diff --git a/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa-streaming_baseline_lora_agg_fc4_s2048/README.md b/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa-streaming_baseline_lora_agg_fc4_s2048/README.md
new file mode 100644
index 0000000000000000000000000000000000000000..d4576fe074287232d3836bf69c21d3f2593290d9
--- /dev/null
+++ b/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa-streaming_baseline_lora_agg_fc4_s2048/README.md
@@ -0,0 +1,9 @@
+---
+library_name: peft
+---
+## Training procedure
+
+### Framework versions
+
+
+- PEFT 0.4.0
diff --git a/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa-streaming_baseline_lora_agg_fc4_s2048/adapter_config.json b/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa-streaming_baseline_lora_agg_fc4_s2048/adapter_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..286aafcf7ccd318e080dba3bacb341fe7046f2e3
--- /dev/null
+++ b/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa-streaming_baseline_lora_agg_fc4_s2048/adapter_config.json
@@ -0,0 +1,26 @@
+{
+  "auto_mapping": null,
+  "base_model_name_or_path": "/data/wiedmann/hub/models--lmms-lab--llava-onevision-qwen2-7b-ov/snapshots/0b07bf7565e244cf4f39982249eafe8cd799d6dd",
+  "bias": "none",
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "lora_alpha": 256,
+  "lora_dropout": 0.05,
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "r": 128,
+  "revision": null,
+  "target_modules": [
+    "up_proj",
+    "v_proj",
+    "k_proj",
+    "o_proj",
+    "q_proj",
+    "gate_proj",
+    "down_proj"
+  ],
+  "task_type": "CAUSAL_LM"
+}
\ No newline at end of file
diff --git a/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa-streaming_baseline_lora_agg_fc4_s2048/adapter_model.bin b/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa-streaming_baseline_lora_agg_fc4_s2048/adapter_model.bin
new file mode 100644
index 0000000000000000000000000000000000000000..8fadb42db6fe54344c408d61139718ef54d766af
--- /dev/null
+++ b/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa-streaming_baseline_lora_agg_fc4_s2048/adapter_model.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3cf5202f06cfe045e2437983d0655b3d319505b7e55e5b07b3dfa2e81f74213e
+size 692127130
diff --git a/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa-streaming_baseline_lora_agg_fc4_s2048/config.json b/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa-streaming_baseline_lora_agg_fc4_s2048/config.json
new file mode 100644
index 0000000000000000000000000000000000000000..524c06f1795a643f88d925d43550e0a0d49eb19f
--- /dev/null
+++ b/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa-streaming_baseline_lora_agg_fc4_s2048/config.json
@@ -0,0 +1,67 @@
+{
+  "_name_or_path": "/data/wiedmann/hub/models--lmms-lab--llava-onevision-qwen2-7b-ov/snapshots/0b07bf7565e244cf4f39982249eafe8cd799d6dd",
+  "add_faster_video": false,
+  "add_time_instruction": false,
+  "architectures": [
+    "LlavaQwenForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 151643,
+  "eos_token_id": 151645,
+  "faster_token_stride": 10,
+  "force_sample": false,
+  "hidden_act": "silu",
+  "hidden_size": 3584,
+  "image_aspect_ratio": "square",
+  "image_crop_resolution": null,
+  "image_grid_pinpoints": null,
+  "image_split_resolution": null,
+  "image_token_index": 151646,
+  "initializer_range": 0.02,
+  "intermediate_size": 18944,
+  "max_position_embeddings": 32768,
+  "max_window_layers": 28,
+  "mm_hidden_size": 1152,
+  "mm_newline_position": "no_token",
+  "mm_patch_merge_type": "spatial_unpad",
+  "mm_projector_lr": null,
+  "mm_projector_type": "mlp2x_gelu",
+  "mm_resampler_type": "streaming_agg",
+  "mm_spatial_pool_mode": "bilinear",
+  "mm_spatial_pool_stride": null,
+  "mm_streaming_frames_per_chunk": 4,
+  "mm_streaming_input_dim": 1152,
+  "mm_streaming_num_heads": 8,
+  "mm_streaming_num_layers": 4,
+  "mm_streaming_num_state_tokens": 2048,
+  "mm_streaming_patches_per_frame": 729,
+  "mm_streaming_state_dim": 1152,
+  "mm_streaming_vision_chunk_size": 8,
+  "mm_tunable_parts": "mm_mlp_adapter,mm_vision_resampler",
+  "mm_use_im_patch_token": false,
+  "mm_use_im_start_end": false,
+  "mm_vision_select_feature": "patch",
+  "mm_vision_select_layer": -2,
+  "mm_vision_tower": "google/siglip-so400m-patch14-384",
+  "mm_vision_tower_lr": null,
+  "model_type": "llava_qwen",
+  "num_attention_heads": 28,
+  "num_hidden_layers": 28,
+  "num_key_value_heads": 4,
+  "pos_skipping_range": 4096,
+  "rms_norm_eps": 1e-06,
+  "rope_scaling": null,
+  "rope_theta": 1000000.0,
+  "sliding_window": 131072,
+  "tie_word_embeddings": false,
+  "tokenizer_model_max_length": 8192,
+  "tokenizer_padding_side": "right",
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.40.0.dev0",
+  "use_cache": true,
+  "use_mm_proj": true,
+  "use_pos_skipping": false,
+  "use_sliding_window": false,
+  "vision_tower_pretrained": null,
+  "vocab_size": 152064
+}
diff --git a/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa-streaming_baseline_lora_agg_fc4_s2048/generation_config.json b/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa-streaming_baseline_lora_agg_fc4_s2048/generation_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..19a297221acb87418d4388a3decef2282c6d7316
--- /dev/null
+++ b/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa-streaming_baseline_lora_agg_fc4_s2048/generation_config.json
@@ -0,0 +1,14 @@
+{
+  "bos_token_id": 151643,
+  "do_sample": true,
+  "eos_token_id": [
+    151645,
+    151643
+  ],
+  "pad_token_id": 151643,
+  "repetition_penalty": 1.05,
+  "temperature": 0.7,
+  "top_k": 20,
+  "top_p": 0.8,
+  "transformers_version": "4.40.0.dev0"
+}
diff --git a/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa-streaming_baseline_lora_agg_fc4_s2048/non_lora_trainables.bin b/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa-streaming_baseline_lora_agg_fc4_s2048/non_lora_trainables.bin
new file mode 100644
index 0000000000000000000000000000000000000000..f940bfd16b7afc3bf208e644d68dd928fa0f5438
--- /dev/null
+++ b/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa-streaming_baseline_lora_agg_fc4_s2048/non_lora_trainables.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1e6b77bee65b1ee2acaf54000f4efe77755b71b066a16e63f5bc97a132a26ee5
+size 211441788
diff --git a/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa-streaming_baseline_lora_agg_fc4_s2048/trainer_state.json b/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa-streaming_baseline_lora_agg_fc4_s2048/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..f51b8d458debc028f096bb6fa6f53b950ffa61e3
--- /dev/null
+++ b/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa-streaming_baseline_lora_agg_fc4_s2048/trainer_state.json
@@ -0,0 +1,7492 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9999413730433253,
+  "eval_steps": 500,
+  "global_step": 1066,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 53.56119918823242,
+      "learning_rate": 3.125e-06,
+      "loss": 1.2329,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 44.24412155151367,
+      "learning_rate": 6.25e-06,
+      "loss": 1.253,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 31.141427993774414,
+      "learning_rate": 9.375000000000001e-06,
+      "loss": 1.1976,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.549352645874023,
+      "learning_rate": 1.25e-05,
+      "loss": 0.9174,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.200796127319336,
+      "learning_rate": 1.5625e-05,
+      "loss": 0.9193,
+      "step": 5
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.501561164855957,
+      "learning_rate": 1.8750000000000002e-05,
+      "loss": 0.7185,
+      "step": 6
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.535277366638184,
+      "learning_rate": 2.1875e-05,
+      "loss": 0.599,
+      "step": 7
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7859258651733398,
+      "learning_rate": 2.5e-05,
+      "loss": 0.4878,
+      "step": 8
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.49070405960083,
+      "learning_rate": 2.8125000000000003e-05,
+      "loss": 0.4525,
+      "step": 9
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8042423725128174,
+      "learning_rate": 3.125e-05,
+      "loss": 0.3429,
+      "step": 10
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8497580885887146,
+      "learning_rate": 3.4375e-05,
+      "loss": 0.3007,
+      "step": 11
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7737765312194824,
+      "learning_rate": 3.7500000000000003e-05,
+      "loss": 0.4551,
+      "step": 12
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9109475016593933,
+      "learning_rate": 4.0625000000000005e-05,
+      "loss": 0.5592,
+      "step": 13
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6791670918464661,
+      "learning_rate": 4.375e-05,
+      "loss": 0.3611,
+      "step": 14
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7185907363891602,
+      "learning_rate": 4.6875e-05,
+      "loss": 0.4888,
+      "step": 15
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7224956750869751,
+      "learning_rate": 5e-05,
+      "loss": 0.3629,
+      "step": 16
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5531966090202332,
+      "learning_rate": 5.3125000000000004e-05,
+      "loss": 0.2789,
+      "step": 17
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7150623202323914,
+      "learning_rate": 5.6250000000000005e-05,
+      "loss": 0.3422,
+      "step": 18
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.4981580078601837,
+      "learning_rate": 5.9375e-05,
+      "loss": 0.3128,
+      "step": 19
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.712741494178772,
+      "learning_rate": 6.25e-05,
+      "loss": 0.4357,
+      "step": 20
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5813599824905396,
+      "learning_rate": 6.562500000000001e-05,
+      "loss": 0.3287,
+      "step": 21
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6859356760978699,
+      "learning_rate": 6.875e-05,
+      "loss": 0.4216,
+      "step": 22
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6142020225524902,
+      "learning_rate": 7.1875e-05,
+      "loss": 0.3745,
+      "step": 23
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5768157243728638,
+      "learning_rate": 7.500000000000001e-05,
+      "loss": 0.3246,
+      "step": 24
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5831692218780518,
+      "learning_rate": 7.8125e-05,
+      "loss": 0.3969,
+      "step": 25
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5747249126434326,
+      "learning_rate": 8.125000000000001e-05,
+      "loss": 0.4105,
+      "step": 26
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5389623641967773,
+      "learning_rate": 8.4375e-05,
+      "loss": 0.3935,
+      "step": 27
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6806145310401917,
+      "learning_rate": 8.75e-05,
+      "loss": 0.3193,
+      "step": 28
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5326066613197327,
+      "learning_rate": 9.062500000000001e-05,
+      "loss": 0.3592,
+      "step": 29
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6331548690795898,
+      "learning_rate": 9.375e-05,
+      "loss": 0.4431,
+      "step": 30
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5740853548049927,
+      "learning_rate": 9.687500000000001e-05,
+      "loss": 0.3317,
+      "step": 31
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4643768072128296,
+      "learning_rate": 0.0001,
+      "loss": 0.2877,
+      "step": 32
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5114359259605408,
+      "learning_rate": 9.999976921990784e-05,
+      "loss": 0.314,
+      "step": 33
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5951298475265503,
+      "learning_rate": 9.999907688176173e-05,
+      "loss": 0.4149,
+      "step": 34
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.47754591703414917,
+      "learning_rate": 9.999792299195278e-05,
+      "loss": 0.2769,
+      "step": 35
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5500280857086182,
+      "learning_rate": 9.999630756113278e-05,
+      "loss": 0.3778,
+      "step": 36
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.48305949568748474,
+      "learning_rate": 9.999423060421411e-05,
+      "loss": 0.3141,
+      "step": 37
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4953254461288452,
+      "learning_rate": 9.999169214036958e-05,
+      "loss": 0.3525,
+      "step": 38
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.47390633821487427,
+      "learning_rate": 9.998869219303227e-05,
+      "loss": 0.3096,
+      "step": 39
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5356336236000061,
+      "learning_rate": 9.998523078989529e-05,
+      "loss": 0.3706,
+      "step": 40
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.46947455406188965,
+      "learning_rate": 9.998130796291156e-05,
+      "loss": 0.3649,
+      "step": 41
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5029218196868896,
+      "learning_rate": 9.997692374829352e-05,
+      "loss": 0.3149,
+      "step": 42
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.274848699569702,
+      "learning_rate": 9.997207818651274e-05,
+      "loss": 0.2937,
+      "step": 43
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4324891269207001,
+      "learning_rate": 9.996677132229957e-05,
+      "loss": 0.3225,
+      "step": 44
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4642087519168854,
+      "learning_rate": 9.996100320464274e-05,
+      "loss": 0.3577,
+      "step": 45
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4427544176578522,
+      "learning_rate": 9.995477388678897e-05,
+      "loss": 0.2711,
+      "step": 46
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.503923773765564,
+      "learning_rate": 9.994808342624234e-05,
+      "loss": 0.3,
+      "step": 47
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4020770788192749,
+      "learning_rate": 9.994093188476382e-05,
+      "loss": 0.2837,
+      "step": 48
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4590403437614441,
+      "learning_rate": 9.993331932837079e-05,
+      "loss": 0.3554,
+      "step": 49
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4729478359222412,
+      "learning_rate": 9.992524582733629e-05,
+      "loss": 0.276,
+      "step": 50
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4242483675479889,
+      "learning_rate": 9.991671145618846e-05,
+      "loss": 0.2385,
+      "step": 51
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4706641733646393,
+      "learning_rate": 9.99077162937098e-05,
+      "loss": 0.329,
+      "step": 52
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.486446738243103,
+      "learning_rate": 9.989826042293652e-05,
+      "loss": 0.3326,
+      "step": 53
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4750838875770569,
+      "learning_rate": 9.988834393115767e-05,
+      "loss": 0.3182,
+      "step": 54
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5082477331161499,
+      "learning_rate": 9.987796690991439e-05,
+      "loss": 0.3621,
+      "step": 55
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.46366146206855774,
+      "learning_rate": 9.98671294549991e-05,
+      "loss": 0.317,
+      "step": 56
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.43866562843322754,
+      "learning_rate": 9.985583166645455e-05,
+      "loss": 0.2456,
+      "step": 57
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.41880330443382263,
+      "learning_rate": 9.98440736485729e-05,
+      "loss": 0.2903,
+      "step": 58
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.40110716223716736,
+      "learning_rate": 9.983185550989487e-05,
+      "loss": 0.2215,
+      "step": 59
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.46664467453956604,
+      "learning_rate": 9.981917736320851e-05,
+      "loss": 0.3576,
+      "step": 60
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.36552098393440247,
+      "learning_rate": 9.980603932554845e-05,
+      "loss": 0.2001,
+      "step": 61
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.40886834263801575,
+      "learning_rate": 9.979244151819453e-05,
+      "loss": 0.247,
+      "step": 62
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4742773473262787,
+      "learning_rate": 9.97783840666709e-05,
+      "loss": 0.3561,
+      "step": 63
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.44993096590042114,
+      "learning_rate": 9.976386710074478e-05,
+      "loss": 0.3031,
+      "step": 64
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.42588794231414795,
+      "learning_rate": 9.974889075442521e-05,
+      "loss": 0.2303,
+      "step": 65
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4792451858520508,
+      "learning_rate": 9.97334551659619e-05,
+      "loss": 0.3047,
+      "step": 66
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.45113521814346313,
+      "learning_rate": 9.971756047784393e-05,
+      "loss": 0.3148,
+      "step": 67
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.43510547280311584,
+      "learning_rate": 9.970120683679838e-05,
+      "loss": 0.3894,
+      "step": 68
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4710015058517456,
+      "learning_rate": 9.968439439378905e-05,
+      "loss": 0.3901,
+      "step": 69
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4542221426963806,
+      "learning_rate": 9.966712330401504e-05,
+      "loss": 0.3633,
+      "step": 70
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5098312497138977,
+      "learning_rate": 9.964939372690926e-05,
+      "loss": 0.4011,
+      "step": 71
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7176426649093628,
+      "learning_rate": 9.96312058261371e-05,
+      "loss": 0.3472,
+      "step": 72
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.44791948795318604,
+      "learning_rate": 9.961255976959473e-05,
+      "loss": 0.3242,
+      "step": 73
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.49342313408851624,
+      "learning_rate": 9.959345572940771e-05,
+      "loss": 0.4155,
+      "step": 74
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.38717854022979736,
+      "learning_rate": 9.957389388192935e-05,
+      "loss": 0.2929,
+      "step": 75
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4250815808773041,
+      "learning_rate": 9.9553874407739e-05,
+      "loss": 0.2688,
+      "step": 76
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3985184133052826,
+      "learning_rate": 9.953339749164057e-05,
+      "loss": 0.3616,
+      "step": 77
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4453914165496826,
+      "learning_rate": 9.951246332266057e-05,
+      "loss": 0.3602,
+      "step": 78
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.46241435408592224,
+      "learning_rate": 9.949107209404665e-05,
+      "loss": 0.4017,
+      "step": 79
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.40397438406944275,
+      "learning_rate": 9.946922400326554e-05,
+      "loss": 0.3773,
+      "step": 80
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.459163635969162,
+      "learning_rate": 9.944691925200145e-05,
+      "loss": 0.3643,
+      "step": 81
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.44683462381362915,
+      "learning_rate": 9.942415804615406e-05,
+      "loss": 0.3826,
+      "step": 82
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.481871098279953,
+      "learning_rate": 9.940094059583671e-05,
+      "loss": 0.3892,
+      "step": 83
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.42870786786079407,
+      "learning_rate": 9.937726711537442e-05,
+      "loss": 0.2709,
+      "step": 84
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.41368356347084045,
+      "learning_rate": 9.93531378233019e-05,
+      "loss": 0.3394,
+      "step": 85
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.48373904824256897,
+      "learning_rate": 9.932855294236154e-05,
+      "loss": 0.3624,
+      "step": 86
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4892710745334625,
+      "learning_rate": 9.930351269950143e-05,
+      "loss": 0.406,
+      "step": 87
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4693341851234436,
+      "learning_rate": 9.927801732587312e-05,
+      "loss": 0.3009,
+      "step": 88
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3869418501853943,
+      "learning_rate": 9.925206705682962e-05,
+      "loss": 0.2619,
+      "step": 89
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5536014437675476,
+      "learning_rate": 9.92256621319231e-05,
+      "loss": 0.4697,
+      "step": 90
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4700610935688019,
+      "learning_rate": 9.919880279490286e-05,
+      "loss": 0.3152,
+      "step": 91
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.35885682702064514,
+      "learning_rate": 9.917148929371288e-05,
+      "loss": 0.1736,
+      "step": 92
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.48592039942741394,
+      "learning_rate": 9.914372188048964e-05,
+      "loss": 0.3746,
+      "step": 93
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.37780046463012695,
+      "learning_rate": 9.911550081155983e-05,
+      "loss": 0.2652,
+      "step": 94
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.48445338010787964,
+      "learning_rate": 9.908682634743784e-05,
+      "loss": 0.3928,
+      "step": 95
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.41280871629714966,
+      "learning_rate": 9.905769875282352e-05,
+      "loss": 0.2897,
+      "step": 96
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4266187250614166,
+      "learning_rate": 9.902811829659961e-05,
+      "loss": 0.3058,
+      "step": 97
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.49948224425315857,
+      "learning_rate": 9.899808525182935e-05,
+      "loss": 0.3773,
+      "step": 98
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3814532160758972,
+      "learning_rate": 9.896759989575386e-05,
+      "loss": 0.2951,
+      "step": 99
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4628780782222748,
+      "learning_rate": 9.893666250978971e-05,
+      "loss": 0.3758,
+      "step": 100
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3314332664012909,
+      "learning_rate": 9.890527337952617e-05,
+      "loss": 0.202,
+      "step": 101
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.43076321482658386,
+      "learning_rate": 9.887343279472272e-05,
+      "loss": 0.3034,
+      "step": 102
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4612520933151245,
+      "learning_rate": 9.884114104930628e-05,
+      "loss": 0.3215,
+      "step": 103
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3903290033340454,
+      "learning_rate": 9.880839844136854e-05,
+      "loss": 0.2625,
+      "step": 104
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4545579254627228,
+      "learning_rate": 9.877520527316317e-05,
+      "loss": 0.398,
+      "step": 105
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3861580193042755,
+      "learning_rate": 9.874156185110306e-05,
+      "loss": 0.3056,
+      "step": 106
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.48816314339637756,
+      "learning_rate": 9.870746848575751e-05,
+      "loss": 0.3215,
+      "step": 107
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4321497976779938,
+      "learning_rate": 9.86729254918493e-05,
+      "loss": 0.2767,
+      "step": 108
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.39948931336402893,
+      "learning_rate": 9.863793318825186e-05,
+      "loss": 0.2968,
+      "step": 109
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4157947301864624,
+      "learning_rate": 9.860249189798627e-05,
+      "loss": 0.4069,
+      "step": 110
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.35165318846702576,
+      "learning_rate": 9.856660194821829e-05,
+      "loss": 0.2218,
+      "step": 111
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4359529912471771,
+      "learning_rate": 9.853026367025535e-05,
+      "loss": 0.3191,
+      "step": 112
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3869144320487976,
+      "learning_rate": 9.849347739954352e-05,
+      "loss": 0.2881,
+      "step": 113
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5311065912246704,
+      "learning_rate": 9.845624347566433e-05,
+      "loss": 0.4661,
+      "step": 114
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3647288680076599,
+      "learning_rate": 9.841856224233174e-05,
+      "loss": 0.2192,
+      "step": 115
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.48680463433265686,
+      "learning_rate": 9.83804340473889e-05,
+      "loss": 0.4224,
+      "step": 116
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4152064025402069,
+      "learning_rate": 9.83418592428049e-05,
+      "loss": 0.2983,
+      "step": 117
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4352041780948639,
+      "learning_rate": 9.830283818467163e-05,
+      "loss": 0.3396,
+      "step": 118
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.45859429240226746,
+      "learning_rate": 9.826337123320046e-05,
+      "loss": 0.3314,
+      "step": 119
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.44424307346343994,
+      "learning_rate": 9.822345875271883e-05,
+      "loss": 0.3391,
+      "step": 120
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4700634181499481,
+      "learning_rate": 9.818310111166699e-05,
+      "loss": 0.341,
+      "step": 121
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4977334141731262,
+      "learning_rate": 9.814229868259452e-05,
+      "loss": 0.3589,
+      "step": 122
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4814792275428772,
+      "learning_rate": 9.810105184215699e-05,
+      "loss": 0.4187,
+      "step": 123
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4522244334220886,
+      "learning_rate": 9.805936097111234e-05,
+      "loss": 0.2712,
+      "step": 124
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5123592019081116,
+      "learning_rate": 9.801722645431754e-05,
+      "loss": 0.3991,
+      "step": 125
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.42592427134513855,
+      "learning_rate": 9.797464868072488e-05,
+      "loss": 0.3108,
+      "step": 126
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4312685430049896,
+      "learning_rate": 9.793162804337845e-05,
+      "loss": 0.3203,
+      "step": 127
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4493767023086548,
+      "learning_rate": 9.788816493941051e-05,
+      "loss": 0.3489,
+      "step": 128
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.45616286993026733,
+      "learning_rate": 9.784425977003784e-05,
+      "loss": 0.3597,
+      "step": 129
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4272081255912781,
+      "learning_rate": 9.779991294055802e-05,
+      "loss": 0.3742,
+      "step": 130
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3702671229839325,
+      "learning_rate": 9.775512486034563e-05,
+      "loss": 0.2558,
+      "step": 131
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.582704484462738,
+      "learning_rate": 9.770989594284857e-05,
+      "loss": 0.5771,
+      "step": 132
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4729974567890167,
+      "learning_rate": 9.766422660558421e-05,
+      "loss": 0.3888,
+      "step": 133
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3848215341567993,
+      "learning_rate": 9.761811727013548e-05,
+      "loss": 0.2774,
+      "step": 134
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3735813796520233,
+      "learning_rate": 9.757156836214706e-05,
+      "loss": 0.3127,
+      "step": 135
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3998161554336548,
+      "learning_rate": 9.752458031132141e-05,
+      "loss": 0.3537,
+      "step": 136
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3877810537815094,
+      "learning_rate": 9.747715355141478e-05,
+      "loss": 0.2334,
+      "step": 137
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.36280906200408936,
+      "learning_rate": 9.742928852023325e-05,
+      "loss": 0.325,
+      "step": 138
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.43097636103630066,
+      "learning_rate": 9.73809856596287e-05,
+      "loss": 0.3603,
+      "step": 139
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4008861482143402,
+      "learning_rate": 9.733224541549464e-05,
+      "loss": 0.3164,
+      "step": 140
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.43347349762916565,
+      "learning_rate": 9.728306823776221e-05,
+      "loss": 0.3332,
+      "step": 141
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4022083878517151,
+      "learning_rate": 9.723345458039594e-05,
+      "loss": 0.3515,
+      "step": 142
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3428991436958313,
+      "learning_rate": 9.718340490138965e-05,
+      "loss": 0.1898,
+      "step": 143
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.45103755593299866,
+      "learning_rate": 9.713291966276206e-05,
+      "loss": 0.3417,
+      "step": 144
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4582728445529938,
+      "learning_rate": 9.708199933055272e-05,
+      "loss": 0.4211,
+      "step": 145
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.48150041699409485,
+      "learning_rate": 9.70306443748176e-05,
+      "loss": 0.4027,
+      "step": 146
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3726791441440582,
+      "learning_rate": 9.697885526962474e-05,
+      "loss": 0.2824,
+      "step": 147
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.41239234805107117,
+      "learning_rate": 9.692663249304992e-05,
+      "loss": 0.3291,
+      "step": 148
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.48422345519065857,
+      "learning_rate": 9.687397652717223e-05,
+      "loss": 0.464,
+      "step": 149
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4588571786880493,
+      "learning_rate": 9.682088785806963e-05,
+      "loss": 0.3762,
+      "step": 150
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3948071002960205,
+      "learning_rate": 9.67673669758144e-05,
+      "loss": 0.2757,
+      "step": 151
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.42584747076034546,
+      "learning_rate": 9.671341437446877e-05,
+      "loss": 0.3365,
+      "step": 152
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5137955546379089,
+      "learning_rate": 9.665903055208014e-05,
+      "loss": 0.4396,
+      "step": 153
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.42933642864227295,
+      "learning_rate": 9.660421601067666e-05,
+      "loss": 0.3334,
+      "step": 154
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.36864256858825684,
+      "learning_rate": 9.654897125626252e-05,
+      "loss": 0.2852,
+      "step": 155
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.37826991081237793,
+      "learning_rate": 9.649329679881334e-05,
+      "loss": 0.2815,
+      "step": 156
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3164992332458496,
+      "learning_rate": 9.643719315227133e-05,
+      "loss": 0.1955,
+      "step": 157
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.43146809935569763,
+      "learning_rate": 9.63806608345407e-05,
+      "loss": 0.3581,
+      "step": 158
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4671701490879059,
+      "learning_rate": 9.632370036748279e-05,
+      "loss": 0.3826,
+      "step": 159
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4189399480819702,
+      "learning_rate": 9.626631227691127e-05,
+      "loss": 0.2845,
+      "step": 160
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.38566893339157104,
+      "learning_rate": 9.62084970925873e-05,
+      "loss": 0.245,
+      "step": 161
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4231869876384735,
+      "learning_rate": 9.615025534821462e-05,
+      "loss": 0.2744,
+      "step": 162
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4824386239051819,
+      "learning_rate": 9.609158758143464e-05,
+      "loss": 0.3822,
+      "step": 163
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4428365230560303,
+      "learning_rate": 9.603249433382144e-05,
+      "loss": 0.2882,
+      "step": 164
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4528397023677826,
+      "learning_rate": 9.597297615087685e-05,
+      "loss": 0.268,
+      "step": 165
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3863323926925659,
+      "learning_rate": 9.591303358202535e-05,
+      "loss": 0.2234,
+      "step": 166
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4577626883983612,
+      "learning_rate": 9.585266718060897e-05,
+      "loss": 0.354,
+      "step": 167
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4039238691329956,
+      "learning_rate": 9.579187750388227e-05,
+      "loss": 0.2139,
+      "step": 168
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.43405696749687195,
+      "learning_rate": 9.573066511300714e-05,
+      "loss": 0.3414,
+      "step": 169
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4332447052001953,
+      "learning_rate": 9.566903057304764e-05,
+      "loss": 0.3628,
+      "step": 170
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4371618628501892,
+      "learning_rate": 9.560697445296474e-05,
+      "loss": 0.3303,
+      "step": 171
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5302167534828186,
+      "learning_rate": 9.554449732561113e-05,
+      "loss": 0.3516,
+      "step": 172
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4591212272644043,
+      "learning_rate": 9.548159976772592e-05,
+      "loss": 0.3495,
+      "step": 173
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.47322651743888855,
+      "learning_rate": 9.541828235992926e-05,
+      "loss": 0.3561,
+      "step": 174
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.38110026717185974,
+      "learning_rate": 9.535454568671704e-05,
+      "loss": 0.2689,
+      "step": 175
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.443582147359848,
+      "learning_rate": 9.529039033645548e-05,
+      "loss": 0.2448,
+      "step": 176
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.37832385301589966,
+      "learning_rate": 9.522581690137567e-05,
+      "loss": 0.305,
+      "step": 177
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3523258566856384,
+      "learning_rate": 9.516082597756815e-05,
+      "loss": 0.2324,
+      "step": 178
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4567958414554596,
+      "learning_rate": 9.509541816497737e-05,
+      "loss": 0.4185,
+      "step": 179
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3902451694011688,
+      "learning_rate": 9.50295940673962e-05,
+      "loss": 0.2496,
+      "step": 180
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.43655723333358765,
+      "learning_rate": 9.496335429246026e-05,
+      "loss": 0.391,
+      "step": 181
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.419964462518692,
+      "learning_rate": 9.489669945164242e-05,
+      "loss": 0.3263,
+      "step": 182
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3563212454319,
+      "learning_rate": 9.482963016024709e-05,
+      "loss": 0.224,
+      "step": 183
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3605343997478485,
+      "learning_rate": 9.476214703740454e-05,
+      "loss": 0.2404,
+      "step": 184
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5000762343406677,
+      "learning_rate": 9.469425070606524e-05,
+      "loss": 0.5458,
+      "step": 185
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.40717747807502747,
+      "learning_rate": 9.462594179299406e-05,
+      "loss": 0.2314,
+      "step": 186
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.41368338465690613,
+      "learning_rate": 9.45572209287645e-05,
+      "loss": 0.3461,
+      "step": 187
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5594714879989624,
+      "learning_rate": 9.44880887477528e-05,
+      "loss": 0.3762,
+      "step": 188
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4597151279449463,
+      "learning_rate": 9.441854588813228e-05,
+      "loss": 0.2904,
+      "step": 189
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.39208316802978516,
+      "learning_rate": 9.43485929918672e-05,
+      "loss": 0.323,
+      "step": 190
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5003756284713745,
+      "learning_rate": 9.427823070470699e-05,
+      "loss": 0.4213,
+      "step": 191
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3722604811191559,
+      "learning_rate": 9.420745967618026e-05,
+      "loss": 0.2739,
+      "step": 192
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.30610206723213196,
+      "learning_rate": 9.413628055958878e-05,
+      "loss": 0.1887,
+      "step": 193
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4051414430141449,
+      "learning_rate": 9.406469401200151e-05,
+      "loss": 0.3377,
+      "step": 194
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.45952826738357544,
+      "learning_rate": 9.399270069424842e-05,
+      "loss": 0.3374,
+      "step": 195
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.45123541355133057,
+      "learning_rate": 9.392030127091452e-05,
+      "loss": 0.2862,
+      "step": 196
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3941093981266022,
+      "learning_rate": 9.384749641033359e-05,
+      "loss": 0.2781,
+      "step": 197
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.45097070932388306,
+      "learning_rate": 9.377428678458214e-05,
+      "loss": 0.3586,
+      "step": 198
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4081740379333496,
+      "learning_rate": 9.370067306947316e-05,
+      "loss": 0.274,
+      "step": 199
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.47927120327949524,
+      "learning_rate": 9.362665594454984e-05,
+      "loss": 0.4323,
+      "step": 200
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4218602478504181,
+      "learning_rate": 9.355223609307933e-05,
+      "loss": 0.2805,
+      "step": 201
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.47772476077079773,
+      "learning_rate": 9.347741420204643e-05,
+      "loss": 0.361,
+      "step": 202
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4543375074863434,
+      "learning_rate": 9.340219096214727e-05,
+      "loss": 0.2929,
+      "step": 203
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4603528082370758,
+      "learning_rate": 9.33265670677829e-05,
+      "loss": 0.364,
+      "step": 204
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4432274103164673,
+      "learning_rate": 9.325054321705289e-05,
+      "loss": 0.3073,
+      "step": 205
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.43326976895332336,
+      "learning_rate": 9.317412011174886e-05,
+      "loss": 0.3113,
+      "step": 206
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4768969714641571,
+      "learning_rate": 9.309729845734813e-05,
+      "loss": 0.4033,
+      "step": 207
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.36311060190200806,
+      "learning_rate": 9.302007896300698e-05,
+      "loss": 0.2685,
+      "step": 208
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.41629478335380554,
+      "learning_rate": 9.29424623415543e-05,
+      "loss": 0.3202,
+      "step": 209
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.47300565242767334,
+      "learning_rate": 9.286444930948496e-05,
+      "loss": 0.361,
+      "step": 210
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.363154798746109,
+      "learning_rate": 9.278604058695313e-05,
+      "loss": 0.2366,
+      "step": 211
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4519854784011841,
+      "learning_rate": 9.270723689776568e-05,
+      "loss": 0.3207,
+      "step": 212
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4606704115867615,
+      "learning_rate": 9.262803896937555e-05,
+      "loss": 0.4074,
+      "step": 213
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.42994555830955505,
+      "learning_rate": 9.254844753287493e-05,
+      "loss": 0.277,
+      "step": 214
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4647696912288666,
+      "learning_rate": 9.24684633229886e-05,
+      "loss": 0.3115,
+      "step": 215
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5536186099052429,
+      "learning_rate": 9.238808707806706e-05,
+      "loss": 0.3567,
+      "step": 216
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.48201286792755127,
+      "learning_rate": 9.230731954007983e-05,
+      "loss": 0.369,
+      "step": 217
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5291519165039062,
+      "learning_rate": 9.222616145460849e-05,
+      "loss": 0.421,
+      "step": 218
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4186636507511139,
+      "learning_rate": 9.214461357083985e-05,
+      "loss": 0.2158,
+      "step": 219
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3635309636592865,
+      "learning_rate": 9.206267664155907e-05,
+      "loss": 0.2536,
+      "step": 220
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4199702739715576,
+      "learning_rate": 9.198035142314259e-05,
+      "loss": 0.3763,
+      "step": 221
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4509720504283905,
+      "learning_rate": 9.189763867555129e-05,
+      "loss": 0.3784,
+      "step": 222
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.42019206285476685,
+      "learning_rate": 9.181453916232339e-05,
+      "loss": 0.2547,
+      "step": 223
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4254032373428345,
+      "learning_rate": 9.173105365056742e-05,
+      "loss": 0.3493,
+      "step": 224
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.47555145621299744,
+      "learning_rate": 9.164718291095515e-05,
+      "loss": 0.375,
+      "step": 225
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.40298861265182495,
+      "learning_rate": 9.156292771771446e-05,
+      "loss": 0.3356,
+      "step": 226
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.379826158285141,
+      "learning_rate": 9.14782888486222e-05,
+      "loss": 0.3092,
+      "step": 227
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.38476112484931946,
+      "learning_rate": 9.1393267084997e-05,
+      "loss": 0.2987,
+      "step": 228
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.46264320611953735,
+      "learning_rate": 9.130786321169209e-05,
+      "loss": 0.3771,
+      "step": 229
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.41812068223953247,
+      "learning_rate": 9.122207801708802e-05,
+      "loss": 0.3283,
+      "step": 230
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3706599771976471,
+      "learning_rate": 9.113591229308538e-05,
+      "loss": 0.2443,
+      "step": 231
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4221385419368744,
+      "learning_rate": 9.104936683509755e-05,
+      "loss": 0.3107,
+      "step": 232
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3780944347381592,
+      "learning_rate": 9.096244244204324e-05,
+      "loss": 0.3095,
+      "step": 233
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4322282075881958,
+      "learning_rate": 9.087513991633924e-05,
+      "loss": 0.3187,
+      "step": 234
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.42131513357162476,
+      "learning_rate": 9.078746006389298e-05,
+      "loss": 0.3539,
+      "step": 235
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.32596921920776367,
+      "learning_rate": 9.069940369409499e-05,
+      "loss": 0.1888,
+      "step": 236
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4986932873725891,
+      "learning_rate": 9.061097161981159e-05,
+      "loss": 0.406,
+      "step": 237
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4771842360496521,
+      "learning_rate": 9.052216465737726e-05,
+      "loss": 0.3748,
+      "step": 238
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.46478331089019775,
+      "learning_rate": 9.043298362658714e-05,
+      "loss": 0.36,
+      "step": 239
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.36818572878837585,
+      "learning_rate": 9.034342935068952e-05,
+      "loss": 0.1987,
+      "step": 240
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4378790259361267,
+      "learning_rate": 9.025350265637815e-05,
+      "loss": 0.3699,
+      "step": 241
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.46986663341522217,
+      "learning_rate": 9.016320437378466e-05,
+      "loss": 0.3178,
+      "step": 242
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.44161608815193176,
+      "learning_rate": 9.007253533647089e-05,
+      "loss": 0.3466,
+      "step": 243
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4717167913913727,
+      "learning_rate": 8.998149638142119e-05,
+      "loss": 0.3324,
+      "step": 244
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5091121792793274,
+      "learning_rate": 8.98900883490347e-05,
+      "loss": 0.4375,
+      "step": 245
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3929200768470764,
+      "learning_rate": 8.979831208311758e-05,
+      "loss": 0.2908,
+      "step": 246
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.46935197710990906,
+      "learning_rate": 8.970616843087524e-05,
+      "loss": 0.3222,
+      "step": 247
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3767223656177521,
+      "learning_rate": 8.96136582429045e-05,
+      "loss": 0.2201,
+      "step": 248
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.46659228205680847,
+      "learning_rate": 8.952078237318575e-05,
+      "loss": 0.3446,
+      "step": 249
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4290735423564911,
+      "learning_rate": 8.942754167907507e-05,
+      "loss": 0.2723,
+      "step": 250
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.45299649238586426,
+      "learning_rate": 8.933393702129628e-05,
+      "loss": 0.2574,
+      "step": 251
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4417344629764557,
+      "learning_rate": 8.923996926393305e-05,
+      "loss": 0.4022,
+      "step": 252
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4162667393684387,
+      "learning_rate": 8.91456392744209e-05,
+      "loss": 0.3696,
+      "step": 253
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.47140949964523315,
+      "learning_rate": 8.905094792353917e-05,
+      "loss": 0.3376,
+      "step": 254
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4572603702545166,
+      "learning_rate": 8.895589608540297e-05,
+      "loss": 0.3583,
+      "step": 255
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.35900965332984924,
+      "learning_rate": 8.886048463745525e-05,
+      "loss": 0.1709,
+      "step": 256
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4216223359107971,
+      "learning_rate": 8.876471446045847e-05,
+      "loss": 0.299,
+      "step": 257
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.41913264989852905,
+      "learning_rate": 8.866858643848665e-05,
+      "loss": 0.2399,
+      "step": 258
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3832896053791046,
+      "learning_rate": 8.857210145891715e-05,
+      "loss": 0.2869,
+      "step": 259
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3874155879020691,
+      "learning_rate": 8.847526041242246e-05,
+      "loss": 0.1823,
+      "step": 260
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.37168049812316895,
+      "learning_rate": 8.8378064192962e-05,
+      "loss": 0.239,
+      "step": 261
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.43271905183792114,
+      "learning_rate": 8.82805136977739e-05,
+      "loss": 0.3633,
+      "step": 262
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.48894739151000977,
+      "learning_rate": 8.818260982736661e-05,
+      "loss": 0.3543,
+      "step": 263
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3934341371059418,
+      "learning_rate": 8.808435348551071e-05,
+      "loss": 0.2181,
+      "step": 264
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.36207613348960876,
+      "learning_rate": 8.798574557923053e-05,
+      "loss": 0.2183,
+      "step": 265
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3730492889881134,
+      "learning_rate": 8.788678701879573e-05,
+      "loss": 0.31,
+      "step": 266
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.36091452836990356,
+      "learning_rate": 8.778747871771292e-05,
+      "loss": 0.2373,
+      "step": 267
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4387395679950714,
+      "learning_rate": 8.768782159271727e-05,
+      "loss": 0.3354,
+      "step": 268
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3900839686393738,
+      "learning_rate": 8.758781656376398e-05,
+      "loss": 0.2464,
+      "step": 269
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4690368175506592,
+      "learning_rate": 8.748746455401986e-05,
+      "loss": 0.3327,
+      "step": 270
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4869069755077362,
+      "learning_rate": 8.738676648985476e-05,
+      "loss": 0.3151,
+      "step": 271
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.42331576347351074,
+      "learning_rate": 8.7285723300833e-05,
+      "loss": 0.3736,
+      "step": 272
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.39506635069847107,
+      "learning_rate": 8.718433591970485e-05,
+      "loss": 0.2875,
+      "step": 273
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4790605902671814,
+      "learning_rate": 8.708260528239788e-05,
+      "loss": 0.3883,
+      "step": 274
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.408903568983078,
+      "learning_rate": 8.698053232800832e-05,
+      "loss": 0.2666,
+      "step": 275
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.42906439304351807,
+      "learning_rate": 8.68781179987924e-05,
+      "loss": 0.361,
+      "step": 276
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3851872384548187,
+      "learning_rate": 8.677536324015765e-05,
+      "loss": 0.2827,
+      "step": 277
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.40262821316719055,
+      "learning_rate": 8.667226900065419e-05,
+      "loss": 0.3229,
+      "step": 278
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4792792797088623,
+      "learning_rate": 8.656883623196592e-05,
+      "loss": 0.3196,
+      "step": 279
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4359915256500244,
+      "learning_rate": 8.646506588890183e-05,
+      "loss": 0.303,
+      "step": 280
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.49512985348701477,
+      "learning_rate": 8.636095892938707e-05,
+      "loss": 0.3536,
+      "step": 281
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4611496925354004,
+      "learning_rate": 8.62565163144542e-05,
+      "loss": 0.4029,
+      "step": 282
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.40325888991355896,
+      "learning_rate": 8.615173900823426e-05,
+      "loss": 0.2887,
+      "step": 283
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.40099236369132996,
+      "learning_rate": 8.60466279779479e-05,
+      "loss": 0.2887,
+      "step": 284
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3835624158382416,
+      "learning_rate": 8.594118419389647e-05,
+      "loss": 0.33,
+      "step": 285
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.43427804112434387,
+      "learning_rate": 8.583540862945301e-05,
+      "loss": 0.3084,
+      "step": 286
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3395395278930664,
+      "learning_rate": 8.57293022610533e-05,
+      "loss": 0.2129,
+      "step": 287
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.40782567858695984,
+      "learning_rate": 8.562286606818684e-05,
+      "loss": 0.2399,
+      "step": 288
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5204586982727051,
+      "learning_rate": 8.55161010333878e-05,
+      "loss": 0.5042,
+      "step": 289
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5167766809463501,
+      "learning_rate": 8.540900814222598e-05,
+      "loss": 0.3576,
+      "step": 290
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.49364760518074036,
+      "learning_rate": 8.530158838329765e-05,
+      "loss": 0.3282,
+      "step": 291
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5482507348060608,
+      "learning_rate": 8.519384274821649e-05,
+      "loss": 0.4414,
+      "step": 292
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.39613139629364014,
+      "learning_rate": 8.508577223160442e-05,
+      "loss": 0.2545,
+      "step": 293
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4539114534854889,
+      "learning_rate": 8.497737783108238e-05,
+      "loss": 0.3221,
+      "step": 294
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.40794727206230164,
+      "learning_rate": 8.486866054726114e-05,
+      "loss": 0.2796,
+      "step": 295
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.40882447361946106,
+      "learning_rate": 8.475962138373213e-05,
+      "loss": 0.2639,
+      "step": 296
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4423540234565735,
+      "learning_rate": 8.465026134705805e-05,
+      "loss": 0.3054,
+      "step": 297
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4468096196651459,
+      "learning_rate": 8.454058144676366e-05,
+      "loss": 0.2876,
+      "step": 298
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.39803358912467957,
+      "learning_rate": 8.443058269532651e-05,
+      "loss": 0.3147,
+      "step": 299
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4247153103351593,
+      "learning_rate": 8.432026610816745e-05,
+      "loss": 0.2773,
+      "step": 300
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.700711965560913,
+      "learning_rate": 8.420963270364137e-05,
+      "loss": 0.3315,
+      "step": 301
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4842532277107239,
+      "learning_rate": 8.409868350302774e-05,
+      "loss": 0.3274,
+      "step": 302
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5232789516448975,
+      "learning_rate": 8.398741953052127e-05,
+      "loss": 0.2746,
+      "step": 303
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5666669011116028,
+      "learning_rate": 8.387584181322233e-05,
+      "loss": 0.304,
+      "step": 304
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5082411170005798,
+      "learning_rate": 8.376395138112754e-05,
+      "loss": 0.35,
+      "step": 305
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.41404369473457336,
+      "learning_rate": 8.365174926712032e-05,
+      "loss": 0.3421,
+      "step": 306
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.44390758872032166,
+      "learning_rate": 8.353923650696118e-05,
+      "loss": 0.3472,
+      "step": 307
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4442753195762634,
+      "learning_rate": 8.342641413927837e-05,
+      "loss": 0.3628,
+      "step": 308
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4544326961040497,
+      "learning_rate": 8.331328320555812e-05,
+      "loss": 0.3457,
+      "step": 309
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5052947402000427,
+      "learning_rate": 8.319984475013512e-05,
+      "loss": 0.3999,
+      "step": 310
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4428708553314209,
+      "learning_rate": 8.308609982018286e-05,
+      "loss": 0.3385,
+      "step": 311
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3840278089046478,
+      "learning_rate": 8.297204946570398e-05,
+      "loss": 0.2749,
+      "step": 312
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.42293739318847656,
+      "learning_rate": 8.285769473952052e-05,
+      "loss": 0.2628,
+      "step": 313
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6979223489761353,
+      "learning_rate": 8.274303669726426e-05,
+      "loss": 0.318,
+      "step": 314
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.45772454142570496,
+      "learning_rate": 8.262807639736692e-05,
+      "loss": 0.4072,
+      "step": 315
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.47757357358932495,
+      "learning_rate": 8.251281490105045e-05,
+      "loss": 0.3767,
+      "step": 316
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.38831740617752075,
+      "learning_rate": 8.239725327231721e-05,
+      "loss": 0.3062,
+      "step": 317
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.40905988216400146,
+      "learning_rate": 8.228139257794012e-05,
+      "loss": 0.3033,
+      "step": 318
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.42288145422935486,
+      "learning_rate": 8.216523388745287e-05,
+      "loss": 0.3485,
+      "step": 319
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4163537919521332,
+      "learning_rate": 8.204877827313997e-05,
+      "loss": 0.3293,
+      "step": 320
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3951532244682312,
+      "learning_rate": 8.193202681002692e-05,
+      "loss": 0.3387,
+      "step": 321
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2811068594455719,
+      "learning_rate": 8.181498057587027e-05,
+      "loss": 0.1691,
+      "step": 322
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3517349362373352,
+      "learning_rate": 8.169764065114764e-05,
+      "loss": 0.273,
+      "step": 323
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4180085062980652,
+      "learning_rate": 8.158000811904778e-05,
+      "loss": 0.2496,
+      "step": 324
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.41197821497917175,
+      "learning_rate": 8.146208406546053e-05,
+      "loss": 0.3327,
+      "step": 325
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4668791592121124,
+      "learning_rate": 8.134386957896688e-05,
+      "loss": 0.4066,
+      "step": 326
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4602564871311188,
+      "learning_rate": 8.122536575082882e-05,
+      "loss": 0.4185,
+      "step": 327
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4547500014305115,
+      "learning_rate": 8.110657367497933e-05,
+      "loss": 0.2798,
+      "step": 328
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.34806668758392334,
+      "learning_rate": 8.098749444801224e-05,
+      "loss": 0.235,
+      "step": 329
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4281531870365143,
+      "learning_rate": 8.08681291691722e-05,
+      "loss": 0.3517,
+      "step": 330
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.43151262402534485,
+      "learning_rate": 8.074847894034434e-05,
+      "loss": 0.2808,
+      "step": 331
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4350490868091583,
+      "learning_rate": 8.062854486604435e-05,
+      "loss": 0.2601,
+      "step": 332
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.39234495162963867,
+      "learning_rate": 8.050832805340806e-05,
+      "loss": 0.2806,
+      "step": 333
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4032362103462219,
+      "learning_rate": 8.038782961218136e-05,
+      "loss": 0.3327,
+      "step": 334
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5096771717071533,
+      "learning_rate": 8.026705065470996e-05,
+      "loss": 0.3412,
+      "step": 335
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4058472514152527,
+      "learning_rate": 8.014599229592894e-05,
+      "loss": 0.3071,
+      "step": 336
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.454199880361557,
+      "learning_rate": 8.002465565335271e-05,
+      "loss": 0.2554,
+      "step": 337
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4330787658691406,
+      "learning_rate": 7.990304184706455e-05,
+      "loss": 0.2798,
+      "step": 338
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.42377325892448425,
+      "learning_rate": 7.978115199970621e-05,
+      "loss": 0.3761,
+      "step": 339
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3846011459827423,
+      "learning_rate": 7.965898723646776e-05,
+      "loss": 0.301,
+      "step": 340
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5546414256095886,
+      "learning_rate": 7.953654868507699e-05,
+      "loss": 0.4261,
+      "step": 341
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4243573248386383,
+      "learning_rate": 7.941383747578912e-05,
+      "loss": 0.338,
+      "step": 342
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.41652941703796387,
+      "learning_rate": 7.929085474137629e-05,
+      "loss": 0.3427,
+      "step": 343
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.36712539196014404,
+      "learning_rate": 7.91676016171172e-05,
+      "loss": 0.3195,
+      "step": 344
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4737776815891266,
+      "learning_rate": 7.904407924078654e-05,
+      "loss": 0.4586,
+      "step": 345
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4451066553592682,
+      "learning_rate": 7.892028875264451e-05,
+      "loss": 0.3344,
+      "step": 346
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3951125144958496,
+      "learning_rate": 7.879623129542633e-05,
+      "loss": 0.2924,
+      "step": 347
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.42933401465415955,
+      "learning_rate": 7.867190801433166e-05,
+      "loss": 0.3728,
+      "step": 348
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3992808759212494,
+      "learning_rate": 7.854732005701402e-05,
+      "loss": 0.3446,
+      "step": 349
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.39772018790245056,
+      "learning_rate": 7.842246857357023e-05,
+      "loss": 0.2887,
+      "step": 350
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.33922725915908813,
+      "learning_rate": 7.829735471652978e-05,
+      "loss": 0.2267,
+      "step": 351
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4072892367839813,
+      "learning_rate": 7.817197964084411e-05,
+      "loss": 0.2689,
+      "step": 352
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.32967281341552734,
+      "learning_rate": 7.804634450387616e-05,
+      "loss": 0.1847,
+      "step": 353
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.42083194851875305,
+      "learning_rate": 7.792045046538941e-05,
+      "loss": 0.3112,
+      "step": 354
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8271592855453491,
+      "learning_rate": 7.77942986875374e-05,
+      "loss": 0.434,
+      "step": 355
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.40372174978256226,
+      "learning_rate": 7.766789033485287e-05,
+      "loss": 0.3318,
+      "step": 356
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.43399181962013245,
+      "learning_rate": 7.75412265742371e-05,
+      "loss": 0.3854,
+      "step": 357
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4727342128753662,
+      "learning_rate": 7.741430857494904e-05,
+      "loss": 0.318,
+      "step": 358
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.43026426434516907,
+      "learning_rate": 7.728713750859458e-05,
+      "loss": 0.3445,
+      "step": 359
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4170074164867401,
+      "learning_rate": 7.715971454911577e-05,
+      "loss": 0.3448,
+      "step": 360
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.44696909189224243,
+      "learning_rate": 7.703204087277988e-05,
+      "loss": 0.3512,
+      "step": 361
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.36404502391815186,
+      "learning_rate": 7.690411765816864e-05,
+      "loss": 0.2587,
+      "step": 362
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4109265208244324,
+      "learning_rate": 7.677594608616729e-05,
+      "loss": 0.307,
+      "step": 363
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3944926857948303,
+      "learning_rate": 7.66475273399537e-05,
+      "loss": 0.2815,
+      "step": 364
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.39314037561416626,
+      "learning_rate": 7.651886260498751e-05,
+      "loss": 0.269,
+      "step": 365
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4431683421134949,
+      "learning_rate": 7.638995306899908e-05,
+      "loss": 0.3221,
+      "step": 366
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3648792505264282,
+      "learning_rate": 7.626079992197857e-05,
+      "loss": 0.278,
+      "step": 367
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.42419299483299255,
+      "learning_rate": 7.613140435616503e-05,
+      "loss": 0.3199,
+      "step": 368
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4009580612182617,
+      "learning_rate": 7.600176756603525e-05,
+      "loss": 0.2531,
+      "step": 369
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3897629380226135,
+      "learning_rate": 7.587189074829284e-05,
+      "loss": 0.3039,
+      "step": 370
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.33926069736480713,
+      "learning_rate": 7.57417751018572e-05,
+      "loss": 0.2319,
+      "step": 371
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.37864574790000916,
+      "learning_rate": 7.561142182785233e-05,
+      "loss": 0.2429,
+      "step": 372
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4201953113079071,
+      "learning_rate": 7.548083212959588e-05,
+      "loss": 0.3103,
+      "step": 373
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.44398027658462524,
+      "learning_rate": 7.535000721258791e-05,
+      "loss": 0.2644,
+      "step": 374
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.43067774176597595,
+      "learning_rate": 7.521894828449994e-05,
+      "loss": 0.3685,
+      "step": 375
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4563278555870056,
+      "learning_rate": 7.508765655516358e-05,
+      "loss": 0.3404,
+      "step": 376
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3916566073894501,
+      "learning_rate": 7.495613323655953e-05,
+      "loss": 0.3062,
+      "step": 377
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4126124680042267,
+      "learning_rate": 7.482437954280635e-05,
+      "loss": 0.2608,
+      "step": 378
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.44294944405555725,
+      "learning_rate": 7.469239669014923e-05,
+      "loss": 0.3678,
+      "step": 379
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4036175608634949,
+      "learning_rate": 7.456018589694873e-05,
+      "loss": 0.2877,
+      "step": 380
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5387861728668213,
+      "learning_rate": 7.442774838366965e-05,
+      "loss": 0.3948,
+      "step": 381
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.28586551547050476,
+      "learning_rate": 7.429508537286963e-05,
+      "loss": 0.1913,
+      "step": 382
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.43656590580940247,
+      "learning_rate": 7.416219808918794e-05,
+      "loss": 0.3534,
+      "step": 383
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4448322653770447,
+      "learning_rate": 7.402908775933419e-05,
+      "loss": 0.3686,
+      "step": 384
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.45444345474243164,
+      "learning_rate": 7.389575561207692e-05,
+      "loss": 0.4593,
+      "step": 385
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4999052882194519,
+      "learning_rate": 7.376220287823236e-05,
+      "loss": 0.3738,
+      "step": 386
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4560929536819458,
+      "learning_rate": 7.3628430790653e-05,
+      "loss": 0.3286,
+      "step": 387
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4010728597640991,
+      "learning_rate": 7.349444058421619e-05,
+      "loss": 0.2886,
+      "step": 388
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.39646634459495544,
+      "learning_rate": 7.336023349581287e-05,
+      "loss": 0.2403,
+      "step": 389
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.36710843443870544,
+      "learning_rate": 7.322581076433596e-05,
+      "loss": 0.2863,
+      "step": 390
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.49684014916419983,
+      "learning_rate": 7.309117363066912e-05,
+      "loss": 0.4096,
+      "step": 391
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.42137226462364197,
+      "learning_rate": 7.295632333767513e-05,
+      "loss": 0.3551,
+      "step": 392
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3678036630153656,
+      "learning_rate": 7.28212611301845e-05,
+      "loss": 0.2341,
+      "step": 393
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.43836355209350586,
+      "learning_rate": 7.2685988254984e-05,
+      "loss": 0.364,
+      "step": 394
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4423292279243469,
+      "learning_rate": 7.255050596080509e-05,
+      "loss": 0.3527,
+      "step": 395
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2825776934623718,
+      "learning_rate": 7.241481549831243e-05,
+      "loss": 0.1635,
+      "step": 396
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4246549606323242,
+      "learning_rate": 7.22789181200923e-05,
+      "loss": 0.3436,
+      "step": 397
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4510544240474701,
+      "learning_rate": 7.214281508064107e-05,
+      "loss": 0.3554,
+      "step": 398
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.36521950364112854,
+      "learning_rate": 7.200650763635366e-05,
+      "loss": 0.2404,
+      "step": 399
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.45220625400543213,
+      "learning_rate": 7.186999704551181e-05,
+      "loss": 0.3145,
+      "step": 400
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.44834861159324646,
+      "learning_rate": 7.173328456827263e-05,
+      "loss": 0.3142,
+      "step": 401
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.31816065311431885,
+      "learning_rate": 7.15963714666568e-05,
+      "loss": 0.1982,
+      "step": 402
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.424333781003952,
+      "learning_rate": 7.145925900453709e-05,
+      "loss": 0.3539,
+      "step": 403
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3757897913455963,
+      "learning_rate": 7.132194844762654e-05,
+      "loss": 0.2778,
+      "step": 404
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4314412474632263,
+      "learning_rate": 7.118444106346687e-05,
+      "loss": 0.3578,
+      "step": 405
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.42340290546417236,
+      "learning_rate": 7.104673812141675e-05,
+      "loss": 0.3659,
+      "step": 406
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4033401906490326,
+      "learning_rate": 7.090884089264011e-05,
+      "loss": 0.2682,
+      "step": 407
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.35764697194099426,
+      "learning_rate": 7.077075065009433e-05,
+      "loss": 0.231,
+      "step": 408
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.43582141399383545,
+      "learning_rate": 7.063246866851858e-05,
+      "loss": 0.2786,
+      "step": 409
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.529912531375885,
+      "learning_rate": 7.049399622442198e-05,
+      "loss": 0.3742,
+      "step": 410
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4656778872013092,
+      "learning_rate": 7.035533459607189e-05,
+      "loss": 0.3333,
+      "step": 411
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.40634432435035706,
+      "learning_rate": 7.021648506348204e-05,
+      "loss": 0.3145,
+      "step": 412
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4188442826271057,
+      "learning_rate": 7.007744890840073e-05,
+      "loss": 0.2721,
+      "step": 413
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.44638150930404663,
+      "learning_rate": 6.993822741429907e-05,
+      "loss": 0.3794,
+      "step": 414
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.36345523595809937,
+      "learning_rate": 6.979882186635897e-05,
+      "loss": 0.2157,
+      "step": 415
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.45502936840057373,
+      "learning_rate": 6.965923355146147e-05,
+      "loss": 0.3469,
+      "step": 416
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.37767454981803894,
+      "learning_rate": 6.951946375817474e-05,
+      "loss": 0.2824,
+      "step": 417
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4633556008338928,
+      "learning_rate": 6.937951377674221e-05,
+      "loss": 0.2924,
+      "step": 418
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.43885359168052673,
+      "learning_rate": 6.923938489907066e-05,
+      "loss": 0.3259,
+      "step": 419
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4263167083263397,
+      "learning_rate": 6.909907841871829e-05,
+      "loss": 0.2902,
+      "step": 420
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.42423805594444275,
+      "learning_rate": 6.895859563088283e-05,
+      "loss": 0.3641,
+      "step": 421
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3689592778682709,
+      "learning_rate": 6.881793783238948e-05,
+      "loss": 0.2492,
+      "step": 422
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.47483524680137634,
+      "learning_rate": 6.867710632167903e-05,
+      "loss": 0.3558,
+      "step": 423
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.36466506123542786,
+      "learning_rate": 6.853610239879586e-05,
+      "loss": 0.2739,
+      "step": 424
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.35030409693717957,
+      "learning_rate": 6.839492736537588e-05,
+      "loss": 0.1946,
+      "step": 425
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.46116241812705994,
+      "learning_rate": 6.82535825246346e-05,
+      "loss": 0.2689,
+      "step": 426
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.47451573610305786,
+      "learning_rate": 6.811206918135502e-05,
+      "loss": 0.3971,
+      "step": 427
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5025789141654968,
+      "learning_rate": 6.797038864187564e-05,
+      "loss": 0.351,
+      "step": 428
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.37761175632476807,
+      "learning_rate": 6.782854221407838e-05,
+      "loss": 0.2974,
+      "step": 429
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.44670093059539795,
+      "learning_rate": 6.768653120737652e-05,
+      "loss": 0.3839,
+      "step": 430
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4108104109764099,
+      "learning_rate": 6.754435693270258e-05,
+      "loss": 0.2815,
+      "step": 431
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.45673632621765137,
+      "learning_rate": 6.740202070249623e-05,
+      "loss": 0.3454,
+      "step": 432
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4945111572742462,
+      "learning_rate": 6.725952383069222e-05,
+      "loss": 0.4538,
+      "step": 433
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.43138283491134644,
+      "learning_rate": 6.711686763270818e-05,
+      "loss": 0.3702,
+      "step": 434
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4254596531391144,
+      "learning_rate": 6.697405342543258e-05,
+      "loss": 0.3817,
+      "step": 435
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.36549097299575806,
+      "learning_rate": 6.683108252721238e-05,
+      "loss": 0.3039,
+      "step": 436
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4733044505119324,
+      "learning_rate": 6.668795625784113e-05,
+      "loss": 0.4458,
+      "step": 437
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.35971495509147644,
+      "learning_rate": 6.654467593854657e-05,
+      "loss": 0.2318,
+      "step": 438
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4079780578613281,
+      "learning_rate": 6.640124289197845e-05,
+      "loss": 0.3086,
+      "step": 439
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.42413631081581116,
+      "learning_rate": 6.625765844219652e-05,
+      "loss": 0.3687,
+      "step": 440
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.39297375082969666,
+      "learning_rate": 6.611392391465802e-05,
+      "loss": 0.297,
+      "step": 441
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4315466284751892,
+      "learning_rate": 6.597004063620567e-05,
+      "loss": 0.3317,
+      "step": 442
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.29865145683288574,
+      "learning_rate": 6.582600993505534e-05,
+      "loss": 0.1866,
+      "step": 443
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.47077304124832153,
+      "learning_rate": 6.568183314078376e-05,
+      "loss": 0.4135,
+      "step": 444
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3485214412212372,
+      "learning_rate": 6.553751158431627e-05,
+      "loss": 0.2308,
+      "step": 445
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3782065808773041,
+      "learning_rate": 6.539304659791456e-05,
+      "loss": 0.2667,
+      "step": 446
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3845309019088745,
+      "learning_rate": 6.524843951516434e-05,
+      "loss": 0.2734,
+      "step": 447
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4229350686073303,
+      "learning_rate": 6.510369167096308e-05,
+      "loss": 0.2934,
+      "step": 448
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3936381936073303,
+      "learning_rate": 6.495880440150756e-05,
+      "loss": 0.2677,
+      "step": 449
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.39204105734825134,
+      "learning_rate": 6.481377904428171e-05,
+      "loss": 0.26,
+      "step": 450
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.41730165481567383,
+      "learning_rate": 6.466861693804413e-05,
+      "loss": 0.3263,
+      "step": 451
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4841783344745636,
+      "learning_rate": 6.45233194228158e-05,
+      "loss": 0.4078,
+      "step": 452
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.43934983015060425,
+      "learning_rate": 6.437788783986766e-05,
+      "loss": 0.2471,
+      "step": 453
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.43099477887153625,
+      "learning_rate": 6.42323235317083e-05,
+      "loss": 0.3084,
+      "step": 454
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4412413537502289,
+      "learning_rate": 6.408662784207149e-05,
+      "loss": 0.291,
+      "step": 455
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4185587763786316,
+      "learning_rate": 6.394080211590381e-05,
+      "loss": 0.2864,
+      "step": 456
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.35048070549964905,
+      "learning_rate": 6.379484769935223e-05,
+      "loss": 0.2278,
+      "step": 457
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4162713587284088,
+      "learning_rate": 6.364876593975173e-05,
+      "loss": 0.2714,
+      "step": 458
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.34501340985298157,
+      "learning_rate": 6.350255818561277e-05,
+      "loss": 0.2115,
+      "step": 459
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4311944246292114,
+      "learning_rate": 6.335622578660889e-05,
+      "loss": 0.3105,
+      "step": 460
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3668000102043152,
+      "learning_rate": 6.320977009356431e-05,
+      "loss": 0.2445,
+      "step": 461
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.49384593963623047,
+      "learning_rate": 6.306319245844133e-05,
+      "loss": 0.3664,
+      "step": 462
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4154500663280487,
+      "learning_rate": 6.291649423432799e-05,
+      "loss": 0.2472,
+      "step": 463
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3742479681968689,
+      "learning_rate": 6.276967677542542e-05,
+      "loss": 0.2682,
+      "step": 464
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.41379591822624207,
+      "learning_rate": 6.262274143703554e-05,
+      "loss": 0.3027,
+      "step": 465
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.42273882031440735,
+      "learning_rate": 6.24756895755484e-05,
+      "loss": 0.2976,
+      "step": 466
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.36084771156311035,
+      "learning_rate": 6.232852254842962e-05,
+      "loss": 0.2578,
+      "step": 467
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4634334444999695,
+      "learning_rate": 6.218124171420806e-05,
+      "loss": 0.3502,
+      "step": 468
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3936538100242615,
+      "learning_rate": 6.203384843246307e-05,
+      "loss": 0.3405,
+      "step": 469
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3866216838359833,
+      "learning_rate": 6.188634406381207e-05,
+      "loss": 0.33,
+      "step": 470
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.48050788044929504,
+      "learning_rate": 6.173872996989793e-05,
+      "loss": 0.3355,
+      "step": 471
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3211848735809326,
+      "learning_rate": 6.159100751337642e-05,
+      "loss": 0.204,
+      "step": 472
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3827769160270691,
+      "learning_rate": 6.144317805790361e-05,
+      "loss": 0.2832,
+      "step": 473
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3828567564487457,
+      "learning_rate": 6.129524296812335e-05,
+      "loss": 0.252,
+      "step": 474
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3291488289833069,
+      "learning_rate": 6.114720360965453e-05,
+      "loss": 0.1864,
+      "step": 475
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4140992760658264,
+      "learning_rate": 6.099906134907868e-05,
+      "loss": 0.2806,
+      "step": 476
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.436565101146698,
+      "learning_rate": 6.085081755392714e-05,
+      "loss": 0.3529,
+      "step": 477
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4582252502441406,
+      "learning_rate": 6.07024735926686e-05,
+      "loss": 0.3275,
+      "step": 478
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4374159276485443,
+      "learning_rate": 6.055403083469637e-05,
+      "loss": 0.3755,
+      "step": 479
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.489577978849411,
+      "learning_rate": 6.04054906503158e-05,
+      "loss": 0.408,
+      "step": 480
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.46644705533981323,
+      "learning_rate": 6.0256854410731565e-05,
+      "loss": 0.3451,
+      "step": 481
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.46568021178245544,
+      "learning_rate": 6.010812348803509e-05,
+      "loss": 0.3739,
+      "step": 482
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.45645302534103394,
+      "learning_rate": 5.99592992551918e-05,
+      "loss": 0.4199,
+      "step": 483
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.416567325592041,
+      "learning_rate": 5.9810383086028535e-05,
+      "loss": 0.3734,
+      "step": 484
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3670683801174164,
+      "learning_rate": 5.9661376355220734e-05,
+      "loss": 0.2927,
+      "step": 485
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.47614961862564087,
+      "learning_rate": 5.9512280438279914e-05,
+      "loss": 0.414,
+      "step": 486
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3908349871635437,
+      "learning_rate": 5.936309671154084e-05,
+      "loss": 0.337,
+      "step": 487
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4337731599807739,
+      "learning_rate": 5.9213826552148886e-05,
+      "loss": 0.3428,
+      "step": 488
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.40524789690971375,
+      "learning_rate": 5.906447133804731e-05,
+      "loss": 0.3534,
+      "step": 489
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3826107382774353,
+      "learning_rate": 5.891503244796448e-05,
+      "loss": 0.2569,
+      "step": 490
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3548087775707245,
+      "learning_rate": 5.8765511261401254e-05,
+      "loss": 0.2976,
+      "step": 491
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3621930778026581,
+      "learning_rate": 5.861590915861817e-05,
+      "loss": 0.2355,
+      "step": 492
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3902803361415863,
+      "learning_rate": 5.846622752062268e-05,
+      "loss": 0.2642,
+      "step": 493
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4269382059574127,
+      "learning_rate": 5.831646772915651e-05,
+      "loss": 0.3481,
+      "step": 494
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4209762513637543,
+      "learning_rate": 5.816663116668276e-05,
+      "loss": 0.303,
+      "step": 495
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.39953798055648804,
+      "learning_rate": 5.801671921637328e-05,
+      "loss": 0.3168,
+      "step": 496
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4956504702568054,
+      "learning_rate": 5.786673326209584e-05,
+      "loss": 0.404,
+      "step": 497
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.38132691383361816,
+      "learning_rate": 5.7716674688401286e-05,
+      "loss": 0.2658,
+      "step": 498
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4979095458984375,
+      "learning_rate": 5.756654488051091e-05,
+      "loss": 0.391,
+      "step": 499
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4013850688934326,
+      "learning_rate": 5.7416345224303524e-05,
+      "loss": 0.2979,
+      "step": 500
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3869968056678772,
+      "learning_rate": 5.7266077106302785e-05,
+      "loss": 0.2835,
+      "step": 501
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3484272062778473,
+      "learning_rate": 5.7115741913664264e-05,
+      "loss": 0.2679,
+      "step": 502
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3580166697502136,
+      "learning_rate": 5.696534103416276e-05,
+      "loss": 0.2667,
+      "step": 503
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4448302090167999,
+      "learning_rate": 5.6814875856179414e-05,
+      "loss": 0.3617,
+      "step": 504
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5003171563148499,
+      "learning_rate": 5.666434776868895e-05,
+      "loss": 0.3631,
+      "step": 505
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.46512144804000854,
+      "learning_rate": 5.651375816124679e-05,
+      "loss": 0.3196,
+      "step": 506
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4460465610027313,
+      "learning_rate": 5.636310842397629e-05,
+      "loss": 0.3175,
+      "step": 507
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.43357494473457336,
+      "learning_rate": 5.621239994755583e-05,
+      "loss": 0.3122,
+      "step": 508
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4276319146156311,
+      "learning_rate": 5.606163412320608e-05,
+      "loss": 0.2931,
+      "step": 509
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4117043614387512,
+      "learning_rate": 5.5910812342677065e-05,
+      "loss": 0.2695,
+      "step": 510
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3576218783855438,
+      "learning_rate": 5.575993599823536e-05,
+      "loss": 0.2507,
+      "step": 511
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.35004672408103943,
+      "learning_rate": 5.560900648265124e-05,
+      "loss": 0.1801,
+      "step": 512
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.49767759442329407,
+      "learning_rate": 5.545802518918579e-05,
+      "loss": 0.4032,
+      "step": 513
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5268928408622742,
+      "learning_rate": 5.5306993511578096e-05,
+      "loss": 0.4158,
+      "step": 514
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.46446239948272705,
+      "learning_rate": 5.515591284403234e-05,
+      "loss": 0.3088,
+      "step": 515
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.39802390336990356,
+      "learning_rate": 5.5004784581204927e-05,
+      "loss": 0.301,
+      "step": 516
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3617722690105438,
+      "learning_rate": 5.485361011819164e-05,
+      "loss": 0.2405,
+      "step": 517
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4040432274341583,
+      "learning_rate": 5.4702390850514726e-05,
+      "loss": 0.2692,
+      "step": 518
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.41277679800987244,
+      "learning_rate": 5.455112817411006e-05,
+      "loss": 0.3128,
+      "step": 519
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.41128095984458923,
+      "learning_rate": 5.4399823485314226e-05,
+      "loss": 0.2589,
+      "step": 520
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.44558587670326233,
+      "learning_rate": 5.4248478180851604e-05,
+      "loss": 0.3861,
+      "step": 521
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.36561065912246704,
+      "learning_rate": 5.409709365782154e-05,
+      "loss": 0.2454,
+      "step": 522
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.46217024326324463,
+      "learning_rate": 5.3945671313685386e-05,
+      "loss": 0.3887,
+      "step": 523
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.37300512194633484,
+      "learning_rate": 5.379421254625366e-05,
+      "loss": 0.2722,
+      "step": 524
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.39892271161079407,
+      "learning_rate": 5.364271875367311e-05,
+      "loss": 0.2965,
+      "step": 525
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.43289822340011597,
+      "learning_rate": 5.3491191334413746e-05,
+      "loss": 0.3766,
+      "step": 526
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.37322935461997986,
+      "learning_rate": 5.3339631687256084e-05,
+      "loss": 0.2401,
+      "step": 527
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.401250422000885,
+      "learning_rate": 5.318804121127807e-05,
+      "loss": 0.2671,
+      "step": 528
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.45596757531166077,
+      "learning_rate": 5.3036421305842276e-05,
+      "loss": 0.3871,
+      "step": 529
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.35917335748672485,
+      "learning_rate": 5.288477337058293e-05,
+      "loss": 0.2398,
+      "step": 530
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4593126177787781,
+      "learning_rate": 5.273309880539301e-05,
+      "loss": 0.3669,
+      "step": 531
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5227428078651428,
+      "learning_rate": 5.258139901041132e-05,
+      "loss": 0.3048,
+      "step": 532
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.31588104367256165,
+      "learning_rate": 5.242967538600957e-05,
+      "loss": 0.1823,
+      "step": 533
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4211719036102295,
+      "learning_rate": 5.227792933277943e-05,
+      "loss": 0.3316,
+      "step": 534
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3306739926338196,
+      "learning_rate": 5.212616225151965e-05,
+      "loss": 0.2222,
+      "step": 535
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.43753111362457275,
+      "learning_rate": 5.197437554322304e-05,
+      "loss": 0.4076,
+      "step": 536
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4082089364528656,
+      "learning_rate": 5.182257060906365e-05,
+      "loss": 0.2929,
+      "step": 537
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4506302773952484,
+      "learning_rate": 5.167074885038373e-05,
+      "loss": 0.3252,
+      "step": 538
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.35362350940704346,
+      "learning_rate": 5.151891166868086e-05,
+      "loss": 0.2435,
+      "step": 539
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5236488580703735,
+      "learning_rate": 5.1367060465595006e-05,
+      "loss": 0.4728,
+      "step": 540
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3949301242828369,
+      "learning_rate": 5.121519664289553e-05,
+      "loss": 0.2987,
+      "step": 541
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4690653383731842,
+      "learning_rate": 5.106332160246834e-05,
+      "loss": 0.409,
+      "step": 542
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3642612099647522,
+      "learning_rate": 5.0911436746302834e-05,
+      "loss": 0.2921,
+      "step": 543
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.372464120388031,
+      "learning_rate": 5.075954347647909e-05,
+      "loss": 0.2481,
+      "step": 544
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.37555721402168274,
+      "learning_rate": 5.0607643195154796e-05,
+      "loss": 0.2836,
+      "step": 545
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3466416001319885,
+      "learning_rate": 5.045573730455241e-05,
+      "loss": 0.2854,
+      "step": 546
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3371579349040985,
+      "learning_rate": 5.030382720694612e-05,
+      "loss": 0.2514,
+      "step": 547
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.42777353525161743,
+      "learning_rate": 5.0151914304649015e-05,
+      "loss": 0.3554,
+      "step": 548
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.34071090817451477,
+      "learning_rate": 5e-05,
+      "loss": 0.2285,
+      "step": 549
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.43033647537231445,
+      "learning_rate": 4.984808569535101e-05,
+      "loss": 0.3356,
+      "step": 550
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.40992996096611023,
+      "learning_rate": 4.969617279305388e-05,
+      "loss": 0.3154,
+      "step": 551
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3767620623111725,
+      "learning_rate": 4.954426269544761e-05,
+      "loss": 0.2635,
+      "step": 552
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.40804165601730347,
+      "learning_rate": 4.939235680484522e-05,
+      "loss": 0.3069,
+      "step": 553
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3589411675930023,
+      "learning_rate": 4.924045652352092e-05,
+      "loss": 0.24,
+      "step": 554
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.38028955459594727,
+      "learning_rate": 4.908856325369718e-05,
+      "loss": 0.2939,
+      "step": 555
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3343428671360016,
+      "learning_rate": 4.893667839753168e-05,
+      "loss": 0.1953,
+      "step": 556
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4250152111053467,
+      "learning_rate": 4.878480335710448e-05,
+      "loss": 0.357,
+      "step": 557
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.2950625419616699,
+      "learning_rate": 4.8632939534405006e-05,
+      "loss": 0.1712,
+      "step": 558
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3887781798839569,
+      "learning_rate": 4.8481088331319146e-05,
+      "loss": 0.3143,
+      "step": 559
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.35813620686531067,
+      "learning_rate": 4.832925114961629e-05,
+      "loss": 0.2471,
+      "step": 560
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4876914322376251,
+      "learning_rate": 4.817742939093635e-05,
+      "loss": 0.4083,
+      "step": 561
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.43107870221138,
+      "learning_rate": 4.8025624456776966e-05,
+      "loss": 0.273,
+      "step": 562
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3670273423194885,
+      "learning_rate": 4.787383774848037e-05,
+      "loss": 0.2949,
+      "step": 563
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.47010329365730286,
+      "learning_rate": 4.772207066722057e-05,
+      "loss": 0.3449,
+      "step": 564
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4678891897201538,
+      "learning_rate": 4.757032461399044e-05,
+      "loss": 0.3232,
+      "step": 565
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.38481488823890686,
+      "learning_rate": 4.7418600989588694e-05,
+      "loss": 0.2349,
+      "step": 566
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3220939636230469,
+      "learning_rate": 4.726690119460701e-05,
+      "loss": 0.191,
+      "step": 567
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.39618757367134094,
+      "learning_rate": 4.7115226629417075e-05,
+      "loss": 0.3071,
+      "step": 568
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4422629773616791,
+      "learning_rate": 4.6963578694157736e-05,
+      "loss": 0.2635,
+      "step": 569
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4440760314464569,
+      "learning_rate": 4.681195878872194e-05,
+      "loss": 0.3334,
+      "step": 570
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3999294340610504,
+      "learning_rate": 4.666036831274392e-05,
+      "loss": 0.2625,
+      "step": 571
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.43602943420410156,
+      "learning_rate": 4.6508808665586265e-05,
+      "loss": 0.3042,
+      "step": 572
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.33296841382980347,
+      "learning_rate": 4.635728124632692e-05,
+      "loss": 0.2561,
+      "step": 573
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.43026474118232727,
+      "learning_rate": 4.6205787453746336e-05,
+      "loss": 0.284,
+      "step": 574
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.34284353256225586,
+      "learning_rate": 4.605432868631462e-05,
+      "loss": 0.2312,
+      "step": 575
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5019268989562988,
+      "learning_rate": 4.590290634217848e-05,
+      "loss": 0.3643,
+      "step": 576
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4448760151863098,
+      "learning_rate": 4.57515218191484e-05,
+      "loss": 0.3368,
+      "step": 577
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3717033863067627,
+      "learning_rate": 4.5600176514685786e-05,
+      "loss": 0.2658,
+      "step": 578
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4158264994621277,
+      "learning_rate": 4.5448871825889946e-05,
+      "loss": 0.3105,
+      "step": 579
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4152980446815491,
+      "learning_rate": 4.52976091494853e-05,
+      "loss": 0.2573,
+      "step": 580
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3900246322154999,
+      "learning_rate": 4.514638988180837e-05,
+      "loss": 0.2733,
+      "step": 581
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4747683107852936,
+      "learning_rate": 4.4995215418795085e-05,
+      "loss": 0.3899,
+      "step": 582
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4682185649871826,
+      "learning_rate": 4.484408715596768e-05,
+      "loss": 0.301,
+      "step": 583
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.41452059149742126,
+      "learning_rate": 4.4693006488421915e-05,
+      "loss": 0.2765,
+      "step": 584
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4349689185619354,
+      "learning_rate": 4.454197481081422e-05,
+      "loss": 0.2998,
+      "step": 585
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.39544400572776794,
+      "learning_rate": 4.439099351734878e-05,
+      "loss": 0.2777,
+      "step": 586
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4015561044216156,
+      "learning_rate": 4.4240064001764646e-05,
+      "loss": 0.2579,
+      "step": 587
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.353808730840683,
+      "learning_rate": 4.4089187657322953e-05,
+      "loss": 0.2631,
+      "step": 588
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.35133761167526245,
+      "learning_rate": 4.393836587679394e-05,
+      "loss": 0.2423,
+      "step": 589
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.47139087319374084,
+      "learning_rate": 4.3787600052444174e-05,
+      "loss": 0.3469,
+      "step": 590
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.47959962487220764,
+      "learning_rate": 4.363689157602373e-05,
+      "loss": 0.4231,
+      "step": 591
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3742324709892273,
+      "learning_rate": 4.348624183875322e-05,
+      "loss": 0.2906,
+      "step": 592
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3570776879787445,
+      "learning_rate": 4.333565223131107e-05,
+      "loss": 0.2154,
+      "step": 593
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.42536646127700806,
+      "learning_rate": 4.318512414382059e-05,
+      "loss": 0.3376,
+      "step": 594
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.40395909547805786,
+      "learning_rate": 4.3034658965837255e-05,
+      "loss": 0.1824,
+      "step": 595
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.42154961824417114,
+      "learning_rate": 4.288425808633575e-05,
+      "loss": 0.3297,
+      "step": 596
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3924674391746521,
+      "learning_rate": 4.273392289369723e-05,
+      "loss": 0.277,
+      "step": 597
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.36088064312934875,
+      "learning_rate": 4.258365477569648e-05,
+      "loss": 0.2243,
+      "step": 598
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4456118941307068,
+      "learning_rate": 4.2433455119489105e-05,
+      "loss": 0.317,
+      "step": 599
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.44901955127716064,
+      "learning_rate": 4.228332531159871e-05,
+      "loss": 0.3589,
+      "step": 600
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.47233232855796814,
+      "learning_rate": 4.2133266737904176e-05,
+      "loss": 0.2958,
+      "step": 601
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.43617701530456543,
+      "learning_rate": 4.1983280783626724e-05,
+      "loss": 0.3634,
+      "step": 602
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4319142699241638,
+      "learning_rate": 4.183336883331723e-05,
+      "loss": 0.3653,
+      "step": 603
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.43837970495224,
+      "learning_rate": 4.1683532270843504e-05,
+      "loss": 0.3446,
+      "step": 604
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.41829127073287964,
+      "learning_rate": 4.153377247937732e-05,
+      "loss": 0.3062,
+      "step": 605
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.41083550453186035,
+      "learning_rate": 4.138409084138185e-05,
+      "loss": 0.3098,
+      "step": 606
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4804052412509918,
+      "learning_rate": 4.1234488738598744e-05,
+      "loss": 0.2789,
+      "step": 607
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3317751884460449,
+      "learning_rate": 4.108496755203553e-05,
+      "loss": 0.2116,
+      "step": 608
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.40524640679359436,
+      "learning_rate": 4.0935528661952716e-05,
+      "loss": 0.3023,
+      "step": 609
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.41121968626976013,
+      "learning_rate": 4.0786173447851126e-05,
+      "loss": 0.3146,
+      "step": 610
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.385101854801178,
+      "learning_rate": 4.063690328845916e-05,
+      "loss": 0.2655,
+      "step": 611
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3425019681453705,
+      "learning_rate": 4.04877195617201e-05,
+      "loss": 0.2553,
+      "step": 612
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4267544448375702,
+      "learning_rate": 4.033862364477927e-05,
+      "loss": 0.3549,
+      "step": 613
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3356296420097351,
+      "learning_rate": 4.0189616913971484e-05,
+      "loss": 0.2249,
+      "step": 614
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4114169478416443,
+      "learning_rate": 4.0040700744808204e-05,
+      "loss": 0.3824,
+      "step": 615
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3610970675945282,
+      "learning_rate": 3.989187651196493e-05,
+      "loss": 0.2309,
+      "step": 616
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4254869520664215,
+      "learning_rate": 3.974314558926844e-05,
+      "loss": 0.3008,
+      "step": 617
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.36089685559272766,
+      "learning_rate": 3.9594509349684216e-05,
+      "loss": 0.2992,
+      "step": 618
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.40409085154533386,
+      "learning_rate": 3.9445969165303647e-05,
+      "loss": 0.2812,
+      "step": 619
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.458126425743103,
+      "learning_rate": 3.929752640733141e-05,
+      "loss": 0.3108,
+      "step": 620
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.41229763627052307,
+      "learning_rate": 3.914918244607287e-05,
+      "loss": 0.2853,
+      "step": 621
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3905296325683594,
+      "learning_rate": 3.900093865092134e-05,
+      "loss": 0.2386,
+      "step": 622
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3707413077354431,
+      "learning_rate": 3.885279639034546e-05,
+      "loss": 0.2428,
+      "step": 623
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4487210810184479,
+      "learning_rate": 3.870475703187667e-05,
+      "loss": 0.2722,
+      "step": 624
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.45813697576522827,
+      "learning_rate": 3.855682194209639e-05,
+      "loss": 0.2897,
+      "step": 625
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.31196317076683044,
+      "learning_rate": 3.840899248662358e-05,
+      "loss": 0.2186,
+      "step": 626
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.40307724475860596,
+      "learning_rate": 3.8261270030102084e-05,
+      "loss": 0.2858,
+      "step": 627
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4220031797885895,
+      "learning_rate": 3.8113655936187947e-05,
+      "loss": 0.3035,
+      "step": 628
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.39654064178466797,
+      "learning_rate": 3.796615156753696e-05,
+      "loss": 0.3392,
+      "step": 629
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4065658748149872,
+      "learning_rate": 3.7818758285791955e-05,
+      "loss": 0.3046,
+      "step": 630
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3877567648887634,
+      "learning_rate": 3.767147745157039e-05,
+      "loss": 0.3327,
+      "step": 631
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5222271680831909,
+      "learning_rate": 3.7524310424451635e-05,
+      "loss": 0.3651,
+      "step": 632
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3869018852710724,
+      "learning_rate": 3.7377258562964454e-05,
+      "loss": 0.3138,
+      "step": 633
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.42713502049446106,
+      "learning_rate": 3.723032322457458e-05,
+      "loss": 0.3476,
+      "step": 634
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.47757217288017273,
+      "learning_rate": 3.708350576567204e-05,
+      "loss": 0.4056,
+      "step": 635
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.37596315145492554,
+      "learning_rate": 3.6936807541558674e-05,
+      "loss": 0.2626,
+      "step": 636
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.48347583413124084,
+      "learning_rate": 3.6790229906435705e-05,
+      "loss": 0.4516,
+      "step": 637
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3164611756801605,
+      "learning_rate": 3.664377421339111e-05,
+      "loss": 0.2061,
+      "step": 638
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4118727445602417,
+      "learning_rate": 3.6497441814387247e-05,
+      "loss": 0.2964,
+      "step": 639
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3826994001865387,
+      "learning_rate": 3.6351234060248286e-05,
+      "loss": 0.2836,
+      "step": 640
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.44861793518066406,
+      "learning_rate": 3.6205152300647784e-05,
+      "loss": 0.3768,
+      "step": 641
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3583225905895233,
+      "learning_rate": 3.605919788409621e-05,
+      "loss": 0.2999,
+      "step": 642
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3917998671531677,
+      "learning_rate": 3.591337215792852e-05,
+      "loss": 0.3231,
+      "step": 643
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3839118182659149,
+      "learning_rate": 3.5767676468291713e-05,
+      "loss": 0.2315,
+      "step": 644
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4764842987060547,
+      "learning_rate": 3.562211216013235e-05,
+      "loss": 0.4668,
+      "step": 645
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.38360172510147095,
+      "learning_rate": 3.5476680577184206e-05,
+      "loss": 0.325,
+      "step": 646
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.33429163694381714,
+      "learning_rate": 3.533138306195588e-05,
+      "loss": 0.2305,
+      "step": 647
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.44550594687461853,
+      "learning_rate": 3.5186220955718306e-05,
+      "loss": 0.2942,
+      "step": 648
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3722698390483856,
+      "learning_rate": 3.5041195598492446e-05,
+      "loss": 0.3052,
+      "step": 649
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.37912797927856445,
+      "learning_rate": 3.489630832903694e-05,
+      "loss": 0.2645,
+      "step": 650
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.46016430854797363,
+      "learning_rate": 3.475156048483567e-05,
+      "loss": 0.3226,
+      "step": 651
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4096508324146271,
+      "learning_rate": 3.460695340208546e-05,
+      "loss": 0.3527,
+      "step": 652
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.39146289229393005,
+      "learning_rate": 3.446248841568375e-05,
+      "loss": 0.275,
+      "step": 653
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.41832536458969116,
+      "learning_rate": 3.431816685921625e-05,
+      "loss": 0.3173,
+      "step": 654
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.35429659485816956,
+      "learning_rate": 3.417399006494466e-05,
+      "loss": 0.2347,
+      "step": 655
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.36290469765663147,
+      "learning_rate": 3.402995936379433e-05,
+      "loss": 0.2451,
+      "step": 656
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.38287097215652466,
+      "learning_rate": 3.3886076085341986e-05,
+      "loss": 0.2538,
+      "step": 657
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3295687735080719,
+      "learning_rate": 3.37423415578035e-05,
+      "loss": 0.2345,
+      "step": 658
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.37062516808509827,
+      "learning_rate": 3.3598757108021546e-05,
+      "loss": 0.2345,
+      "step": 659
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3951757848262787,
+      "learning_rate": 3.345532406145345e-05,
+      "loss": 0.2836,
+      "step": 660
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.35152289271354675,
+      "learning_rate": 3.331204374215888e-05,
+      "loss": 0.2325,
+      "step": 661
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.36952489614486694,
+      "learning_rate": 3.316891747278761e-05,
+      "loss": 0.2382,
+      "step": 662
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.45323944091796875,
+      "learning_rate": 3.302594657456744e-05,
+      "loss": 0.375,
+      "step": 663
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.4516223073005676,
+      "learning_rate": 3.288313236729183e-05,
+      "loss": 0.3336,
+      "step": 664
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.4407421350479126,
+      "learning_rate": 3.274047616930781e-05,
+      "loss": 0.3417,
+      "step": 665
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.41437798738479614,
+      "learning_rate": 3.259797929750378e-05,
+      "loss": 0.2577,
+      "step": 666
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.38213199377059937,
+      "learning_rate": 3.245564306729744e-05,
+      "loss": 0.2891,
+      "step": 667
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3836421072483063,
+      "learning_rate": 3.231346879262349e-05,
+      "loss": 0.2795,
+      "step": 668
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3943834900856018,
+      "learning_rate": 3.217145778592162e-05,
+      "loss": 0.3158,
+      "step": 669
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4441621005535126,
+      "learning_rate": 3.202961135812437e-05,
+      "loss": 0.3423,
+      "step": 670
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4147234559059143,
+      "learning_rate": 3.1887930818644996e-05,
+      "loss": 0.3495,
+      "step": 671
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.41143837571144104,
+      "learning_rate": 3.1746417475365405e-05,
+      "loss": 0.2875,
+      "step": 672
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.46894603967666626,
+      "learning_rate": 3.1605072634624125e-05,
+      "loss": 0.4085,
+      "step": 673
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.42437368631362915,
+      "learning_rate": 3.146389760120416e-05,
+      "loss": 0.2643,
+      "step": 674
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4013984203338623,
+      "learning_rate": 3.132289367832097e-05,
+      "loss": 0.2912,
+      "step": 675
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.45246797800064087,
+      "learning_rate": 3.118206216761053e-05,
+      "loss": 0.3124,
+      "step": 676
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.377215713262558,
+      "learning_rate": 3.104140436911719e-05,
+      "loss": 0.1922,
+      "step": 677
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.39179039001464844,
+      "learning_rate": 3.0900921581281725e-05,
+      "loss": 0.3054,
+      "step": 678
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.42899250984191895,
+      "learning_rate": 3.076061510092935e-05,
+      "loss": 0.3311,
+      "step": 679
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.38909435272216797,
+      "learning_rate": 3.062048622325779e-05,
+      "loss": 0.3041,
+      "step": 680
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.41301432251930237,
+      "learning_rate": 3.0480536241825263e-05,
+      "loss": 0.2429,
+      "step": 681
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.43931370973587036,
+      "learning_rate": 3.034076644853853e-05,
+      "loss": 0.2854,
+      "step": 682
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.45271268486976624,
+      "learning_rate": 3.0201178133641038e-05,
+      "loss": 0.3819,
+      "step": 683
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3465835154056549,
+      "learning_rate": 3.0061772585700953e-05,
+      "loss": 0.2388,
+      "step": 684
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.39036041498184204,
+      "learning_rate": 2.992255109159926e-05,
+      "loss": 0.2508,
+      "step": 685
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.39026275277137756,
+      "learning_rate": 2.9783514936517965e-05,
+      "loss": 0.3071,
+      "step": 686
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.40448805689811707,
+      "learning_rate": 2.9644665403928117e-05,
+      "loss": 0.2968,
+      "step": 687
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.47924384474754333,
+      "learning_rate": 2.950600377557804e-05,
+      "loss": 0.4256,
+      "step": 688
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4134678542613983,
+      "learning_rate": 2.9367531331481436e-05,
+      "loss": 0.2633,
+      "step": 689
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4081580638885498,
+      "learning_rate": 2.9229249349905684e-05,
+      "loss": 0.3271,
+      "step": 690
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.386106938123703,
+      "learning_rate": 2.909115910735991e-05,
+      "loss": 0.2604,
+      "step": 691
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.34313780069351196,
+      "learning_rate": 2.895326187858326e-05,
+      "loss": 0.238,
+      "step": 692
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.47729331254959106,
+      "learning_rate": 2.881555893653314e-05,
+      "loss": 0.4743,
+      "step": 693
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.40500473976135254,
+      "learning_rate": 2.8678051552373487e-05,
+      "loss": 0.2936,
+      "step": 694
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3432915508747101,
+      "learning_rate": 2.854074099546291e-05,
+      "loss": 0.2432,
+      "step": 695
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4056801199913025,
+      "learning_rate": 2.8403628533343206e-05,
+      "loss": 0.2039,
+      "step": 696
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3884841203689575,
+      "learning_rate": 2.826671543172738e-05,
+      "loss": 0.3125,
+      "step": 697
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.34453776478767395,
+      "learning_rate": 2.8130002954488183e-05,
+      "loss": 0.2515,
+      "step": 698
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3966999650001526,
+      "learning_rate": 2.799349236364634e-05,
+      "loss": 0.3374,
+      "step": 699
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.40421062707901,
+      "learning_rate": 2.7857184919358937e-05,
+      "loss": 0.3175,
+      "step": 700
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3502410650253296,
+      "learning_rate": 2.7721081879907718e-05,
+      "loss": 0.2444,
+      "step": 701
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.499380886554718,
+      "learning_rate": 2.7585184501687577e-05,
+      "loss": 0.3606,
+      "step": 702
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4188764989376068,
+      "learning_rate": 2.74494940391949e-05,
+      "loss": 0.359,
+      "step": 703
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.42254406213760376,
+      "learning_rate": 2.731401174501601e-05,
+      "loss": 0.3138,
+      "step": 704
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.35356250405311584,
+      "learning_rate": 2.7178738869815506e-05,
+      "loss": 0.2556,
+      "step": 705
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.34609168767929077,
+      "learning_rate": 2.7043676662324878e-05,
+      "loss": 0.2317,
+      "step": 706
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4097319543361664,
+      "learning_rate": 2.69088263693309e-05,
+      "loss": 0.3008,
+      "step": 707
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.37519168853759766,
+      "learning_rate": 2.6774189235664026e-05,
+      "loss": 0.2497,
+      "step": 708
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.47046300768852234,
+      "learning_rate": 2.663976650418715e-05,
+      "loss": 0.4079,
+      "step": 709
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.4490043818950653,
+      "learning_rate": 2.650555941578381e-05,
+      "loss": 0.409,
+      "step": 710
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.39349454641342163,
+      "learning_rate": 2.6371569209347014e-05,
+      "loss": 0.2863,
+      "step": 711
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.4129186272621155,
+      "learning_rate": 2.6237797121767634e-05,
+      "loss": 0.3697,
+      "step": 712
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3730863332748413,
+      "learning_rate": 2.6104244387923082e-05,
+      "loss": 0.2673,
+      "step": 713
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.329772412776947,
+      "learning_rate": 2.5970912240665813e-05,
+      "loss": 0.1977,
+      "step": 714
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3878955543041229,
+      "learning_rate": 2.5837801910812053e-05,
+      "loss": 0.2983,
+      "step": 715
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3468000888824463,
+      "learning_rate": 2.5704914627130374e-05,
+      "loss": 0.2667,
+      "step": 716
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.4333817958831787,
+      "learning_rate": 2.5572251616330373e-05,
+      "loss": 0.3788,
+      "step": 717
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3754556477069855,
+      "learning_rate": 2.5439814103051284e-05,
+      "loss": 0.2182,
+      "step": 718
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2976343035697937,
+      "learning_rate": 2.530760330985079e-05,
+      "loss": 0.178,
+      "step": 719
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4496738314628601,
+      "learning_rate": 2.517562045719367e-05,
+      "loss": 0.3968,
+      "step": 720
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3929567039012909,
+      "learning_rate": 2.504386676344047e-05,
+      "loss": 0.3243,
+      "step": 721
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.40064507722854614,
+      "learning_rate": 2.4912343444836445e-05,
+      "loss": 0.3359,
+      "step": 722
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4283524751663208,
+      "learning_rate": 2.4781051715500076e-05,
+      "loss": 0.3824,
+      "step": 723
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3501032888889313,
+      "learning_rate": 2.46499927874121e-05,
+      "loss": 0.214,
+      "step": 724
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3761228024959564,
+      "learning_rate": 2.4519167870404125e-05,
+      "loss": 0.2473,
+      "step": 725
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4208412170410156,
+      "learning_rate": 2.4388578172147675e-05,
+      "loss": 0.3179,
+      "step": 726
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.38908398151397705,
+      "learning_rate": 2.4258224898142807e-05,
+      "loss": 0.2357,
+      "step": 727
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4253256022930145,
+      "learning_rate": 2.4128109251707155e-05,
+      "loss": 0.3923,
+      "step": 728
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.38555988669395447,
+      "learning_rate": 2.399823243396476e-05,
+      "loss": 0.2836,
+      "step": 729
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.39008665084838867,
+      "learning_rate": 2.3868595643834995e-05,
+      "loss": 0.2493,
+      "step": 730
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.409845232963562,
+      "learning_rate": 2.373920007802144e-05,
+      "loss": 0.3188,
+      "step": 731
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3560733199119568,
+      "learning_rate": 2.361004693100094e-05,
+      "loss": 0.2564,
+      "step": 732
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.40539562702178955,
+      "learning_rate": 2.3481137395012513e-05,
+      "loss": 0.2965,
+      "step": 733
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.38528934121131897,
+      "learning_rate": 2.3352472660046295e-05,
+      "loss": 0.2615,
+      "step": 734
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.37828436493873596,
+      "learning_rate": 2.322405391383273e-05,
+      "loss": 0.303,
+      "step": 735
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4200487732887268,
+      "learning_rate": 2.3095882341831372e-05,
+      "loss": 0.3986,
+      "step": 736
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.37558552622795105,
+      "learning_rate": 2.296795912722014e-05,
+      "loss": 0.3008,
+      "step": 737
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.35239487886428833,
+      "learning_rate": 2.284028545088423e-05,
+      "loss": 0.2579,
+      "step": 738
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3456938862800598,
+      "learning_rate": 2.2712862491405436e-05,
+      "loss": 0.2618,
+      "step": 739
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3525884747505188,
+      "learning_rate": 2.258569142505098e-05,
+      "loss": 0.2489,
+      "step": 740
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.41400423645973206,
+      "learning_rate": 2.2458773425762912e-05,
+      "loss": 0.2781,
+      "step": 741
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.45001018047332764,
+      "learning_rate": 2.2332109665147127e-05,
+      "loss": 0.3845,
+      "step": 742
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.40422332286834717,
+      "learning_rate": 2.2205701312462617e-05,
+      "loss": 0.351,
+      "step": 743
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.39631444215774536,
+      "learning_rate": 2.2079549534610606e-05,
+      "loss": 0.2952,
+      "step": 744
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3830812871456146,
+      "learning_rate": 2.1953655496123853e-05,
+      "loss": 0.2824,
+      "step": 745
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4185529351234436,
+      "learning_rate": 2.1828020359155905e-05,
+      "loss": 0.2806,
+      "step": 746
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3633481562137604,
+      "learning_rate": 2.1702645283470236e-05,
+      "loss": 0.2395,
+      "step": 747
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4514968693256378,
+      "learning_rate": 2.1577531426429782e-05,
+      "loss": 0.3592,
+      "step": 748
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.33255240321159363,
+      "learning_rate": 2.1452679942985993e-05,
+      "loss": 0.2244,
+      "step": 749
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4340452253818512,
+      "learning_rate": 2.132809198566837e-05,
+      "loss": 0.3434,
+      "step": 750
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4698704183101654,
+      "learning_rate": 2.1203768704573672e-05,
+      "loss": 0.3705,
+      "step": 751
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3831236660480499,
+      "learning_rate": 2.1079711247355505e-05,
+      "loss": 0.2703,
+      "step": 752
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.44130879640579224,
+      "learning_rate": 2.095592075921347e-05,
+      "loss": 0.2404,
+      "step": 753
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3395165503025055,
+      "learning_rate": 2.08323983828828e-05,
+      "loss": 0.1883,
+      "step": 754
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.43158164620399475,
+      "learning_rate": 2.0709145258623704e-05,
+      "loss": 0.3831,
+      "step": 755
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.40739014744758606,
+      "learning_rate": 2.0586162524210895e-05,
+      "loss": 0.2756,
+      "step": 756
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4256882071495056,
+      "learning_rate": 2.0463451314923015e-05,
+      "loss": 0.322,
+      "step": 757
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.43485283851623535,
+      "learning_rate": 2.0341012763532243e-05,
+      "loss": 0.3444,
+      "step": 758
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.36340105533599854,
+      "learning_rate": 2.021884800029379e-05,
+      "loss": 0.2585,
+      "step": 759
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.39190852642059326,
+      "learning_rate": 2.009695815293548e-05,
+      "loss": 0.294,
+      "step": 760
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4237244427204132,
+      "learning_rate": 1.9975344346647297e-05,
+      "loss": 0.31,
+      "step": 761
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3808172345161438,
+      "learning_rate": 1.9854007704071064e-05,
+      "loss": 0.2931,
+      "step": 762
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3244672119617462,
+      "learning_rate": 1.973294934529007e-05,
+      "loss": 0.2037,
+      "step": 763
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3610917925834656,
+      "learning_rate": 1.961217038781863e-05,
+      "loss": 0.2589,
+      "step": 764
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.4579406976699829,
+      "learning_rate": 1.9491671946591962e-05,
+      "loss": 0.3183,
+      "step": 765
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.4583299160003662,
+      "learning_rate": 1.9371455133955675e-05,
+      "loss": 0.349,
+      "step": 766
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.39002010226249695,
+      "learning_rate": 1.925152105965567e-05,
+      "loss": 0.2644,
+      "step": 767
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.35974547266960144,
+      "learning_rate": 1.9131870830827818e-05,
+      "loss": 0.2283,
+      "step": 768
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.4203035831451416,
+      "learning_rate": 1.9012505551987765e-05,
+      "loss": 0.2639,
+      "step": 769
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3690374791622162,
+      "learning_rate": 1.8893426325020686e-05,
+      "loss": 0.2513,
+      "step": 770
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.42623621225357056,
+      "learning_rate": 1.8774634249171185e-05,
+      "loss": 0.3978,
+      "step": 771
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3870992958545685,
+      "learning_rate": 1.8656130421033123e-05,
+      "loss": 0.2935,
+      "step": 772
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.35730135440826416,
+      "learning_rate": 1.8537915934539486e-05,
+      "loss": 0.1643,
+      "step": 773
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3966630697250366,
+      "learning_rate": 1.841999188095224e-05,
+      "loss": 0.3057,
+      "step": 774
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.4256872832775116,
+      "learning_rate": 1.830235934885237e-05,
+      "loss": 0.3644,
+      "step": 775
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.35789552330970764,
+      "learning_rate": 1.818501942412975e-05,
+      "loss": 0.2894,
+      "step": 776
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.40891268849372864,
+      "learning_rate": 1.8067973189973075e-05,
+      "loss": 0.3312,
+      "step": 777
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.4185810387134552,
+      "learning_rate": 1.7951221726860045e-05,
+      "loss": 0.3332,
+      "step": 778
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.4217431843280792,
+      "learning_rate": 1.7834766112547142e-05,
+      "loss": 0.3096,
+      "step": 779
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.38097184896469116,
+      "learning_rate": 1.771860742205988e-05,
+      "loss": 0.3022,
+      "step": 780
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.4239228665828705,
+      "learning_rate": 1.7602746727682796e-05,
+      "loss": 0.3631,
+      "step": 781
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.36988747119903564,
+      "learning_rate": 1.7487185098949565e-05,
+      "loss": 0.276,
+      "step": 782
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.4268799424171448,
+      "learning_rate": 1.7371923602633078e-05,
+      "loss": 0.3585,
+      "step": 783
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.42823129892349243,
+      "learning_rate": 1.725696330273575e-05,
+      "loss": 0.368,
+      "step": 784
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.39439141750335693,
+      "learning_rate": 1.7142305260479474e-05,
+      "loss": 0.3613,
+      "step": 785
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.39605358242988586,
+      "learning_rate": 1.7027950534296027e-05,
+      "loss": 0.3753,
+      "step": 786
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3405487537384033,
+      "learning_rate": 1.6913900179817144e-05,
+      "loss": 0.2243,
+      "step": 787
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3856514096260071,
+      "learning_rate": 1.6800155249864896e-05,
+      "loss": 0.3305,
+      "step": 788
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3783203661441803,
+      "learning_rate": 1.668671679444192e-05,
+      "loss": 0.2892,
+      "step": 789
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.41972652077674866,
+      "learning_rate": 1.6573585860721646e-05,
+      "loss": 0.3498,
+      "step": 790
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.37746182084083557,
+      "learning_rate": 1.646076349303884e-05,
+      "loss": 0.277,
+      "step": 791
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3971934914588928,
+      "learning_rate": 1.63482507328797e-05,
+      "loss": 0.2944,
+      "step": 792
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2976275384426117,
+      "learning_rate": 1.6236048618872456e-05,
+      "loss": 0.2105,
+      "step": 793
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5275766849517822,
+      "learning_rate": 1.6124158186777676e-05,
+      "loss": 0.3172,
+      "step": 794
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.42608052492141724,
+      "learning_rate": 1.6012580469478743e-05,
+      "loss": 0.3299,
+      "step": 795
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.32864508032798767,
+      "learning_rate": 1.5901316496972262e-05,
+      "loss": 0.1943,
+      "step": 796
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3908816874027252,
+      "learning_rate": 1.5790367296358644e-05,
+      "loss": 0.2985,
+      "step": 797
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3235936760902405,
+      "learning_rate": 1.5679733891832556e-05,
+      "loss": 0.2088,
+      "step": 798
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.38437268137931824,
+      "learning_rate": 1.55694173046735e-05,
+      "loss": 0.2778,
+      "step": 799
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.36689984798431396,
+      "learning_rate": 1.5459418553236343e-05,
+      "loss": 0.2769,
+      "step": 800
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3990126848220825,
+      "learning_rate": 1.5349738652941968e-05,
+      "loss": 0.3203,
+      "step": 801
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3759137690067291,
+      "learning_rate": 1.5240378616267886e-05,
+      "loss": 0.2586,
+      "step": 802
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.38042765855789185,
+      "learning_rate": 1.5131339452738863e-05,
+      "loss": 0.3687,
+      "step": 803
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.37907275557518005,
+      "learning_rate": 1.5022622168917649e-05,
+      "loss": 0.272,
+      "step": 804
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3538564443588257,
+      "learning_rate": 1.4914227768395595e-05,
+      "loss": 0.2581,
+      "step": 805
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.39212673902511597,
+      "learning_rate": 1.4806157251783515e-05,
+      "loss": 0.291,
+      "step": 806
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.36042389273643494,
+      "learning_rate": 1.4698411616702356e-05,
+      "loss": 0.2512,
+      "step": 807
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.387197881937027,
+      "learning_rate": 1.4590991857774038e-05,
+      "loss": 0.3171,
+      "step": 808
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.38724789023399353,
+      "learning_rate": 1.4483898966612209e-05,
+      "loss": 0.3734,
+      "step": 809
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3722423017024994,
+      "learning_rate": 1.437713393181317e-05,
+      "loss": 0.2774,
+      "step": 810
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.40014025568962097,
+      "learning_rate": 1.4270697738946704e-05,
+      "loss": 0.2615,
+      "step": 811
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3334675431251526,
+      "learning_rate": 1.4164591370547004e-05,
+      "loss": 0.2324,
+      "step": 812
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.4055711030960083,
+      "learning_rate": 1.4058815806103542e-05,
+      "loss": 0.2911,
+      "step": 813
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.40564215183258057,
+      "learning_rate": 1.3953372022052107e-05,
+      "loss": 0.3312,
+      "step": 814
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.34772953391075134,
+      "learning_rate": 1.3848260991765755e-05,
+      "loss": 0.2875,
+      "step": 815
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.4844672381877899,
+      "learning_rate": 1.3743483685545811e-05,
+      "loss": 0.4066,
+      "step": 816
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.42509323358535767,
+      "learning_rate": 1.363904107061294e-05,
+      "loss": 0.3193,
+      "step": 817
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.43162110447883606,
+      "learning_rate": 1.3534934111098179e-05,
+      "loss": 0.3408,
+      "step": 818
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3848356306552887,
+      "learning_rate": 1.3431163768034077e-05,
+      "loss": 0.314,
+      "step": 819
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.35395869612693787,
+      "learning_rate": 1.3327730999345817e-05,
+      "loss": 0.2277,
+      "step": 820
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.35360512137413025,
+      "learning_rate": 1.3224636759842363e-05,
+      "loss": 0.2947,
+      "step": 821
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3542770445346832,
+      "learning_rate": 1.3121882001207614e-05,
+      "loss": 0.223,
+      "step": 822
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.38950300216674805,
+      "learning_rate": 1.3019467671991692e-05,
+      "loss": 0.2419,
+      "step": 823
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.4121183454990387,
+      "learning_rate": 1.2917394717602121e-05,
+      "loss": 0.3552,
+      "step": 824
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.34702908992767334,
+      "learning_rate": 1.2815664080295159e-05,
+      "loss": 0.2483,
+      "step": 825
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.34193891286849976,
+      "learning_rate": 1.2714276699166994e-05,
+      "loss": 0.2278,
+      "step": 826
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.37185028195381165,
+      "learning_rate": 1.261323351014525e-05,
+      "loss": 0.2543,
+      "step": 827
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.31859350204467773,
+      "learning_rate": 1.251253544598014e-05,
+      "loss": 0.2054,
+      "step": 828
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.35622653365135193,
+      "learning_rate": 1.241218343623602e-05,
+      "loss": 0.3048,
+      "step": 829
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3775176405906677,
+      "learning_rate": 1.2312178407282749e-05,
+      "loss": 0.2991,
+      "step": 830
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3728039264678955,
+      "learning_rate": 1.2212521282287092e-05,
+      "loss": 0.3001,
+      "step": 831
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.42662662267684937,
+      "learning_rate": 1.2113212981204292e-05,
+      "loss": 0.3589,
+      "step": 832
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.41523322463035583,
+      "learning_rate": 1.2014254420769466e-05,
+      "loss": 0.3247,
+      "step": 833
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3754156231880188,
+      "learning_rate": 1.1915646514489292e-05,
+      "loss": 0.2881,
+      "step": 834
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.39000576734542847,
+      "learning_rate": 1.1817390172633403e-05,
+      "loss": 0.3189,
+      "step": 835
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3429880738258362,
+      "learning_rate": 1.1719486302226118e-05,
+      "loss": 0.2378,
+      "step": 836
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5210950374603271,
+      "learning_rate": 1.1621935807038003e-05,
+      "loss": 0.4818,
+      "step": 837
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.41101348400115967,
+      "learning_rate": 1.152473958757756e-05,
+      "loss": 0.3383,
+      "step": 838
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3532755970954895,
+      "learning_rate": 1.1427898541082855e-05,
+      "loss": 0.2378,
+      "step": 839
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.4082874655723572,
+      "learning_rate": 1.133141356151336e-05,
+      "loss": 0.3243,
+      "step": 840
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.40283286571502686,
+      "learning_rate": 1.123528553954154e-05,
+      "loss": 0.3047,
+      "step": 841
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.30116912722587585,
+      "learning_rate": 1.1139515362544755e-05,
+      "loss": 0.1996,
+      "step": 842
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.4485073685646057,
+      "learning_rate": 1.1044103914597031e-05,
+      "loss": 0.3775,
+      "step": 843
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.46858519315719604,
+      "learning_rate": 1.0949052076460853e-05,
+      "loss": 0.4176,
+      "step": 844
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.4420608878135681,
+      "learning_rate": 1.085436072557911e-05,
+      "loss": 0.3277,
+      "step": 845
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3578694462776184,
+      "learning_rate": 1.0760030736066951e-05,
+      "loss": 0.2433,
+      "step": 846
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3760528564453125,
+      "learning_rate": 1.0666062978703733e-05,
+      "loss": 0.2819,
+      "step": 847
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.37020227313041687,
+      "learning_rate": 1.0572458320924943e-05,
+      "loss": 0.2603,
+      "step": 848
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4318196773529053,
+      "learning_rate": 1.0479217626814253e-05,
+      "loss": 0.3103,
+      "step": 849
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.31711700558662415,
+      "learning_rate": 1.0386341757095502e-05,
+      "loss": 0.1956,
+      "step": 850
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4092193841934204,
+      "learning_rate": 1.0293831569124774e-05,
+      "loss": 0.3291,
+      "step": 851
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4069688618183136,
+      "learning_rate": 1.0201687916882418e-05,
+      "loss": 0.3463,
+      "step": 852
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.35108041763305664,
+      "learning_rate": 1.0109911650965314e-05,
+      "loss": 0.2141,
+      "step": 853
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5507402420043945,
+      "learning_rate": 1.0018503618578818e-05,
+      "loss": 0.521,
+      "step": 854
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.40518254041671753,
+      "learning_rate": 9.927464663529118e-06,
+      "loss": 0.2749,
+      "step": 855
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4341948926448822,
+      "learning_rate": 9.836795626215356e-06,
+      "loss": 0.3684,
+      "step": 856
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.36623236536979675,
+      "learning_rate": 9.746497343621857e-06,
+      "loss": 0.296,
+      "step": 857
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.33644017577171326,
+      "learning_rate": 9.656570649310481e-06,
+      "loss": 0.2323,
+      "step": 858
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3651730418205261,
+      "learning_rate": 9.567016373412857e-06,
+      "loss": 0.3039,
+      "step": 859
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.46405401825904846,
+      "learning_rate": 9.477835342622759e-06,
+      "loss": 0.3854,
+      "step": 860
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.44806769490242004,
+      "learning_rate": 9.389028380188419e-06,
+      "loss": 0.3988,
+      "step": 861
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3805658519268036,
+      "learning_rate": 9.300596305905013e-06,
+      "loss": 0.2861,
+      "step": 862
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.39312711358070374,
+      "learning_rate": 9.212539936107029e-06,
+      "loss": 0.2729,
+      "step": 863
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.34711524844169617,
+      "learning_rate": 9.124860083660769e-06,
+      "loss": 0.2602,
+      "step": 864
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.33725085854530334,
+      "learning_rate": 9.037557557956766e-06,
+      "loss": 0.1977,
+      "step": 865
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.4098747670650482,
+      "learning_rate": 8.950633164902467e-06,
+      "loss": 0.3187,
+      "step": 866
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.418828547000885,
+      "learning_rate": 8.86408770691462e-06,
+      "loss": 0.2814,
+      "step": 867
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.33042341470718384,
+      "learning_rate": 8.777921982911996e-06,
+      "loss": 0.2307,
+      "step": 868
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3797352612018585,
+      "learning_rate": 8.692136788307903e-06,
+      "loss": 0.2654,
+      "step": 869
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3360169529914856,
+      "learning_rate": 8.606732915003002e-06,
+      "loss": 0.2483,
+      "step": 870
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.36900776624679565,
+      "learning_rate": 8.521711151377803e-06,
+      "loss": 0.2554,
+      "step": 871
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.4058022201061249,
+      "learning_rate": 8.437072282285535e-06,
+      "loss": 0.318,
+      "step": 872
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3715745210647583,
+      "learning_rate": 8.35281708904485e-06,
+      "loss": 0.2381,
+      "step": 873
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.39454740285873413,
+      "learning_rate": 8.268946349432582e-06,
+      "loss": 0.3084,
+      "step": 874
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.42099666595458984,
+      "learning_rate": 8.185460837676612e-06,
+      "loss": 0.292,
+      "step": 875
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.45269036293029785,
+      "learning_rate": 8.102361324448715e-06,
+      "loss": 0.2864,
+      "step": 876
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.41998958587646484,
+      "learning_rate": 8.019648576857425e-06,
+      "loss": 0.2996,
+      "step": 877
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.4255835711956024,
+      "learning_rate": 7.937323358440935e-06,
+      "loss": 0.3387,
+      "step": 878
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.352857381105423,
+      "learning_rate": 7.85538642916015e-06,
+      "loss": 0.2711,
+      "step": 879
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3418557047843933,
+      "learning_rate": 7.773838545391515e-06,
+      "loss": 0.2004,
+      "step": 880
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3853307068347931,
+      "learning_rate": 7.692680459920188e-06,
+      "loss": 0.2871,
+      "step": 881
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.34913262724876404,
+      "learning_rate": 7.6119129219329395e-06,
+      "loss": 0.228,
+      "step": 882
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.42278945446014404,
+      "learning_rate": 7.5315366770114195e-06,
+      "loss": 0.3265,
+      "step": 883
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.4618382453918457,
+      "learning_rate": 7.4515524671250725e-06,
+      "loss": 0.2927,
+      "step": 884
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3939862549304962,
+      "learning_rate": 7.371961030624452e-06,
+      "loss": 0.2553,
+      "step": 885
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.37247055768966675,
+      "learning_rate": 7.292763102234329e-06,
+      "loss": 0.278,
+      "step": 886
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3525467813014984,
+      "learning_rate": 7.213959413046894e-06,
+      "loss": 0.2045,
+      "step": 887
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.34763529896736145,
+      "learning_rate": 7.135550690515052e-06,
+      "loss": 0.272,
+      "step": 888
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.37334975600242615,
+      "learning_rate": 7.057537658445701e-06,
+      "loss": 0.2354,
+      "step": 889
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3018919825553894,
+      "learning_rate": 6.979921036993042e-06,
+      "loss": 0.1974,
+      "step": 890
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.42736706137657166,
+      "learning_rate": 6.902701542651874e-06,
+      "loss": 0.2821,
+      "step": 891
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.33115968108177185,
+      "learning_rate": 6.825879888251135e-06,
+      "loss": 0.179,
+      "step": 892
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.38003167510032654,
+      "learning_rate": 6.749456782947122e-06,
+      "loss": 0.2484,
+      "step": 893
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.40489086508750916,
+      "learning_rate": 6.6734329322171165e-06,
+      "loss": 0.3016,
+      "step": 894
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3853268027305603,
+      "learning_rate": 6.597809037852726e-06,
+      "loss": 0.3321,
+      "step": 895
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3887777626514435,
+      "learning_rate": 6.522585797953579e-06,
+      "loss": 0.2402,
+      "step": 896
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.46126478910446167,
+      "learning_rate": 6.447763906920679e-06,
+      "loss": 0.3896,
+      "step": 897
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4578423798084259,
+      "learning_rate": 6.373344055450165e-06,
+      "loss": 0.3362,
+      "step": 898
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.35297033190727234,
+      "learning_rate": 6.2993269305268495e-06,
+      "loss": 0.2309,
+      "step": 899
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.41612932085990906,
+      "learning_rate": 6.2257132154178665e-06,
+      "loss": 0.3451,
+      "step": 900
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.41109976172447205,
+      "learning_rate": 6.152503589666425e-06,
+      "loss": 0.3227,
+      "step": 901
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.41766732931137085,
+      "learning_rate": 6.079698729085498e-06,
+      "loss": 0.3057,
+      "step": 902
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.29895317554473877,
+      "learning_rate": 6.007299305751585e-06,
+      "loss": 0.1809,
+      "step": 903
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.35016998648643494,
+      "learning_rate": 5.935305987998496e-06,
+      "loss": 0.2466,
+      "step": 904
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3651289939880371,
+      "learning_rate": 5.863719440411214e-06,
+      "loss": 0.2383,
+      "step": 905
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.37916186451911926,
+      "learning_rate": 5.792540323819751e-06,
+      "loss": 0.3321,
+      "step": 906
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3627545237541199,
+      "learning_rate": 5.721769295293034e-06,
+      "loss": 0.2442,
+      "step": 907
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.4374503791332245,
+      "learning_rate": 5.651407008132809e-06,
+      "loss": 0.3614,
+      "step": 908
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.36250773072242737,
+      "learning_rate": 5.5814541118677284e-06,
+      "loss": 0.2129,
+      "step": 909
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.41063565015792847,
+      "learning_rate": 5.5119112522471924e-06,
+      "loss": 0.3507,
+      "step": 910
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.37300723791122437,
+      "learning_rate": 5.442779071235516e-06,
+      "loss": 0.252,
+      "step": 911
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.33876121044158936,
+      "learning_rate": 5.374058207005944e-06,
+      "loss": 0.2375,
+      "step": 912
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.38935375213623047,
+      "learning_rate": 5.305749293934764e-06,
+      "loss": 0.2552,
+      "step": 913
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.42013880610466003,
+      "learning_rate": 5.237852962595469e-06,
+      "loss": 0.2975,
+      "step": 914
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3413439393043518,
+      "learning_rate": 5.170369839752925e-06,
+      "loss": 0.2017,
+      "step": 915
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3880120813846588,
+      "learning_rate": 5.1033005483575925e-06,
+      "loss": 0.2561,
+      "step": 916
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.40935495495796204,
+      "learning_rate": 5.036645707539745e-06,
+      "loss": 0.2604,
+      "step": 917
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.4027320444583893,
+      "learning_rate": 4.9704059326038055e-06,
+      "loss": 0.2602,
+      "step": 918
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3935575783252716,
+      "learning_rate": 4.90458183502262e-06,
+      "loss": 0.2899,
+      "step": 919
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.4156310260295868,
+      "learning_rate": 4.839174022431858e-06,
+      "loss": 0.3689,
+      "step": 920
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.4116358160972595,
+      "learning_rate": 4.7741830986243356e-06,
+      "loss": 0.3135,
+      "step": 921
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3684353828430176,
+      "learning_rate": 4.709609663544534e-06,
+      "loss": 0.2824,
+      "step": 922
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3770512044429779,
+      "learning_rate": 4.645454313282965e-06,
+      "loss": 0.2948,
+      "step": 923
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.47205623984336853,
+      "learning_rate": 4.581717640070743e-06,
+      "loss": 0.3442,
+      "step": 924
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.42810696363449097,
+      "learning_rate": 4.5184002322740785e-06,
+      "loss": 0.347,
+      "step": 925
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.31437572836875916,
+      "learning_rate": 4.455502674388873e-06,
+      "loss": 0.2313,
+      "step": 926
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.46491920948028564,
+      "learning_rate": 4.3930255470352736e-06,
+      "loss": 0.3452,
+      "step": 927
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4385717511177063,
+      "learning_rate": 4.330969426952375e-06,
+      "loss": 0.3261,
+      "step": 928
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.396494060754776,
+      "learning_rate": 4.269334886992876e-06,
+      "loss": 0.281,
+      "step": 929
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.38151490688323975,
+      "learning_rate": 4.208122496117744e-06,
+      "loss": 0.2138,
+      "step": 930
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3869413435459137,
+      "learning_rate": 4.147332819391048e-06,
+      "loss": 0.2732,
+      "step": 931
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4076899588108063,
+      "learning_rate": 4.0869664179746694e-06,
+      "loss": 0.2687,
+      "step": 932
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.35533949732780457,
+      "learning_rate": 4.027023849123157e-06,
+      "loss": 0.2326,
+      "step": 933
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3902985751628876,
+      "learning_rate": 3.967505666178556e-06,
+      "loss": 0.2674,
+      "step": 934
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3649982213973999,
+      "learning_rate": 3.908412418565371e-06,
+      "loss": 0.2548,
+      "step": 935
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.35166463255882263,
+      "learning_rate": 3.849744651785381e-06,
+      "loss": 0.246,
+      "step": 936
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3820720314979553,
+      "learning_rate": 3.7915029074126974e-06,
+      "loss": 0.3574,
+      "step": 937
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.4962540566921234,
+      "learning_rate": 3.7336877230887246e-06,
+      "loss": 0.4256,
+      "step": 938
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.4633179008960724,
+      "learning_rate": 3.676299632517216e-06,
+      "loss": 0.421,
+      "step": 939
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.48635226488113403,
+      "learning_rate": 3.619339165459307e-06,
+      "loss": 0.3362,
+      "step": 940
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.35124728083610535,
+      "learning_rate": 3.562806847728678e-06,
+      "loss": 0.2662,
+      "step": 941
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.4047708213329315,
+      "learning_rate": 3.5067032011866783e-06,
+      "loss": 0.3547,
+      "step": 942
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.30400896072387695,
+      "learning_rate": 3.4510287437374835e-06,
+      "loss": 0.1609,
+      "step": 943
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3707870841026306,
+      "learning_rate": 3.3957839893233536e-06,
+      "loss": 0.2381,
+      "step": 944
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3312855064868927,
+      "learning_rate": 3.340969447919873e-06,
+      "loss": 0.246,
+      "step": 945
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3357097804546356,
+      "learning_rate": 3.286585625531241e-06,
+      "loss": 0.2066,
+      "step": 946
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.33315232396125793,
+      "learning_rate": 3.232633024185583e-06,
+      "loss": 0.2208,
+      "step": 947
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3273695707321167,
+      "learning_rate": 3.1791121419303794e-06,
+      "loss": 0.2188,
+      "step": 948
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.40958088636398315,
+      "learning_rate": 3.1260234728277717e-06,
+      "loss": 0.2762,
+      "step": 949
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3858436644077301,
+      "learning_rate": 3.0733675069500865e-06,
+      "loss": 0.2863,
+      "step": 950
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.4404066503047943,
+      "learning_rate": 3.0211447303752695e-06,
+      "loss": 0.2966,
+      "step": 951
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3974824845790863,
+      "learning_rate": 2.9693556251824185e-06,
+      "loss": 0.2889,
+      "step": 952
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.4611467719078064,
+      "learning_rate": 2.9180006694472906e-06,
+      "loss": 0.4199,
+      "step": 953
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.4060475826263428,
+      "learning_rate": 2.867080337237954e-06,
+      "loss": 0.3649,
+      "step": 954
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.43869730830192566,
+      "learning_rate": 2.8165950986103805e-06,
+      "loss": 0.3723,
+      "step": 955
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4404228925704956,
+      "learning_rate": 2.7665454196040664e-06,
+      "loss": 0.3284,
+      "step": 956
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.40427085757255554,
+      "learning_rate": 2.716931762237801e-06,
+      "loss": 0.2748,
+      "step": 957
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.41850516200065613,
+      "learning_rate": 2.667754584505372e-06,
+      "loss": 0.3376,
+      "step": 958
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.49761441349983215,
+      "learning_rate": 2.6190143403713174e-06,
+      "loss": 0.3464,
+      "step": 959
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.42231816053390503,
+      "learning_rate": 2.5707114797667465e-06,
+      "loss": 0.2808,
+      "step": 960
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4168257713317871,
+      "learning_rate": 2.522846448585231e-06,
+      "loss": 0.3563,
+      "step": 961
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.44141292572021484,
+      "learning_rate": 2.4754196886785986e-06,
+      "loss": 0.2718,
+      "step": 962
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3264952600002289,
+      "learning_rate": 2.4284316378529404e-06,
+      "loss": 0.2626,
+      "step": 963
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3807451128959656,
+      "learning_rate": 2.3818827298645207e-06,
+      "loss": 0.2457,
+      "step": 964
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3312126696109772,
+      "learning_rate": 2.335773394415802e-06,
+      "loss": 0.2161,
+      "step": 965
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3949142396450043,
+      "learning_rate": 2.2901040571514322e-06,
+      "loss": 0.2985,
+      "step": 966
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.4175260066986084,
+      "learning_rate": 2.2448751396543787e-06,
+      "loss": 0.2932,
+      "step": 967
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.33168932795524597,
+      "learning_rate": 2.2000870594419908e-06,
+      "loss": 0.1803,
+      "step": 968
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3617693781852722,
+      "learning_rate": 2.155740229962161e-06,
+      "loss": 0.2597,
+      "step": 969
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3556663990020752,
+      "learning_rate": 2.1118350605894955e-06,
+      "loss": 0.3051,
+      "step": 970
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.36153483390808105,
+      "learning_rate": 2.068371956621562e-06,
+      "loss": 0.2594,
+      "step": 971
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.42352989315986633,
+      "learning_rate": 2.0253513192751373e-06,
+      "loss": 0.3176,
+      "step": 972
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.38619571924209595,
+      "learning_rate": 1.982773545682459e-06,
+      "loss": 0.291,
+      "step": 973
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3379153311252594,
+      "learning_rate": 1.9406390288876586e-06,
+      "loss": 0.2215,
+      "step": 974
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3745526671409607,
+      "learning_rate": 1.8989481578430223e-06,
+      "loss": 0.243,
+      "step": 975
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4201480448246002,
+      "learning_rate": 1.8577013174054857e-06,
+      "loss": 0.3248,
+      "step": 976
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.40764811635017395,
+      "learning_rate": 1.8168988883330185e-06,
+      "loss": 0.3478,
+      "step": 977
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.46223610639572144,
+      "learning_rate": 1.7765412472811771e-06,
+      "loss": 0.4414,
+      "step": 978
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3968757092952728,
+      "learning_rate": 1.7366287667995417e-06,
+      "loss": 0.2254,
+      "step": 979
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.37118399143218994,
+      "learning_rate": 1.697161815328363e-06,
+      "loss": 0.2569,
+      "step": 980
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3366474509239197,
+      "learning_rate": 1.6581407571951092e-06,
+      "loss": 0.2767,
+      "step": 981
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.35800257325172424,
+      "learning_rate": 1.6195659526111185e-06,
+      "loss": 0.2592,
+      "step": 982
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.41821491718292236,
+      "learning_rate": 1.5814377576682527e-06,
+      "loss": 0.3047,
+      "step": 983
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.35340431332588196,
+      "learning_rate": 1.5437565243356656e-06,
+      "loss": 0.1947,
+      "step": 984
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4180166721343994,
+      "learning_rate": 1.5065226004564893e-06,
+      "loss": 0.3728,
+      "step": 985
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3838116526603699,
+      "learning_rate": 1.4697363297446477e-06,
+      "loss": 0.2781,
+      "step": 986
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.4529761075973511,
+      "learning_rate": 1.4333980517817203e-06,
+      "loss": 0.3834,
+      "step": 987
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.34723663330078125,
+      "learning_rate": 1.3975081020137392e-06,
+      "loss": 0.1994,
+      "step": 988
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5195287466049194,
+      "learning_rate": 1.3620668117481472e-06,
+      "loss": 0.3468,
+      "step": 989
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.37333688139915466,
+      "learning_rate": 1.3270745081506997e-06,
+      "loss": 0.3263,
+      "step": 990
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5042914152145386,
+      "learning_rate": 1.292531514242501e-06,
+      "loss": 0.4647,
+      "step": 991
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.4056055247783661,
+      "learning_rate": 1.2584381488969454e-06,
+      "loss": 0.297,
+      "step": 992
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3029918968677521,
+      "learning_rate": 1.2247947268368364e-06,
+      "loss": 0.2179,
+      "step": 993
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.390737384557724,
+      "learning_rate": 1.191601558631461e-06,
+      "loss": 0.257,
+      "step": 994
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.35613298416137695,
+      "learning_rate": 1.1588589506937198e-06,
+      "loss": 0.2713,
+      "step": 995
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5220173597335815,
+      "learning_rate": 1.126567205277279e-06,
+      "loss": 0.3513,
+      "step": 996
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3554520308971405,
+      "learning_rate": 1.094726620473835e-06,
+      "loss": 0.2744,
+      "step": 997
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.4104813039302826,
+      "learning_rate": 1.0633374902103088e-06,
+      "loss": 0.3098,
+      "step": 998
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3933941423892975,
+      "learning_rate": 1.0324001042461395e-06,
+      "loss": 0.2521,
+      "step": 999
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.4088434875011444,
+      "learning_rate": 1.0019147481706625e-06,
+      "loss": 0.3249,
+      "step": 1000
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.4932601749897003,
+      "learning_rate": 9.718817034003901e-07,
+      "loss": 0.4119,
+      "step": 1001
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.33127981424331665,
+      "learning_rate": 9.423012471764914e-07,
+      "loss": 0.1965,
+      "step": 1002
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.35370326042175293,
+      "learning_rate": 9.131736525621603e-07,
+      "loss": 0.2197,
+      "step": 1003
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.40955832600593567,
+      "learning_rate": 8.844991884401854e-07,
+      "loss": 0.2728,
+      "step": 1004
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3946859538555145,
+      "learning_rate": 8.56278119510362e-07,
+      "loss": 0.2659,
+      "step": 1005
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.4458279609680176,
+      "learning_rate": 8.285107062871333e-07,
+      "loss": 0.348,
+      "step": 1006
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.35338565707206726,
+      "learning_rate": 8.011972050971483e-07,
+      "loss": 0.2703,
+      "step": 1007
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.39034345746040344,
+      "learning_rate": 7.74337868076902e-07,
+      "loss": 0.3125,
+      "step": 1008
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.44654595851898193,
+      "learning_rate": 7.479329431703985e-07,
+      "loss": 0.399,
+      "step": 1009
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.4880700409412384,
+      "learning_rate": 7.21982674126881e-07,
+      "loss": 0.3532,
+      "step": 1010
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.35642099380493164,
+      "learning_rate": 6.964873004985717e-07,
+      "loss": 0.2302,
+      "step": 1011
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.4704366624355316,
+      "learning_rate": 6.714470576384579e-07,
+      "loss": 0.3791,
+      "step": 1012
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3759339153766632,
+      "learning_rate": 6.468621766981154e-07,
+      "loss": 0.2714,
+      "step": 1013
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3735077381134033,
+      "learning_rate": 6.227328846255931e-07,
+      "loss": 0.2643,
+      "step": 1014
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.36363470554351807,
+      "learning_rate": 5.990594041632991e-07,
+      "loss": 0.2521,
+      "step": 1015
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3581157624721527,
+      "learning_rate": 5.758419538459459e-07,
+      "loss": 0.268,
+      "step": 1016
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.4589844048023224,
+      "learning_rate": 5.530807479985633e-07,
+      "loss": 0.3436,
+      "step": 1017
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.32665497064590454,
+      "learning_rate": 5.307759967344672e-07,
+      "loss": 0.2617,
+      "step": 1018
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4270218312740326,
+      "learning_rate": 5.089279059533658e-07,
+      "loss": 0.3261,
+      "step": 1019
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4280019998550415,
+      "learning_rate": 4.87536677339423e-07,
+      "loss": 0.348,
+      "step": 1020
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3628242313861847,
+      "learning_rate": 4.666025083594483e-07,
+      "loss": 0.2567,
+      "step": 1021
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.40238940715789795,
+      "learning_rate": 4.461255922609986e-07,
+      "loss": 0.301,
+      "step": 1022
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4348450005054474,
+      "learning_rate": 4.261061180706627e-07,
+      "loss": 0.3572,
+      "step": 1023
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3945859968662262,
+      "learning_rate": 4.065442705922906e-07,
+      "loss": 0.3631,
+      "step": 1024
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4079526662826538,
+      "learning_rate": 3.8744023040528374e-07,
+      "loss": 0.2975,
+      "step": 1025
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4182088375091553,
+      "learning_rate": 3.687941738629186e-07,
+      "loss": 0.3489,
+      "step": 1026
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3973821997642517,
+      "learning_rate": 3.5060627309074224e-07,
+      "loss": 0.2775,
+      "step": 1027
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.406399667263031,
+      "learning_rate": 3.3287669598497383e-07,
+      "loss": 0.2776,
+      "step": 1028
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.44541242718696594,
+      "learning_rate": 3.156056062109503e-07,
+      "loss": 0.3415,
+      "step": 1029
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3318565785884857,
+      "learning_rate": 2.987931632016272e-07,
+      "loss": 0.2179,
+      "step": 1030
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4106653332710266,
+      "learning_rate": 2.824395221560805e-07,
+      "loss": 0.2856,
+      "step": 1031
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3946784436702728,
+      "learning_rate": 2.665448340381016e-07,
+      "loss": 0.2658,
+      "step": 1032
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3629414141178131,
+      "learning_rate": 2.511092455747932e-07,
+      "loss": 0.2869,
+      "step": 1033
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.37011194229125977,
+      "learning_rate": 2.361328992552314e-07,
+      "loss": 0.2481,
+      "step": 1034
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.47178444266319275,
+      "learning_rate": 2.2161593332910013e-07,
+      "loss": 0.3622,
+      "step": 1035
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.44122573733329773,
+      "learning_rate": 2.0755848180547543e-07,
+      "loss": 0.3903,
+      "step": 1036
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.32538658380508423,
+      "learning_rate": 1.9396067445155986e-07,
+      "loss": 0.2214,
+      "step": 1037
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.43575894832611084,
+      "learning_rate": 1.8082263679148337e-07,
+      "loss": 0.3991,
+      "step": 1038
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.42153045535087585,
+      "learning_rate": 1.681444901051432e-07,
+      "loss": 0.3468,
+      "step": 1039
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.39070653915405273,
+      "learning_rate": 1.5592635142709367e-07,
+      "loss": 0.2821,
+      "step": 1040
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.373835951089859,
+      "learning_rate": 1.4416833354546356e-07,
+      "loss": 0.2638,
+      "step": 1041
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.38927844166755676,
+      "learning_rate": 1.328705450009071e-07,
+      "loss": 0.2571,
+      "step": 1042
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.4438738226890564,
+      "learning_rate": 1.2203309008561592e-07,
+      "loss": 0.3562,
+      "step": 1043
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.4671138525009155,
+      "learning_rate": 1.1165606884234181e-07,
+      "loss": 0.3204,
+      "step": 1044
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.43793073296546936,
+      "learning_rate": 1.0173957706348659e-07,
+      "loss": 0.3208,
+      "step": 1045
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.40344715118408203,
+      "learning_rate": 9.228370629019711e-08,
+      "loss": 0.2873,
+      "step": 1046
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.425571471452713,
+      "learning_rate": 8.328854381154938e-08,
+      "loss": 0.3374,
+      "step": 1047
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.3397400677204132,
+      "learning_rate": 7.475417266371576e-08,
+      "loss": 0.2717,
+      "step": 1048
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.3563021719455719,
+      "learning_rate": 6.668067162921566e-08,
+      "loss": 0.2625,
+      "step": 1049
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.39349859952926636,
+      "learning_rate": 5.906811523618272e-08,
+      "loss": 0.2895,
+      "step": 1050
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.44752970337867737,
+      "learning_rate": 5.191657375767656e-08,
+      "loss": 0.3835,
+      "step": 1051
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.40532049536705017,
+      "learning_rate": 4.522611321103321e-08,
+      "loss": 0.2817,
+      "step": 1052
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.35706475377082825,
+      "learning_rate": 3.8996795357254535e-08,
+      "loss": 0.2191,
+      "step": 1053
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5339099168777466,
+      "learning_rate": 3.322867770044202e-08,
+      "loss": 0.3686,
+      "step": 1054
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.47053080797195435,
+      "learning_rate": 2.792181348726941e-08,
+      "loss": 0.3318,
+      "step": 1055
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.4522939920425415,
+      "learning_rate": 2.3076251706477536e-08,
+      "loss": 0.3358,
+      "step": 1056
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.46299952268600464,
+      "learning_rate": 1.869203708843581e-08,
+      "loss": 0.3769,
+      "step": 1057
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3941827714443207,
+      "learning_rate": 1.476921010471477e-08,
+      "loss": 0.3189,
+      "step": 1058
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3861558139324188,
+      "learning_rate": 1.1307806967741919e-08,
+      "loss": 0.2869,
+      "step": 1059
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.32701894640922546,
+      "learning_rate": 8.307859630429793e-09,
+      "loss": 0.2303,
+      "step": 1060
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.44485971331596375,
+      "learning_rate": 5.7693957858984125e-09,
+      "loss": 0.3537,
+      "step": 1061
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.4371122419834137,
+      "learning_rate": 3.6924388672254785e-09,
+      "loss": 0.3147,
+      "step": 1062
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.37674111127853394,
+      "learning_rate": 2.0770080472298783e-09,
+      "loss": 0.286,
+      "step": 1063
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.48630061745643616,
+      "learning_rate": 9.231182382773984e-10,
+      "loss": 0.3817,
+      "step": 1064
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.37980979681015015,
+      "learning_rate": 2.307800921641512e-10,
+      "loss": 0.283,
+      "step": 1065
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.4119957387447357,
+      "learning_rate": 0.0,
+      "loss": 0.2733,
+      "step": 1066
+    },
+    {
+      "epoch": 1.0,
+      "step": 1066,
+      "total_flos": 1.171159722670162e+17,
+      "train_loss": 0.3121360269205655,
+      "train_runtime": 51953.7782,
+      "train_samples_per_second": 0.657,
+      "train_steps_per_second": 0.021
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 1066,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 100,
+  "total_flos": 1.171159722670162e+17,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}