rate": 3.959687277149283e-06,
+      "loss": 0.1127,
+      "step": 2029
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4804538068281452,
+      "learning_rate": 3.955162184843625e-06,
+      "loss": 0.1281,
+      "step": 2030
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5231994878403236,
+      "learning_rate": 3.950637987220669e-06,
+      "loss": 0.1136,
+      "step": 2031
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.37016488981910206,
+      "learning_rate": 3.94611468815443e-06,
+      "loss": 0.0686,
+      "step": 2032
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.26871161684579375,
+      "learning_rate": 3.9415922915181595e-06,
+      "loss": 0.0305,
+      "step": 2033
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.37256629953416204,
+      "learning_rate": 3.937070801184333e-06,
+      "loss": 0.0885,
+      "step": 2034
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3738101422663222,
+      "learning_rate": 3.932550221024651e-06,
+      "loss": 0.1334,
+      "step": 2035
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6029645029931219,
+      "learning_rate": 3.928030554910037e-06,
+      "loss": 0.136,
+      "step": 2036
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3554442316959907,
+      "learning_rate": 3.9235118067106255e-06,
+      "loss": 0.0938,
+      "step": 2037
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3235074713065256,
+      "learning_rate": 3.918993980295774e-06,
+      "loss": 0.0579,
+      "step": 2038
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4983600522619609,
+      "learning_rate": 3.914477079534041e-06,
+      "loss": 0.1103,
+      "step": 2039
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.39863078748668757,
+      "learning_rate": 3.9099611082932e-06,
+      "loss": 0.0977,
+      "step": 2040
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.28027634060902773,
+      "learning_rate": 3.9054460704402246e-06,
+      "loss": 0.0636,
+      "step": 2041
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3253098584369501,
+      "learning_rate": 3.90093196984129e-06,
+      "loss": 0.0623,
+      "step": 2042
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6229740514530065,
+      "learning_rate": 3.8964188103617685e-06,
+      "loss": 0.0998,
+      "step": 2043
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.48643580185941077,
+      "learning_rate": 3.89190659586623e-06,
+      "loss": 0.112,
+      "step": 2044
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.28979063553797807,
+      "learning_rate": 3.887395330218429e-06,
+      "loss": 0.0687,
+      "step": 2045
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.535227874364318,
+      "learning_rate": 3.882885017281312e-06,
+      "loss": 0.1024,
+      "step": 2046
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5717070347300525,
+      "learning_rate": 3.87837566091701e-06,
+      "loss": 0.1171,
+      "step": 2047
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.31107606437867763,
+      "learning_rate": 3.87386726498683e-06,
+      "loss": 0.0771,
+      "step": 2048
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.23525510475650052,
+      "learning_rate": 3.869359833351263e-06,
+      "loss": 0.0406,
+      "step": 2049
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5419064036518809,
+      "learning_rate": 3.8648533698699695e-06,
+      "loss": 0.1364,
+      "step": 2050
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4316380279177835,
+      "learning_rate": 3.8603478784017845e-06,
+      "loss": 0.0964,
+      "step": 2051
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.41189041144702077,
+      "learning_rate": 3.855843362804707e-06,
+      "loss": 0.0968,
+      "step": 2052
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5044277218391547,
+      "learning_rate": 3.851339826935904e-06,
+      "loss": 0.1223,
+      "step": 2053
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.47221302636165313,
+      "learning_rate": 3.8468372746517e-06,
+      "loss": 0.0965,
+      "step": 2054
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.44530159406031433,
+      "learning_rate": 3.842335709807582e-06,
+      "loss": 0.0863,
+      "step": 2055
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.586630340755237,
+      "learning_rate": 3.8378351362581844e-06,
+      "loss": 0.1075,
+      "step": 2056
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2809788979845124,
+      "learning_rate": 3.833335557857302e-06,
+      "loss": 0.0637,
+      "step": 2057
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.34647577554375586,
+      "learning_rate": 3.828836978457868e-06,
+      "loss": 0.0897,
+      "step": 2058
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3246781426625446,
+      "learning_rate": 3.824339401911967e-06,
+      "loss": 0.0655,
+      "step": 2059
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4237243542390314,
+      "learning_rate": 3.819842832070822e-06,
+      "loss": 0.0668,
+      "step": 2060
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.49527451270191114,
+      "learning_rate": 3.815347272784795e-06,
+      "loss": 0.0912,
+      "step": 2061
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2359837032835801,
+      "learning_rate": 3.810852727903381e-06,
+      "loss": 0.0651,
+      "step": 2062
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7228845888573183,
+      "learning_rate": 3.806359201275209e-06,
+      "loss": 0.1581,
+      "step": 2063
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3270182167549771,
+      "learning_rate": 3.8018666967480333e-06,
+      "loss": 0.0881,
+      "step": 2064
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.28759702320644553,
+      "learning_rate": 3.7973752181687336e-06,
+      "loss": 0.0687,
+      "step": 2065
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.29294071260960564,
+      "learning_rate": 3.7928847693833136e-06,
+      "loss": 0.032,
+      "step": 2066
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.37243145893279406,
+      "learning_rate": 3.7883953542368917e-06,
+      "loss": 0.066,
+      "step": 2067
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4596837041749248,
+      "learning_rate": 3.7839069765737024e-06,
+      "loss": 0.0874,
+      "step": 2068
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5068728692270521,
+      "learning_rate": 3.7794196402370962e-06,
+      "loss": 0.0876,
+      "step": 2069
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4575890766694308,
+      "learning_rate": 3.774933349069524e-06,
+      "loss": 0.1353,
+      "step": 2070
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5473491177990236,
+      "learning_rate": 3.7704481069125486e-06,
+      "loss": 0.1561,
+      "step": 2071
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5012890458465599,
+      "learning_rate": 3.7659639176068287e-06,
+      "loss": 0.1215,
+      "step": 2072
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4131563875581746,
+      "learning_rate": 3.761480784992127e-06,
+      "loss": 0.0933,
+      "step": 2073
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.36967504989043437,
+      "learning_rate": 3.756998712907297e-06,
+      "loss": 0.0918,
+      "step": 2074
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.47895162415247106,
+      "learning_rate": 3.752517705190287e-06,
+      "loss": 0.1049,
+      "step": 2075
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7174889321607305,
+      "learning_rate": 3.748037765678132e-06,
+      "loss": 0.1359,
+      "step": 2076
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2949635943803785,
+      "learning_rate": 3.743558898206955e-06,
+      "loss": 0.0727,
+      "step": 2077
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4494486899447512,
+      "learning_rate": 3.7390811066119552e-06,
+      "loss": 0.1244,
+      "step": 2078
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3855162317520461,
+      "learning_rate": 3.734604394727419e-06,
+      "loss": 0.0733,
+      "step": 2079
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.26065067980205636,
+      "learning_rate": 3.7301287663867002e-06,
+      "loss": 0.0759,
+      "step": 2080
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4370956810185193,
+      "learning_rate": 3.7256542254222307e-06,
+      "loss": 0.0851,
+      "step": 2081
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.34142654390983773,
+      "learning_rate": 3.7211807756655065e-06,
+      "loss": 0.0413,
+      "step": 2082
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2160190652719215,
+      "learning_rate": 3.7167084209470938e-06,
+      "loss": 0.0621,
+      "step": 2083
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7083407811772259,
+      "learning_rate": 3.7122371650966188e-06,
+      "loss": 0.1387,
+      "step": 2084
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.27300976717333647,
+      "learning_rate": 3.7077670119427644e-06,
+      "loss": 0.0587,
+      "step": 2085
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.47681884685724285,
+      "learning_rate": 3.703297965313275e-06,
+      "loss": 0.092,
+      "step": 2086
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6321938008876815,
+      "learning_rate": 3.6988300290349414e-06,
+      "loss": 0.1317,
+      "step": 2087
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2016558933938644,
+      "learning_rate": 3.694363206933609e-06,
+      "loss": 0.031,
+      "step": 2088
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3052856384218095,
+      "learning_rate": 3.6898975028341636e-06,
+      "loss": 0.0803,
+      "step": 2089
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4115355894563856,
+      "learning_rate": 3.68543292056054e-06,
+      "loss": 0.0701,
+      "step": 2090
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4755149189406155,
+      "learning_rate": 3.680969463935706e-06,
+      "loss": 0.0824,
+      "step": 2091
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.44893729851957614,
+      "learning_rate": 3.676507136781672e-06,
+      "loss": 0.1018,
+      "step": 2092
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.26040225037583753,
+      "learning_rate": 3.6720459429194743e-06,
+      "loss": 0.0463,
+      "step": 2093
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.35296792517971115,
+      "learning_rate": 3.6675858861691848e-06,
+      "loss": 0.0993,
+      "step": 2094
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3225183136762917,
+      "learning_rate": 3.6631269703498974e-06,
+      "loss": 0.0726,
+      "step": 2095
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2979364180447945,
+      "learning_rate": 3.6586691992797318e-06,
+      "loss": 0.0702,
+      "step": 2096
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.30125386069696103,
+      "learning_rate": 3.6542125767758263e-06,
+      "loss": 0.0664,
+      "step": 2097
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.21973910774686486,
+      "learning_rate": 3.649757106654337e-06,
+      "loss": 0.053,
+      "step": 2098
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.36142441994666674,
+      "learning_rate": 3.6453027927304313e-06,
+      "loss": 0.062,
+      "step": 2099
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3179639445416395,
+      "learning_rate": 3.6408496388182857e-06,
+      "loss": 0.0921,
+      "step": 2100
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.37353396499274105,
+      "learning_rate": 3.636397648731088e-06,
+      "loss": 0.0799,
+      "step": 2101
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3391512824605172,
+      "learning_rate": 3.631946826281024e-06,
+      "loss": 0.0753,
+      "step": 2102
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5814300713636016,
+      "learning_rate": 3.627497175279285e-06,
+      "loss": 0.1163,
+      "step": 2103
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.45910829179398005,
+      "learning_rate": 3.6230486995360524e-06,
+      "loss": 0.0694,
+      "step": 2104
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5575080215268169,
+      "learning_rate": 3.61860140286051e-06,
+      "loss": 0.114,
+      "step": 2105
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4390168966721899,
+      "learning_rate": 3.6141552890608224e-06,
+      "loss": 0.1202,
+      "step": 2106
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3800681344027182,
+      "learning_rate": 3.6097103619441505e-06,
+      "loss": 0.0884,
+      "step": 2107
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2101455047264348,
+      "learning_rate": 3.6052666253166306e-06,
+      "loss": 0.0475,
+      "step": 2108
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.29183422779381585,
+      "learning_rate": 3.6008240829833873e-06,
+      "loss": 0.0611,
+      "step": 2109
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.31568111702329904,
+      "learning_rate": 3.596382738748516e-06,
+      "loss": 0.072,
+      "step": 2110
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.28554512922628006,
+      "learning_rate": 3.591942596415092e-06,
+      "loss": 0.0637,
+      "step": 2111
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3857264661032645,
+      "learning_rate": 3.5875036597851553e-06,
+      "loss": 0.091,
+      "step": 2112
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7699789192606351,
+      "learning_rate": 3.5830659326597194e-06,
+      "loss": 0.0688,
+      "step": 2113
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5046650432948563,
+      "learning_rate": 3.578629418838757e-06,
+      "loss": 0.0845,
+      "step": 2114
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3232395398561082,
+      "learning_rate": 3.574194122121207e-06,
+      "loss": 0.0571,
+      "step": 2115
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.28310953285578133,
+      "learning_rate": 3.5697600463049626e-06,
+      "loss": 0.0781,
+      "step": 2116
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3089597690100642,
+      "learning_rate": 3.5653271951868695e-06,
+      "loss": 0.0877,
+      "step": 2117
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3348014203597471,
+      "learning_rate": 3.5608955725627315e-06,
+      "loss": 0.0544,
+      "step": 2118
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8053852606312272,
+      "learning_rate": 3.556465182227293e-06,
+      "loss": 0.1304,
+      "step": 2119
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5324945990152041,
+      "learning_rate": 3.55203602797425e-06,
+      "loss": 0.1281,
+      "step": 2120
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.20157718142726572,
+      "learning_rate": 3.5476081135962335e-06,
+      "loss": 0.0506,
+      "step": 2121
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3980883363467904,
+      "learning_rate": 3.5431814428848195e-06,
+      "loss": 0.086,
+      "step": 2122
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6848620674157934,
+      "learning_rate": 3.538756019630513e-06,
+      "loss": 0.1021,
+      "step": 2123
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.24304849186337293,
+      "learning_rate": 3.534331847622754e-06,
+      "loss": 0.0496,
+      "step": 2124
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6450678304980599,
+      "learning_rate": 3.52990893064991e-06,
+      "loss": 0.1042,
+      "step": 2125
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6959247328834508,
+      "learning_rate": 3.525487272499277e-06,
+      "loss": 0.1424,
+      "step": 2126
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5588994073367515,
+      "learning_rate": 3.5210668769570665e-06,
+      "loss": 0.1053,
+      "step": 2127
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.48674739074483947,
+      "learning_rate": 3.516647747808417e-06,
+      "loss": 0.0928,
+      "step": 2128
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.35485867318752906,
+      "learning_rate": 3.5122298888373742e-06,
+      "loss": 0.1112,
+      "step": 2129
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6166926786366669,
+      "learning_rate": 3.5078133038269034e-06,
+      "loss": 0.1541,
+      "step": 2130
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.42556221404167865,
+      "learning_rate": 3.503397996558874e-06,
+      "loss": 0.089,
+      "step": 2131
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7448092765410511,
+      "learning_rate": 3.4989839708140655e-06,
+      "loss": 0.1801,
+      "step": 2132
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.27981953417812216,
+      "learning_rate": 3.4945712303721558e-06,
+      "loss": 0.0749,
+      "step": 2133
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.20423481686160258,
+      "learning_rate": 3.490159779011724e-06,
+      "loss": 0.0564,
+      "step": 2134
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.43105399089281066,
+      "learning_rate": 3.4857496205102475e-06,
+      "loss": 0.0595,
+      "step": 2135
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4278895694978233,
+      "learning_rate": 3.481340758644092e-06,
+      "loss": 0.1133,
+      "step": 2136
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5413177432687966,
+      "learning_rate": 3.4769331971885188e-06,
+      "loss": 0.1183,
+      "step": 2137
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.295541889286931,
+      "learning_rate": 3.4725269399176693e-06,
+      "loss": 0.082,
+      "step": 2138
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3775732730926014,
+      "learning_rate": 3.468121990604574e-06,
+      "loss": 0.0612,
+      "step": 2139
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.48139120350221437,
+      "learning_rate": 3.463718353021138e-06,
+      "loss": 0.0895,
+      "step": 2140
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3874337431938493,
+      "learning_rate": 3.45931603093815e-06,
+      "loss": 0.0955,
+      "step": 2141
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.43567749982613474,
+      "learning_rate": 3.4549150281252635e-06,
+      "loss": 0.0573,
+      "step": 2142
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.272848024064493,
+      "learning_rate": 3.45051534835101e-06,
+      "loss": 0.0628,
+      "step": 2143
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.35791505390401157,
+      "learning_rate": 3.4461169953827846e-06,
+      "loss": 0.0724,
+      "step": 2144
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4601821466912252,
+      "learning_rate": 3.441719972986846e-06,
+      "loss": 0.1207,
+      "step": 2145
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6526346090023954,
+      "learning_rate": 3.437324284928314e-06,
+      "loss": 0.1118,
+      "step": 2146
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6249212896997014,
+      "learning_rate": 3.4329299349711687e-06,
+      "loss": 0.1339,
+      "step": 2147
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.47480506648556675,
+      "learning_rate": 3.4285369268782383e-06,
+      "loss": 0.1276,
+      "step": 2148
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6535323858490625,
+      "learning_rate": 3.4241452644112085e-06,
+      "loss": 0.1101,
+      "step": 2149
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4758298024601593,
+      "learning_rate": 3.4197549513306076e-06,
+      "loss": 0.1161,
+      "step": 2150
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2615492377403989,
+      "learning_rate": 3.4153659913958116e-06,
+      "loss": 0.0548,
+      "step": 2151
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.31850448121328606,
+      "learning_rate": 3.4109783883650373e-06,
+      "loss": 0.069,
+      "step": 2152
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.4594313366056636,
+      "learning_rate": 3.4065921459953365e-06,
+      "loss": 0.0795,
+      "step": 2153
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.4106609140190271,
+      "learning_rate": 3.4022072680426027e-06,
+      "loss": 0.0744,
+      "step": 2154
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.4732180186440131,
+      "learning_rate": 3.3978237582615535e-06,
+      "loss": 0.0687,
+      "step": 2155
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5910776142453107,
+      "learning_rate": 3.3934416204057396e-06,
+      "loss": 0.1225,
+      "step": 2156
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.4936430605684156,
+      "learning_rate": 3.3890608582275355e-06,
+      "loss": 0.0754,
+      "step": 2157
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.583288419944622,
+      "learning_rate": 3.384681475478139e-06,
+      "loss": 0.1412,
+      "step": 2158
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5287228494839469,
+      "learning_rate": 3.3803034759075624e-06,
+      "loss": 0.104,
+      "step": 2159
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3975751459912013,
+      "learning_rate": 3.37592686326464e-06,
+      "loss": 0.0774,
+      "step": 2160
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.43572649667643254,
+      "learning_rate": 3.371551641297014e-06,
+      "loss": 0.0825,
+      "step": 2161
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3196413576474667,
+      "learning_rate": 3.367177813751137e-06,
+      "loss": 0.0629,
+      "step": 2162
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.21128788446811378,
+      "learning_rate": 3.3628053843722674e-06,
+      "loss": 0.0559,
+      "step": 2163
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.27702982294106715,
+      "learning_rate": 3.3584343569044673e-06,
+      "loss": 0.0851,
+      "step": 2164
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3875813962199568,
+      "learning_rate": 3.3540647350905985e-06,
+      "loss": 0.0622,
+      "step": 2165
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.4497287018982541,
+      "learning_rate": 3.3496965226723156e-06,
+      "loss": 0.1178,
+      "step": 2166
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.31694673104908516,
+      "learning_rate": 3.3453297233900713e-06,
+      "loss": 0.0649,
+      "step": 2167
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.4098826564932513,
+      "learning_rate": 3.340964340983104e-06,
+      "loss": 0.0967,
+      "step": 2168
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.26200242125710105,
+      "learning_rate": 3.336600379189444e-06,
+      "loss": 0.0531,
+      "step": 2169
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.435971327270325,
+      "learning_rate": 3.3322378417458985e-06,
+      "loss": 0.1184,
+      "step": 2170
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3718658211792619,
+      "learning_rate": 3.327876732388061e-06,
+      "loss": 0.0503,
+      "step": 2171
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3756197979693583,
+      "learning_rate": 3.323517054850298e-06,
+      "loss": 0.1149,
+      "step": 2172
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3569703926571223,
+      "learning_rate": 3.3191588128657537e-06,
+      "loss": 0.088,
+      "step": 2173
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.38788936508162164,
+      "learning_rate": 3.314802010166337e-06,
+      "loss": 0.1116,
+      "step": 2174
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8353656774664352,
+      "learning_rate": 3.3104466504827327e-06,
+      "loss": 0.1362,
+      "step": 2175
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.4300659430461634,
+      "learning_rate": 3.3060927375443845e-06,
+      "loss": 0.0702,
+      "step": 2176
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2865548349541641,
+      "learning_rate": 3.3017402750794976e-06,
+      "loss": 0.0845,
+      "step": 2177
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9610315165190803,
+      "learning_rate": 3.2973892668150365e-06,
+      "loss": 0.1596,
+      "step": 2178
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.429895758130708,
+      "learning_rate": 3.293039716476719e-06,
+      "loss": 0.1016,
+      "step": 2179
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.367772969826401,
+      "learning_rate": 3.288691627789017e-06,
+      "loss": 0.0625,
+      "step": 2180
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3544993583005483,
+      "learning_rate": 3.2843450044751468e-06,
+      "loss": 0.0913,
+      "step": 2181
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6121129591343168,
+      "learning_rate": 3.279999850257076e-06,
+      "loss": 0.1027,
+      "step": 2182
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3790910928707313,
+      "learning_rate": 3.275656168855506e-06,
+      "loss": 0.0752,
+      "step": 2183
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.44063285211477354,
+      "learning_rate": 3.271313963989886e-06,
+      "loss": 0.083,
+      "step": 2184
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.43900728889382,
+      "learning_rate": 3.2669732393783944e-06,
+      "loss": 0.0853,
+      "step": 2185
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.46661852778740603,
+      "learning_rate": 3.262633998737943e-06,
+      "loss": 0.1349,
+      "step": 2186
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.4818974737283349,
+      "learning_rate": 3.258296245784176e-06,
+      "loss": 0.0918,
+      "step": 2187
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3958916826996518,
+      "learning_rate": 3.253959984231461e-06,
+      "loss": 0.062,
+      "step": 2188
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8232773554824766,
+      "learning_rate": 3.24962521779289e-06,
+      "loss": 0.0629,
+      "step": 2189
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4250745879875681,
+      "learning_rate": 3.2452919501802714e-06,
+      "loss": 0.0653,
+      "step": 2190
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4639581230707936,
+      "learning_rate": 3.240960185104137e-06,
+      "loss": 0.1221,
+      "step": 2191
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6113470051884101,
+      "learning_rate": 3.2366299262737246e-06,
+      "loss": 0.1047,
+      "step": 2192
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.17230213607604017,
+      "learning_rate": 3.232301177396987e-06,
+      "loss": 0.0503,
+      "step": 2193
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5241393634202148,
+      "learning_rate": 3.227973942180581e-06,
+      "loss": 0.1253,
+      "step": 2194
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.32115684457500066,
+      "learning_rate": 3.223648224329872e-06,
+      "loss": 0.084,
+      "step": 2195
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.35818981969753083,
+      "learning_rate": 3.219324027548918e-06,
+      "loss": 0.0801,
+      "step": 2196
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4356633496199792,
+      "learning_rate": 3.215001355540483e-06,
+      "loss": 0.0964,
+      "step": 2197
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5290389502712896,
+      "learning_rate": 3.2106802120060197e-06,
+      "loss": 0.0986,
+      "step": 2198
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7535690752406217,
+      "learning_rate": 3.206360600645676e-06,
+      "loss": 0.1215,
+      "step": 2199
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.33861654307401245,
+      "learning_rate": 3.202042525158284e-06,
+      "loss": 0.072,
+      "step": 2200
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5242903634387209,
+      "learning_rate": 3.197725989241364e-06,
+      "loss": 0.0874,
+      "step": 2201
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.46914209904777004,
+      "learning_rate": 3.1934109965911166e-06,
+      "loss": 0.102,
+      "step": 2202
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7636962947952328,
+      "learning_rate": 3.1890975509024174e-06,
+      "loss": 0.1623,
+      "step": 2203
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3609409113896043,
+      "learning_rate": 3.184785655868825e-06,
+      "loss": 0.0725,
+      "step": 2204
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5225538568770066,
+      "learning_rate": 3.180475315182563e-06,
+      "loss": 0.1297,
+      "step": 2205
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.46137712922314433,
+      "learning_rate": 3.1761665325345285e-06,
+      "loss": 0.0935,
+      "step": 2206
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2626466733851561,
+      "learning_rate": 3.1718593116142814e-06,
+      "loss": 0.0317,
+      "step": 2207
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.30894544511701955,
+      "learning_rate": 3.167553656110048e-06,
+      "loss": 0.0945,
+      "step": 2208
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.41646567609931173,
+      "learning_rate": 3.1632495697087105e-06,
+      "loss": 0.1083,
+      "step": 2209
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5049042907399488,
+      "learning_rate": 3.1589470560958104e-06,
+      "loss": 0.104,
+      "step": 2210
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.21207648860814676,
+      "learning_rate": 3.1546461189555388e-06,
+      "loss": 0.0315,
+      "step": 2211
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.39901941438593314,
+      "learning_rate": 3.1503467619707407e-06,
+      "loss": 0.0805,
+      "step": 2212
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.383123390502709,
+      "learning_rate": 3.146048988822905e-06,
+      "loss": 0.0615,
+      "step": 2213
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3750701945577319,
+      "learning_rate": 3.1417528031921686e-06,
+      "loss": 0.0791,
+      "step": 2214
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.24813876823351655,
+      "learning_rate": 3.1374582087573026e-06,
+      "loss": 0.0447,
+      "step": 2215
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2256052598341187,
+      "learning_rate": 3.133165209195722e-06,
+      "loss": 0.048,
+      "step": 2216
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.264399776068169,
+      "learning_rate": 3.1288738081834734e-06,
+      "loss": 0.0658,
+      "step": 2217
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.42808072206628417,
+      "learning_rate": 3.124584009395232e-06,
+      "loss": 0.0862,
+      "step": 2218
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3066496864333556,
+      "learning_rate": 3.1202958165043053e-06,
+      "loss": 0.0546,
+      "step": 2219
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4165697205719138,
+      "learning_rate": 3.1160092331826235e-06,
+      "loss": 0.1064,
+      "step": 2220
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.43532672421992785,
+      "learning_rate": 3.11172426310074e-06,
+      "loss": 0.1159,
+      "step": 2221
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.31113532383650017,
+      "learning_rate": 3.107440909927824e-06,
+      "loss": 0.0882,
+      "step": 2222
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3267355531934448,
+      "learning_rate": 3.1031591773316636e-06,
+      "loss": 0.0765,
+      "step": 2223
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.41591711380489454,
+      "learning_rate": 3.0988790689786563e-06,
+      "loss": 0.0722,
+      "step": 2224
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2688967848001076,
+      "learning_rate": 3.0946005885338116e-06,
+      "loss": 0.0406,
+      "step": 2225
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.515598215348305,
+      "learning_rate": 3.090323739660742e-06,
+      "loss": 0.1035,
+      "step": 2226
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3830700355400384,
+      "learning_rate": 3.0860485260216656e-06,
+      "loss": 0.0794,
+      "step": 2227
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4657863791560485,
+      "learning_rate": 3.081774951277397e-06,
+      "loss": 0.1038,
+      "step": 2228
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3447325976522972,
+      "learning_rate": 3.077503019087352e-06,
+      "loss": 0.0988,
+      "step": 2229
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.45516248898991346,
+      "learning_rate": 3.073232733109536e-06,
+      "loss": 0.1096,
+      "step": 2230
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5297690308234168,
+      "learning_rate": 3.068964097000546e-06,
+      "loss": 0.0732,
+      "step": 2231
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.29911509616116244,
+      "learning_rate": 3.064697114415567e-06,
+      "loss": 0.0619,
+      "step": 2232
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1595404958128304,
+      "learning_rate": 3.060431789008368e-06,
+      "loss": 0.098,
+      "step": 2233
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4594281388434096,
+      "learning_rate": 3.056168124431298e-06,
+      "loss": 0.1078,
+      "step": 2234
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4139696114976327,
+      "learning_rate": 3.0519061243352833e-06,
+      "loss": 0.105,
+      "step": 2235
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.320372204104932,
+      "learning_rate": 3.0476457923698272e-06,
+      "loss": 0.0844,
+      "step": 2236
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4529211258444078,
+      "learning_rate": 3.043387132183002e-06,
+      "loss": 0.1162,
+      "step": 2237
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3511034466577657,
+      "learning_rate": 3.039130147421452e-06,
+      "loss": 0.0839,
+      "step": 2238
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.33845853053609387,
+      "learning_rate": 3.0348748417303826e-06,
+      "loss": 0.0539,
+      "step": 2239
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6523629577883808,
+      "learning_rate": 3.0306212187535653e-06,
+      "loss": 0.1134,
+      "step": 2240
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2719026631173136,
+      "learning_rate": 3.026369282133327e-06,
+      "loss": 0.067,
+      "step": 2241
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4498298121008999,
+      "learning_rate": 3.022119035510554e-06,
+      "loss": 0.1098,
+      "step": 2242
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5022798077012678,
+      "learning_rate": 3.017870482524683e-06,
+      "loss": 0.1272,
+      "step": 2243
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2807535534499567,
+      "learning_rate": 3.0136236268137032e-06,
+      "loss": 0.0726,
+      "step": 2244
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5121126276985974,
+      "learning_rate": 3.0093784720141456e-06,
+      "loss": 0.0913,
+      "step": 2245
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.26096862249592634,
+      "learning_rate": 3.005135021761091e-06,
+      "loss": 0.0724,
+      "step": 2246
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7872548946211441,
+      "learning_rate": 3.000893279688155e-06,
+      "loss": 0.1343,
+      "step": 2247
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3225511049616021,
+      "learning_rate": 2.996653249427496e-06,
+      "loss": 0.0654,
+      "step": 2248
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2511025912885566,
+      "learning_rate": 2.992414934609799e-06,
+      "loss": 0.0384,
+      "step": 2249
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2524792538389449,
+      "learning_rate": 2.988178338864289e-06,
+      "loss": 0.0614,
+      "step": 2250
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5460979812134171,
+      "learning_rate": 2.9839434658187128e-06,
+      "loss": 0.1359,
+      "step": 2251
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3406985894983384,
+      "learning_rate": 2.9797103190993424e-06,
+      "loss": 0.0801,
+      "step": 2252
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.22919616400573684,
+      "learning_rate": 2.975478902330976e-06,
+      "loss": 0.0541,
+      "step": 2253
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4635200241574937,
+      "learning_rate": 2.9712492191369245e-06,
+      "loss": 0.0847,
+      "step": 2254
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6917975599765706,
+      "learning_rate": 2.9670212731390202e-06,
+      "loss": 0.1315,
+      "step": 2255
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2251775767650042,
+      "learning_rate": 2.9627950679576023e-06,
+      "loss": 0.0472,
+      "step": 2256
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6397004555752241,
+      "learning_rate": 2.958570607211525e-06,
+      "loss": 0.1295,
+      "step": 2257
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4237224899475914,
+      "learning_rate": 2.9543478945181425e-06,
+      "loss": 0.0561,
+      "step": 2258
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5939801535434834,
+      "learning_rate": 2.9501269334933193e-06,
+      "loss": 0.1463,
+      "step": 2259
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7527823900971068,
+      "learning_rate": 2.945907727751412e-06,
+      "loss": 0.101,
+      "step": 2260
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3487757755684883,
+      "learning_rate": 2.9416902809052817e-06,
+      "loss": 0.066,
+      "step": 2261
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5378627638414805,
+      "learning_rate": 2.9374745965662784e-06,
+      "loss": 0.0837,
+      "step": 2262
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.760392574774493,
+      "learning_rate": 2.933260678344245e-06,
+      "loss": 0.1198,
+      "step": 2263
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.32810580236675957,
+      "learning_rate": 2.9290485298475115e-06,
+      "loss": 0.1006,
+      "step": 2264
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2822204050564391,
+      "learning_rate": 2.924838154682893e-06,
+      "loss": 0.0705,
+      "step": 2265
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5341164505024624,
+      "learning_rate": 2.920629556455684e-06,
+      "loss": 0.0847,
+      "step": 2266
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3743597615263854,
+      "learning_rate": 2.916422738769663e-06,
+      "loss": 0.0841,
+      "step": 2267
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7059118764674842,
+      "learning_rate": 2.912217705227075e-06,
+      "loss": 0.1319,
+      "step": 2268
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.35156988006072726,
+      "learning_rate": 2.908014459428644e-06,
+      "loss": 0.0672,
+      "step": 2269
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3914407366252117,
+      "learning_rate": 2.9038130049735634e-06,
+      "loss": 0.0769,
+      "step": 2270
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2595686182686397,
+      "learning_rate": 2.899613345459491e-06,
+      "loss": 0.0632,
+      "step": 2271
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2833529473405738,
+      "learning_rate": 2.8954154844825433e-06,
+      "loss": 0.0643,
+      "step": 2272
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.47269186350984344,
+      "learning_rate": 2.8912194256373034e-06,
+      "loss": 0.1092,
+      "step": 2273
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.46887301792249453,
+      "learning_rate": 2.8870251725168086e-06,
+      "loss": 0.0981,
+      "step": 2274
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.29914548278429276,
+      "learning_rate": 2.882832728712551e-06,
+      "loss": 0.0647,
+      "step": 2275
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.582651436150537,
+      "learning_rate": 2.8786420978144696e-06,
+      "loss": 0.1176,
+      "step": 2276
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.37627485850216297,
+      "learning_rate": 2.8744532834109556e-06,
+      "loss": 0.0807,
+      "step": 2277
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9227049972258796,
+      "learning_rate": 2.870266289088842e-06,
+      "loss": 0.0877,
+      "step": 2278
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.35587640966716405,
+      "learning_rate": 2.866081118433408e-06,
+      "loss": 0.0694,
+      "step": 2279
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.462732949217908,
+      "learning_rate": 2.8618977750283605e-06,
+      "loss": 0.0932,
+      "step": 2280
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5029208423475465,
+      "learning_rate": 2.857716262455854e-06,
+      "loss": 0.1059,
+      "step": 2281
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.25775354993098337,
+      "learning_rate": 2.8535365842964713e-06,
+      "loss": 0.0536,
+      "step": 2282
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6197971299923342,
+      "learning_rate": 2.849358744129218e-06,
+      "loss": 0.0819,
+      "step": 2283
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.32764532195894824,
+      "learning_rate": 2.845182745531534e-06,
+      "loss": 0.1028,
+      "step": 2284
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.540918453714262,
+      "learning_rate": 2.841008592079281e-06,
+      "loss": 0.1116,
+      "step": 2285
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.49124903855960494,
+      "learning_rate": 2.8368362873467335e-06,
+      "loss": 0.0937,
+      "step": 2286
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5143010565119079,
+      "learning_rate": 2.832665834906593e-06,
+      "loss": 0.0796,
+      "step": 2287
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.30782139294150074,
+      "learning_rate": 2.8284972383299694e-06,
+      "loss": 0.0743,
+      "step": 2288
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.42864511735546884,
+      "learning_rate": 2.8243305011863843e-06,
+      "loss": 0.083,
+      "step": 2289
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.38034608340888637,
+      "learning_rate": 2.8201656270437662e-06,
+      "loss": 0.0827,
+      "step": 2290
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4534563520465355,
+      "learning_rate": 2.81600261946845e-06,
+      "loss": 0.095,
+      "step": 2291
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3548952684845326,
+      "learning_rate": 2.8118414820251707e-06,
+      "loss": 0.0987,
+      "step": 2292
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.28516233145113645,
+      "learning_rate": 2.8076822182770656e-06,
+      "loss": 0.09,
+      "step": 2293
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.42426040806675386,
+      "learning_rate": 2.8035248317856615e-06,
+      "loss": 0.0991,
+      "step": 2294
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3887180211646909,
+      "learning_rate": 2.7993693261108823e-06,
+      "loss": 0.1032,
+      "step": 2295
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5686559544022615,
+      "learning_rate": 2.7952157048110406e-06,
+      "loss": 0.1218,
+      "step": 2296
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.29805688973147837,
+      "learning_rate": 2.791063971442836e-06,
+      "loss": 0.0733,
+      "step": 2297
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.355908684610805,
+      "learning_rate": 2.7869141295613472e-06,
+      "loss": 0.0913,
+      "step": 2298
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5548207284391696,
+      "learning_rate": 2.782766182720038e-06,
+      "loss": 0.1288,
+      "step": 2299
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5217284728277821,
+      "learning_rate": 2.7786201344707487e-06,
+      "loss": 0.1233,
+      "step": 2300
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4866602616651964,
+      "learning_rate": 2.774475988363694e-06,
+      "loss": 0.1037,
+      "step": 2301
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2904837015128024,
+      "learning_rate": 2.770333747947455e-06,
+      "loss": 0.0466,
+      "step": 2302
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.33017920417950397,
+      "learning_rate": 2.7661934167689887e-06,
+      "loss": 0.1245,
+      "step": 2303
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3270457335369551,
+      "learning_rate": 2.762054998373613e-06,
+      "loss": 0.0462,
+      "step": 2304
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.36402043517321275,
+      "learning_rate": 2.7579184963050056e-06,
+      "loss": 0.094,
+      "step": 2305
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.23328938667347496,
+      "learning_rate": 2.753783914105208e-06,
+      "loss": 0.0527,
+      "step": 2306
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4116614024354482,
+      "learning_rate": 2.749651255314616e-06,
+      "loss": 0.0804,
+      "step": 2307
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.35467948540901745,
+      "learning_rate": 2.7455205234719797e-06,
+      "loss": 0.1148,
+      "step": 2308
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3075646107856484,
+      "learning_rate": 2.741391722114394e-06,
+      "loss": 0.0664,
+      "step": 2309
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5760189831599258,
+      "learning_rate": 2.7372648547773063e-06,
+      "loss": 0.083,
+      "step": 2310
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4691585993981161,
+      "learning_rate": 2.733139924994505e-06,
+      "loss": 0.0907,
+      "step": 2311
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4152150789814372,
+      "learning_rate": 2.729016936298124e-06,
+      "loss": 0.1374,
+      "step": 2312
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.37046898927892935,
+      "learning_rate": 2.7248958922186263e-06,
+      "loss": 0.1009,
+      "step": 2313
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5507208814623268,
+      "learning_rate": 2.720776796284818e-06,
+      "loss": 0.1398,
+      "step": 2314
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.45677895593757234,
+      "learning_rate": 2.716659652023833e-06,
+      "loss": 0.0814,
+      "step": 2315
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4696547355345099,
+      "learning_rate": 2.7125444629611376e-06,
+      "loss": 0.1145,
+      "step": 2316
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.262415017630031,
+      "learning_rate": 2.7084312326205164e-06,
+      "loss": 0.0743,
+      "step": 2317
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.34551270325499794,
+      "learning_rate": 2.704319964524085e-06,
+      "loss": 0.1069,
+      "step": 2318
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.42502155045237316,
+      "learning_rate": 2.700210662192276e-06,
+      "loss": 0.1037,
+      "step": 2319
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4374639820148765,
+      "learning_rate": 2.6961033291438343e-06,
+      "loss": 0.0675,
+      "step": 2320
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.36941815964521674,
+      "learning_rate": 2.6919979688958255e-06,
+      "loss": 0.0922,
+      "step": 2321
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.35262405423393256,
+      "learning_rate": 2.6878945849636206e-06,
+      "loss": 0.0954,
+      "step": 2322
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4429212320097036,
+      "learning_rate": 2.683793180860905e-06,
+      "loss": 0.1116,
+      "step": 2323
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4694217281221106,
+      "learning_rate": 2.6796937600996587e-06,
+      "loss": 0.0957,
+      "step": 2324
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3089160314002098,
+      "learning_rate": 2.6755963261901706e-06,
+      "loss": 0.0707,
+      "step": 2325
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8337121284683863,
+      "learning_rate": 2.671500882641027e-06,
+      "loss": 0.1809,
+      "step": 2326
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.35030972559754975,
+      "learning_rate": 2.6674074329591127e-06,
+      "loss": 0.0749,
+      "step": 2327
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6028853006788526,
+      "learning_rate": 2.6633159806495967e-06,
+      "loss": 0.1404,
+      "step": 2328
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.43589668081556143,
+      "learning_rate": 2.6592265292159446e-06,
+      "loss": 0.0852,
+      "step": 2329
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5621747812600117,
+      "learning_rate": 2.655139082159908e-06,
+      "loss": 0.117,
+      "step": 2330
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5545220254320469,
+      "learning_rate": 2.6510536429815224e-06,
+      "loss": 0.0997,
+      "step": 2331
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.32132232425096724,
+      "learning_rate": 2.6469702151791e-06,
+      "loss": 0.0761,
+      "step": 2332
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.34028601038042017,
+      "learning_rate": 2.6428888022492335e-06,
+      "loss": 0.0973,
+      "step": 2333
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3065868212890701,
+      "learning_rate": 2.6388094076867916e-06,
+      "loss": 0.0992,
+      "step": 2334
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.31547156332143655,
+      "learning_rate": 2.634732034984915e-06,
+      "loss": 0.097,
+      "step": 2335
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.39473869436419473,
+      "learning_rate": 2.6306566876350072e-06,
+      "loss": 0.0886,
+      "step": 2336
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.4191660642892399,
+      "learning_rate": 2.6265833691267438e-06,
+      "loss": 0.0876,
+      "step": 2337
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.4452460547410256,
+      "learning_rate": 2.622512082948063e-06,
+      "loss": 0.0907,
+      "step": 2338
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3975032750118112,
+      "learning_rate": 2.6184428325851576e-06,
+      "loss": 0.0554,
+      "step": 2339
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.4047627494777148,
+      "learning_rate": 2.6143756215224803e-06,
+      "loss": 0.1015,
+      "step": 2340
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5047816744255545,
+      "learning_rate": 2.6103104532427392e-06,
+      "loss": 0.115,
+      "step": 2341
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.42429120474061843,
+      "learning_rate": 2.606247331226892e-06,
+      "loss": 0.0852,
+      "step": 2342
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5137784528616098,
+      "learning_rate": 2.6021862589541413e-06,
+      "loss": 0.0951,
+      "step": 2343
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6173231686309556,
+      "learning_rate": 2.5981272399019384e-06,
+      "loss": 0.1037,
+      "step": 2344
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.4402885133628901,
+      "learning_rate": 2.594070277545975e-06,
+      "loss": 0.1001,
+      "step": 2345
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.47376847868750077,
+      "learning_rate": 2.590015375360183e-06,
+      "loss": 0.094,
+      "step": 2346
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6065813528577138,
+      "learning_rate": 2.585962536816725e-06,
+      "loss": 0.1145,
+      "step": 2347
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.33636633356163,
+      "learning_rate": 2.581911765386004e-06,
+      "loss": 0.0589,
+      "step": 2348
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3881472042788369,
+      "learning_rate": 2.5778630645366477e-06,
+      "loss": 0.084,
+      "step": 2349
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5420653870323854,
+      "learning_rate": 2.5738164377355148e-06,
+      "loss": 0.0987,
+      "step": 2350
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5276164136281538,
+      "learning_rate": 2.569771888447682e-06,
+      "loss": 0.0973,
+      "step": 2351
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3759638370555702,
+      "learning_rate": 2.5657294201364526e-06,
+      "loss": 0.0692,
+      "step": 2352
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5737221650371205,
+      "learning_rate": 2.561689036263347e-06,
+      "loss": 0.1292,
+      "step": 2353
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.616287565521532,
+      "learning_rate": 2.557650740288098e-06,
+      "loss": 0.1382,
+      "step": 2354
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.33382954054747793,
+      "learning_rate": 2.5536145356686528e-06,
+      "loss": 0.0863,
+      "step": 2355
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5897948544876979,
+      "learning_rate": 2.5495804258611667e-06,
+      "loss": 0.1484,
+      "step": 2356
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2847946190578612,
+      "learning_rate": 2.545548414320006e-06,
+      "loss": 0.0624,
+      "step": 2357
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3132650762979806,
+      "learning_rate": 2.541518504497731e-06,
+      "loss": 0.0455,
+      "step": 2358
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5106073690364898,
+      "learning_rate": 2.5374906998451094e-06,
+      "loss": 0.0776,
+      "step": 2359
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5190182152117772,
+      "learning_rate": 2.5334650038111045e-06,
+      "loss": 0.0939,
+      "step": 2360
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.32729181393164075,
+      "learning_rate": 2.5294414198428764e-06,
+      "loss": 0.0756,
+      "step": 2361
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.30135334398108327,
+      "learning_rate": 2.525419951385769e-06,
+      "loss": 0.0653,
+      "step": 2362
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5707745801369596,
+      "learning_rate": 2.5214006018833225e-06,
+      "loss": 0.0994,
+      "step": 2363
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5057873530785689,
+      "learning_rate": 2.5173833747772614e-06,
+      "loss": 0.0739,
+      "step": 2364
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.46545610006026966,
+      "learning_rate": 2.5133682735074904e-06,
+      "loss": 0.1025,
+      "step": 2365
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.37332773727241425,
+      "learning_rate": 2.5093553015120937e-06,
+      "loss": 0.0291,
+      "step": 2366
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.46681952599272564,
+      "learning_rate": 2.5053444622273336e-06,
+      "loss": 0.0605,
+      "step": 2367
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5989662159459328,
+      "learning_rate": 2.5013357590876496e-06,
+      "loss": 0.1448,
+      "step": 2368
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4341230373512625,
+      "learning_rate": 2.4973291955256427e-06,
+      "loss": 0.1006,
+      "step": 2369
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.29255441779312863,
+      "learning_rate": 2.4933247749720912e-06,
+      "loss": 0.0699,
+      "step": 2370
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.23899279770949503,
+      "learning_rate": 2.4893225008559335e-06,
+      "loss": 0.0459,
+      "step": 2371
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6315854138492232,
+      "learning_rate": 2.4853223766042737e-06,
+      "loss": 0.0829,
+      "step": 2372
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.573114555979581,
+      "learning_rate": 2.4813244056423692e-06,
+      "loss": 0.0958,
+      "step": 2373
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.44635644161624893,
+      "learning_rate": 2.477328591393638e-06,
+      "loss": 0.0758,
+      "step": 2374
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4551954292578954,
+      "learning_rate": 2.4733349372796506e-06,
+      "loss": 0.0987,
+      "step": 2375
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.31487377604106886,
+      "learning_rate": 2.4693434467201293e-06,
+      "loss": 0.0774,
+      "step": 2376
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6363468084043964,
+      "learning_rate": 2.4653541231329387e-06,
+      "loss": 0.1264,
+      "step": 2377
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.321675658045344,
+      "learning_rate": 2.461366969934092e-06,
+      "loss": 0.0885,
+      "step": 2378
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6557954818417866,
+      "learning_rate": 2.457381990537743e-06,
+      "loss": 0.0874,
+      "step": 2379
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.31224448558205475,
+      "learning_rate": 2.4533991883561868e-06,
+      "loss": 0.0708,
+      "step": 2380
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3424106441845974,
+      "learning_rate": 2.4494185667998473e-06,
+      "loss": 0.0562,
+      "step": 2381
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.26569710307490246,
+      "learning_rate": 2.4454401292772874e-06,
+      "loss": 0.0753,
+      "step": 2382
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.34777337003477676,
+      "learning_rate": 2.4414638791951972e-06,
+      "loss": 0.0545,
+      "step": 2383
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3597508592783276,
+      "learning_rate": 2.437489819958396e-06,
+      "loss": 0.0544,
+      "step": 2384
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4589742365365593,
+      "learning_rate": 2.4335179549698233e-06,
+      "loss": 0.068,
+      "step": 2385
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5046809887906143,
+      "learning_rate": 2.429548287630542e-06,
+      "loss": 0.1102,
+      "step": 2386
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.31299804205560783,
+      "learning_rate": 2.425580821339733e-06,
+      "loss": 0.0433,
+      "step": 2387
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.26008699604595936,
+      "learning_rate": 2.421615559494693e-06,
+      "loss": 0.0552,
+      "step": 2388
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6297760568594132,
+      "learning_rate": 2.417652505490831e-06,
+      "loss": 0.1029,
+      "step": 2389
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4213946805287631,
+      "learning_rate": 2.4136916627216656e-06,
+      "loss": 0.0934,
+      "step": 2390
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.390487020380129,
+      "learning_rate": 2.4097330345788184e-06,
+      "loss": 0.058,
+      "step": 2391
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7391042862420466,
+      "learning_rate": 2.4057766244520185e-06,
+      "loss": 0.1124,
+      "step": 2392
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.316328593496384,
+      "learning_rate": 2.4018224357290952e-06,
+      "loss": 0.0647,
+      "step": 2393
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5438667023288013,
+      "learning_rate": 2.3978704717959777e-06,
+      "loss": 0.1048,
+      "step": 2394
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.464045075823488,
+      "learning_rate": 2.393920736036683e-06,
+      "loss": 0.0812,
+      "step": 2395
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.30258637973305913,
+      "learning_rate": 2.3899732318333275e-06,
+      "loss": 0.073,
+      "step": 2396
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.30139675813601924,
+      "learning_rate": 2.3860279625661125e-06,
+      "loss": 0.0825,
+      "step": 2397
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6560813055550506,
+      "learning_rate": 2.3820849316133303e-06,
+      "loss": 0.102,
+      "step": 2398
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7185847389274075,
+      "learning_rate": 2.3781441423513497e-06,
+      "loss": 0.1199,
+      "step": 2399
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2171484352781935,
+      "learning_rate": 2.374205598154624e-06,
+      "loss": 0.0439,
+      "step": 2400
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.575515360659286,
+      "learning_rate": 2.3702693023956853e-06,
+      "loss": 0.0949,
+      "step": 2401
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.47632807746805766,
+      "learning_rate": 2.3663352584451394e-06,
+      "loss": 0.0777,
+      "step": 2402
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5535319658893976,
+      "learning_rate": 2.36240346967166e-06,
+      "loss": 0.0616,
+      "step": 2403
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.980873831478703,
+      "learning_rate": 2.3584739394419953e-06,
+      "loss": 0.1417,
+      "step": 2404
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.33218187698802715,
+      "learning_rate": 2.3545466711209585e-06,
+      "loss": 0.0918,
+      "step": 2405
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3762678115177452,
+      "learning_rate": 2.3506216680714204e-06,
+      "loss": 0.0813,
+      "step": 2406
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.44929720481790353,
+      "learning_rate": 2.3466989336543196e-06,
+      "loss": 0.1212,
+      "step": 2407
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4881987037263634,
+      "learning_rate": 2.342778471228648e-06,
+      "loss": 0.0781,
+      "step": 2408
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.37359657823670966,
+      "learning_rate": 2.3388602841514542e-06,
+      "loss": 0.0694,
+      "step": 2409
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.45748035159517075,
+      "learning_rate": 2.3349443757778346e-06,
+      "loss": 0.0742,
+      "step": 2410
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3542304760666429,
+      "learning_rate": 2.3310307494609374e-06,
+      "loss": 0.0905,
+      "step": 2411
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5466585898427205,
+      "learning_rate": 2.3271194085519565e-06,
+      "loss": 0.087,
+      "step": 2412
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6021142086632345,
+      "learning_rate": 2.323210356400131e-06,
+      "loss": 0.1261,
+      "step": 2413
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.47097148145120954,
+      "learning_rate": 2.3193035963527345e-06,
+      "loss": 0.1298,
+      "step": 2414
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.21159878346072342,
+      "learning_rate": 2.315399131755081e-06,
+      "loss": 0.0548,
+      "step": 2415
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3219946254576339,
+      "learning_rate": 2.3114969659505197e-06,
+      "loss": 0.0839,
+      "step": 2416
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2530687573818007,
+      "learning_rate": 2.307597102280434e-06,
+      "loss": 0.0695,
+      "step": 2417
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5291999227454721,
+      "learning_rate": 2.303699544084227e-06,
+      "loss": 0.0776,
+      "step": 2418
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3675315539376898,
+      "learning_rate": 2.2998042946993355e-06,
+      "loss": 0.0819,
+      "step": 2419
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.33178216560844864,
+      "learning_rate": 2.2959113574612204e-06,
+      "loss": 0.0737,
+      "step": 2420
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.28283196379546766,
+      "learning_rate": 2.2920207357033535e-06,
+      "loss": 0.0645,
+      "step": 2421
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3188858291542534,
+      "learning_rate": 2.2881324327572336e-06,
+      "loss": 0.0709,
+      "step": 2422
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3255783974618697,
+      "learning_rate": 2.2842464519523692e-06,
+      "loss": 0.0771,
+      "step": 2423
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5174623259760657,
+      "learning_rate": 2.2803627966162833e-06,
+      "loss": 0.0921,
+      "step": 2424
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.48022556995191873,
+      "learning_rate": 2.2764814700745025e-06,
+      "loss": 0.0902,
+      "step": 2425
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.624728299602598,
+      "learning_rate": 2.272602475650563e-06,
+      "loss": 0.098,
+      "step": 2426
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.24497852595002126,
+      "learning_rate": 2.2687258166660055e-06,
+      "loss": 0.0602,
+      "step": 2427
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5642659374889866,
+      "learning_rate": 2.2648514964403685e-06,
+      "loss": 0.0993,
+      "step": 2428
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4953658521208724,
+      "learning_rate": 2.260979518291186e-06,
+      "loss": 0.0739,
+      "step": 2429
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4888118571508726,
+      "learning_rate": 2.25710988553399e-06,
+      "loss": 0.0698,
+      "step": 2430
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.34404975575745816,
+      "learning_rate": 2.253242601482303e-06,
+      "loss": 0.0618,
+      "step": 2431
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5492917518585991,
+      "learning_rate": 2.2493776694476376e-06,
+      "loss": 0.1028,
+      "step": 2432
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3704229676752792,
+      "learning_rate": 2.245515092739488e-06,
+      "loss": 0.0656,
+      "step": 2433
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.26439338016716446,
+      "learning_rate": 2.241654874665336e-06,
+      "loss": 0.0818,
+      "step": 2434
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.36120165175789193,
+      "learning_rate": 2.2377970185306424e-06,
+      "loss": 0.0798,
+      "step": 2435
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.42707191592250243,
+      "learning_rate": 2.233941527638848e-06,
+      "loss": 0.0978,
+      "step": 2436
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2885428286429536,
+      "learning_rate": 2.23008840529136e-06,
+      "loss": 0.068,
+      "step": 2437
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5509856220690513,
+      "learning_rate": 2.2262376547875665e-06,
+      "loss": 0.1508,
+      "step": 2438
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3659854627199059,
+      "learning_rate": 2.222389279424823e-06,
+      "loss": 0.0588,
+      "step": 2439
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3313354240544666,
+      "learning_rate": 2.2185432824984455e-06,
+      "loss": 0.0627,
+      "step": 2440
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.510604467026776,
+      "learning_rate": 2.214699667301719e-06,
+      "loss": 0.1026,
+      "step": 2441
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4047591204620746,
+      "learning_rate": 2.2108584371258875e-06,
+      "loss": 0.0535,
+      "step": 2442
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4356057849030091,
+      "learning_rate": 2.207019595260154e-06,
+      "loss": 0.0863,
+      "step": 2443
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.39166512422836935,
+      "learning_rate": 2.2031831449916716e-06,
+      "loss": 0.0915,
+      "step": 2444
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4891860123956304,
+      "learning_rate": 2.1993490896055514e-06,
+      "loss": 0.1222,
+      "step": 2445
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.29687866285229086,
+      "learning_rate": 2.19551743238485e-06,
+      "loss": 0.0742,
+      "step": 2446
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5393478987922212,
+      "learning_rate": 2.191688176610575e-06,
+      "loss": 0.1401,
+      "step": 2447
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3816829342957274,
+      "learning_rate": 2.1878613255616693e-06,
+      "loss": 0.0704,
+      "step": 2448
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.26128072920496465,
+      "learning_rate": 2.1840368825150243e-06,
+      "loss": 0.069,
+      "step": 2449
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.42514841528539127,
+      "learning_rate": 2.1802148507454675e-06,
+      "loss": 0.075,
+      "step": 2450
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.356997507982979,
+      "learning_rate": 2.1763952335257622e-06,
+      "loss": 0.0736,
+      "step": 2451
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3273788304692756,
+      "learning_rate": 2.172578034126599e-06,
+      "loss": 0.084,
+      "step": 2452
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2841996763065209,
+      "learning_rate": 2.1687632558166055e-06,
+      "loss": 0.0719,
+      "step": 2453
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3224434379173398,
+      "learning_rate": 2.1649509018623334e-06,
+      "loss": 0.0895,
+      "step": 2454
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0355583441148388,
+      "learning_rate": 2.1611409755282542e-06,
+      "loss": 0.1385,
+      "step": 2455
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3504069134414007,
+      "learning_rate": 2.157333480076767e-06,
+      "loss": 0.086,
+      "step": 2456
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.38088794633133544,
+      "learning_rate": 2.1535284187681866e-06,
+      "loss": 0.0982,
+      "step": 2457
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.24279834366214054,
+      "learning_rate": 2.149725794860745e-06,
+      "loss": 0.041,
+      "step": 2458
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.26569461877107126,
+      "learning_rate": 2.1459256116105815e-06,
+      "loss": 0.0595,
+      "step": 2459
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5549982536064857,
+      "learning_rate": 2.1421278722717524e-06,
+      "loss": 0.0644,
+      "step": 2460
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.31356757379161765,
+      "learning_rate": 2.1383325800962187e-06,
+      "loss": 0.0639,
+      "step": 2461
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4153190752587748,
+      "learning_rate": 2.1345397383338466e-06,
+      "loss": 0.076,
+      "step": 2462
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6208101669149098,
+      "learning_rate": 2.1307493502324007e-06,
+      "loss": 0.1198,
+      "step": 2463
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.476865981895542,
+      "learning_rate": 2.1269614190375477e-06,
+      "loss": 0.0803,
+      "step": 2464
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3277038078003154,
+      "learning_rate": 2.123175947992851e-06,
+      "loss": 0.0765,
+      "step": 2465
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.34938994313075933,
+      "learning_rate": 2.119392940339767e-06,
+      "loss": 0.0716,
+      "step": 2466
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.45887568894719905,
+      "learning_rate": 2.1156123993176398e-06,
+      "loss": 0.0843,
+      "step": 2467
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5093631762615431,
+      "learning_rate": 2.1118343281637042e-06,
+      "loss": 0.0825,
+      "step": 2468
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.260934900231959,
+      "learning_rate": 2.1080587301130795e-06,
+      "loss": 0.0546,
+      "step": 2469
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4705286799504309,
+      "learning_rate": 2.1042856083987694e-06,
+      "loss": 0.1055,
+      "step": 2470
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5109345615450195,
+      "learning_rate": 2.1005149662516517e-06,
+      "loss": 0.1004,
+      "step": 2471
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.27146351549399605,
+      "learning_rate": 2.096746806900485e-06,
+      "loss": 0.0617,
+      "step": 2472
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6455686348885666,
+      "learning_rate": 2.0929811335719037e-06,
+      "loss": 0.1418,
+      "step": 2473
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.33708652872222555,
+      "learning_rate": 2.0892179494904076e-06,
+      "loss": 0.0685,
+      "step": 2474
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.37054245087436677,
+      "learning_rate": 2.085457257878369e-06,
+      "loss": 0.0427,
+      "step": 2475
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2960352781684179,
+      "learning_rate": 2.0816990619560264e-06,
+      "loss": 0.0489,
+      "step": 2476
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5256612762305234,
+      "learning_rate": 2.077943364941481e-06,
+      "loss": 0.0941,
+      "step": 2477
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.583439390213951,
+      "learning_rate": 2.07419017005069e-06,
+      "loss": 0.114,
+      "step": 2478
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4838188104210831,
+      "learning_rate": 2.0704394804974732e-06,
+      "loss": 0.0782,
+      "step": 2479
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.516294044621413,
+      "learning_rate": 2.0666912994935034e-06,
+      "loss": 0.131,
+      "step": 2480
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5609974918952368,
+      "learning_rate": 2.0629456302483068e-06,
+      "loss": 0.1124,
+      "step": 2481
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.39316597613776283,
+      "learning_rate": 2.059202475969254e-06,
+      "loss": 0.0864,
+      "step": 2482
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5216492741893036,
+      "learning_rate": 2.0554618398615674e-06,
+      "loss": 0.1055,
+      "step": 2483
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.304966561786873,
+      "learning_rate": 2.0517237251283096e-06,
+      "loss": 0.0662,
+      "step": 2484
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3776164971594819,
+      "learning_rate": 2.0479881349703885e-06,
+      "loss": 0.09,
+      "step": 2485
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3502129832354215,
+      "learning_rate": 2.0442550725865434e-06,
+      "loss": 0.0872,
+      "step": 2486
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2679011132795074,
+      "learning_rate": 2.0405245411733548e-06,
+      "loss": 0.0731,
+      "step": 2487
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5659730523629836,
+      "learning_rate": 2.036796543925236e-06,
+      "loss": 0.1215,
+      "step": 2488
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.26544866911837645,
+      "learning_rate": 2.0330710840344257e-06,
+      "loss": 0.0544,
+      "step": 2489
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.426339964067119,
+      "learning_rate": 2.0293481646909934e-06,
+      "loss": 0.0761,
+      "step": 2490
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.18642279906911316,
+      "learning_rate": 2.025627789082833e-06,
+      "loss": 0.0311,
+      "step": 2491
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.573801440813466,
+      "learning_rate": 2.021909960395661e-06,
+      "loss": 0.108,
+      "step": 2492
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3164197409282355,
+      "learning_rate": 2.0181946818130073e-06,
+      "loss": 0.0743,
+      "step": 2493
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.28230931710733365,
+      "learning_rate": 2.0144819565162265e-06,
+      "loss": 0.0707,
+      "step": 2494
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2776470109975148,
+      "learning_rate": 2.010771787684484e-06,
+      "loss": 0.0731,
+      "step": 2495
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.49469879552584534,
+      "learning_rate": 2.0070641784947516e-06,
+      "loss": 0.0919,
+      "step": 2496
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4661691610632084,
+      "learning_rate": 2.0033591321218136e-06,
+      "loss": 0.0767,
+      "step": 2497
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.47424428622317366,
+      "learning_rate": 1.999656651738259e-06,
+      "loss": 0.0427,
+      "step": 2498
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.39612728899221655,
+      "learning_rate": 1.9959567405144825e-06,
+      "loss": 0.0935,
+      "step": 2499
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3083803254418959,
+      "learning_rate": 1.9922594016186716e-06,
+      "loss": 0.0647,
+      "step": 2500
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.44479842277769166,
+      "learning_rate": 1.9885646382168165e-06,
+      "loss": 0.0866,
+      "step": 2501
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.662541347825402,
+      "learning_rate": 1.984872453472702e-06,
+      "loss": 0.1136,
+      "step": 2502
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3126659292319549,
+      "learning_rate": 1.9811828505479046e-06,
+      "loss": 0.0607,
+      "step": 2503
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5218609887142224,
+      "learning_rate": 1.9774958326017867e-06,
+      "loss": 0.0864,
+      "step": 2504
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.40934548562644574,
+      "learning_rate": 1.9738114027915007e-06,
+      "loss": 0.0929,
+      "step": 2505
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.29386181225783486,
+      "learning_rate": 1.9701295642719836e-06,
+      "loss": 0.0707,
+      "step": 2506
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.495108064171112,
+      "learning_rate": 1.9664503201959483e-06,
+      "loss": 0.0769,
+      "step": 2507
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.28809812690886766,
+      "learning_rate": 1.9627736737138914e-06,
+      "loss": 0.0457,
+      "step": 2508
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.38777632566314457,
+      "learning_rate": 1.9590996279740837e-06,
+      "loss": 0.0858,
+      "step": 2509
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.4790592223308331,
+      "learning_rate": 1.9554281861225694e-06,
+      "loss": 0.1072,
+      "step": 2510
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.4780180012717242,
+      "learning_rate": 1.9517593513031596e-06,
+      "loss": 0.09,
+      "step": 2511
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.37817152714820146,
+      "learning_rate": 1.948093126657437e-06,
+      "loss": 0.112,
+      "step": 2512
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.38182769355388146,
+      "learning_rate": 1.944429515324749e-06,
+      "loss": 0.0734,
+      "step": 2513
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.4388759296646012,
+      "learning_rate": 1.940768520442204e-06,
+      "loss": 0.0856,
+      "step": 2514
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.25436412200026715,
+      "learning_rate": 1.9371101451446685e-06,
+      "loss": 0.0573,
+      "step": 2515
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5931175586972784,
+      "learning_rate": 1.9334543925647676e-06,
+      "loss": 0.1285,
+      "step": 2516
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5238592325729333,
+      "learning_rate": 1.929801265832882e-06,
+      "loss": 0.1078,
+      "step": 2517
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.35228657987798356,
+      "learning_rate": 1.9261507680771435e-06,
+      "loss": 0.0739,
+      "step": 2518
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.4872524333058193,
+      "learning_rate": 1.9225029024234283e-06,
+      "loss": 0.11,
+      "step": 2519
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3897387784932764,
+      "learning_rate": 1.9188576719953635e-06,
+      "loss": 0.1008,
+      "step": 2520
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3625974938173454,
+      "learning_rate": 1.9152150799143205e-06,
+      "loss": 0.0941,
+      "step": 2521
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.282595284072661,
+      "learning_rate": 1.911575129299406e-06,
+      "loss": 0.0852,
+      "step": 2522
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8018839374923361,
+      "learning_rate": 1.9079378232674694e-06,
+      "loss": 0.1595,
+      "step": 2523
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.087252907192483,
+      "learning_rate": 1.904303164933095e-06,
+      "loss": 0.0973,
+      "step": 2524
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5931258894394622,
+      "learning_rate": 1.9006711574086006e-06,
+      "loss": 0.1251,
+      "step": 2525
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5540071030415847,
+      "learning_rate": 1.8970418038040294e-06,
+      "loss": 0.1252,
+      "step": 2526
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.36403783381932675,
+      "learning_rate": 1.8934151072271573e-06,
+      "loss": 0.1087,
+      "step": 2527
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.4849985474767569,
+      "learning_rate": 1.8897910707834827e-06,
+      "loss": 0.1204,
+      "step": 2528
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5593547938359725,
+      "learning_rate": 1.886169697576229e-06,
+      "loss": 0.1585,
+      "step": 2529
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5711921629050263,
+      "learning_rate": 1.8825509907063328e-06,
+      "loss": 0.0873,
+      "step": 2530
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3440541183818768,
+      "learning_rate": 1.8789349532724538e-06,
+      "loss": 0.023,
+      "step": 2531
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.42056902290070336,
+      "learning_rate": 1.875321588370963e-06,
+      "loss": 0.0972,
+      "step": 2532
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5454169841754962,
+      "learning_rate": 1.871710899095946e-06,
+      "loss": 0.121,
+      "step": 2533
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6885413968558864,
+      "learning_rate": 1.8681028885391905e-06,
+      "loss": 0.0916,
+      "step": 2534
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.35303891701106427,
+      "learning_rate": 1.8644975597901977e-06,
+      "loss": 0.0948,
+      "step": 2535
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.37936485228068184,
+      "learning_rate": 1.8608949159361678e-06,
+      "loss": 0.0715,
+      "step": 2536
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2558958213488092,
+      "learning_rate": 1.8572949600620065e-06,
+      "loss": 0.0613,
+      "step": 2537
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.479947753306411,
+      "learning_rate": 1.8536976952503117e-06,
+      "loss": 0.1268,
+      "step": 2538
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6083090733298329,
+      "learning_rate": 1.8501031245813812e-06,
+      "loss": 0.1272,
+      "step": 2539
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.42330265842017006,
+      "learning_rate": 1.8465112511332068e-06,
+      "loss": 0.1344,
+      "step": 2540
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.22908849834727446,
+      "learning_rate": 1.8429220779814654e-06,
+      "loss": 0.0506,
+      "step": 2541
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.458204674075358,
+      "learning_rate": 1.8393356081995268e-06,
+      "loss": 0.1042,
+      "step": 2542
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.45881719631610296,
+      "learning_rate": 1.8357518448584438e-06,
+      "loss": 0.1073,
+      "step": 2543
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5492049235325289,
+      "learning_rate": 1.8321707910269543e-06,
+      "loss": 0.107,
+      "step": 2544
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3855242374478838,
+      "learning_rate": 1.8285924497714702e-06,
+      "loss": 0.0735,
+      "step": 2545
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6934325475627683,
+      "learning_rate": 1.825016824156086e-06,
+      "loss": 0.0981,
+      "step": 2546
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5807867962657384,
+      "learning_rate": 1.821443917242569e-06,
+      "loss": 0.1551,
+      "step": 2547
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3945357997814155,
+      "learning_rate": 1.81787373209036e-06,
+      "loss": 0.1006,
+      "step": 2548
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7882629236339732,
+      "learning_rate": 1.814306271756565e-06,
+      "loss": 0.1191,
+      "step": 2549
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2720043278141304,
+      "learning_rate": 1.8107415392959615e-06,
+      "loss": 0.0473,
+      "step": 2550
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.27683130579675463,
+      "learning_rate": 1.8071795377609886e-06,
+      "loss": 0.0551,
+      "step": 2551
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6525786380014297,
+      "learning_rate": 1.803620270201749e-06,
+      "loss": 0.1733,
+      "step": 2552
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.4844662615536015,
+      "learning_rate": 1.8000637396660008e-06,
+      "loss": 0.0837,
+      "step": 2553
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.50943545728967,
+      "learning_rate": 1.796509949199161e-06,
+      "loss": 0.108,
+      "step": 2554
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.38158254719622997,
+      "learning_rate": 1.7929589018443016e-06,
+      "loss": 0.0937,
+      "step": 2555
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.28357806200464436,
+      "learning_rate": 1.7894106006421413e-06,
+      "loss": 0.0663,
+      "step": 2556
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0821239900705801,
+      "learning_rate": 1.7858650486310514e-06,
+      "loss": 0.1272,
+      "step": 2557
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.48254464476777337,
+      "learning_rate": 1.782322248847047e-06,
+      "loss": 0.1075,
+      "step": 2558
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.4295029014847295,
+      "learning_rate": 1.778782204323789e-06,
+      "loss": 0.1138,
+      "step": 2559
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.4753672790765889,
+      "learning_rate": 1.7752449180925746e-06,
+      "loss": 0.1429,
+      "step": 2560
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.32048825085718274,
+      "learning_rate": 1.7717103931823427e-06,
+      "loss": 0.0665,
+      "step": 2561
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.614838330598215,
+      "learning_rate": 1.7681786326196665e-06,
+      "loss": 0.1012,
+      "step": 2562
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.41390375713775546,
+      "learning_rate": 1.7646496394287539e-06,
+      "loss": 0.0989,
+      "step": 2563
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5144818871779464,
+      "learning_rate": 1.7611234166314383e-06,
+      "loss": 0.1237,
+      "step": 2564
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5951516361411493,
+      "learning_rate": 1.7575999672471866e-06,
+      "loss": 0.1359,
+      "step": 2565
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.4378968878280546,
+      "learning_rate": 1.7540792942930867e-06,
+      "loss": 0.0957,
+      "step": 2566
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3247311437329441,
+      "learning_rate": 1.750561400783854e-06,
+      "loss": 0.0907,
+      "step": 2567
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3465510699911141,
+      "learning_rate": 1.7470462897318158e-06,
+      "loss": 0.0722,
+      "step": 2568
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.37834003760338325,
+      "learning_rate": 1.743533964146924e-06,
+      "loss": 0.0758,
+      "step": 2569
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.4777199453984389,
+      "learning_rate": 1.7400244270367429e-06,
+      "loss": 0.09,
+      "step": 2570
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.43798325624572815,
+      "learning_rate": 1.73651768140645e-06,
+      "loss": 0.0629,
+      "step": 2571
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5783430551555623,
+      "learning_rate": 1.7330137302588296e-06,
+      "loss": 0.0874,
+      "step": 2572
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.525030866571335,
+      "learning_rate": 1.7295125765942756e-06,
+      "loss": 0.1089,
+      "step": 2573
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.24218658107287364,
+      "learning_rate": 1.7260142234107884e-06,
+      "loss": 0.0503,
+      "step": 2574
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0969761535113411,
+      "learning_rate": 1.7225186737039639e-06,
+      "loss": 0.06,
+      "step": 2575
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.28402634767935336,
+      "learning_rate": 1.7190259304670038e-06,
+      "loss": 0.075,
+      "step": 2576
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5489601653379527,
+      "learning_rate": 1.7155359966907031e-06,
+      "loss": 0.1,
+      "step": 2577
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.47282715089742283,
+      "learning_rate": 1.7120488753634557e-06,
+      "loss": 0.0822,
+      "step": 2578
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5390442319351164,
+      "learning_rate": 1.7085645694712389e-06,
+      "loss": 0.0481,
+      "step": 2579
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.4174594573499582,
+      "learning_rate": 1.7050830819976266e-06,
+      "loss": 0.1074,
+      "step": 2580
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.36223049253439804,
+      "learning_rate": 1.701604415923776e-06,
+      "loss": 0.0666,
+      "step": 2581
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.33486703002453844,
+      "learning_rate": 1.698128574228432e-06,
+      "loss": 0.0927,
+      "step": 2582
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.30391198102043865,
+      "learning_rate": 1.6946555598879138e-06,
+      "loss": 0.0504,
+      "step": 2583
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3586138390448678,
+      "learning_rate": 1.6911853758761266e-06,
+      "loss": 0.0703,
+      "step": 2584
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.4143796337534252,
+      "learning_rate": 1.6877180251645487e-06,
+      "loss": 0.0652,
+      "step": 2585
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5238135076267086,
+      "learning_rate": 1.6842535107222341e-06,
+      "loss": 0.1455,
+      "step": 2586
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5336304169816062,
+      "learning_rate": 1.6807918355158047e-06,
+      "loss": 0.1551,
+      "step": 2587
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.4029823891279307,
+      "learning_rate": 1.6773330025094548e-06,
+      "loss": 0.0804,
+      "step": 2588
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3002081949142874,
+      "learning_rate": 1.6738770146649452e-06,
+      "loss": 0.0664,
+      "step": 2589
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.30477472734831845,
+      "learning_rate": 1.6704238749415958e-06,
+      "loss": 0.0567,
+      "step": 2590
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2851830159767868,
+      "learning_rate": 1.6669735862962921e-06,
+      "loss": 0.0558,
+      "step": 2591
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.24950351066145945,
+      "learning_rate": 1.6635261516834772e-06,
+      "loss": 0.0734,
+      "step": 2592
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6762482960226738,
+      "learning_rate": 1.6600815740551506e-06,
+      "loss": 0.1358,
+      "step": 2593
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.44321232980048486,
+      "learning_rate": 1.6566398563608632e-06,
+      "loss": 0.0553,
+      "step": 2594
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.33784167126787296,
+      "learning_rate": 1.653201001547719e-06,
+      "loss": 0.0742,
+      "step": 2595
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.30014213361415476,
+      "learning_rate": 1.649765012560371e-06,
+      "loss": 0.076,
+      "step": 2596
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.45292133178369026,
+      "learning_rate": 1.6463318923410183e-06,
+      "loss": 0.0869,
+      "step": 2597
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6184536306772263,
+      "learning_rate": 1.6429016438294e-06,
+      "loss": 0.0879,
+      "step": 2598
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.495024335209103,
+      "learning_rate": 1.6394742699627974e-06,
+      "loss": 0.0912,
+      "step": 2599
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.38168924285329525,
+      "learning_rate": 1.6360497736760383e-06,
+      "loss": 0.0553,
+      "step": 2600
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.4806005582767,
+      "learning_rate": 1.6326281579014735e-06,
+      "loss": 0.0978,
+      "step": 2601
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.29568765197846747,
+      "learning_rate": 1.629209425568996e-06,
+      "loss": 0.0691,
+      "step": 2602
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.37174080360595985,
+      "learning_rate": 1.625793579606027e-06,
+      "loss": 0.0904,
+      "step": 2603
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6006605808807327,
+      "learning_rate": 1.6223806229375182e-06,
+      "loss": 0.095,
+      "step": 2604
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6992857364133147,
+      "learning_rate": 1.6189705584859422e-06,
+      "loss": 0.1207,
+      "step": 2605
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2736248669303282,
+      "learning_rate": 1.6155633891712996e-06,
+      "loss": 0.0585,
+      "step": 2606
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7911493275423609,
+      "learning_rate": 1.6121591179111123e-06,
+      "loss": 0.137,
+      "step": 2607
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.625722077009125,
+      "learning_rate": 1.6087577476204157e-06,
+      "loss": 0.1127,
+      "step": 2608
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.42699049142940376,
+      "learning_rate": 1.6053592812117664e-06,
+      "loss": 0.0883,
+      "step": 2609
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.660223327398503,
+      "learning_rate": 1.6019637215952322e-06,
+      "loss": 0.1435,
+      "step": 2610
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.31709778065732325,
+      "learning_rate": 1.5985710716783936e-06,
+      "loss": 0.0784,
+      "step": 2611
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.48417144427391934,
+      "learning_rate": 1.5951813343663348e-06,
+      "loss": 0.104,
+      "step": 2612
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.25883797075508036,
+      "learning_rate": 1.5917945125616501e-06,
+      "loss": 0.0387,
+      "step": 2613
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.34870233021124686,
+      "learning_rate": 1.588410609164438e-06,
+      "loss": 0.0815,
+      "step": 2614
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5309655676696245,
+      "learning_rate": 1.5850296270722965e-06,
+      "loss": 0.0907,
+      "step": 2615
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.42249690499901665,
+      "learning_rate": 1.5816515691803191e-06,
+      "loss": 0.0891,
+      "step": 2616
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3646035291729176,
+      "learning_rate": 1.5782764383811005e-06,
+      "loss": 0.081,
+      "step": 2617
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2889192192777551,
+      "learning_rate": 1.5749042375647261e-06,
+      "loss": 0.0622,
+      "step": 2618
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4339159041669743,
+      "learning_rate": 1.5715349696187747e-06,
+      "loss": 0.1031,
+      "step": 2619
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3869929594165701,
+      "learning_rate": 1.568168637428309e-06,
+      "loss": 0.0828,
+      "step": 2620
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.47578851341027734,
+      "learning_rate": 1.5648052438758815e-06,
+      "loss": 0.0915,
+      "step": 2621
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2456645792508491,
+      "learning_rate": 1.5614447918415293e-06,
+      "loss": 0.0696,
+      "step": 2622
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4599299314364116,
+      "learning_rate": 1.5580872842027695e-06,
+      "loss": 0.0874,
+      "step": 2623
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4053157697879296,
+      "learning_rate": 1.554732723834595e-06,
+      "loss": 0.0957,
+      "step": 2624
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.20457843183326466,
+      "learning_rate": 1.5513811136094786e-06,
+      "loss": 0.0411,
+      "step": 2625
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.26154833938049865,
+      "learning_rate": 1.5480324563973682e-06,
+      "loss": 0.052,
+      "step": 2626
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3119819877266621,
+      "learning_rate": 1.544686755065677e-06,
+      "loss": 0.0575,
+      "step": 2627
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4989784511629701,
+      "learning_rate": 1.5413440124792927e-06,
+      "loss": 0.113,
+      "step": 2628
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4346739099405508,
+      "learning_rate": 1.5380042315005678e-06,
+      "loss": 0.0725,
+      "step": 2629
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6709816425775076,
+      "learning_rate": 1.5346674149893204e-06,
+      "loss": 0.1315,
+      "step": 2630
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9662191359456356,
+      "learning_rate": 1.5313335658028243e-06,
+      "loss": 0.1427,
+      "step": 2631
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5578108335434087,
+      "learning_rate": 1.5280026867958186e-06,
+      "loss": 0.1261,
+      "step": 2632
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5173757681366414,
+      "learning_rate": 1.5246747808204975e-06,
+      "loss": 0.1205,
+      "step": 2633
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.24859148570689224,
+      "learning_rate": 1.52134985072651e-06,
+      "loss": 0.0643,
+      "step": 2634
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.1771976967966618,
+      "learning_rate": 1.5180278993609527e-06,
+      "loss": 0.0391,
+      "step": 2635
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.29787364829888674,
+      "learning_rate": 1.5147089295683764e-06,
+      "loss": 0.0654,
+      "step": 2636
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.38107131734558986,
+      "learning_rate": 1.5113929441907765e-06,
+      "loss": 0.1127,
+      "step": 2637
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2732293353965668,
+      "learning_rate": 1.508079946067596e-06,
+      "loss": 0.0593,
+      "step": 2638
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.33349814069711353,
+      "learning_rate": 1.5047699380357134e-06,
+      "loss": 0.0625,
+      "step": 2639
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.38864589430654867,
+      "learning_rate": 1.5014629229294525e-06,
+      "loss": 0.0924,
+      "step": 2640
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3331088969764812,
+      "learning_rate": 1.4981589035805744e-06,
+      "loss": 0.07,
+      "step": 2641
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.41250917034250284,
+      "learning_rate": 1.4948578828182702e-06,
+      "loss": 0.1086,
+      "step": 2642
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.36533927786997383,
+      "learning_rate": 1.4915598634691676e-06,
+      "loss": 0.0921,
+      "step": 2643
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8328973222067414,
+      "learning_rate": 1.4882648483573225e-06,
+      "loss": 0.0829,
+      "step": 2644
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3114824784127253,
+      "learning_rate": 1.4849728403042213e-06,
+      "loss": 0.0738,
+      "step": 2645
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5497525059585238,
+      "learning_rate": 1.4816838421287693e-06,
+      "loss": 0.0961,
+      "step": 2646
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5519244639558591,
+      "learning_rate": 1.4783978566472996e-06,
+      "loss": 0.0901,
+      "step": 2647
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.39311287049845434,
+      "learning_rate": 1.4751148866735643e-06,
+      "loss": 0.0456,
+      "step": 2648
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.4883511060389514,
+      "learning_rate": 1.4718349350187345e-06,
+      "loss": 0.069,
+      "step": 2649
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5873910505805034,
+      "learning_rate": 1.4685580044913921e-06,
+      "loss": 0.0668,
+      "step": 2650
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.43547640692040174,
+      "learning_rate": 1.4652840978975375e-06,
+      "loss": 0.0586,
+      "step": 2651
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.4276000956755098,
+      "learning_rate": 1.4620132180405788e-06,
+      "loss": 0.0707,
+      "step": 2652
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.37816521473506937,
+      "learning_rate": 1.4587453677213348e-06,
+      "loss": 0.1042,
+      "step": 2653
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.39808430833250824,
+      "learning_rate": 1.4554805497380259e-06,
+      "loss": 0.0614,
+      "step": 2654
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7520466931538227,
+      "learning_rate": 1.4522187668862797e-06,
+      "loss": 0.1079,
+      "step": 2655
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.34863470137890984,
+      "learning_rate": 1.4489600219591248e-06,
+      "loss": 0.0877,
+      "step": 2656
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6076465010479657,
+      "learning_rate": 1.4457043177469853e-06,
+      "loss": 0.1332,
+      "step": 2657
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6521913189689925,
+      "learning_rate": 1.4424516570376839e-06,
+      "loss": 0.0818,
+      "step": 2658
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.39413640679159595,
+      "learning_rate": 1.4392020426164372e-06,
+      "loss": 0.0552,
+      "step": 2659
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.470630071495147,
+      "learning_rate": 1.4359554772658551e-06,
+      "loss": 0.0789,
+      "step": 2660
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.4960749156861106,
+      "learning_rate": 1.4327119637659303e-06,
+      "loss": 0.1069,
+      "step": 2661
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5254663328203152,
+      "learning_rate": 1.4294715048940488e-06,
+      "loss": 0.0772,
+      "step": 2662
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.31937275119922653,
+      "learning_rate": 1.4262341034249788e-06,
+      "loss": 0.0673,
+      "step": 2663
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5169102555885261,
+      "learning_rate": 1.4229997621308706e-06,
+      "loss": 0.1103,
+      "step": 2664
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.37006622430508596,
+      "learning_rate": 1.419768483781252e-06,
+      "loss": 0.0499,
+      "step": 2665
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2966919863409953,
+      "learning_rate": 1.4165402711430304e-06,
+      "loss": 0.0706,
+      "step": 2666
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5666527249676943,
+      "learning_rate": 1.4133151269804873e-06,
+      "loss": 0.1068,
+      "step": 2667
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3540776948842063,
+      "learning_rate": 1.4100930540552793e-06,
+      "loss": 0.0777,
+      "step": 2668
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.4122198000786065,
+      "learning_rate": 1.4068740551264271e-06,
+      "loss": 0.0776,
+      "step": 2669
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5090814806700498,
+      "learning_rate": 1.4036581329503245e-06,
+      "loss": 0.0764,
+      "step": 2670
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.320180169936241,
+      "learning_rate": 1.4004452902807287e-06,
+      "loss": 0.095,
+      "step": 2671
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.4390591330106556,
+      "learning_rate": 1.3972355298687618e-06,
+      "loss": 0.0891,
+      "step": 2672
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.34451407300811754,
+      "learning_rate": 1.3940288544629016e-06,
+      "loss": 0.0808,
+      "step": 2673
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.28701435077378057,
+      "learning_rate": 1.39082526680899e-06,
+      "loss": 0.0826,
+      "step": 2674
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3053081800023214,
+      "learning_rate": 1.3876247696502238e-06,
+      "loss": 0.0716,
+      "step": 2675
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5194267560372032,
+      "learning_rate": 1.3844273657271484e-06,
+      "loss": 0.081,
+      "step": 2676
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.387820089754962,
+      "learning_rate": 1.3812330577776673e-06,
+      "loss": 0.0796,
+      "step": 2677
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3732618211107229,
+      "learning_rate": 1.3780418485370284e-06,
+      "loss": 0.0913,
+      "step": 2678
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.26776979055421385,
+      "learning_rate": 1.3748537407378304e-06,
+      "loss": 0.0673,
+      "step": 2679
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5440561516501896,
+      "learning_rate": 1.3716687371100096e-06,
+      "loss": 0.0989,
+      "step": 2680
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9010406860477679,
+      "learning_rate": 1.368486840380851e-06,
+      "loss": 0.1329,
+      "step": 2681
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3394260000622673,
+      "learning_rate": 1.3653080532749752e-06,
+      "loss": 0.1006,
+      "step": 2682
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6340984530569764,
+      "learning_rate": 1.3621323785143442e-06,
+      "loss": 0.1268,
+      "step": 2683
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2631174349729944,
+      "learning_rate": 1.358959818818248e-06,
+      "loss": 0.0731,
+      "step": 2684
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.35241267836190354,
+      "learning_rate": 1.355790376903315e-06,
+      "loss": 0.0659,
+      "step": 2685
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.28811053331300185,
+      "learning_rate": 1.3526240554835017e-06,
+      "loss": 0.0658,
+      "step": 2686
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6916709139010689,
+      "learning_rate": 1.3494608572700957e-06,
+      "loss": 0.1236,
+      "step": 2687
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.32963589759427525,
+      "learning_rate": 1.3463007849717035e-06,
+      "loss": 0.0517,
+      "step": 2688
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5085408235852549,
+      "learning_rate": 1.343143841294261e-06,
+      "loss": 0.0887,
+      "step": 2689
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.4967939438389619,
+      "learning_rate": 1.3399900289410245e-06,
+      "loss": 0.0866,
+      "step": 2690
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.29057950619970113,
+      "learning_rate": 1.3368393506125644e-06,
+      "loss": 0.048,
+      "step": 2691
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3555035605248169,
+      "learning_rate": 1.3336918090067719e-06,
+      "loss": 0.0808,
+      "step": 2692
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3354657285502833,
+      "learning_rate": 1.3305474068188511e-06,
+      "loss": 0.0741,
+      "step": 2693
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.49188656269912656,
+      "learning_rate": 1.3274061467413197e-06,
+      "loss": 0.1053,
+      "step": 2694
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.544736178863369,
+      "learning_rate": 1.3242680314639995e-06,
+      "loss": 0.1306,
+      "step": 2695
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3153382932370989,
+      "learning_rate": 1.3211330636740238e-06,
+      "loss": 0.0634,
+      "step": 2696
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.709153732467823,
+      "learning_rate": 1.31800124605583e-06,
+      "loss": 0.1374,
+      "step": 2697
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.35550602803712794,
+      "learning_rate": 1.31487258129116e-06,
+      "loss": 0.0894,
+      "step": 2698
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.4916818164442743,
+      "learning_rate": 1.3117470720590502e-06,
+      "loss": 0.1085,
+      "step": 2699
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7291601707842152,
+      "learning_rate": 1.30862472103584e-06,
+      "loss": 0.168,
+      "step": 2700
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.4091074135324916,
+      "learning_rate": 1.305505530895163e-06,
+      "loss": 0.0869,
+      "step": 2701
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.589806944907326,
+      "learning_rate": 1.3023895043079476e-06,
+      "loss": 0.1101,
+      "step": 2702
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.4937937398097772,
+      "learning_rate": 1.2992766439424087e-06,
+      "loss": 0.1005,
+      "step": 2703
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.25658833971200207,
+      "learning_rate": 1.2961669524640546e-06,
+      "loss": 0.0493,
+      "step": 2704
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.628315964705664,
+      "learning_rate": 1.2930604325356793e-06,
+      "loss": 0.0886,
+      "step": 2705
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7835084995549616,
+      "learning_rate": 1.2899570868173594e-06,
+      "loss": 0.1345,
+      "step": 2706
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3607217679322699,
+      "learning_rate": 1.2868569179664553e-06,
+      "loss": 0.0909,
+      "step": 2707
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.4026200411189742,
+      "learning_rate": 1.2837599286376068e-06,
+      "loss": 0.1031,
+      "step": 2708
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.42305516993564185,
+      "learning_rate": 1.2806661214827286e-06,
+      "loss": 0.0747,
+      "step": 2709
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.710974738363163,
+      "learning_rate": 1.277575499151013e-06,
+      "loss": 0.0961,
+      "step": 2710
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.28907775814873893,
+      "learning_rate": 1.2744880642889252e-06,
+      "loss": 0.0978,
+      "step": 2711
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.38625794475406194,
+      "learning_rate": 1.2714038195402012e-06,
+      "loss": 0.0776,
+      "step": 2712
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.4519225312945511,
+      "learning_rate": 1.2683227675458425e-06,
+      "loss": 0.1095,
+      "step": 2713
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.40309345074738945,
+      "learning_rate": 1.2652449109441201e-06,
+      "loss": 0.0916,
+      "step": 2714
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3047906846411138,
+      "learning_rate": 1.2621702523705676e-06,
+      "loss": 0.072,
+      "step": 2715
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6843646586237986,
+      "learning_rate": 1.2590987944579808e-06,
+      "loss": 0.087,
+      "step": 2716
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.41241601177748055,
+      "learning_rate": 1.2560305398364114e-06,
+      "loss": 0.0708,
+      "step": 2717
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5304004162365995,
+      "learning_rate": 1.2529654911331728e-06,
+      "loss": 0.1086,
+      "step": 2718
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.44528453113911076,
+      "learning_rate": 1.2499036509728307e-06,
+      "loss": 0.0896,
+      "step": 2719
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7207219633116309,
+      "learning_rate": 1.2468450219772054e-06,
+      "loss": 0.1107,
+      "step": 2720
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3741186812084746,
+      "learning_rate": 1.2437896067653638e-06,
+      "loss": 0.1031,
+      "step": 2721
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6283162647257376,
+      "learning_rate": 1.2407374079536239e-06,
+      "loss": 0.1327,
+      "step": 2722
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6475774510135026,
+      "learning_rate": 1.2376884281555485e-06,
+      "loss": 0.1085,
+      "step": 2723
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.35222383764753795,
+      "learning_rate": 1.234642669981946e-06,
+      "loss": 0.086,
+      "step": 2724
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6922336687152492,
+      "learning_rate": 1.2316001360408614e-06,
+      "loss": 0.1373,
+      "step": 2725
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.44725196472203255,
+      "learning_rate": 1.2285608289375833e-06,
+      "loss": 0.0615,
+      "step": 2726
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3213998336436028,
+      "learning_rate": 1.2255247512746372e-06,
+      "loss": 0.0642,
+      "step": 2727
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5776731153630892,
+      "learning_rate": 1.2224919056517786e-06,
+      "loss": 0.098,
+      "step": 2728
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.40259312510282574,
+      "learning_rate": 1.2194622946660001e-06,
+      "loss": 0.0948,
+      "step": 2729
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.686430169671714,
+      "learning_rate": 1.2164359209115235e-06,
+      "loss": 0.1243,
+      "step": 2730
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7633711815377928,
+      "learning_rate": 1.2134127869797991e-06,
+      "loss": 0.1514,
+      "step": 2731
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5600011653870505,
+      "learning_rate": 1.210392895459499e-06,
+      "loss": 0.1149,
+      "step": 2732
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2967848509388518,
+      "learning_rate": 1.2073762489365242e-06,
+      "loss": 0.076,
+      "step": 2733
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.4489149559111999,
+      "learning_rate": 1.2043628499939935e-06,
+      "loss": 0.0992,
+      "step": 2734
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.46295267641964,
+      "learning_rate": 1.2013527012122477e-06,
+      "loss": 0.0889,
+      "step": 2735
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8328140275817446,
+      "learning_rate": 1.1983458051688406e-06,
+      "loss": 0.0965,
+      "step": 2736
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.41845875458614895,
+      "learning_rate": 1.1953421644385444e-06,
+      "loss": 0.0948,
+      "step": 2737
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.49620564930349254,
+      "learning_rate": 1.192341781593342e-06,
+      "loss": 0.0816,
+      "step": 2738
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.38504477754165445,
+      "learning_rate": 1.1893446592024283e-06,
+      "loss": 0.1031,
+      "step": 2739
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.39016458551382455,
+      "learning_rate": 1.1863507998322022e-06,
+      "loss": 0.069,
+      "step": 2740
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.22827157128847111,
+      "learning_rate": 1.1833602060462728e-06,
+      "loss": 0.0586,
+      "step": 2741
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8874348290726536,
+      "learning_rate": 1.1803728804054531e-06,
+      "loss": 0.1065,
+      "step": 2742
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3131888683682014,
+      "learning_rate": 1.177388825467753e-06,
+      "loss": 0.0751,
+      "step": 2743
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5194046067680366,
+      "learning_rate": 1.1744080437883859e-06,
+      "loss": 0.0593,
+      "step": 2744
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.27003515092720987,
+      "learning_rate": 1.1714305379197616e-06,
+      "loss": 0.0821,
+      "step": 2745
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.4808487609610235,
+      "learning_rate": 1.1684563104114855e-06,
+      "loss": 0.0853,
+      "step": 2746
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5099143639466569,
+      "learning_rate": 1.165485363810352e-06,
+      "loss": 0.1072,
+      "step": 2747
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.319803676365498,
+      "learning_rate": 1.16251770066035e-06,
+      "loss": 0.0704,
+      "step": 2748
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.365630626273584,
+      "learning_rate": 1.1595533235026553e-06,
+      "loss": 0.0597,
+      "step": 2749
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3697889630605743,
+      "learning_rate": 1.1565922348756324e-06,
+      "loss": 0.1002,
+      "step": 2750
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3913128920524271,
+      "learning_rate": 1.1536344373148245e-06,
+      "loss": 0.0925,
+      "step": 2751
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5866744171195533,
+      "learning_rate": 1.1506799333529612e-06,
+      "loss": 0.1068,
+      "step": 2752
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3769226940647506,
+      "learning_rate": 1.1477287255199505e-06,
+      "loss": 0.0579,
+      "step": 2753
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2894931067593348,
+      "learning_rate": 1.1447808163428797e-06,
+      "loss": 0.0452,
+      "step": 2754
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.43340750097569425,
+      "learning_rate": 1.1418362083460067e-06,
+      "loss": 0.1164,
+      "step": 2755
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6733334804824626,
+      "learning_rate": 1.1388949040507675e-06,
+      "loss": 0.1172,
+      "step": 2756
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3527328121841945,
+      "learning_rate": 1.1359569059757687e-06,
+      "loss": 0.0743,
+      "step": 2757
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.60009730725847,
+      "learning_rate": 1.133022216636781e-06,
+      "loss": 0.0965,
+      "step": 2758
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.620437087256445,
+      "learning_rate": 1.130090838546748e-06,
+      "loss": 0.1133,
+      "step": 2759
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3329959959511501,
+      "learning_rate": 1.1271627742157743e-06,
+      "loss": 0.0662,
+      "step": 2760
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.37892833642644586,
+      "learning_rate": 1.1242380261511304e-06,
+      "loss": 0.0842,
+      "step": 2761
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3817881696008621,
+      "learning_rate": 1.121316596857241e-06,
+      "loss": 0.0553,
+      "step": 2762
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.24201730005673097,
+      "learning_rate": 1.1183984888356941e-06,
+      "loss": 0.0628,
+      "step": 2763
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.35583340694806587,
+      "learning_rate": 1.1154837045852323e-06,
+      "loss": 0.0974,
+      "step": 2764
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.34386430017657676,
+      "learning_rate": 1.1125722466017547e-06,
+      "loss": 0.0806,
+      "step": 2765
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5910163295261405,
+      "learning_rate": 1.1096641173783051e-06,
+      "loss": 0.0937,
+      "step": 2766
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.4034154149418227,
+      "learning_rate": 1.1067593194050836e-06,
+      "loss": 0.1148,
+      "step": 2767
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5189484582631975,
+      "learning_rate": 1.1038578551694356e-06,
+      "loss": 0.0668,
+      "step": 2768
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3060186904872248,
+      "learning_rate": 1.100959727155853e-06,
+      "loss": 0.0667,
+      "step": 2769
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.30294171126259456,
+      "learning_rate": 1.0980649378459668e-06,
+      "loss": 0.0731,
+      "step": 2770
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3874341784461755,
+      "learning_rate": 1.095173489718554e-06,
+      "loss": 0.09,
+      "step": 2771
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.434440144321733,
+      "learning_rate": 1.092285385249528e-06,
+      "loss": 0.0855,
+      "step": 2772
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6359252146047165,
+      "learning_rate": 1.0894006269119418e-06,
+      "loss": 0.1389,
+      "step": 2773
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.4430630849907701,
+      "learning_rate": 1.0865192171759775e-06,
+      "loss": 0.0776,
+      "step": 2774
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.43476186080572843,
+      "learning_rate": 1.083641158508955e-06,
+      "loss": 0.0824,
+      "step": 2775
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1742104617023283,
+      "learning_rate": 1.0807664533753254e-06,
+      "loss": 0.0867,
+      "step": 2776
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.334150985656815,
+      "learning_rate": 1.077895104236662e-06,
+      "loss": 0.085,
+      "step": 2777
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.24993114212790044,
+      "learning_rate": 1.0750271135516699e-06,
+      "loss": 0.0591,
+      "step": 2778
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.36716776788229694,
+      "learning_rate": 1.0721624837761768e-06,
+      "loss": 0.0766,
+      "step": 2779
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.45376370633964364,
+      "learning_rate": 1.0693012173631346e-06,
+      "loss": 0.0637,
+      "step": 2780
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.437864184116154,
+      "learning_rate": 1.0664433167626093e-06,
+      "loss": 0.0825,
+      "step": 2781
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.370380840422749,
+      "learning_rate": 1.06358878442179e-06,
+      "loss": 0.0934,
+      "step": 2782
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2662176212691549,
+      "learning_rate": 1.0607376227849814e-06,
+      "loss": 0.0615,
+      "step": 2783
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3817764874201513,
+      "learning_rate": 1.0578898342936005e-06,
+      "loss": 0.1027,
+      "step": 2784
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2218744021163004,
+      "learning_rate": 1.055045421386175e-06,
+      "loss": 0.0579,
+      "step": 2785
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4552660320663978,
+      "learning_rate": 1.0522043864983428e-06,
+      "loss": 0.0851,
+      "step": 2786
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3729685725361397,
+      "learning_rate": 1.049366732062851e-06,
+      "loss": 0.0912,
+      "step": 2787
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7396419953001301,
+      "learning_rate": 1.0465324605095518e-06,
+      "loss": 0.1302,
+      "step": 2788
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3208893902931182,
+      "learning_rate": 1.043701574265396e-06,
+      "loss": 0.0734,
+      "step": 2789
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4076777130770347,
+      "learning_rate": 1.0408740757544416e-06,
+      "loss": 0.1091,
+      "step": 2790
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.33703014908359386,
+      "learning_rate": 1.0380499673978445e-06,
+      "loss": 0.0313,
+      "step": 2791
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.20277498351388953,
+      "learning_rate": 1.0352292516138534e-06,
+      "loss": 0.0632,
+      "step": 2792
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.36471426024703923,
+      "learning_rate": 1.0324119308178166e-06,
+      "loss": 0.0855,
+      "step": 2793
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2687099659071702,
+      "learning_rate": 1.0295980074221734e-06,
+      "loss": 0.0535,
+      "step": 2794
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6404241662040591,
+      "learning_rate": 1.0267874838364561e-06,
+      "loss": 0.0761,
+      "step": 2795
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3952340607681106,
+      "learning_rate": 1.0239803624672812e-06,
+      "loss": 0.0924,
+      "step": 2796
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.34656500673155227,
+      "learning_rate": 1.0211766457183559e-06,
+      "loss": 0.0727,
+      "step": 2797
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.620946135308157,
+      "learning_rate": 1.0183763359904709e-06,
+      "loss": 0.0991,
+      "step": 2798
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3575695250399971,
+      "learning_rate": 1.0155794356815008e-06,
+      "loss": 0.049,
+      "step": 2799
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.43396464675005203,
+      "learning_rate": 1.012785947186397e-06,
+      "loss": 0.0992,
+      "step": 2800
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5674349734622195,
+      "learning_rate": 1.0099958728971927e-06,
+      "loss": 0.0978,
+      "step": 2801
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5453517603974076,
+      "learning_rate": 1.007209215202997e-06,
+      "loss": 0.0801,
+      "step": 2802
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2618303304569744,
+      "learning_rate": 1.0044259764899944e-06,
+      "loss": 0.0616,
+      "step": 2803
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.27981141303459556,
+      "learning_rate": 1.001646159141438e-06,
+      "loss": 0.0566,
+      "step": 2804
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.34849611796154445,
+      "learning_rate": 9.988697655376544e-07,
+      "loss": 0.0783,
+      "step": 2805
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.37035816726159837,
+      "learning_rate": 9.960967980560381e-07,
+      "loss": 0.0831,
+      "step": 2806
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9066553222055533,
+      "learning_rate": 9.933272590710508e-07,
+      "loss": 0.1539,
+      "step": 2807
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5795957607561886,
+      "learning_rate": 9.905611509542152e-07,
+      "loss": 0.0639,
+      "step": 2808
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3869834235549946,
+      "learning_rate": 9.877984760741178e-07,
+      "loss": 0.107,
+      "step": 2809
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4471845905449251,
+      "learning_rate": 9.850392367964085e-07,
+      "loss": 0.0869,
+      "step": 2810
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3496629403015358,
+      "learning_rate": 9.82283435483788e-07,
+      "loss": 0.07,
+      "step": 2811
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3740387045768065,
+      "learning_rate": 9.795310744960208e-07,
+      "loss": 0.0695,
+      "step": 2812
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7292364365628232,
+      "learning_rate": 9.767821561899228e-07,
+      "loss": 0.1467,
+      "step": 2813
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4994691523909614,
+      "learning_rate": 9.740366829193587e-07,
+      "loss": 0.0868,
+      "step": 2814
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.38711287893410873,
+      "learning_rate": 9.71294657035247e-07,
+      "loss": 0.0938,
+      "step": 2815
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.24490457740145347,
+      "learning_rate": 9.685560808855544e-07,
+      "loss": 0.0408,
+      "step": 2816
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5419294497471646,
+      "learning_rate": 9.658209568152916e-07,
+      "loss": 0.1329,
+      "step": 2817
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.45460720715635144,
+      "learning_rate": 9.630892871665133e-07,
+      "loss": 0.111,
+      "step": 2818
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.46255716715647505,
+      "learning_rate": 9.60361074278317e-07,
+      "loss": 0.0727,
+      "step": 2819
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.4061108346522638,
+      "learning_rate": 9.576363204868417e-07,
+      "loss": 0.0513,
+      "step": 2820
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6744081389247782,
+      "learning_rate": 9.549150281252633e-07,
+      "loss": 0.122,
+      "step": 2821
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.4360874287989033,
+      "learning_rate": 9.521971995237911e-07,
+      "loss": 0.1174,
+      "step": 2822
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2643017902959821,
+      "learning_rate": 9.494828370096715e-07,
+      "loss": 0.0643,
+      "step": 2823
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5564283230338363,
+      "learning_rate": 9.467719429071831e-07,
+      "loss": 0.099,
+      "step": 2824
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5505495465238062,
+      "learning_rate": 9.440645195376341e-07,
+      "loss": 0.0761,
+      "step": 2825
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.30881846748015823,
+      "learning_rate": 9.413605692193584e-07,
+      "loss": 0.0837,
+      "step": 2826
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.550776009963471,
+      "learning_rate": 9.386600942677182e-07,
+      "loss": 0.0695,
+      "step": 2827
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6265061663190412,
+      "learning_rate": 9.359630969951012e-07,
+      "loss": 0.1013,
+      "step": 2828
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3923154026245669,
+      "learning_rate": 9.332695797109132e-07,
+      "loss": 0.0494,
+      "step": 2829
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5266718204069095,
+      "learning_rate": 9.305795447215827e-07,
+      "loss": 0.1391,
+      "step": 2830
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.34860035589051297,
+      "learning_rate": 9.278929943305564e-07,
+      "loss": 0.0686,
+      "step": 2831
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5124762529182002,
+      "learning_rate": 9.252099308382978e-07,
+      "loss": 0.0879,
+      "step": 2832
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5698073998785758,
+      "learning_rate": 9.225303565422817e-07,
+      "loss": 0.0959,
+      "step": 2833
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.31438160295090767,
+      "learning_rate": 9.198542737369976e-07,
+      "loss": 0.0622,
+      "step": 2834
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6483003468673872,
+      "learning_rate": 9.171816847139447e-07,
+      "loss": 0.0963,
+      "step": 2835
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.41165188927663543,
+      "learning_rate": 9.145125917616327e-07,
+      "loss": 0.0749,
+      "step": 2836
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.4264223221349069,
+      "learning_rate": 9.118469971655713e-07,
+      "loss": 0.0961,
+      "step": 2837
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.4301376449654001,
+      "learning_rate": 9.091849032082812e-07,
+      "loss": 0.1109,
+      "step": 2838
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.576459655052675,
+      "learning_rate": 9.06526312169283e-07,
+      "loss": 0.1437,
+      "step": 2839
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.1932529341020626,
+      "learning_rate": 9.038712263250982e-07,
+      "loss": 0.0204,
+      "step": 2840
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3484701734942682,
+      "learning_rate": 9.012196479492452e-07,
+      "loss": 0.0963,
+      "step": 2841
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6629192358786643,
+      "learning_rate": 8.985715793122407e-07,
+      "loss": 0.1323,
+      "step": 2842
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3534822491024164,
+      "learning_rate": 8.959270226815975e-07,
+      "loss": 0.085,
+      "step": 2843
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.19751266473335766,
+      "learning_rate": 8.932859803218164e-07,
+      "loss": 0.0493,
+      "step": 2844
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3329729549344692,
+      "learning_rate": 8.906484544943933e-07,
+      "loss": 0.0901,
+      "step": 2845
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3335366646419601,
+      "learning_rate": 8.880144474578112e-07,
+      "loss": 0.0798,
+      "step": 2846
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5108606944550274,
+      "learning_rate": 8.853839614675419e-07,
+      "loss": 0.061,
+      "step": 2847
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.4623437829139172,
+      "learning_rate": 8.827569987760376e-07,
+      "loss": 0.0543,
+      "step": 2848
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6842205149737411,
+      "learning_rate": 8.801335616327378e-07,
+      "loss": 0.0805,
+      "step": 2849
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3178162051179874,
+      "learning_rate": 8.775136522840622e-07,
+      "loss": 0.0655,
+      "step": 2850
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5848830188373092,
+      "learning_rate": 8.748972729734106e-07,
+      "loss": 0.1373,
+      "step": 2851
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.4276547095637747,
+      "learning_rate": 8.722844259411556e-07,
+      "loss": 0.0956,
+      "step": 2852
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3585215992558086,
+      "learning_rate": 8.696751134246495e-07,
+      "loss": 0.0691,
+      "step": 2853
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.40274959039113967,
+      "learning_rate": 8.670693376582178e-07,
+      "loss": 0.0698,
+      "step": 2854
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.4248036823633964,
+      "learning_rate": 8.64467100873157e-07,
+      "loss": 0.0524,
+      "step": 2855
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3126984161887551,
+      "learning_rate": 8.618684052977305e-07,
+      "loss": 0.06,
+      "step": 2856
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.39044460133519976,
+      "learning_rate": 8.592732531571729e-07,
+      "loss": 0.0793,
+      "step": 2857
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.25780148761497845,
+      "learning_rate": 8.566816466736839e-07,
+      "loss": 0.0702,
+      "step": 2858
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.4383894880905106,
+      "learning_rate": 8.540935880664264e-07,
+      "loss": 0.0866,
+      "step": 2859
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7016608638537952,
+      "learning_rate": 8.515090795515247e-07,
+      "loss": 0.1351,
+      "step": 2860
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5702103761570979,
+      "learning_rate": 8.489281233420638e-07,
+      "loss": 0.0721,
+      "step": 2861
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5198763043413318,
+      "learning_rate": 8.463507216480888e-07,
+      "loss": 0.1277,
+      "step": 2862
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.30581526914788437,
+      "learning_rate": 8.437768766765975e-07,
+      "loss": 0.0745,
+      "step": 2863
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.701597939475846,
+      "learning_rate": 8.412065906315442e-07,
+      "loss": 0.1264,
+      "step": 2864
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.33795592011986664,
+      "learning_rate": 8.386398657138356e-07,
+      "loss": 0.1102,
+      "step": 2865
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.32917807335753546,
+      "learning_rate": 8.360767041213313e-07,
+      "loss": 0.0669,
+      "step": 2866
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.502718544738482,
+      "learning_rate": 8.335171080488335e-07,
+      "loss": 0.1058,
+      "step": 2867
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3955852977521546,
+      "learning_rate": 8.309610796880974e-07,
+      "loss": 0.0872,
+      "step": 2868
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2774387218426999,
+      "learning_rate": 8.284086212278203e-07,
+      "loss": 0.0546,
+      "step": 2869
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5783304515013379,
+      "learning_rate": 8.258597348536452e-07,
+      "loss": 0.1034,
+      "step": 2870
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.4458221235508557,
+      "learning_rate": 8.23314422748151e-07,
+      "loss": 0.1201,
+      "step": 2871
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.49341453415108794,
+      "learning_rate": 8.207726870908606e-07,
+      "loss": 0.1385,
+      "step": 2872
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3487671009900291,
+      "learning_rate": 8.182345300582345e-07,
+      "loss": 0.0707,
+      "step": 2873
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5522184102356418,
+      "learning_rate": 8.156999538236677e-07,
+      "loss": 0.097,
+      "step": 2874
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6330713775200014,
+      "learning_rate": 8.131689605574867e-07,
+      "loss": 0.1066,
+      "step": 2875
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3924522621969745,
+      "learning_rate": 8.106415524269534e-07,
+      "loss": 0.0787,
+      "step": 2876
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3224278855296004,
+      "learning_rate": 8.081177315962601e-07,
+      "loss": 0.0524,
+      "step": 2877
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.47868413859796966,
+      "learning_rate": 8.055975002265232e-07,
+      "loss": 0.1041,
+      "step": 2878
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.45484670557724943,
+      "learning_rate": 8.030808604757894e-07,
+      "loss": 0.0981,
+      "step": 2879
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.32337393598167563,
+      "learning_rate": 8.005678144990281e-07,
+      "loss": 0.0843,
+      "step": 2880
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.283998253163553,
+      "learning_rate": 7.980583644481332e-07,
+      "loss": 0.0691,
+      "step": 2881
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.43132136388324704,
+      "learning_rate": 7.95552512471916e-07,
+      "loss": 0.0579,
+      "step": 2882
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.48200739667118,
+      "learning_rate": 7.9305026071611e-07,
+      "loss": 0.078,
+      "step": 2883
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.43047973112848487,
+      "learning_rate": 7.905516113233652e-07,
+      "loss": 0.0943,
+      "step": 2884
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5279160649395449,
+      "learning_rate": 7.880565664332473e-07,
+      "loss": 0.097,
+      "step": 2885
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3443432371418148,
+      "learning_rate": 7.85565128182233e-07,
+      "loss": 0.0776,
+      "step": 2886
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.56806116067874,
+      "learning_rate": 7.830772987037127e-07,
+      "loss": 0.1325,
+      "step": 2887
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3497432221638485,
+      "learning_rate": 7.805930801279871e-07,
+      "loss": 0.0982,
+      "step": 2888
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6869156690954531,
+      "learning_rate": 7.781124745822649e-07,
+      "loss": 0.1377,
+      "step": 2889
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3190067588408853,
+      "learning_rate": 7.756354841906582e-07,
+      "loss": 0.0571,
+      "step": 2890
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.2603420227812946,
+      "learning_rate": 7.731621110741871e-07,
+      "loss": 0.0723,
+      "step": 2891
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.39173247556403507,
+      "learning_rate": 7.70692357350774e-07,
+      "loss": 0.101,
+      "step": 2892
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.42356015496705873,
+      "learning_rate": 7.682262251352379e-07,
+      "loss": 0.0798,
+      "step": 2893
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3480654007170492,
+      "learning_rate": 7.657637165393011e-07,
+      "loss": 0.0873,
+      "step": 2894
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5106815053077439,
+      "learning_rate": 7.633048336715815e-07,
+      "loss": 0.097,
+      "step": 2895
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.4017028545099516,
+      "learning_rate": 7.608495786375941e-07,
+      "loss": 0.0739,
+      "step": 2896
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3780038758141077,
+      "learning_rate": 7.583979535397428e-07,
+      "loss": 0.0761,
+      "step": 2897
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.29093877108629496,
+      "learning_rate": 7.55949960477328e-07,
+      "loss": 0.0532,
+      "step": 2898
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3795134425124431,
+      "learning_rate": 7.535056015465381e-07,
+      "loss": 0.0698,
+      "step": 2899
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3404338539562814,
+      "learning_rate": 7.510648788404501e-07,
+      "loss": 0.0645,
+      "step": 2900
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.275210598494758,
+      "learning_rate": 7.486277944490256e-07,
+      "loss": 0.0678,
+      "step": 2901
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3411486611699063,
+      "learning_rate": 7.461943504591129e-07,
+      "loss": 0.0709,
+      "step": 2902
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.4445929264562924,
+      "learning_rate": 7.437645489544426e-07,
+      "loss": 0.0811,
+      "step": 2903
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.4442615605419769,
+      "learning_rate": 7.41338392015627e-07,
+      "loss": 0.0826,
+      "step": 2904
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.26484249727441866,
+      "learning_rate": 7.389158817201541e-07,
+      "loss": 0.0425,
+      "step": 2905
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6063348479835404,
+      "learning_rate": 7.364970201423938e-07,
+      "loss": 0.0882,
+      "step": 2906
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7413241247014105,
+      "learning_rate": 7.340818093535884e-07,
+      "loss": 0.1124,
+      "step": 2907
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.4527187327280281,
+      "learning_rate": 7.316702514218577e-07,
+      "loss": 0.0566,
+      "step": 2908
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.34783032196224395,
+      "learning_rate": 7.292623484121875e-07,
+      "loss": 0.0829,
+      "step": 2909
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.426314062370534,
+      "learning_rate": 7.268581023864402e-07,
+      "loss": 0.0817,
+      "step": 2910
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5627538496006483,
+      "learning_rate": 7.244575154033451e-07,
+      "loss": 0.0899,
+      "step": 2911
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.39816441651303663,
+      "learning_rate": 7.220605895184946e-07,
+      "loss": 0.0862,
+      "step": 2912
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.47072614754320086,
+      "learning_rate": 7.196673267843507e-07,
+      "loss": 0.09,
+      "step": 2913
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.322931872595961,
+      "learning_rate": 7.172777292502369e-07,
+      "loss": 0.0585,
+      "step": 2914
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.78148842567208,
+      "learning_rate": 7.148917989623388e-07,
+      "loss": 0.0871,
+      "step": 2915
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3639198711606861,
+      "learning_rate": 7.125095379636998e-07,
+      "loss": 0.111,
+      "step": 2916
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5604431493609869,
+      "learning_rate": 7.101309482942226e-07,
+      "loss": 0.1237,
+      "step": 2917
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3332096412037399,
+      "learning_rate": 7.077560319906696e-07,
+      "loss": 0.063,
+      "step": 2918
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.46800171988972583,
+      "learning_rate": 7.053847910866513e-07,
+      "loss": 0.1035,
+      "step": 2919
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.97911375149235,
+      "learning_rate": 7.030172276126351e-07,
+      "loss": 0.1079,
+      "step": 2920
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3779810733674687,
+      "learning_rate": 7.00653343595939e-07,
+      "loss": 0.0664,
+      "step": 2921
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.37881221141973503,
+      "learning_rate": 6.982931410607307e-07,
+      "loss": 0.046,
+      "step": 2922
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4587799402449723,
+      "learning_rate": 6.959366220280228e-07,
+      "loss": 0.0659,
+      "step": 2923
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3527440578984237,
+      "learning_rate": 6.935837885156765e-07,
+      "loss": 0.0605,
+      "step": 2924
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0532685327962303,
+      "learning_rate": 6.912346425383964e-07,
+      "loss": 0.0526,
+      "step": 2925
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.36622834853909303,
+      "learning_rate": 6.888891861077301e-07,
+      "loss": 0.0827,
+      "step": 2926
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.43796984358312346,
+      "learning_rate": 6.865474212320638e-07,
+      "loss": 0.1017,
+      "step": 2927
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.38462566932411085,
+      "learning_rate": 6.842093499166241e-07,
+      "loss": 0.0894,
+      "step": 2928
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.381073163413314,
+      "learning_rate": 6.818749741634773e-07,
+      "loss": 0.0975,
+      "step": 2929
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4702508945259308,
+      "learning_rate": 6.795442959715192e-07,
+      "loss": 0.0499,
+      "step": 2930
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4515601651038705,
+      "learning_rate": 6.772173173364843e-07,
+      "loss": 0.069,
+      "step": 2931
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.27852764893931564,
+      "learning_rate": 6.748940402509386e-07,
+      "loss": 0.0861,
+      "step": 2932
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.25234917442123944,
+      "learning_rate": 6.725744667042778e-07,
+      "loss": 0.0495,
+      "step": 2933
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4160714519667449,
+      "learning_rate": 6.702585986827248e-07,
+      "loss": 0.0546,
+      "step": 2934
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3131089315605834,
+      "learning_rate": 6.679464381693324e-07,
+      "loss": 0.0504,
+      "step": 2935
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3734963943172259,
+      "learning_rate": 6.656379871439761e-07,
+      "loss": 0.1037,
+      "step": 2936
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3673991544751655,
+      "learning_rate": 6.633332475833581e-07,
+      "loss": 0.0817,
+      "step": 2937
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.25188060698912723,
+      "learning_rate": 6.610322214609982e-07,
+      "loss": 0.0328,
+      "step": 2938
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7295652337957627,
+      "learning_rate": 6.587349107472401e-07,
+      "loss": 0.1042,
+      "step": 2939
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3767937513035894,
+      "learning_rate": 6.564413174092443e-07,
+      "loss": 0.0985,
+      "step": 2940
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2979979711754606,
+      "learning_rate": 6.541514434109902e-07,
+      "loss": 0.0681,
+      "step": 2941
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.37209889854597344,
+      "learning_rate": 6.518652907132683e-07,
+      "loss": 0.0644,
+      "step": 2942
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6889404031047961,
+      "learning_rate": 6.49582861273686e-07,
+      "loss": 0.0871,
+      "step": 2943
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.38580637690100655,
+      "learning_rate": 6.473041570466631e-07,
+      "loss": 0.0988,
+      "step": 2944
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.421392756298022,
+      "learning_rate": 6.450291799834257e-07,
+      "loss": 0.065,
+      "step": 2945
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.37457346311721995,
+      "learning_rate": 6.427579320320116e-07,
+      "loss": 0.0814,
+      "step": 2946
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4362013612573185,
+      "learning_rate": 6.404904151372649e-07,
+      "loss": 0.0775,
+      "step": 2947
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.31466494537643575,
+      "learning_rate": 6.382266312408347e-07,
+      "loss": 0.0603,
+      "step": 2948
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.34666952369419024,
+      "learning_rate": 6.359665822811722e-07,
+      "loss": 0.055,
+      "step": 2949
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5589321142596126,
+      "learning_rate": 6.337102701935322e-07,
+      "loss": 0.1242,
+      "step": 2950
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.35686956590657676,
+      "learning_rate": 6.31457696909969e-07,
+      "loss": 0.0774,
+      "step": 2951
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6666184390323098,
+      "learning_rate": 6.292088643593359e-07,
+      "loss": 0.08,
+      "step": 2952
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.35116524371456787,
+      "learning_rate": 6.269637744672813e-07,
+      "loss": 0.061,
+      "step": 2953
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3684925250204042,
+      "learning_rate": 6.24722429156251e-07,
+      "loss": 0.0846,
+      "step": 2954
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.49795523364453675,
+      "learning_rate": 6.224848303454828e-07,
+      "loss": 0.1028,
+      "step": 2955
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.46939181035728417,
+      "learning_rate": 6.202509799510086e-07,
+      "loss": 0.0856,
+      "step": 2956
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4001252277978988,
+      "learning_rate": 6.180208798856463e-07,
+      "loss": 0.0928,
+      "step": 2957
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.1974884057510179,
+      "learning_rate": 6.157945320590076e-07,
+      "loss": 0.036,
+      "step": 2958
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.311837737945852,
+      "learning_rate": 6.135719383774869e-07,
+      "loss": 0.0563,
+      "step": 2959
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5622996076001634,
+      "learning_rate": 6.11353100744268e-07,
+      "loss": 0.1094,
+      "step": 2960
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.40436782668162075,
+      "learning_rate": 6.091380210593145e-07,
+      "loss": 0.0886,
+      "step": 2961
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.48470414454889127,
+      "learning_rate": 6.069267012193735e-07,
+      "loss": 0.1034,
+      "step": 2962
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.559384360048733,
+      "learning_rate": 6.047191431179755e-07,
+      "loss": 0.1052,
+      "step": 2963
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2985281023166187,
+      "learning_rate": 6.025153486454238e-07,
+      "loss": 0.0798,
+      "step": 2964
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.39376381859130344,
+      "learning_rate": 6.003153196888045e-07,
+      "loss": 0.0917,
+      "step": 2965
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.41963611920222765,
+      "learning_rate": 5.981190581319773e-07,
+      "loss": 0.0977,
+      "step": 2966
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.4950705151532809,
+      "learning_rate": 5.959265658555769e-07,
+      "loss": 0.099,
+      "step": 2967
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8195548106339083,
+      "learning_rate": 5.937378447370068e-07,
+      "loss": 0.1482,
+      "step": 2968
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6608634549739837,
+      "learning_rate": 5.915528966504453e-07,
+      "loss": 0.0977,
+      "step": 2969
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.42591936441814277,
+      "learning_rate": 5.893717234668383e-07,
+      "loss": 0.1271,
+      "step": 2970
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3180127669552666,
+      "learning_rate": 5.87194327053901e-07,
+      "loss": 0.0767,
+      "step": 2971
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.26117207730592873,
+      "learning_rate": 5.850207092761107e-07,
+      "loss": 0.0673,
+      "step": 2972
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.24778096320432705,
+      "learning_rate": 5.828508719947123e-07,
+      "loss": 0.0669,
+      "step": 2973
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5001788346295303,
+      "learning_rate": 5.806848170677132e-07,
+      "loss": 0.0862,
+      "step": 2974
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2825532774947254,
+      "learning_rate": 5.785225463498828e-07,
+      "loss": 0.0458,
+      "step": 2975
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.46497002069784876,
+      "learning_rate": 5.763640616927458e-07,
+      "loss": 0.0806,
+      "step": 2976
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3106646283993831,
+      "learning_rate": 5.742093649445901e-07,
+      "loss": 0.0609,
+      "step": 2977
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3883572399341558,
+      "learning_rate": 5.720584579504579e-07,
+      "loss": 0.0712,
+      "step": 2978
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.4657594143842723,
+      "learning_rate": 5.699113425521452e-07,
+      "loss": 0.0978,
+      "step": 2979
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.40682139926705757,
+      "learning_rate": 5.67768020588203e-07,
+      "loss": 0.0858,
+      "step": 2980
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6267414290126576,
+      "learning_rate": 5.656284938939332e-07,
+      "loss": 0.1127,
+      "step": 2981
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3272751913041415,
+      "learning_rate": 5.634927643013899e-07,
+      "loss": 0.0651,
+      "step": 2982
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5510241979365119,
+      "learning_rate": 5.613608336393711e-07,
+      "loss": 0.0865,
+      "step": 2983
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.386240817194325,
+      "learning_rate": 5.592327037334261e-07,
+      "loss": 0.0884,
+      "step": 2984
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.4599662420251313,
+      "learning_rate": 5.571083764058482e-07,
+      "loss": 0.079,
+      "step": 2985
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.4086491572266885,
+      "learning_rate": 5.549878534756759e-07,
+      "loss": 0.0952,
+      "step": 2986
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3835854137487857,
+      "learning_rate": 5.528711367586864e-07,
+      "loss": 0.1178,
+      "step": 2987
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5224166158794535,
+      "learning_rate": 5.507582280674012e-07,
+      "loss": 0.0866,
+      "step": 2988
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.626662330089551,
+      "learning_rate": 5.486491292110796e-07,
+      "loss": 0.1309,
+      "step": 2989
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.49030610702431343,
+      "learning_rate": 5.465438419957209e-07,
+      "loss": 0.1277,
+      "step": 2990
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.28726618585301955,
+      "learning_rate": 5.444423682240558e-07,
+      "loss": 0.0756,
+      "step": 2991
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.31634354240371376,
+      "learning_rate": 5.423447096955531e-07,
+      "loss": 0.0793,
+      "step": 2992
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5936135712888403,
+      "learning_rate": 5.402508682064151e-07,
+      "loss": 0.06,
+      "step": 2993
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.458107465639092,
+      "learning_rate": 5.381608455495724e-07,
+      "loss": 0.0792,
+      "step": 2994
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.4006113400235521,
+      "learning_rate": 5.360746435146885e-07,
+      "loss": 0.0728,
+      "step": 2995
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.33429001845102757,
+      "learning_rate": 5.339922638881545e-07,
+      "loss": 0.0788,
+      "step": 2996
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.4748180018707747,
+      "learning_rate": 5.319137084530896e-07,
+      "loss": 0.0764,
+      "step": 2997
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6211664302256894,
+      "learning_rate": 5.298389789893343e-07,
+      "loss": 0.1283,
+      "step": 2998
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.34368775168904847,
+      "learning_rate": 5.277680772734578e-07,
+      "loss": 0.0571,
+      "step": 2999
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3828658272501252,
+      "learning_rate": 5.257010050787487e-07,
+      "loss": 0.1067,
+      "step": 3000
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3007889578238413,
+      "learning_rate": 5.236377641752189e-07,
+      "loss": 0.1028,
+      "step": 3001
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.4207146551719682,
+      "learning_rate": 5.215783563295956e-07,
+      "loss": 0.1075,
+      "step": 3002
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3473750429740535,
+      "learning_rate": 5.195227833053273e-07,
+      "loss": 0.0845,
+      "step": 3003
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.43525881622309176,
+      "learning_rate": 5.174710468625782e-07,
+      "loss": 0.0893,
+      "step": 3004
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.29119338252675736,
+      "learning_rate": 5.154231487582273e-07,
+      "loss": 0.073,
+      "step": 3005
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.379529649945552,
+      "learning_rate": 5.133790907458652e-07,
+      "loss": 0.0527,
+      "step": 3006
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.35829786161691707,
+      "learning_rate": 5.113388745757953e-07,
+      "loss": 0.0494,
+      "step": 3007
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3496510390374235,
+      "learning_rate": 5.093025019950321e-07,
+      "loss": 0.0776,
+      "step": 3008
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.45181850433930404,
+      "learning_rate": 5.072699747472987e-07,
+      "loss": 0.0769,
+      "step": 3009
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.4085586014306148,
+      "learning_rate": 5.05241294573024e-07,
+      "loss": 0.1025,
+      "step": 3010
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.562440220038349,
+      "learning_rate": 5.032164632093439e-07,
+      "loss": 0.0699,
+      "step": 3011
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5193491169755903,
+      "learning_rate": 5.011954823900994e-07,
+      "loss": 0.1309,
+      "step": 3012
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.31955848751933863,
+      "learning_rate": 4.991783538458312e-07,
+      "loss": 0.0653,
+      "step": 3013
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3471447646597856,
+      "learning_rate": 4.971650793037852e-07,
+      "loss": 0.0895,
+      "step": 3014
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.49495044323085113,
+      "learning_rate": 4.951556604879049e-07,
+      "loss": 0.1276,
+      "step": 3015
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5339757502143804,
+      "learning_rate": 4.931500991188331e-07,
+      "loss": 0.0685,
+      "step": 3016
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.400482854213073,
+      "learning_rate": 4.911483969139086e-07,
+      "loss": 0.0941,
+      "step": 3017
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.32685845079501163,
+      "learning_rate": 4.891505555871656e-07,
+      "loss": 0.0708,
+      "step": 3018
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.4032628689364356,
+      "learning_rate": 4.871565768493341e-07,
+      "loss": 0.044,
+      "step": 3019
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.23237142422515322,
+      "learning_rate": 4.851664624078356e-07,
+      "loss": 0.0498,
+      "step": 3020
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.443930951871871,
+      "learning_rate": 4.831802139667807e-07,
+      "loss": 0.1177,
+      "step": 3021
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5498928475161263,
+      "learning_rate": 4.811978332269718e-07,
+      "loss": 0.1225,
+      "step": 3022
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5755441923444777,
+      "learning_rate": 4.792193218858998e-07,
+      "loss": 0.0892,
+      "step": 3023
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.4011200933670287,
+      "learning_rate": 4.772446816377408e-07,
+      "loss": 0.0976,
+      "step": 3024
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.4912835564287914,
+      "learning_rate": 4.7527391417335647e-07,
+      "loss": 0.106,
+      "step": 3025
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.48304625141090346,
+      "learning_rate": 4.733070211802926e-07,
+      "loss": 0.0892,
+      "step": 3026
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.4162973613416196,
+      "learning_rate": 4.7134400434277915e-07,
+      "loss": 0.0785,
+      "step": 3027
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6162478027910685,
+      "learning_rate": 4.693848653417216e-07,
+      "loss": 0.081,
+      "step": 3028
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4033790143022547,
+      "learning_rate": 4.6742960585471066e-07,
+      "loss": 0.0828,
+      "step": 3029
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4833368881485651,
+      "learning_rate": 4.654782275560127e-07,
+      "loss": 0.0988,
+      "step": 3030
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5348699015250554,
+      "learning_rate": 4.6353073211656886e-07,
+      "loss": 0.1089,
+      "step": 3031
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.41834073298044805,
+      "learning_rate": 4.615871212039991e-07,
+      "loss": 0.0831,
+      "step": 3032
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.306588739500313,
+      "learning_rate": 4.5964739648259426e-07,
+      "loss": 0.0717,
+      "step": 3033
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2856522381562219,
+      "learning_rate": 4.577115596133197e-07,
+      "loss": 0.0585,
+      "step": 3034
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5410819550718816,
+      "learning_rate": 4.5577961225380886e-07,
+      "loss": 0.1096,
+      "step": 3035
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.427611744954672,
+      "learning_rate": 4.538515560583673e-07,
+      "loss": 0.0558,
+      "step": 3036
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.970277134252806,
+      "learning_rate": 4.519273926779666e-07,
+      "loss": 0.1109,
+      "step": 3037
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5654875809557892,
+      "learning_rate": 4.5000712376024826e-07,
+      "loss": 0.1136,
+      "step": 3038
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.47511571753867105,
+      "learning_rate": 4.480907509495136e-07,
+      "loss": 0.1209,
+      "step": 3039
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.35770035937260286,
+      "learning_rate": 4.4617827588673167e-07,
+      "loss": 0.0822,
+      "step": 3040
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.37123942578895763,
+      "learning_rate": 4.442697002095342e-07,
+      "loss": 0.0961,
+      "step": 3041
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3367041494830472,
+      "learning_rate": 4.423650255522127e-07,
+      "loss": 0.0938,
+      "step": 3042
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.48564194331727517,
+      "learning_rate": 4.404642535457165e-07,
+      "loss": 0.0626,
+      "step": 3043
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6338948134739405,
+      "learning_rate": 4.3856738581765703e-07,
+      "loss": 0.1139,
+      "step": 3044
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8487780302819357,
+      "learning_rate": 4.3667442399229985e-07,
+      "loss": 0.1229,
+      "step": 3045
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4002482373146379,
+      "learning_rate": 4.347853696905657e-07,
+      "loss": 0.0483,
+      "step": 3046
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5192388658579471,
+      "learning_rate": 4.329002245300307e-07,
+      "loss": 0.1042,
+      "step": 3047
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.378020294946494,
+      "learning_rate": 4.3101899012492343e-07,
+      "loss": 0.1018,
+      "step": 3048
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7343787532260014,
+      "learning_rate": 4.2914166808612436e-07,
+      "loss": 0.1131,
+      "step": 3049
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8289430771710176,
+      "learning_rate": 4.272682600211608e-07,
+      "loss": 0.1076,
+      "step": 3050
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.35776484260402475,
+      "learning_rate": 4.2539876753421194e-07,
+      "loss": 0.099,
+      "step": 3051
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2510200087052072,
+      "learning_rate": 4.2353319222610265e-07,
+      "loss": 0.0545,
+      "step": 3052
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6291154821345246,
+      "learning_rate": 4.2167153569430465e-07,
+      "loss": 0.1009,
+      "step": 3053
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9616173056350652,
+      "learning_rate": 4.1981379953293155e-07,
+      "loss": 0.0995,
+      "step": 3054
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4222400099548638,
+      "learning_rate": 4.179599853327426e-07,
+      "loss": 0.0789,
+      "step": 3055
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.24391964187242662,
+      "learning_rate": 4.1611009468113806e-07,
+      "loss": 0.0304,
+      "step": 3056
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4679791473568192,
+      "learning_rate": 4.1426412916215795e-07,
+      "loss": 0.062,
+      "step": 3057
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5744433048375208,
+      "learning_rate": 4.1242209035648075e-07,
+      "loss": 0.0991,
+      "step": 3058
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3636600923831621,
+      "learning_rate": 4.1058397984142405e-07,
+      "loss": 0.0799,
+      "step": 3059
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5308972871030511,
+      "learning_rate": 4.0874979919094004e-07,
+      "loss": 0.1245,
+      "step": 3060
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5242302791407181,
+      "learning_rate": 4.06919549975619e-07,
+      "loss": 0.114,
+      "step": 3061
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.1964078610675712,
+      "learning_rate": 4.050932337626795e-07,
+      "loss": 0.0487,
+      "step": 3062
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5361090093876446,
+      "learning_rate": 4.032708521159762e-07,
+      "loss": 0.0785,
+      "step": 3063
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5596155053367649,
+      "learning_rate": 4.014524065959952e-07,
+      "loss": 0.1127,
+      "step": 3064
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2744212241668214,
+      "learning_rate": 3.996378987598487e-07,
+      "loss": 0.0507,
+      "step": 3065
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6448222861745297,
+      "learning_rate": 3.9782733016128006e-07,
+      "loss": 0.1036,
+      "step": 3066
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.27275233150122863,
+      "learning_rate": 3.960207023506579e-07,
+      "loss": 0.0345,
+      "step": 3067
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.35955531617117187,
+      "learning_rate": 3.9421801687497873e-07,
+      "loss": 0.0943,
+      "step": 3068
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.47438097218592085,
+      "learning_rate": 3.924192752778588e-07,
+      "loss": 0.0928,
+      "step": 3069
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5002486621511697,
+      "learning_rate": 3.906244790995423e-07,
+      "loss": 0.0986,
+      "step": 3070
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.30826085522781366,
+      "learning_rate": 3.8883362987689134e-07,
+      "loss": 0.0383,
+      "step": 3071
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.34721910690891533,
+      "learning_rate": 3.8704672914339104e-07,
+      "loss": 0.0656,
+      "step": 3072
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2324282059512588,
+      "learning_rate": 3.852637784291424e-07,
+      "loss": 0.0428,
+      "step": 3073
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5372066165704869,
+      "learning_rate": 3.83484779260867e-07,
+      "loss": 0.0941,
+      "step": 3074
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.38097292742712063,
+      "learning_rate": 3.8170973316190074e-07,
+      "loss": 0.0893,
+      "step": 3075
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.4076076612996428,
+      "learning_rate": 3.799386416521966e-07,
+      "loss": 0.1051,
+      "step": 3076
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.664951475549366,
+      "learning_rate": 3.781715062483188e-07,
+      "loss": 0.1276,
+      "step": 3077
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.4028588925299585,
+      "learning_rate": 3.7640832846344565e-07,
+      "loss": 0.1067,
+      "step": 3078
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.4714048602084793,
+      "learning_rate": 3.746491098073668e-07,
+      "loss": 0.0732,
+      "step": 3079
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5652779938474279,
+      "learning_rate": 3.728938517864794e-07,
+      "loss": 0.1264,
+      "step": 3080
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5414256824462692,
+      "learning_rate": 3.7114255590379234e-07,
+      "loss": 0.0912,
+      "step": 3081
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.48296713636325733,
+      "learning_rate": 3.693952236589199e-07,
+      "loss": 0.0659,
+      "step": 3082
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.4225849330260804,
+      "learning_rate": 3.6765185654808357e-07,
+      "loss": 0.1001,
+      "step": 3083
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3123657679015782,
+      "learning_rate": 3.659124560641064e-07,
+      "loss": 0.0785,
+      "step": 3084
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.4664992130368141,
+      "learning_rate": 3.641770236964193e-07,
+      "loss": 0.073,
+      "step": 3085
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.24769732912850836,
+      "learning_rate": 3.6244556093105285e-07,
+      "loss": 0.0468,
+      "step": 3086
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.4856025683797443,
+      "learning_rate": 3.60718069250639e-07,
+      "loss": 0.0854,
+      "step": 3087
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.389891640849095,
+      "learning_rate": 3.5899455013440777e-07,
+      "loss": 0.0805,
+      "step": 3088
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.4865077865198653,
+      "learning_rate": 3.5727500505819023e-07,
+      "loss": 0.0806,
+      "step": 3089
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3325968133787995,
+      "learning_rate": 3.555594354944125e-07,
+      "loss": 0.0802,
+      "step": 3090
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.48090649055260104,
+      "learning_rate": 3.5384784291209896e-07,
+      "loss": 0.1066,
+      "step": 3091
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2947187497781367,
+      "learning_rate": 3.521402287768644e-07,
+      "loss": 0.0674,
+      "step": 3092
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.27166754855906533,
+      "learning_rate": 3.504365945509203e-07,
+      "loss": 0.051,
+      "step": 3093
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.44337475956250444,
+      "learning_rate": 3.4873694169306915e-07,
+      "loss": 0.1323,
+      "step": 3094
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.615474051294563,
+      "learning_rate": 3.4704127165870514e-07,
+      "loss": 0.101,
+      "step": 3095
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6088624752605024,
+      "learning_rate": 3.453495858998102e-07,
+      "loss": 0.1513,
+      "step": 3096
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.327733742344023,
+      "learning_rate": 3.4366188586495543e-07,
+      "loss": 0.0626,
+      "step": 3097
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6320767703722748,
+      "learning_rate": 3.419781729993005e-07,
+      "loss": 0.0704,
+      "step": 3098
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3548027088632169,
+      "learning_rate": 3.402984487445876e-07,
+      "loss": 0.0667,
+      "step": 3099
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.717913730591346,
+      "learning_rate": 3.386227145391463e-07,
+      "loss": 0.1242,
+      "step": 3100
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5855094279203372,
+      "learning_rate": 3.369509718178887e-07,
+      "loss": 0.0948,
+      "step": 3101
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.35693098689849084,
+      "learning_rate": 3.352832220123098e-07,
+      "loss": 0.0782,
+      "step": 3102
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.27597356385516486,
+      "learning_rate": 3.336194665504833e-07,
+      "loss": 0.0567,
+      "step": 3103
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.41699629268519856,
+      "learning_rate": 3.319597068570646e-07,
+      "loss": 0.0834,
+      "step": 3104
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.4326512457963444,
+      "learning_rate": 3.303039443532874e-07,
+      "loss": 0.1048,
+      "step": 3105
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.32748288346108995,
+      "learning_rate": 3.2865218045696256e-07,
+      "loss": 0.085,
+      "step": 3106
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.43981623675700016,
+      "learning_rate": 3.2700441658247484e-07,
+      "loss": 0.0773,
+      "step": 3107
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.46051008363093954,
+      "learning_rate": 3.2536065414078724e-07,
+      "loss": 0.0918,
+      "step": 3108
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5280500620248085,
+      "learning_rate": 3.237208945394343e-07,
+      "loss": 0.0928,
+      "step": 3109
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.611026107449624,
+      "learning_rate": 3.220851391825247e-07,
+      "loss": 0.1458,
+      "step": 3110
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3527668707160877,
+      "learning_rate": 3.204533894707346e-07,
+      "loss": 0.1001,
+      "step": 3111
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3904280066329125,
+      "learning_rate": 3.18825646801314e-07,
+      "loss": 0.1256,
+      "step": 3112
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2625376396953162,
+      "learning_rate": 3.172019125680814e-07,
+      "loss": 0.0593,
+      "step": 3113
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.32005754213725546,
+      "learning_rate": 3.1558218816142015e-07,
+      "loss": 0.0783,
+      "step": 3114
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3776388385135846,
+      "learning_rate": 3.1396647496828245e-07,
+      "loss": 0.0692,
+      "step": 3115
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.34578952977950755,
+      "learning_rate": 3.12354774372185e-07,
+      "loss": 0.0886,
+      "step": 3116
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.4051418101171471,
+      "learning_rate": 3.107470877532093e-07,
+      "loss": 0.0967,
+      "step": 3117
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.40334526262593723,
+      "learning_rate": 3.0914341648799805e-07,
+      "loss": 0.0892,
+      "step": 3118
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5334893902788516,
+      "learning_rate": 3.0754376194975676e-07,
+      "loss": 0.1091,
+      "step": 3119
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5427067548895909,
+      "learning_rate": 3.0594812550825194e-07,
+      "loss": 0.0923,
+      "step": 3120
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.4861593840855422,
+      "learning_rate": 3.0435650852980947e-07,
+      "loss": 0.0918,
+      "step": 3121
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5309956283735078,
+      "learning_rate": 3.0276891237731085e-07,
+      "loss": 0.1044,
+      "step": 3122
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6687981611951446,
+      "learning_rate": 3.0118533841019814e-07,
+      "loss": 0.1141,
+      "step": 3123
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.44634404114813325,
+      "learning_rate": 2.996057879844666e-07,
+      "loss": 0.085,
+      "step": 3124
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3994425433847164,
+      "learning_rate": 2.980302624526693e-07,
+      "loss": 0.0844,
+      "step": 3125
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.37461657518182806,
+      "learning_rate": 2.964587631639082e-07,
+      "loss": 0.112,
+      "step": 3126
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.405516764115973,
+      "learning_rate": 2.948912914638413e-07,
+      "loss": 0.091,
+      "step": 3127
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.29298656641682885,
+      "learning_rate": 2.933278486946772e-07,
+      "loss": 0.059,
+      "step": 3128
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.37281228634911845,
+      "learning_rate": 2.917684361951728e-07,
+      "loss": 0.1029,
+      "step": 3129
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.4085533070703712,
+      "learning_rate": 2.902130553006366e-07,
+      "loss": 0.0585,
+      "step": 3130
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5666335396313852,
+      "learning_rate": 2.886617073429232e-07,
+      "loss": 0.1125,
+      "step": 3131
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.37081535824124257,
+      "learning_rate": 2.8711439365043394e-07,
+      "loss": 0.0937,
+      "step": 3132
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.35616624004800407,
+      "learning_rate": 2.8557111554811555e-07,
+      "loss": 0.098,
+      "step": 3133
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6015374036138982,
+      "learning_rate": 2.840318743574599e-07,
+      "loss": 0.1356,
+      "step": 3134
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3933359066546871,
+      "learning_rate": 2.8249667139650215e-07,
+      "loss": 0.0774,
+      "step": 3135
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.26557566617258205,
+      "learning_rate": 2.809655079798179e-07,
+      "loss": 0.0665,
+      "step": 3136
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3922047326006346,
+      "learning_rate": 2.7943838541852563e-07,
+      "loss": 0.0743,
+      "step": 3137
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2195651730364043,
+      "learning_rate": 2.7791530502028263e-07,
+      "loss": 0.0692,
+      "step": 3138
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.33700214535511575,
+      "learning_rate": 2.763962680892862e-07,
+      "loss": 0.0839,
+      "step": 3139
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3916410619215585,
+      "learning_rate": 2.748812759262687e-07,
+      "loss": 0.0832,
+      "step": 3140
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3581159465712483,
+      "learning_rate": 2.7337032982850177e-07,
+      "loss": 0.0552,
+      "step": 3141
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.41164379015605307,
+      "learning_rate": 2.7186343108979106e-07,
+      "loss": 0.077,
+      "step": 3142
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.38996126568477396,
+      "learning_rate": 2.7036058100047723e-07,
+      "loss": 0.0579,
+      "step": 3143
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.33228438098231394,
+      "learning_rate": 2.688617808474331e-07,
+      "loss": 0.0754,
+      "step": 3144
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4607563379599993,
+      "learning_rate": 2.6736703191406366e-07,
+      "loss": 0.0854,
+      "step": 3145
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.39145788545942767,
+      "learning_rate": 2.658763354803062e-07,
+      "loss": 0.0962,
+      "step": 3146
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.25668093475194925,
+      "learning_rate": 2.643896928226275e-07,
+      "loss": 0.0746,
+      "step": 3147
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.46525565113147777,
+      "learning_rate": 2.62907105214022e-07,
+      "loss": 0.0799,
+      "step": 3148
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4859299085580294,
+      "learning_rate": 2.614285739240119e-07,
+      "loss": 0.0721,
+      "step": 3149
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3738322315883588,
+      "learning_rate": 2.599541002186479e-07,
+      "loss": 0.0693,
+      "step": 3150
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4962416550518549,
+      "learning_rate": 2.5848368536050437e-07,
+      "loss": 0.1139,
+      "step": 3151
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4352815152144947,
+      "learning_rate": 2.570173306086804e-07,
+      "loss": 0.1183,
+      "step": 3152
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.22482784421018465,
+      "learning_rate": 2.5555503721879924e-07,
+      "loss": 0.0386,
+      "step": 3153
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.607644090721368,
+      "learning_rate": 2.5409680644300716e-07,
+      "loss": 0.1257,
+      "step": 3154
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.752535300456928,
+      "learning_rate": 2.5264263952996915e-07,
+      "loss": 0.1057,
+      "step": 3155
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4914630908602349,
+      "learning_rate": 2.5119253772487137e-07,
+      "loss": 0.0979,
+      "step": 3156
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7996483933800769,
+      "learning_rate": 2.497465022694207e-07,
+      "loss": 0.1053,
+      "step": 3157
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4469903126921449,
+      "learning_rate": 2.483045344018403e-07,
+      "loss": 0.1086,
+      "step": 3158
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.35171566577316166,
+      "learning_rate": 2.468666353568705e-07,
+      "loss": 0.0604,
+      "step": 3159
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.24449597579980728,
+      "learning_rate": 2.4543280636576795e-07,
+      "loss": 0.0561,
+      "step": 3160
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.40197464292987445,
+      "learning_rate": 2.4400304865630443e-07,
+      "loss": 0.0779,
+      "step": 3161
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4963289897989669,
+      "learning_rate": 2.425773634527656e-07,
+      "loss": 0.1029,
+      "step": 3162
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.474085929931268,
+      "learning_rate": 2.411557519759483e-07,
+      "loss": 0.0881,
+      "step": 3163
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5514450672530389,
+      "learning_rate": 2.397382154431621e-07,
+      "loss": 0.0865,
+      "step": 3164
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2864026441455067,
+      "learning_rate": 2.3832475506822937e-07,
+      "loss": 0.0848,
+      "step": 3165
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4852871054394636,
+      "learning_rate": 2.3691537206147752e-07,
+      "loss": 0.0837,
+      "step": 3166
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5361997184288021,
+      "learning_rate": 2.3551006762974615e-07,
+      "loss": 0.1035,
+      "step": 3167
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3813576245138359,
+      "learning_rate": 2.3410884297638214e-07,
+      "loss": 0.0962,
+      "step": 3168
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.42105703849591514,
+      "learning_rate": 2.3271169930123794e-07,
+      "loss": 0.0903,
+      "step": 3169
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5499075512988395,
+      "learning_rate": 2.3131863780067043e-07,
+      "loss": 0.117,
+      "step": 3170
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5624073866287035,
+      "learning_rate": 2.2992965966754378e-07,
+      "loss": 0.1106,
+      "step": 3171
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2762019032785361,
+      "learning_rate": 2.2854476609122324e-07,
+      "loss": 0.0541,
+      "step": 3172
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3713599896848042,
+      "learning_rate": 2.2716395825757853e-07,
+      "loss": 0.1105,
+      "step": 3173
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3275689030994533,
+      "learning_rate": 2.2578723734897778e-07,
+      "loss": 0.0701,
+      "step": 3174
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7161842915206217,
+      "learning_rate": 2.2441460454429298e-07,
+      "loss": 0.1375,
+      "step": 3175
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7011596949713271,
+      "learning_rate": 2.230460610188928e-07,
+      "loss": 0.0902,
+      "step": 3176
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.24589768083601782,
+      "learning_rate": 2.216816079446471e-07,
+      "loss": 0.0409,
+      "step": 3177
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.35005746098384194,
+      "learning_rate": 2.2032124648992015e-07,
+      "loss": 0.071,
+      "step": 3178
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.28128887155202625,
+      "learning_rate": 2.189649778195735e-07,
+      "loss": 0.0515,
+      "step": 3179
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6220102521075701,
+      "learning_rate": 2.1761280309496645e-07,
+      "loss": 0.1105,
+      "step": 3180
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5035768917234578,
+      "learning_rate": 2.1626472347394846e-07,
+      "loss": 0.1463,
+      "step": 3181
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3079254944344666,
+      "learning_rate": 2.149207401108666e-07,
+      "loss": 0.0946,
+      "step": 3182
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.35225887804025297,
+      "learning_rate": 2.1358085415655706e-07,
+      "loss": 0.0733,
+      "step": 3183
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.44659481711958765,
+      "learning_rate": 2.1224506675835033e-07,
+      "loss": 0.0954,
+      "step": 3184
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5715611052168519,
+      "learning_rate": 2.109133790600648e-07,
+      "loss": 0.0989,
+      "step": 3185
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.43700031982172266,
+      "learning_rate": 2.0958579220200946e-07,
+      "loss": 0.1155,
+      "step": 3186
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.4062292000931909,
+      "learning_rate": 2.0826230732098273e-07,
+      "loss": 0.0741,
+      "step": 3187
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.42563861553701804,
+      "learning_rate": 2.0694292555026918e-07,
+      "loss": 0.0764,
+      "step": 3188
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5406280370247767,
+      "learning_rate": 2.056276480196401e-07,
+      "loss": 0.0976,
+      "step": 3189
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3888386079015403,
+      "learning_rate": 2.043164758553523e-07,
+      "loss": 0.1083,
+      "step": 3190
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3939488506044353,
+      "learning_rate": 2.0300941018014775e-07,
+      "loss": 0.1011,
+      "step": 3191
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5155873758611413,
+      "learning_rate": 2.0170645211325335e-07,
+      "loss": 0.095,
+      "step": 3192
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3307401139117292,
+      "learning_rate": 2.0040760277037497e-07,
+      "loss": 0.0594,
+      "step": 3193
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5379951042311034,
+      "learning_rate": 1.9911286326370404e-07,
+      "loss": 0.1394,
+      "step": 3194
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2784503017561588,
+      "learning_rate": 1.9782223470191043e-07,
+      "loss": 0.0627,
+      "step": 3195
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7306195747673293,
+      "learning_rate": 1.9653571819014504e-07,
+      "loss": 0.0586,
+      "step": 3196
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3736511783424484,
+      "learning_rate": 1.952533148300373e-07,
+      "loss": 0.0736,
+      "step": 3197
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5612025754942825,
+      "learning_rate": 1.9397502571969372e-07,
+      "loss": 0.1022,
+      "step": 3198
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5941467998400771,
+      "learning_rate": 1.9270085195370048e-07,
+      "loss": 0.1111,
+      "step": 3199
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.30722840404377555,
+      "learning_rate": 1.9143079462311644e-07,
+      "loss": 0.0764,
+      "step": 3200
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2542069273644538,
+      "learning_rate": 1.9016485481547775e-07,
+      "loss": 0.0545,
+      "step": 3201
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2906094176902533,
+      "learning_rate": 1.88903033614794e-07,
+      "loss": 0.0542,
+      "step": 3202
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3805554005266994,
+      "learning_rate": 1.8764533210154866e-07,
+      "loss": 0.0343,
+      "step": 3203
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.47367162811423613,
+      "learning_rate": 1.8639175135269694e-07,
+      "loss": 0.0983,
+      "step": 3204
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2976198291144996,
+      "learning_rate": 1.851422924416657e-07,
+      "loss": 0.0631,
+      "step": 3205
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6087985301665866,
+      "learning_rate": 1.838969564383525e-07,
+      "loss": 0.1004,
+      "step": 3206
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2820442871041826,
+      "learning_rate": 1.8265574440912482e-07,
+      "loss": 0.0835,
+      "step": 3207
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3762518008633823,
+      "learning_rate": 1.8141865741681696e-07,
+      "loss": 0.0632,
+      "step": 3208
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4503721909351601,
+      "learning_rate": 1.801856965207338e-07,
+      "loss": 0.1024,
+      "step": 3209
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4820321119434082,
+      "learning_rate": 1.789568627766447e-07,
+      "loss": 0.083,
+      "step": 3210
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5439868678076599,
+      "learning_rate": 1.7773215723678738e-07,
+      "loss": 0.0971,
+      "step": 3211
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4855445971609712,
+      "learning_rate": 1.7651158094986132e-07,
+      "loss": 0.0758,
+      "step": 3212
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4096660548585762,
+      "learning_rate": 1.7529513496103322e-07,
+      "loss": 0.1115,
+      "step": 3213
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3252218780472281,
+      "learning_rate": 1.7408282031193213e-07,
+      "loss": 0.1095,
+      "step": 3214
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5576200799213603,
+      "learning_rate": 1.7287463804064874e-07,
+      "loss": 0.1319,
+      "step": 3215
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.30832358778532887,
+      "learning_rate": 1.7167058918173552e-07,
+      "loss": 0.0754,
+      "step": 3216
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3434327836068545,
+      "learning_rate": 1.7047067476620605e-07,
+      "loss": 0.0927,
+      "step": 3217
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3785415899616005,
+      "learning_rate": 1.69274895821534e-07,
+      "loss": 0.0739,
+      "step": 3218
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.40350550111390837,
+      "learning_rate": 1.6808325337164976e-07,
+      "loss": 0.0953,
+      "step": 3219
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3168142233658187,
+      "learning_rate": 1.6689574843694433e-07,
+      "loss": 0.0615,
+      "step": 3220
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8192219085259064,
+      "learning_rate": 1.6571238203426431e-07,
+      "loss": 0.1105,
+      "step": 3221
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2060648279584498,
+      "learning_rate": 1.6453315517691304e-07,
+      "loss": 0.1924,
+      "step": 3222
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3242110053967352,
+      "learning_rate": 1.633580688746489e-07,
+      "loss": 0.0782,
+      "step": 3223
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3474303083418852,
+      "learning_rate": 1.6218712413368421e-07,
+      "loss": 0.0964,
+      "step": 3224
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.49573276932793436,
+      "learning_rate": 1.6102032195668639e-07,
+      "loss": 0.117,
+      "step": 3225
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5592060687338686,
+      "learning_rate": 1.5985766334277565e-07,
+      "loss": 0.0706,
+      "step": 3226
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7275520899795545,
+      "learning_rate": 1.5869914928752117e-07,
+      "loss": 0.1233,
+      "step": 3227
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.36510226191341333,
+      "learning_rate": 1.5754478078294667e-07,
+      "loss": 0.1008,
+      "step": 3228
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4295287535206983,
+      "learning_rate": 1.563945588175253e-07,
+      "loss": 0.0951,
+      "step": 3229
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2997562005991779,
+      "learning_rate": 1.5524848437617757e-07,
+      "loss": 0.0754,
+      "step": 3230
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2837746237666502,
+      "learning_rate": 1.5410655844027455e-07,
+      "loss": 0.0645,
+      "step": 3231
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5188797010105655,
+      "learning_rate": 1.529687819876341e-07,
+      "loss": 0.109,
+      "step": 3232
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4537171225884437,
+      "learning_rate": 1.518351559925224e-07,
+      "loss": 0.1158,
+      "step": 3233
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4030401708955543,
+      "learning_rate": 1.5070568142564912e-07,
+      "loss": 0.0894,
+      "step": 3234
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4316463361203101,
+      "learning_rate": 1.4958035925417002e-07,
+      "loss": 0.068,
+      "step": 3235
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4367746398417832,
+      "learning_rate": 1.4845919044168765e-07,
+      "loss": 0.1019,
+      "step": 3236
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.31041811684435033,
+      "learning_rate": 1.4734217594824462e-07,
+      "loss": 0.0764,
+      "step": 3237
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.44291887730989216,
+      "learning_rate": 1.4622931673032802e-07,
+      "loss": 0.1056,
+      "step": 3238
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5296446008071295,
+      "learning_rate": 1.4512061374086673e-07,
+      "loss": 0.0884,
+      "step": 3239
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.38103576880475604,
+      "learning_rate": 1.4401606792923018e-07,
+      "loss": 0.0823,
+      "step": 3240
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.34581749166633186,
+      "learning_rate": 1.4291568024122848e-07,
+      "loss": 0.0737,
+      "step": 3241
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.4918685641044202,
+      "learning_rate": 1.418194516191107e-07,
+      "loss": 0.0802,
+      "step": 3242
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5656988173108841,
+      "learning_rate": 1.4072738300156542e-07,
+      "loss": 0.1261,
+      "step": 3243
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5050725988253879,
+      "learning_rate": 1.3963947532371847e-07,
+      "loss": 0.0749,
+      "step": 3244
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5657532368454418,
+      "learning_rate": 1.3855572951713247e-07,
+      "loss": 0.0827,
+      "step": 3245
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.42601363507525336,
+      "learning_rate": 1.374761465098068e-07,
+      "loss": 0.0928,
+      "step": 3246
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.4157097788132322,
+      "learning_rate": 1.3640072722617582e-07,
+      "loss": 0.0989,
+      "step": 3247
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3528585691193063,
+      "learning_rate": 1.3532947258710905e-07,
+      "loss": 0.0749,
+      "step": 3248
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3565371330254819,
+      "learning_rate": 1.342623835099094e-07,
+      "loss": 0.0968,
+      "step": 3249
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5161753236446334,
+      "learning_rate": 1.3319946090831372e-07,
+      "loss": 0.0847,
+      "step": 3250
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.46508133632526183,
+      "learning_rate": 1.3214070569249005e-07,
+      "loss": 0.0954,
+      "step": 3251
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.4372601058548169,
+      "learning_rate": 1.3108611876903764e-07,
+      "loss": 0.0775,
+      "step": 3252
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.4836003045139583,
+      "learning_rate": 1.3003570104098807e-07,
+      "loss": 0.1008,
+      "step": 3253
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.4218276298255375,
+      "learning_rate": 1.2898945340780234e-07,
+      "loss": 0.0726,
+      "step": 3254
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.39651672447591346,
+      "learning_rate": 1.2794737676536993e-07,
+      "loss": 0.0666,
+      "step": 3255
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.50122331583475,
+      "learning_rate": 1.2690947200600878e-07,
+      "loss": 0.0934,
+      "step": 3256
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.49216263695395385,
+      "learning_rate": 1.258757400184657e-07,
+      "loss": 0.1262,
+      "step": 3257
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.29244085203128617,
+      "learning_rate": 1.2484618168791318e-07,
+      "loss": 0.059,
+      "step": 3258
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.35483072986087194,
+      "learning_rate": 1.2382079789595046e-07,
+      "loss": 0.0608,
+      "step": 3259
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.47992746531576247,
+      "learning_rate": 1.2279958952060133e-07,
+      "loss": 0.0914,
+      "step": 3260
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5153281714533938,
+      "learning_rate": 1.2178255743631573e-07,
+      "loss": 0.0934,
+      "step": 3261
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2966278691312279,
+      "learning_rate": 1.2076970251396593e-07,
+      "loss": 0.0494,
+      "step": 3262
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.4332412938004602,
+      "learning_rate": 1.1976102562084923e-07,
+      "loss": 0.0757,
+      "step": 3263
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2894091835267774,
+      "learning_rate": 1.1875652762068257e-07,
+      "loss": 0.0809,
+      "step": 3264
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.4539431948922057,
+      "learning_rate": 1.1775620937360677e-07,
+      "loss": 0.1014,
+      "step": 3265
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.4326588252693965,
+      "learning_rate": 1.1676007173618386e-07,
+      "loss": 0.0925,
+      "step": 3266
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5796840020659046,
+      "learning_rate": 1.1576811556139379e-07,
+      "loss": 0.1243,
+      "step": 3267
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.4064585936446287,
+      "learning_rate": 1.1478034169863761e-07,
+      "loss": 0.0742,
+      "step": 3268
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7822521092715196,
+      "learning_rate": 1.1379675099373489e-07,
+      "loss": 0.108,
+      "step": 3269
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.34478155691521306,
+      "learning_rate": 1.128173442889241e-07,
+      "loss": 0.0906,
+      "step": 3270
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3409959373970809,
+      "learning_rate": 1.118421224228583e-07,
+      "loss": 0.0794,
+      "step": 3271
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7497633233841792,
+      "learning_rate": 1.1087108623061005e-07,
+      "loss": 0.1793,
+      "step": 3272
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.35876689533953393,
+      "learning_rate": 1.0990423654366589e-07,
+      "loss": 0.0905,
+      "step": 3273
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7473414230592106,
+      "learning_rate": 1.0894157418992913e-07,
+      "loss": 0.1061,
+      "step": 3274
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5565685234768728,
+      "learning_rate": 1.0798309999371537e-07,
+      "loss": 0.0935,
+      "step": 3275
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3876416799604804,
+      "learning_rate": 1.0702881477575589e-07,
+      "loss": 0.1132,
+      "step": 3276
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.43566974002535763,
+      "learning_rate": 1.0607871935319424e-07,
+      "loss": 0.0821,
+      "step": 3277
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.4047787715443791,
+      "learning_rate": 1.0513281453958634e-07,
+      "loss": 0.0936,
+      "step": 3278
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6282803767499353,
+      "learning_rate": 1.0419110114489872e-07,
+      "loss": 0.1171,
+      "step": 3279
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5866470912434046,
+      "learning_rate": 1.0325357997551133e-07,
+      "loss": 0.1036,
+      "step": 3280
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3486570549630734,
+      "learning_rate": 1.0232025183421201e-07,
+      "loss": 0.05,
+      "step": 3281
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3384301582813744,
+      "learning_rate": 1.0139111752019815e-07,
+      "loss": 0.0636,
+      "step": 3282
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3176755796300654,
+      "learning_rate": 1.004661778290783e-07,
+      "loss": 0.0698,
+      "step": 3283
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.4258889461217338,
+      "learning_rate": 9.954543355286728e-08,
+      "loss": 0.0975,
+      "step": 3284
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7024887379112578,
+      "learning_rate": 9.862888547998828e-08,
+      "loss": 0.0737,
+      "step": 3285
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.4206105340505765,
+      "learning_rate": 9.771653439527018e-08,
+      "loss": 0.102,
+      "step": 3286
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.4195246951953242,
+      "learning_rate": 9.680838107994862e-08,
+      "loss": 0.0756,
+      "step": 3287
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3917028594702229,
+      "learning_rate": 9.590442631166596e-08,
+      "loss": 0.083,
+      "step": 3288
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.29266001892902305,
+      "learning_rate": 9.500467086446808e-08,
+      "loss": 0.0437,
+      "step": 3289
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.37412673788755235,
+      "learning_rate": 9.410911550880474e-08,
+      "loss": 0.0734,
+      "step": 3290
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.46499662507825834,
+      "learning_rate": 9.321776101153036e-08,
+      "loss": 0.1064,
+      "step": 3291
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.658447741871222,
+      "learning_rate": 9.2330608135901e-08,
+      "loss": 0.1305,
+      "step": 3292
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6157612226991941,
+      "learning_rate": 9.144765764157626e-08,
+      "loss": 0.1157,
+      "step": 3293
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.36526823528116403,
+      "learning_rate": 9.056891028461634e-08,
+      "loss": 0.0761,
+      "step": 3294
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3600432633643026,
+      "learning_rate": 8.969436681748211e-08,
+      "loss": 0.0747,
+      "step": 3295
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3851054540579154,
+      "learning_rate": 8.882402798903567e-08,
+      "loss": 0.0709,
+      "step": 3296
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.36745096125280796,
+      "learning_rate": 8.795789454453862e-08,
+      "loss": 0.063,
+      "step": 3297
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5655616240414351,
+      "learning_rate": 8.709596722564995e-08,
+      "loss": 0.1035,
+      "step": 3298
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2887872178091258,
+      "learning_rate": 8.623824677042869e-08,
+      "loss": 0.0686,
+      "step": 3299
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.325500457544978,
+      "learning_rate": 8.53847339133318e-08,
+      "loss": 0.0864,
+      "step": 3300
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7261985904500424,
+      "learning_rate": 8.453542938521186e-08,
+      "loss": 0.0871,
+      "step": 3301
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.33562648861250804,
+      "learning_rate": 8.369033391331827e-08,
+      "loss": 0.0856,
+      "step": 3302
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.45394825171012915,
+      "learning_rate": 8.284944822129771e-08,
+      "loss": 0.1497,
+      "step": 3303
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.4725440529741108,
+      "learning_rate": 8.201277302919086e-08,
+      "loss": 0.0777,
+      "step": 3304
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.4171524438210051,
+      "learning_rate": 8.118030905343244e-08,
+      "loss": 0.1102,
+      "step": 3305
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.4799686894869307,
+      "learning_rate": 8.035205700685167e-08,
+      "loss": 0.0815,
+      "step": 3306
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3405687698273351,
+      "learning_rate": 7.952801759867234e-08,
+      "loss": 0.069,
+      "step": 3307
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.33108709791236074,
+      "learning_rate": 7.870819153450948e-08,
+      "loss": 0.0717,
+      "step": 3308
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.34672819397009647,
+      "learning_rate": 7.789257951637097e-08,
+      "loss": 0.065,
+      "step": 3309
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3926708691791063,
+      "learning_rate": 7.708118224265538e-08,
+      "loss": 0.0908,
+      "step": 3310
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3420887632203836,
+      "learning_rate": 7.627400040815414e-08,
+      "loss": 0.0867,
+      "step": 3311
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.4632687188261814,
+      "learning_rate": 7.547103470404715e-08,
+      "loss": 0.0789,
+      "step": 3312
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6222433898869117,
+      "learning_rate": 7.467228581790442e-08,
+      "loss": 0.1371,
+      "step": 3313
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.246291151164361,
+      "learning_rate": 7.38777544336855e-08,
+      "loss": 0.0523,
+      "step": 3314
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3241067907932272,
+      "learning_rate": 7.308744123174006e-08,
+      "loss": 0.0698,
+      "step": 3315
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.4332113502979993,
+      "learning_rate": 7.23013468888023e-08,
+      "loss": 0.059,
+      "step": 3316
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.4236873711937894,
+      "learning_rate": 7.151947207799659e-08,
+      "loss": 0.0946,
+      "step": 3317
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6952784620458924,
+      "learning_rate": 7.074181746883402e-08,
+      "loss": 0.0957,
+      "step": 3318
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.28276571411557155,
+      "learning_rate": 6.996838372721081e-08,
+      "loss": 0.0656,
+      "step": 3319
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5588409740863747,
+      "learning_rate": 6.919917151540944e-08,
+      "loss": 0.1218,
+      "step": 3320
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.4612321770491532,
+      "learning_rate": 6.843418149209746e-08,
+      "loss": 0.0901,
+      "step": 3321
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.4064045617447785,
+      "learning_rate": 6.7673414312327e-08,
+      "loss": 0.0931,
+      "step": 3322
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5689550638979906,
+      "learning_rate": 6.691687062753527e-08,
+      "loss": 0.0915,
+      "step": 3323
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5663080168611111,
+      "learning_rate": 6.616455108554077e-08,
+      "loss": 0.1018,
+      "step": 3324
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8013417354675473,
+      "learning_rate": 6.54164563305465e-08,
+      "loss": 0.1508,
+      "step": 3325
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.4814055110657226,
+      "learning_rate": 6.467258700313783e-08,
+      "loss": 0.0787,
+      "step": 3326
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3147537830167229,
+      "learning_rate": 6.393294374028191e-08,
+      "loss": 0.0732,
+      "step": 3327
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2579453432870022,
+      "learning_rate": 6.319752717532546e-08,
+      "loss": 0.0722,
+      "step": 3328
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.42759856071289837,
+      "learning_rate": 6.246633793799861e-08,
+      "loss": 0.0726,
+      "step": 3329
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.49898341338214997,
+      "learning_rate": 6.173937665440943e-08,
+      "loss": 0.1104,
+      "step": 3330
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6096860816133118,
+      "learning_rate": 6.101664394704776e-08,
+      "loss": 0.102,
+      "step": 3331
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5523672671007938,
+      "learning_rate": 6.029814043478022e-08,
+      "loss": 0.1123,
+      "step": 3332
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5052568663859954,
+      "learning_rate": 5.958386673285299e-08,
+      "loss": 0.1082,
+      "step": 3333
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.29911481746438745,
+      "learning_rate": 5.887382345289183e-08,
+      "loss": 0.0535,
+      "step": 3334
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0355043715505179,
+      "learning_rate": 5.816801120289761e-08,
+      "loss": 0.1142,
+      "step": 3335
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3985300031301009,
+      "learning_rate": 5.7466430587249654e-08,
+      "loss": 0.1013,
+      "step": 3336
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3168866278861751,
+      "learning_rate": 5.676908220670352e-08,
+      "loss": 0.0695,
+      "step": 3337
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.31826716573248615,
+      "learning_rate": 5.607596665839099e-08,
+      "loss": 0.0832,
+      "step": 3338
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.4231131202459236,
+      "learning_rate": 5.538708453581787e-08,
+      "loss": 0.1107,
+      "step": 3339
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5265555752924048,
+      "learning_rate": 5.47024364288673e-08,
+      "loss": 0.1225,
+      "step": 3340
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8986658213989515,
+      "learning_rate": 5.402202292379477e-08,
+      "loss": 0.1519,
+      "step": 3341
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9789127155337204,
+      "learning_rate": 5.334584460323089e-08,
+      "loss": 0.1313,
+      "step": 3342
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.49907376320964203,
+      "learning_rate": 5.267390204617917e-08,
+      "loss": 0.0862,
+      "step": 3343
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5862843627476348,
+      "learning_rate": 5.200619582801714e-08,
+      "loss": 0.1009,
+      "step": 3344
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3247460361294157,
+      "learning_rate": 5.13427265204941e-08,
+      "loss": 0.0713,
+      "step": 3345
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5319999254878629,
+      "learning_rate": 5.068349469173006e-08,
+      "loss": 0.0985,
+      "step": 3346
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3623403173729547,
+      "learning_rate": 5.002850090621847e-08,
+      "loss": 0.0868,
+      "step": 3347
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3339024567836248,
+      "learning_rate": 4.937774572482235e-08,
+      "loss": 0.072,
+      "step": 3348
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.44323003646036496,
+      "learning_rate": 4.8731229704777613e-08,
+      "loss": 0.0858,
+      "step": 3349
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.487973680532394,
+      "learning_rate": 4.808895339968644e-08,
+      "loss": 0.073,
+      "step": 3350
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.37731754414089147,
+      "learning_rate": 4.7450917359524454e-08,
+      "loss": 0.0582,
+      "step": 3351
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6750099556148973,
+      "learning_rate": 4.68171221306335e-08,
+      "loss": 0.091,
+      "step": 3352
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.35522013923132423,
+      "learning_rate": 4.618756825572612e-08,
+      "loss": 0.0586,
+      "step": 3353
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.31009866016671966,
+      "learning_rate": 4.556225627388111e-08,
+      "loss": 0.0742,
+      "step": 3354
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3063728360736794,
+      "learning_rate": 4.4941186720546257e-08,
+      "loss": 0.0668,
+      "step": 3355
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4919264632058933,
+      "learning_rate": 4.4324360127536715e-08,
+      "loss": 0.1048,
+      "step": 3356
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6002799573453805,
+      "learning_rate": 4.371177702303386e-08,
+      "loss": 0.0849,
+      "step": 3357
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.43782652600416505,
+      "learning_rate": 4.310343793158589e-08,
+      "loss": 0.095,
+      "step": 3358
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7871045398422243,
+      "learning_rate": 4.2499343374106106e-08,
+      "loss": 0.057,
+      "step": 3359
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7023890818196845,
+      "learning_rate": 4.189949386787462e-08,
+      "loss": 0.1334,
+      "step": 3360
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3359971779330775,
+      "learning_rate": 4.1303889926534445e-08,
+      "loss": 0.0651,
+      "step": 3361
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5400735310857169,
+      "learning_rate": 4.07125320600954e-08,
+      "loss": 0.1259,
+      "step": 3362
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5405640670288637,
+      "learning_rate": 4.01254207749302e-08,
+      "loss": 0.1642,
+      "step": 3363
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5348866870220175,
+      "learning_rate": 3.9542556573775595e-08,
+      "loss": 0.1004,
+      "step": 3364
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.24910636097863664,
+      "learning_rate": 3.896393995573178e-08,
+      "loss": 0.0616,
+      "step": 3365
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.366873144328219,
+      "learning_rate": 3.8389571416260763e-08,
+      "loss": 0.0898,
+      "step": 3366
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.39985874735107874,
+      "learning_rate": 3.781945144718912e-08,
+      "loss": 0.1133,
+      "step": 3367
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6632584084143394,
+      "learning_rate": 3.725358053670247e-08,
+      "loss": 0.1274,
+      "step": 3368
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3700240104742133,
+      "learning_rate": 3.669195916935042e-08,
+      "loss": 0.0742,
+      "step": 3369
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.37909204222745296,
+      "learning_rate": 3.613458782604329e-08,
+      "loss": 0.1113,
+      "step": 3370
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.22199649197537116,
+      "learning_rate": 3.5581466984051514e-08,
+      "loss": 0.0443,
+      "step": 3371
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.270257742809934,
+      "learning_rate": 3.5032597117005684e-08,
+      "loss": 0.0656,
+      "step": 3372
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.251242812550788,
+      "learning_rate": 3.4487978694897615e-08,
+      "loss": 0.0492,
+      "step": 3373
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3459479416212941,
+      "learning_rate": 3.394761218407705e-08,
+      "loss": 0.0942,
+      "step": 3374
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3633909239035125,
+      "learning_rate": 3.341149804725496e-08,
+      "loss": 0.071,
+      "step": 3375
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.29610327368760175,
+      "learning_rate": 3.2879636743498036e-08,
+      "loss": 0.0658,
+      "step": 3376
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5563475551282037,
+      "learning_rate": 3.235202872823362e-08,
+      "loss": 0.1119,
+      "step": 3377
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.48711414978677237,
+      "learning_rate": 3.1828674453247e-08,
+      "loss": 0.0973,
+      "step": 3378
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7148941865718158,
+      "learning_rate": 3.1309574366680805e-08,
+      "loss": 0.1406,
+      "step": 3379
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3068520706239714,
+      "learning_rate": 3.079472891303337e-08,
+      "loss": 0.104,
+      "step": 3380
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5064547131371717,
+      "learning_rate": 3.0284138533160924e-08,
+      "loss": 0.0996,
+      "step": 3381
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3391909003656588,
+      "learning_rate": 2.977780366427763e-08,
+      "loss": 0.0849,
+      "step": 3382
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4983722811439082,
+      "learning_rate": 2.9275724739951107e-08,
+      "loss": 0.1175,
+      "step": 3383
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.36572172944746717,
+      "learning_rate": 2.8777902190105788e-08,
+      "loss": 0.0485,
+      "step": 3384
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.38564736716257747,
+      "learning_rate": 2.8284336441021797e-08,
+      "loss": 0.1195,
+      "step": 3385
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7374921825896716,
+      "learning_rate": 2.779502791533384e-08,
+      "loss": 0.1746,
+      "step": 3386
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.509760011740364,
+      "learning_rate": 2.73099770320312e-08,
+      "loss": 0.0702,
+      "step": 3387
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.26007978575729523,
+      "learning_rate": 2.6829184206457194e-08,
+      "loss": 0.0854,
+      "step": 3388
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6364074706483942,
+      "learning_rate": 2.635264985030972e-08,
+      "loss": 0.0943,
+      "step": 3389
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.33450087203858087,
+      "learning_rate": 2.5880374371639594e-08,
+      "loss": 0.0835,
+      "step": 3390
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.40708895394714223,
+      "learning_rate": 2.5412358174850547e-08,
+      "loss": 0.0737,
+      "step": 3391
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.43748443437299306,
+      "learning_rate": 2.4948601660699234e-08,
+      "loss": 0.1078,
+      "step": 3392
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5448005776903556,
+      "learning_rate": 2.4489105226295773e-08,
+      "loss": 0.1102,
+      "step": 3393
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3589647034407713,
+      "learning_rate": 2.4033869265102095e-08,
+      "loss": 0.0493,
+      "step": 3394
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4862344371988249,
+      "learning_rate": 2.358289416693027e-08,
+      "loss": 0.1112,
+      "step": 3395
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3247753245697565,
+      "learning_rate": 2.3136180317945845e-08,
+      "loss": 0.0573,
+      "step": 3396
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6654975992640926,
+      "learning_rate": 2.2693728100664503e-08,
+      "loss": 0.1228,
+      "step": 3397
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.39955271232783485,
+      "learning_rate": 2.2255537893953737e-08,
+      "loss": 0.0932,
+      "step": 3398
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3349268757937497,
+      "learning_rate": 2.182161007303063e-08,
+      "loss": 0.0881,
+      "step": 3399
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4246192138527646,
+      "learning_rate": 2.1391945009461844e-08,
+      "loss": 0.1012,
+      "step": 3400
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5570969127838759,
+      "learning_rate": 2.09665430711653e-08,
+      "loss": 0.0847,
+      "step": 3401
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3703196996109987,
+      "learning_rate": 2.0545404622407396e-08,
+      "loss": 0.0733,
+      "step": 3402
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.35354514712258345,
+      "learning_rate": 2.012853002380466e-08,
+      "loss": 0.0385,
+      "step": 3403
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5707313295190611,
+      "learning_rate": 1.9715919632322112e-08,
+      "loss": 0.1097,
+      "step": 3404
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7196503236563584,
+      "learning_rate": 1.930757380127324e-08,
+      "loss": 0.1063,
+      "step": 3405
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.31695260719600904,
+      "learning_rate": 1.890349288031945e-08,
+      "loss": 0.0706,
+      "step": 3406
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5134340109642018,
+      "learning_rate": 1.8503677215470638e-08,
+      "loss": 0.0866,
+      "step": 3407
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3469534044839269,
+      "learning_rate": 1.8108127149085164e-08,
+      "loss": 0.0907,
+      "step": 3408
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.23949055173243536,
+      "learning_rate": 1.7716843019867646e-08,
+      "loss": 0.0583,
+      "step": 3409
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.47941271748203396,
+      "learning_rate": 1.7329825162870073e-08,
+      "loss": 0.0798,
+      "step": 3410
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.48442996156809487,
+      "learning_rate": 1.694707390949124e-08,
+      "loss": 0.1072,
+      "step": 3411
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.45628246575253745,
+      "learning_rate": 1.656858958747676e-08,
+      "loss": 0.1119,
+      "step": 3412
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5146618838428757,
+      "learning_rate": 1.6194372520919044e-08,
+      "loss": 0.0637,
+      "step": 3413
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.4363049530506406,
+      "learning_rate": 1.5824423030255665e-08,
+      "loss": 0.0673,
+      "step": 3414
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.4012837372101367,
+      "learning_rate": 1.545874143226933e-08,
+      "loss": 0.0857,
+      "step": 3415
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.45617617905043395,
+      "learning_rate": 1.509732804009012e-08,
+      "loss": 0.0702,
+      "step": 3416
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.38870183272881115,
+      "learning_rate": 1.474018316319159e-08,
+      "loss": 0.0587,
+      "step": 3417
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.36078498147593285,
+      "learning_rate": 1.4387307107393e-08,
+      "loss": 0.0602,
+      "step": 3418
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6315291875414932,
+      "learning_rate": 1.4038700174857645e-08,
+      "loss": 0.1023,
+      "step": 3419
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.3796270388545187,
+      "learning_rate": 1.3694362664094518e-08,
+      "loss": 0.0815,
+      "step": 3420
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.4198817889154841,
+      "learning_rate": 1.3354294869954987e-08,
+      "loss": 0.0925,
+      "step": 3421
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.34376713163103356,
+      "learning_rate": 1.3018497083636117e-08,
+      "loss": 0.0395,
+      "step": 3422
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.3443455089540214,
+      "learning_rate": 1.268696959267679e-08,
+      "loss": 0.0675,
+      "step": 3423
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5244005996225471,
+      "learning_rate": 1.2359712680961588e-08,
+      "loss": 0.0837,
+      "step": 3424
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.34067582097236015,
+      "learning_rate": 1.2036726628715245e-08,
+      "loss": 0.0595,
+      "step": 3425
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.654286644862362,
+      "learning_rate": 1.1718011712507637e-08,
+      "loss": 0.1032,
+      "step": 3426
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5809334125193262,
+      "learning_rate": 1.140356820525157e-08,
+      "loss": 0.0819,
+      "step": 3427
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.41627985970051984,
+      "learning_rate": 1.1093396376200549e-08,
+      "loss": 0.1077,
+      "step": 3428
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5133411959970431,
+      "learning_rate": 1.078749649095101e-08,
+      "loss": 0.0946,
+      "step": 3429
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.32382094063284705,
+      "learning_rate": 1.0485868811441757e-08,
+      "loss": 0.0535,
+      "step": 3430
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.575475400919534,
+      "learning_rate": 1.018851359595341e-08,
+      "loss": 0.0933,
+      "step": 3431
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.25770742163761506,
+      "learning_rate": 9.895431099107845e-09,
+      "loss": 0.0484,
+      "step": 3432
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6130156592616252,
+      "learning_rate": 9.606621571867647e-09,
+      "loss": 0.0832,
+      "step": 3433
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.40838636752123786,
+      "learning_rate": 9.322085261537772e-09,
+      "loss": 0.079,
+      "step": 3434
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.37592866415572707,
+      "learning_rate": 9.041822411763323e-09,
+      "loss": 0.0919,
+      "step": 3435
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.4348417259620382,
+      "learning_rate": 8.76583326253011e-09,
+      "loss": 0.0669,
+      "step": 3436
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.49535041246122063,
+      "learning_rate": 8.494118050164646e-09,
+      "loss": 0.0851,
+      "step": 3437
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2600484154682323,
+      "learning_rate": 8.22667700733304e-09,
+      "loss": 0.1315,
+      "step": 3438
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.45761398346885324,
+      "learning_rate": 7.963510363042103e-09,
+      "loss": 0.083,
+      "step": 3439
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8751324001451453,
+      "learning_rate": 7.704618342638804e-09,
+      "loss": 0.1298,
+      "step": 3440
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5002611858711786,
+      "learning_rate": 7.450001167809695e-09,
+      "loss": 0.1023,
+      "step": 3441
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.591030695217879,
+      "learning_rate": 7.199659056579822e-09,
+      "loss": 0.1224,
+      "step": 3442
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5605203273893826,
+      "learning_rate": 6.953592223314376e-09,
+      "loss": 0.0871,
+      "step": 3443
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.4037315182990444,
+      "learning_rate": 6.711800878718144e-09,
+      "loss": 0.1001,
+      "step": 3444
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5752254954729186,
+      "learning_rate": 6.4742852298338434e-09,
+      "loss": 0.1079,
+      "step": 3445
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.3958881293253481,
+      "learning_rate": 6.241045480043229e-09,
+      "loss": 0.0721,
+      "step": 3446
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.39285518812159115,
+      "learning_rate": 6.012081829067096e-09,
+      "loss": 0.0866,
+      "step": 3447
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.40275387561550535,
+      "learning_rate": 5.787394472964725e-09,
+      "loss": 0.1046,
+      "step": 3448
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.438490568732715,
+      "learning_rate": 5.566983604133325e-09,
+      "loss": 0.0689,
+      "step": 3449
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.22473269343661573,
+      "learning_rate": 5.35084941130748e-09,
+      "loss": 0.0485,
+      "step": 3450
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6516198767415275,
+      "learning_rate": 5.138992079561367e-09,
+      "loss": 0.1205,
+      "step": 3451
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1032766914286642,
+      "learning_rate": 4.931411790304874e-09,
+      "loss": 0.1663,
+      "step": 3452
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2583927664869538,
+      "learning_rate": 4.728108721288038e-09,
+      "loss": 0.0848,
+      "step": 3453
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3356188968818276,
+      "learning_rate": 4.5290830465954946e-09,
+      "loss": 0.0649,
+      "step": 3454
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.4091658339723283,
+      "learning_rate": 4.334334936652029e-09,
+      "loss": 0.105,
+      "step": 3455
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8466365415029252,
+      "learning_rate": 4.143864558217026e-09,
+      "loss": 0.1017,
+      "step": 3456
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.31266377817181285,
+      "learning_rate": 3.957672074388908e-09,
+      "loss": 0.0763,
+      "step": 3457
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3668851444638233,
+      "learning_rate": 3.775757644601808e-09,
+      "loss": 0.1018,
+      "step": 3458
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.26861290057771103,
+      "learning_rate": 3.5981214246266773e-09,
+      "loss": 0.0509,
+      "step": 3459
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5573965471791562,
+      "learning_rate": 3.4247635665723977e-09,
+      "loss": 0.0934,
+      "step": 3460
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5645186381180453,
+      "learning_rate": 3.255684218882449e-09,
+      "loss": 0.1307,
+      "step": 3461
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.30226112122952675,
+      "learning_rate": 3.0908835263376845e-09,
+      "loss": 0.0591,
+      "step": 3462
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2376879396232217,
+      "learning_rate": 2.930361630055223e-09,
+      "loss": 0.1214,
+      "step": 3463
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.4493319321934206,
+      "learning_rate": 2.7741186674878905e-09,
+      "loss": 0.0721,
+      "step": 3464
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.26721974000859217,
+      "learning_rate": 2.6221547724253337e-09,
+      "loss": 0.0687,
+      "step": 3465
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.38897379648915154,
+      "learning_rate": 2.4744700749923524e-09,
+      "loss": 0.092,
+      "step": 3466
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5296078663197749,
+      "learning_rate": 2.3310647016488996e-09,
+      "loss": 0.124,
+      "step": 3467
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.47833822699019957,
+      "learning_rate": 2.1919387751917485e-09,
+      "loss": 0.0899,
+      "step": 3468
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.4470979908901922,
+      "learning_rate": 2.057092414753381e-09,
+      "loss": 0.0723,
+      "step": 3469
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.35133914818930667,
+      "learning_rate": 1.926525735800877e-09,
+      "loss": 0.0568,
+      "step": 3470
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.27584097224744486,
+      "learning_rate": 1.800238850136471e-09,
+      "loss": 0.0552,
+      "step": 3471
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3517674209525712,
+      "learning_rate": 1.6782318658992159e-09,
+      "loss": 0.0509,
+      "step": 3472
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.4149660412637721,
+      "learning_rate": 1.5605048875610985e-09,
+      "loss": 0.0923,
+      "step": 3473
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6636339719837038,
+      "learning_rate": 1.4470580159314795e-09,
+      "loss": 0.0778,
+      "step": 3474
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5387027081203459,
+      "learning_rate": 1.3378913481526534e-09,
+      "loss": 0.0989,
+      "step": 3475
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.4298688281989752,
+      "learning_rate": 1.2330049777037334e-09,
+      "loss": 0.0847,
+      "step": 3476
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.35398138709642324,
+      "learning_rate": 1.1323989943973213e-09,
+      "loss": 0.0616,
+      "step": 3477
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6395952838124319,
+      "learning_rate": 1.0360734843806175e-09,
+      "loss": 0.0998,
+      "step": 3478
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2491902941704778,
+      "learning_rate": 9.440285301370865e-10,
+      "loss": 0.0559,
+      "step": 3479
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.4139740029135101,
+      "learning_rate": 8.562642104831265e-10,
+      "loss": 0.0547,
+      "step": 3480
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.46733585602740035,
+      "learning_rate": 7.727806005702887e-10,
+      "loss": 0.0776,
+      "step": 3481
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5162713301266479,
+      "learning_rate": 6.935777718847237e-10,
+      "loss": 0.123,
+      "step": 3482
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.4656719111645452,
+      "learning_rate": 6.186557922471803e-10,
+      "loss": 0.0828,
+      "step": 3483
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.4034051607145749,
+      "learning_rate": 5.480147258118962e-10,
+      "loss": 0.0617,
+      "step": 3484
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.3924679536845503,
+      "learning_rate": 4.816546330688177e-10,
+      "loss": 0.0877,
+      "step": 3485
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.36789267903815426,
+      "learning_rate": 4.1957557084082447e-10,
+      "loss": 0.0727,
+      "step": 3486
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.8386319740629019,
+      "learning_rate": 3.61777592285395e-10,
+      "loss": 0.1195,
+      "step": 3487
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.017152865730572,
+      "learning_rate": 3.082607468946064e-10,
+      "loss": 0.0837,
+      "step": 3488
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.38688824453684373,
+      "learning_rate": 2.590250804945793e-10,
+      "loss": 0.0869,
+      "step": 3489
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.4058464346142467,
+      "learning_rate": 2.1407063524436777e-10,
+      "loss": 0.0595,
+      "step": 3490
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5868300953204145,
+      "learning_rate": 1.7339744963873474e-10,
+      "loss": 0.0709,
+      "step": 3491
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5930557259373533,
+      "learning_rate": 1.3700555850537645e-10,
+      "loss": 0.1052,
+      "step": 3492
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.4733487558447307,
+      "learning_rate": 1.0489499300603279e-10,
+      "loss": 0.1137,
+      "step": 3493
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.35835173300491624,
+      "learning_rate": 7.706578063704228e-11,
+      "loss": 0.0936,
+      "step": 3494
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.3587238751658971,
+      "learning_rate": 5.351794522823195e-11,
+      "loss": 0.0553,
+      "step": 3495
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.43350933398175695,
+      "learning_rate": 3.425150694291724e-11,
+      "loss": 0.0858,
+      "step": 3496
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.27176137333516126,
+      "learning_rate": 1.926648227901229e-11,
+      "loss": 0.0606,
+      "step": 3497
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.9522309126437695,
+      "learning_rate": 8.562884067919719e-12,
+      "loss": 0.178,
+      "step": 3498
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.27174219401664274,
+      "learning_rate": 2.140721475085705e-12,
+      "loss": 0.0647,
+      "step": 3499
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.2727162763560935,
+      "learning_rate": 0.0,
+      "loss": 0.0498,
+      "step": 3500
+    },
+    {
+      "epoch": 1.0,
+      "step": 3500,
+      "total_flos": 133144802129920.0,
+      "train_loss": 0.10024488587464606,
+      "train_runtime": 7976.3246,
+      "train_samples_per_second": 0.878,
+      "train_steps_per_second": 0.439
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 3500,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 50000,
+  "total_flos": 133144802129920.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}