diff --git a/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa-streaming_baseline_only_lora_agg_fc4_s2048/README.md b/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa-streaming_baseline_only_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_only_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_only_lora_agg_fc4_s2048/adapter_config.json b/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa-streaming_baseline_only_lora_agg_fc4_s2048/adapter_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..bdf910678210d8ccdf49a4409f54a5815db54f77
--- /dev/null
+++ b/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa-streaming_baseline_only_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": [
+    "down_proj",
+    "v_proj",
+    "q_proj",
+    "k_proj",
+    "o_proj",
+    "gate_proj",
+    "up_proj"
+  ],
+  "task_type": "CAUSAL_LM"
+}
\ No newline at end of file
diff --git a/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa-streaming_baseline_only_lora_agg_fc4_s2048/adapter_model.bin b/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa-streaming_baseline_only_lora_agg_fc4_s2048/adapter_model.bin
new file mode 100644
index 0000000000000000000000000000000000000000..04a05f0c6693085b6b1fe856c3e75d8307fd9be3
--- /dev/null
+++ b/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa-streaming_baseline_only_lora_agg_fc4_s2048/adapter_model.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d7d6c3ea53b723a805f00cdad89353275883987aaba37e6b1eb5c9792ec4fde7
+size 692127130
diff --git a/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa-streaming_baseline_only_lora_agg_fc4_s2048/config.json b/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa-streaming_baseline_only_lora_agg_fc4_s2048/config.json
new file mode 100644
index 0000000000000000000000000000000000000000..296c88047f8e2b39d47783382a227fda663ab28a
--- /dev/null
+++ b/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa-streaming_baseline_only_lora_agg_fc4_s2048/config.json
@@ -0,0 +1,68 @@
+{
+  "_name_or_path": "/data/wiedmann/hub/models--lmms-lab--llava-onevision-qwen2-7b-ov/snapshots/0b07bf7565e244cf4f39982249eafe8cd799d6dd",
+  "add_faster_video": false,
+  "add_time_instruction": false,
+  "architectures": [
+    "LlavaQwenForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 151643,
+  "eos_token_id": 151645,
+  "faster_token_stride": 10,
+  "force_sample": false,
+  "hidden_act": "silu",
+  "hidden_size": 3584,
+  "image_aspect_ratio": "square",
+  "image_crop_resolution": null,
+  "image_grid_pinpoints": null,
+  "image_split_resolution": null,
+  "image_token_index": 151646,
+  "initializer_range": 0.02,
+  "intermediate_size": 18944,
+  "max_position_embeddings": 32768,
+  "max_window_layers": 28,
+  "mm_hidden_size": 1152,
+  "mm_newline_position": "no_token",
+  "mm_patch_merge_type": "spatial_unpad",
+  "mm_projector_lr": null,
+  "mm_projector_type": "mlp2x_gelu",
+  "mm_resampler_type": "streaming_agg",
+  "mm_spatial_pool_mode": "bilinear",
+  "mm_spatial_pool_stride": null,
+  "mm_streaming_frames_per_chunk": 4,
+  "mm_streaming_input_dim": 1152,
+  "mm_streaming_mlp_ratio": 4.0,
+  "mm_streaming_num_heads": 8,
+  "mm_streaming_num_layers": 4,
+  "mm_streaming_num_state_tokens": 2048,
+  "mm_streaming_patches_per_frame": 729,
+  "mm_streaming_state_dim": 1152,
+  "mm_streaming_vision_chunk_size": 8,
+  "mm_tunable_parts": "mm_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_only_lora_agg_fc4_s2048/generation_config.json b/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa-streaming_baseline_only_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_only_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_only_lora_agg_fc4_s2048/non_lora_trainables.bin b/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa-streaming_baseline_only_lora_agg_fc4_s2048/non_lora_trainables.bin
new file mode 100644
index 0000000000000000000000000000000000000000..79e0264811faf21b6cd2191fec21518e015bd0fc
--- /dev/null
+++ b/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa-streaming_baseline_only_lora_agg_fc4_s2048/non_lora_trainables.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:33845fa64adaa245f31c2ee34d19c4324d2bc3d1624b344301c3863c804ae76d
+size 177478484
diff --git a/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa-streaming_baseline_only_lora_agg_fc4_s2048/trainer_state.json b/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa-streaming_baseline_only_lora_agg_fc4_s2048/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..ecbae5742f44c72311dd26e5aefb8e9d6f722d62
--- /dev/null
+++ b/llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-nextqa-streaming_baseline_only_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": 18.367340087890625,
+      "learning_rate": 3.125e-06,
+      "loss": 1.1997,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 20.36041259765625,
+      "learning_rate": 6.25e-06,
+      "loss": 1.2617,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 15.435030937194824,
+      "learning_rate": 9.375000000000001e-06,
+      "loss": 1.1302,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 11.671319007873535,
+      "learning_rate": 1.25e-05,
+      "loss": 0.9212,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.698945045471191,
+      "learning_rate": 1.5625e-05,
+      "loss": 0.8626,
+      "step": 5
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.7610063552856445,
+      "learning_rate": 1.8750000000000002e-05,
+      "loss": 0.6768,
+      "step": 6
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.763170003890991,
+      "learning_rate": 2.1875e-05,
+      "loss": 0.554,
+      "step": 7
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.646397352218628,
+      "learning_rate": 2.5e-05,
+      "loss": 0.4462,
+      "step": 8
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2087757587432861,
+      "learning_rate": 2.8125000000000003e-05,
+      "loss": 0.4335,
+      "step": 9
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0779973268508911,
+      "learning_rate": 3.125e-05,
+      "loss": 0.3426,
+      "step": 10
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8232861757278442,
+      "learning_rate": 3.4375e-05,
+      "loss": 0.2855,
+      "step": 11
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7847572565078735,
+      "learning_rate": 3.7500000000000003e-05,
+      "loss": 0.429,
+      "step": 12
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9708127975463867,
+      "learning_rate": 4.0625000000000005e-05,
+      "loss": 0.5351,
+      "step": 13
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.680749535560608,
+      "learning_rate": 4.375e-05,
+      "loss": 0.3552,
+      "step": 14
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.754399299621582,
+      "learning_rate": 4.6875e-05,
+      "loss": 0.4887,
+      "step": 15
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8019405007362366,
+      "learning_rate": 5e-05,
+      "loss": 0.3746,
+      "step": 16
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.1756885051727295,
+      "learning_rate": 5.3125000000000004e-05,
+      "loss": 0.2939,
+      "step": 17
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1910041570663452,
+      "learning_rate": 5.6250000000000005e-05,
+      "loss": 0.3498,
+      "step": 18
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7563303709030151,
+      "learning_rate": 5.9375e-05,
+      "loss": 0.3151,
+      "step": 19
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6908413767814636,
+      "learning_rate": 6.25e-05,
+      "loss": 0.427,
+      "step": 20
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5701287388801575,
+      "learning_rate": 6.562500000000001e-05,
+      "loss": 0.3341,
+      "step": 21
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6663073301315308,
+      "learning_rate": 6.875e-05,
+      "loss": 0.4151,
+      "step": 22
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.647335410118103,
+      "learning_rate": 7.1875e-05,
+      "loss": 0.3664,
+      "step": 23
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.250497817993164,
+      "learning_rate": 7.500000000000001e-05,
+      "loss": 0.3233,
+      "step": 24
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6220181584358215,
+      "learning_rate": 7.8125e-05,
+      "loss": 0.3933,
+      "step": 25
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6255826354026794,
+      "learning_rate": 8.125000000000001e-05,
+      "loss": 0.4192,
+      "step": 26
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7323064804077148,
+      "learning_rate": 8.4375e-05,
+      "loss": 0.4033,
+      "step": 27
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8183202147483826,
+      "learning_rate": 8.75e-05,
+      "loss": 0.3325,
+      "step": 28
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.56742924451828,
+      "learning_rate": 9.062500000000001e-05,
+      "loss": 0.3638,
+      "step": 29
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6514180898666382,
+      "learning_rate": 9.375e-05,
+      "loss": 0.4459,
+      "step": 30
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5290864109992981,
+      "learning_rate": 9.687500000000001e-05,
+      "loss": 0.3277,
+      "step": 31
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.47609055042266846,
+      "learning_rate": 0.0001,
+      "loss": 0.2922,
+      "step": 32
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5185883641242981,
+      "learning_rate": 9.999976921990784e-05,
+      "loss": 0.3217,
+      "step": 33
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5954408049583435,
+      "learning_rate": 9.999907688176173e-05,
+      "loss": 0.4124,
+      "step": 34
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4328826069831848,
+      "learning_rate": 9.999792299195278e-05,
+      "loss": 0.2708,
+      "step": 35
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.629949152469635,
+      "learning_rate": 9.999630756113278e-05,
+      "loss": 0.3774,
+      "step": 36
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4580020010471344,
+      "learning_rate": 9.999423060421411e-05,
+      "loss": 0.3097,
+      "step": 37
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.48158058524131775,
+      "learning_rate": 9.999169214036958e-05,
+      "loss": 0.3514,
+      "step": 38
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4538012742996216,
+      "learning_rate": 9.998869219303227e-05,
+      "loss": 0.3093,
+      "step": 39
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5379505753517151,
+      "learning_rate": 9.998523078989529e-05,
+      "loss": 0.3777,
+      "step": 40
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.445813924074173,
+      "learning_rate": 9.998130796291156e-05,
+      "loss": 0.3601,
+      "step": 41
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5426644682884216,
+      "learning_rate": 9.997692374829352e-05,
+      "loss": 0.3143,
+      "step": 42
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4094751179218292,
+      "learning_rate": 9.997207818651274e-05,
+      "loss": 0.2912,
+      "step": 43
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4211678206920624,
+      "learning_rate": 9.996677132229957e-05,
+      "loss": 0.3235,
+      "step": 44
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4476563334465027,
+      "learning_rate": 9.996100320464274e-05,
+      "loss": 0.3559,
+      "step": 45
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4389052093029022,
+      "learning_rate": 9.995477388678897e-05,
+      "loss": 0.2772,
+      "step": 46
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5050854086875916,
+      "learning_rate": 9.994808342624234e-05,
+      "loss": 0.2992,
+      "step": 47
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.39739540219306946,
+      "learning_rate": 9.994093188476382e-05,
+      "loss": 0.2832,
+      "step": 48
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4640009105205536,
+      "learning_rate": 9.993331932837079e-05,
+      "loss": 0.355,
+      "step": 49
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5145576596260071,
+      "learning_rate": 9.992524582733629e-05,
+      "loss": 0.2745,
+      "step": 50
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4169548451900482,
+      "learning_rate": 9.991671145618846e-05,
+      "loss": 0.2376,
+      "step": 51
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.46700263023376465,
+      "learning_rate": 9.99077162937098e-05,
+      "loss": 0.3257,
+      "step": 52
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4765835702419281,
+      "learning_rate": 9.989826042293652e-05,
+      "loss": 0.3293,
+      "step": 53
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.48216092586517334,
+      "learning_rate": 9.988834393115767e-05,
+      "loss": 0.3152,
+      "step": 54
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.518074631690979,
+      "learning_rate": 9.987796690991439e-05,
+      "loss": 0.3669,
+      "step": 55
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4637100100517273,
+      "learning_rate": 9.98671294549991e-05,
+      "loss": 0.3148,
+      "step": 56
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.43243342638015747,
+      "learning_rate": 9.985583166645455e-05,
+      "loss": 0.2475,
+      "step": 57
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4122597277164459,
+      "learning_rate": 9.98440736485729e-05,
+      "loss": 0.2878,
+      "step": 58
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4131738841533661,
+      "learning_rate": 9.983185550989487e-05,
+      "loss": 0.2217,
+      "step": 59
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4752655327320099,
+      "learning_rate": 9.981917736320851e-05,
+      "loss": 0.3538,
+      "step": 60
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.37928274273872375,
+      "learning_rate": 9.980603932554845e-05,
+      "loss": 0.1967,
+      "step": 61
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4203208386898041,
+      "learning_rate": 9.979244151819453e-05,
+      "loss": 0.2483,
+      "step": 62
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.49629998207092285,
+      "learning_rate": 9.97783840666709e-05,
+      "loss": 0.3649,
+      "step": 63
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4602612853050232,
+      "learning_rate": 9.976386710074478e-05,
+      "loss": 0.3001,
+      "step": 64
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.43678635358810425,
+      "learning_rate": 9.974889075442521e-05,
+      "loss": 0.2301,
+      "step": 65
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.49575650691986084,
+      "learning_rate": 9.97334551659619e-05,
+      "loss": 0.3064,
+      "step": 66
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4498412609100342,
+      "learning_rate": 9.971756047784393e-05,
+      "loss": 0.3118,
+      "step": 67
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4476684629917145,
+      "learning_rate": 9.970120683679838e-05,
+      "loss": 0.3873,
+      "step": 68
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4587801396846771,
+      "learning_rate": 9.968439439378905e-05,
+      "loss": 0.3894,
+      "step": 69
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4500073790550232,
+      "learning_rate": 9.966712330401504e-05,
+      "loss": 0.3612,
+      "step": 70
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4947085976600647,
+      "learning_rate": 9.964939372690926e-05,
+      "loss": 0.4023,
+      "step": 71
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4781345725059509,
+      "learning_rate": 9.96312058261371e-05,
+      "loss": 0.3429,
+      "step": 72
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.43458712100982666,
+      "learning_rate": 9.961255976959473e-05,
+      "loss": 0.3179,
+      "step": 73
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.49464428424835205,
+      "learning_rate": 9.959345572940771e-05,
+      "loss": 0.4133,
+      "step": 74
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.38269466161727905,
+      "learning_rate": 9.957389388192935e-05,
+      "loss": 0.2905,
+      "step": 75
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.43023020029067993,
+      "learning_rate": 9.9553874407739e-05,
+      "loss": 0.2674,
+      "step": 76
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4165527820587158,
+      "learning_rate": 9.953339749164057e-05,
+      "loss": 0.3692,
+      "step": 77
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.44769737124443054,
+      "learning_rate": 9.951246332266057e-05,
+      "loss": 0.3606,
+      "step": 78
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.45688146352767944,
+      "learning_rate": 9.949107209404665e-05,
+      "loss": 0.4039,
+      "step": 79
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4197036623954773,
+      "learning_rate": 9.946922400326554e-05,
+      "loss": 0.377,
+      "step": 80
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4688505232334137,
+      "learning_rate": 9.944691925200145e-05,
+      "loss": 0.3651,
+      "step": 81
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4570940136909485,
+      "learning_rate": 9.942415804615406e-05,
+      "loss": 0.3847,
+      "step": 82
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4865867495536804,
+      "learning_rate": 9.940094059583671e-05,
+      "loss": 0.3874,
+      "step": 83
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4403970539569855,
+      "learning_rate": 9.937726711537442e-05,
+      "loss": 0.2722,
+      "step": 84
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.41962695121765137,
+      "learning_rate": 9.93531378233019e-05,
+      "loss": 0.341,
+      "step": 85
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5067736506462097,
+      "learning_rate": 9.932855294236154e-05,
+      "loss": 0.3613,
+      "step": 86
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5133152604103088,
+      "learning_rate": 9.930351269950143e-05,
+      "loss": 0.4087,
+      "step": 87
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4906843602657318,
+      "learning_rate": 9.927801732587312e-05,
+      "loss": 0.3011,
+      "step": 88
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3938533067703247,
+      "learning_rate": 9.925206705682962e-05,
+      "loss": 0.2591,
+      "step": 89
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5528268814086914,
+      "learning_rate": 9.92256621319231e-05,
+      "loss": 0.4683,
+      "step": 90
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.47138574719429016,
+      "learning_rate": 9.919880279490286e-05,
+      "loss": 0.3129,
+      "step": 91
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3691083788871765,
+      "learning_rate": 9.917148929371288e-05,
+      "loss": 0.1754,
+      "step": 92
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.49399805068969727,
+      "learning_rate": 9.914372188048964e-05,
+      "loss": 0.3776,
+      "step": 93
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.39202970266342163,
+      "learning_rate": 9.911550081155983e-05,
+      "loss": 0.2701,
+      "step": 94
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4894983768463135,
+      "learning_rate": 9.908682634743784e-05,
+      "loss": 0.3881,
+      "step": 95
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.41826292872428894,
+      "learning_rate": 9.905769875282352e-05,
+      "loss": 0.2901,
+      "step": 96
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.44304174184799194,
+      "learning_rate": 9.902811829659961e-05,
+      "loss": 0.3045,
+      "step": 97
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5131700038909912,
+      "learning_rate": 9.899808525182935e-05,
+      "loss": 0.3788,
+      "step": 98
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3887397348880768,
+      "learning_rate": 9.896759989575386e-05,
+      "loss": 0.2958,
+      "step": 99
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.463975727558136,
+      "learning_rate": 9.893666250978971e-05,
+      "loss": 0.379,
+      "step": 100
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3379839062690735,
+      "learning_rate": 9.890527337952617e-05,
+      "loss": 0.203,
+      "step": 101
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4153408110141754,
+      "learning_rate": 9.887343279472272e-05,
+      "loss": 0.2995,
+      "step": 102
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.46450096368789673,
+      "learning_rate": 9.884114104930628e-05,
+      "loss": 0.3221,
+      "step": 103
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.39753568172454834,
+      "learning_rate": 9.880839844136854e-05,
+      "loss": 0.2657,
+      "step": 104
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.44558021426200867,
+      "learning_rate": 9.877520527316317e-05,
+      "loss": 0.3989,
+      "step": 105
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3921840786933899,
+      "learning_rate": 9.874156185110306e-05,
+      "loss": 0.3058,
+      "step": 106
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4834759533405304,
+      "learning_rate": 9.870746848575751e-05,
+      "loss": 0.3249,
+      "step": 107
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.45938247442245483,
+      "learning_rate": 9.86729254918493e-05,
+      "loss": 0.279,
+      "step": 108
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.39709120988845825,
+      "learning_rate": 9.863793318825186e-05,
+      "loss": 0.2953,
+      "step": 109
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4169032573699951,
+      "learning_rate": 9.860249189798627e-05,
+      "loss": 0.409,
+      "step": 110
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3651694357395172,
+      "learning_rate": 9.856660194821829e-05,
+      "loss": 0.2225,
+      "step": 111
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4336308240890503,
+      "learning_rate": 9.853026367025535e-05,
+      "loss": 0.3209,
+      "step": 112
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3963097631931305,
+      "learning_rate": 9.849347739954352e-05,
+      "loss": 0.2915,
+      "step": 113
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5299807190895081,
+      "learning_rate": 9.845624347566433e-05,
+      "loss": 0.4689,
+      "step": 114
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.37367191910743713,
+      "learning_rate": 9.841856224233174e-05,
+      "loss": 0.2219,
+      "step": 115
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.473271906375885,
+      "learning_rate": 9.83804340473889e-05,
+      "loss": 0.4204,
+      "step": 116
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.41578221321105957,
+      "learning_rate": 9.83418592428049e-05,
+      "loss": 0.3003,
+      "step": 117
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4305388033390045,
+      "learning_rate": 9.830283818467163e-05,
+      "loss": 0.3411,
+      "step": 118
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4396352767944336,
+      "learning_rate": 9.826337123320046e-05,
+      "loss": 0.328,
+      "step": 119
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.44212085008621216,
+      "learning_rate": 9.822345875271883e-05,
+      "loss": 0.343,
+      "step": 120
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4692837595939636,
+      "learning_rate": 9.818310111166699e-05,
+      "loss": 0.3424,
+      "step": 121
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5092015266418457,
+      "learning_rate": 9.814229868259452e-05,
+      "loss": 0.3566,
+      "step": 122
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4727444052696228,
+      "learning_rate": 9.810105184215699e-05,
+      "loss": 0.4161,
+      "step": 123
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.42226845026016235,
+      "learning_rate": 9.805936097111234e-05,
+      "loss": 0.2661,
+      "step": 124
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5039162635803223,
+      "learning_rate": 9.801722645431754e-05,
+      "loss": 0.4034,
+      "step": 125
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4233477711677551,
+      "learning_rate": 9.797464868072488e-05,
+      "loss": 0.3068,
+      "step": 126
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4316577613353729,
+      "learning_rate": 9.793162804337845e-05,
+      "loss": 0.3214,
+      "step": 127
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.44522222876548767,
+      "learning_rate": 9.788816493941051e-05,
+      "loss": 0.346,
+      "step": 128
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.44972801208496094,
+      "learning_rate": 9.784425977003784e-05,
+      "loss": 0.357,
+      "step": 129
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4264127016067505,
+      "learning_rate": 9.779991294055802e-05,
+      "loss": 0.3754,
+      "step": 130
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3643350303173065,
+      "learning_rate": 9.775512486034563e-05,
+      "loss": 0.2567,
+      "step": 131
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5998265147209167,
+      "learning_rate": 9.770989594284857e-05,
+      "loss": 0.5771,
+      "step": 132
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4687020480632782,
+      "learning_rate": 9.766422660558421e-05,
+      "loss": 0.3903,
+      "step": 133
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3895106017589569,
+      "learning_rate": 9.761811727013548e-05,
+      "loss": 0.2792,
+      "step": 134
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.37773844599723816,
+      "learning_rate": 9.757156836214706e-05,
+      "loss": 0.3133,
+      "step": 135
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.39893820881843567,
+      "learning_rate": 9.752458031132141e-05,
+      "loss": 0.3563,
+      "step": 136
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.38628295063972473,
+      "learning_rate": 9.747715355141478e-05,
+      "loss": 0.2313,
+      "step": 137
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3644122779369354,
+      "learning_rate": 9.742928852023325e-05,
+      "loss": 0.3237,
+      "step": 138
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4312066435813904,
+      "learning_rate": 9.73809856596287e-05,
+      "loss": 0.3646,
+      "step": 139
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4073641002178192,
+      "learning_rate": 9.733224541549464e-05,
+      "loss": 0.3162,
+      "step": 140
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4195702373981476,
+      "learning_rate": 9.728306823776221e-05,
+      "loss": 0.3279,
+      "step": 141
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4012298882007599,
+      "learning_rate": 9.723345458039594e-05,
+      "loss": 0.3518,
+      "step": 142
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3413154184818268,
+      "learning_rate": 9.718340490138965e-05,
+      "loss": 0.1902,
+      "step": 143
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.45729732513427734,
+      "learning_rate": 9.713291966276206e-05,
+      "loss": 0.3451,
+      "step": 144
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4502311050891876,
+      "learning_rate": 9.708199933055272e-05,
+      "loss": 0.4213,
+      "step": 145
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.47957178950309753,
+      "learning_rate": 9.70306443748176e-05,
+      "loss": 0.4034,
+      "step": 146
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.37078186869621277,
+      "learning_rate": 9.697885526962474e-05,
+      "loss": 0.2809,
+      "step": 147
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.41138362884521484,
+      "learning_rate": 9.692663249304992e-05,
+      "loss": 0.3335,
+      "step": 148
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4864597022533417,
+      "learning_rate": 9.687397652717223e-05,
+      "loss": 0.4674,
+      "step": 149
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4568451941013336,
+      "learning_rate": 9.682088785806963e-05,
+      "loss": 0.3774,
+      "step": 150
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4001672565937042,
+      "learning_rate": 9.67673669758144e-05,
+      "loss": 0.2757,
+      "step": 151
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.42816731333732605,
+      "learning_rate": 9.671341437446877e-05,
+      "loss": 0.3363,
+      "step": 152
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5091586112976074,
+      "learning_rate": 9.665903055208014e-05,
+      "loss": 0.4422,
+      "step": 153
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4335111975669861,
+      "learning_rate": 9.660421601067666e-05,
+      "loss": 0.3311,
+      "step": 154
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.36849528551101685,
+      "learning_rate": 9.654897125626252e-05,
+      "loss": 0.2864,
+      "step": 155
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.38034534454345703,
+      "learning_rate": 9.649329679881334e-05,
+      "loss": 0.2848,
+      "step": 156
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.30810433626174927,
+      "learning_rate": 9.643719315227133e-05,
+      "loss": 0.1939,
+      "step": 157
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4283190965652466,
+      "learning_rate": 9.63806608345407e-05,
+      "loss": 0.3569,
+      "step": 158
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4550926387310028,
+      "learning_rate": 9.632370036748279e-05,
+      "loss": 0.3769,
+      "step": 159
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4202008545398712,
+      "learning_rate": 9.626631227691127e-05,
+      "loss": 0.288,
+      "step": 160
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3845793306827545,
+      "learning_rate": 9.62084970925873e-05,
+      "loss": 0.2443,
+      "step": 161
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.41995301842689514,
+      "learning_rate": 9.615025534821462e-05,
+      "loss": 0.2772,
+      "step": 162
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4846205711364746,
+      "learning_rate": 9.609158758143464e-05,
+      "loss": 0.3856,
+      "step": 163
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4353344440460205,
+      "learning_rate": 9.603249433382144e-05,
+      "loss": 0.2875,
+      "step": 164
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4365904629230499,
+      "learning_rate": 9.597297615087685e-05,
+      "loss": 0.2667,
+      "step": 165
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.38388529419898987,
+      "learning_rate": 9.591303358202535e-05,
+      "loss": 0.2221,
+      "step": 166
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.45872360467910767,
+      "learning_rate": 9.585266718060897e-05,
+      "loss": 0.3565,
+      "step": 167
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.40336596965789795,
+      "learning_rate": 9.579187750388227e-05,
+      "loss": 0.2167,
+      "step": 168
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.43868064880371094,
+      "learning_rate": 9.573066511300714e-05,
+      "loss": 0.3417,
+      "step": 169
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.43917733430862427,
+      "learning_rate": 9.566903057304764e-05,
+      "loss": 0.3646,
+      "step": 170
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.44352856278419495,
+      "learning_rate": 9.560697445296474e-05,
+      "loss": 0.3252,
+      "step": 171
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5583091378211975,
+      "learning_rate": 9.554449732561113e-05,
+      "loss": 0.351,
+      "step": 172
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.43708336353302,
+      "learning_rate": 9.548159976772592e-05,
+      "loss": 0.3471,
+      "step": 173
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4961797297000885,
+      "learning_rate": 9.541828235992926e-05,
+      "loss": 0.3535,
+      "step": 174
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.380731999874115,
+      "learning_rate": 9.535454568671704e-05,
+      "loss": 0.2705,
+      "step": 175
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.44424885511398315,
+      "learning_rate": 9.529039033645548e-05,
+      "loss": 0.2366,
+      "step": 176
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3777495324611664,
+      "learning_rate": 9.522581690137567e-05,
+      "loss": 0.305,
+      "step": 177
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.35404613614082336,
+      "learning_rate": 9.516082597756815e-05,
+      "loss": 0.2314,
+      "step": 178
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4622362554073334,
+      "learning_rate": 9.509541816497737e-05,
+      "loss": 0.4166,
+      "step": 179
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.41214224696159363,
+      "learning_rate": 9.50295940673962e-05,
+      "loss": 0.253,
+      "step": 180
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.442114919424057,
+      "learning_rate": 9.496335429246026e-05,
+      "loss": 0.3943,
+      "step": 181
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4232344329357147,
+      "learning_rate": 9.489669945164242e-05,
+      "loss": 0.3273,
+      "step": 182
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.35261598229408264,
+      "learning_rate": 9.482963016024709e-05,
+      "loss": 0.222,
+      "step": 183
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3651808500289917,
+      "learning_rate": 9.476214703740454e-05,
+      "loss": 0.244,
+      "step": 184
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.509012758731842,
+      "learning_rate": 9.469425070606524e-05,
+      "loss": 0.5481,
+      "step": 185
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.40018436312675476,
+      "learning_rate": 9.462594179299406e-05,
+      "loss": 0.2319,
+      "step": 186
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4184328615665436,
+      "learning_rate": 9.45572209287645e-05,
+      "loss": 0.3469,
+      "step": 187
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5992292761802673,
+      "learning_rate": 9.44880887477528e-05,
+      "loss": 0.3817,
+      "step": 188
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4589436650276184,
+      "learning_rate": 9.441854588813228e-05,
+      "loss": 0.2914,
+      "step": 189
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.38457149267196655,
+      "learning_rate": 9.43485929918672e-05,
+      "loss": 0.3181,
+      "step": 190
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4888402819633484,
+      "learning_rate": 9.427823070470699e-05,
+      "loss": 0.4259,
+      "step": 191
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.37316766381263733,
+      "learning_rate": 9.420745967618026e-05,
+      "loss": 0.2753,
+      "step": 192
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.308633416891098,
+      "learning_rate": 9.413628055958878e-05,
+      "loss": 0.1875,
+      "step": 193
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.41204187273979187,
+      "learning_rate": 9.406469401200151e-05,
+      "loss": 0.3384,
+      "step": 194
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4597542881965637,
+      "learning_rate": 9.399270069424842e-05,
+      "loss": 0.3385,
+      "step": 195
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.46371549367904663,
+      "learning_rate": 9.392030127091452e-05,
+      "loss": 0.2874,
+      "step": 196
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3977811932563782,
+      "learning_rate": 9.384749641033359e-05,
+      "loss": 0.2797,
+      "step": 197
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4490582346916199,
+      "learning_rate": 9.377428678458214e-05,
+      "loss": 0.356,
+      "step": 198
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.41083958745002747,
+      "learning_rate": 9.370067306947316e-05,
+      "loss": 0.2747,
+      "step": 199
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.48376941680908203,
+      "learning_rate": 9.362665594454984e-05,
+      "loss": 0.4317,
+      "step": 200
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.44248223304748535,
+      "learning_rate": 9.355223609307933e-05,
+      "loss": 0.2799,
+      "step": 201
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4844808578491211,
+      "learning_rate": 9.347741420204643e-05,
+      "loss": 0.3624,
+      "step": 202
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4655194580554962,
+      "learning_rate": 9.340219096214727e-05,
+      "loss": 0.2899,
+      "step": 203
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4689055383205414,
+      "learning_rate": 9.33265670677829e-05,
+      "loss": 0.3721,
+      "step": 204
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4450901746749878,
+      "learning_rate": 9.325054321705289e-05,
+      "loss": 0.3079,
+      "step": 205
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4498708248138428,
+      "learning_rate": 9.317412011174886e-05,
+      "loss": 0.3111,
+      "step": 206
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.48850810527801514,
+      "learning_rate": 9.309729845734813e-05,
+      "loss": 0.4046,
+      "step": 207
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.37347522377967834,
+      "learning_rate": 9.302007896300698e-05,
+      "loss": 0.271,
+      "step": 208
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.42164427042007446,
+      "learning_rate": 9.29424623415543e-05,
+      "loss": 0.3197,
+      "step": 209
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4926708936691284,
+      "learning_rate": 9.286444930948496e-05,
+      "loss": 0.3602,
+      "step": 210
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3644413352012634,
+      "learning_rate": 9.278604058695313e-05,
+      "loss": 0.2342,
+      "step": 211
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.45351725816726685,
+      "learning_rate": 9.270723689776568e-05,
+      "loss": 0.3237,
+      "step": 212
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4711061418056488,
+      "learning_rate": 9.262803896937555e-05,
+      "loss": 0.4067,
+      "step": 213
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.43553972244262695,
+      "learning_rate": 9.254844753287493e-05,
+      "loss": 0.2759,
+      "step": 214
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4814302623271942,
+      "learning_rate": 9.24684633229886e-05,
+      "loss": 0.3125,
+      "step": 215
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5601657032966614,
+      "learning_rate": 9.238808707806706e-05,
+      "loss": 0.3608,
+      "step": 216
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4870937764644623,
+      "learning_rate": 9.230731954007983e-05,
+      "loss": 0.3721,
+      "step": 217
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5292274951934814,
+      "learning_rate": 9.222616145460849e-05,
+      "loss": 0.4209,
+      "step": 218
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.43616461753845215,
+      "learning_rate": 9.214461357083985e-05,
+      "loss": 0.2172,
+      "step": 219
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.361106812953949,
+      "learning_rate": 9.206267664155907e-05,
+      "loss": 0.2494,
+      "step": 220
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4230857491493225,
+      "learning_rate": 9.198035142314259e-05,
+      "loss": 0.3769,
+      "step": 221
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.44878044724464417,
+      "learning_rate": 9.189763867555129e-05,
+      "loss": 0.3755,
+      "step": 222
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5428657531738281,
+      "learning_rate": 9.181453916232339e-05,
+      "loss": 0.2581,
+      "step": 223
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.41663435101509094,
+      "learning_rate": 9.173105365056742e-05,
+      "loss": 0.3474,
+      "step": 224
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4844930171966553,
+      "learning_rate": 9.164718291095515e-05,
+      "loss": 0.3686,
+      "step": 225
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.40770450234413147,
+      "learning_rate": 9.156292771771446e-05,
+      "loss": 0.3386,
+      "step": 226
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.39481139183044434,
+      "learning_rate": 9.14782888486222e-05,
+      "loss": 0.3119,
+      "step": 227
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.38520610332489014,
+      "learning_rate": 9.1393267084997e-05,
+      "loss": 0.3014,
+      "step": 228
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.46185630559921265,
+      "learning_rate": 9.130786321169209e-05,
+      "loss": 0.3699,
+      "step": 229
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4153963029384613,
+      "learning_rate": 9.122207801708802e-05,
+      "loss": 0.3289,
+      "step": 230
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3800516724586487,
+      "learning_rate": 9.113591229308538e-05,
+      "loss": 0.244,
+      "step": 231
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.41449904441833496,
+      "learning_rate": 9.104936683509755e-05,
+      "loss": 0.3108,
+      "step": 232
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.39178889989852905,
+      "learning_rate": 9.096244244204324e-05,
+      "loss": 0.3124,
+      "step": 233
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4312244951725006,
+      "learning_rate": 9.087513991633924e-05,
+      "loss": 0.3155,
+      "step": 234
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.427442729473114,
+      "learning_rate": 9.078746006389298e-05,
+      "loss": 0.3512,
+      "step": 235
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.33089083433151245,
+      "learning_rate": 9.069940369409499e-05,
+      "loss": 0.1927,
+      "step": 236
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4994446635246277,
+      "learning_rate": 9.061097161981159e-05,
+      "loss": 0.4073,
+      "step": 237
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.47397279739379883,
+      "learning_rate": 9.052216465737726e-05,
+      "loss": 0.3754,
+      "step": 238
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4524490535259247,
+      "learning_rate": 9.043298362658714e-05,
+      "loss": 0.3629,
+      "step": 239
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.37869277596473694,
+      "learning_rate": 9.034342935068952e-05,
+      "loss": 0.1984,
+      "step": 240
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4369172155857086,
+      "learning_rate": 9.025350265637815e-05,
+      "loss": 0.3684,
+      "step": 241
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4749734699726105,
+      "learning_rate": 9.016320437378466e-05,
+      "loss": 0.3212,
+      "step": 242
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4339480996131897,
+      "learning_rate": 9.007253533647089e-05,
+      "loss": 0.3463,
+      "step": 243
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4679538309574127,
+      "learning_rate": 8.998149638142119e-05,
+      "loss": 0.326,
+      "step": 244
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.503963053226471,
+      "learning_rate": 8.98900883490347e-05,
+      "loss": 0.4358,
+      "step": 245
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.38353750109672546,
+      "learning_rate": 8.979831208311758e-05,
+      "loss": 0.2923,
+      "step": 246
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4563164710998535,
+      "learning_rate": 8.970616843087524e-05,
+      "loss": 0.3187,
+      "step": 247
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.36256474256515503,
+      "learning_rate": 8.96136582429045e-05,
+      "loss": 0.2175,
+      "step": 248
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4684314429759979,
+      "learning_rate": 8.952078237318575e-05,
+      "loss": 0.3427,
+      "step": 249
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.45074453949928284,
+      "learning_rate": 8.942754167907507e-05,
+      "loss": 0.2726,
+      "step": 250
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5070464015007019,
+      "learning_rate": 8.933393702129628e-05,
+      "loss": 0.2587,
+      "step": 251
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.44061899185180664,
+      "learning_rate": 8.923996926393305e-05,
+      "loss": 0.3961,
+      "step": 252
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4211723208427429,
+      "learning_rate": 8.91456392744209e-05,
+      "loss": 0.3661,
+      "step": 253
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4651309847831726,
+      "learning_rate": 8.905094792353917e-05,
+      "loss": 0.3365,
+      "step": 254
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4560156464576721,
+      "learning_rate": 8.895589608540297e-05,
+      "loss": 0.3558,
+      "step": 255
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3777078688144684,
+      "learning_rate": 8.886048463745525e-05,
+      "loss": 0.175,
+      "step": 256
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4200543463230133,
+      "learning_rate": 8.876471446045847e-05,
+      "loss": 0.2923,
+      "step": 257
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.44682615995407104,
+      "learning_rate": 8.866858643848665e-05,
+      "loss": 0.2414,
+      "step": 258
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3942912817001343,
+      "learning_rate": 8.857210145891715e-05,
+      "loss": 0.282,
+      "step": 259
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.41610509157180786,
+      "learning_rate": 8.847526041242246e-05,
+      "loss": 0.1799,
+      "step": 260
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.39465031027793884,
+      "learning_rate": 8.8378064192962e-05,
+      "loss": 0.2455,
+      "step": 261
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.45565101504325867,
+      "learning_rate": 8.82805136977739e-05,
+      "loss": 0.3674,
+      "step": 262
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4826546013355255,
+      "learning_rate": 8.818260982736661e-05,
+      "loss": 0.3592,
+      "step": 263
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.41021135449409485,
+      "learning_rate": 8.808435348551071e-05,
+      "loss": 0.2217,
+      "step": 264
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3801558315753937,
+      "learning_rate": 8.798574557923053e-05,
+      "loss": 0.2213,
+      "step": 265
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.37459704279899597,
+      "learning_rate": 8.788678701879573e-05,
+      "loss": 0.3109,
+      "step": 266
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.36728647351264954,
+      "learning_rate": 8.778747871771292e-05,
+      "loss": 0.2384,
+      "step": 267
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.44370126724243164,
+      "learning_rate": 8.768782159271727e-05,
+      "loss": 0.3362,
+      "step": 268
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3853534460067749,
+      "learning_rate": 8.758781656376398e-05,
+      "loss": 0.2443,
+      "step": 269
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.45374879240989685,
+      "learning_rate": 8.748746455401986e-05,
+      "loss": 0.3262,
+      "step": 270
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.47816920280456543,
+      "learning_rate": 8.738676648985476e-05,
+      "loss": 0.3123,
+      "step": 271
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4292296767234802,
+      "learning_rate": 8.7285723300833e-05,
+      "loss": 0.3724,
+      "step": 272
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3887421786785126,
+      "learning_rate": 8.718433591970485e-05,
+      "loss": 0.2855,
+      "step": 273
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.49635881185531616,
+      "learning_rate": 8.708260528239788e-05,
+      "loss": 0.3917,
+      "step": 274
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.39939600229263306,
+      "learning_rate": 8.698053232800832e-05,
+      "loss": 0.2616,
+      "step": 275
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.42968565225601196,
+      "learning_rate": 8.68781179987924e-05,
+      "loss": 0.3605,
+      "step": 276
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.385392427444458,
+      "learning_rate": 8.677536324015765e-05,
+      "loss": 0.2805,
+      "step": 277
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.41780418157577515,
+      "learning_rate": 8.667226900065419e-05,
+      "loss": 0.3242,
+      "step": 278
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.48109957575798035,
+      "learning_rate": 8.656883623196592e-05,
+      "loss": 0.3191,
+      "step": 279
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.45149287581443787,
+      "learning_rate": 8.646506588890183e-05,
+      "loss": 0.3069,
+      "step": 280
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.510420024394989,
+      "learning_rate": 8.636095892938707e-05,
+      "loss": 0.3468,
+      "step": 281
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.46956875920295715,
+      "learning_rate": 8.62565163144542e-05,
+      "loss": 0.4075,
+      "step": 282
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.40705329179763794,
+      "learning_rate": 8.615173900823426e-05,
+      "loss": 0.2867,
+      "step": 283
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3982028663158417,
+      "learning_rate": 8.60466279779479e-05,
+      "loss": 0.2893,
+      "step": 284
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.39112627506256104,
+      "learning_rate": 8.594118419389647e-05,
+      "loss": 0.3286,
+      "step": 285
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.439872682094574,
+      "learning_rate": 8.583540862945301e-05,
+      "loss": 0.307,
+      "step": 286
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.34527483582496643,
+      "learning_rate": 8.57293022610533e-05,
+      "loss": 0.2122,
+      "step": 287
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.41789668798446655,
+      "learning_rate": 8.562286606818684e-05,
+      "loss": 0.236,
+      "step": 288
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5295957922935486,
+      "learning_rate": 8.55161010333878e-05,
+      "loss": 0.504,
+      "step": 289
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5455564856529236,
+      "learning_rate": 8.540900814222598e-05,
+      "loss": 0.3537,
+      "step": 290
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5789703726768494,
+      "learning_rate": 8.530158838329765e-05,
+      "loss": 0.3238,
+      "step": 291
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5657013654708862,
+      "learning_rate": 8.519384274821649e-05,
+      "loss": 0.4341,
+      "step": 292
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4126622676849365,
+      "learning_rate": 8.508577223160442e-05,
+      "loss": 0.2597,
+      "step": 293
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.47889232635498047,
+      "learning_rate": 8.497737783108238e-05,
+      "loss": 0.3202,
+      "step": 294
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6022685766220093,
+      "learning_rate": 8.486866054726114e-05,
+      "loss": 0.3014,
+      "step": 295
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4188893437385559,
+      "learning_rate": 8.475962138373213e-05,
+      "loss": 0.2655,
+      "step": 296
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4596685767173767,
+      "learning_rate": 8.465026134705805e-05,
+      "loss": 0.3072,
+      "step": 297
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4453608989715576,
+      "learning_rate": 8.454058144676366e-05,
+      "loss": 0.2954,
+      "step": 298
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.40905341506004333,
+      "learning_rate": 8.443058269532651e-05,
+      "loss": 0.3124,
+      "step": 299
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.43399733304977417,
+      "learning_rate": 8.432026610816745e-05,
+      "loss": 0.2793,
+      "step": 300
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4337409734725952,
+      "learning_rate": 8.420963270364137e-05,
+      "loss": 0.2741,
+      "step": 301
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4746047854423523,
+      "learning_rate": 8.409868350302774e-05,
+      "loss": 0.3285,
+      "step": 302
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.45777222514152527,
+      "learning_rate": 8.398741953052127e-05,
+      "loss": 0.281,
+      "step": 303
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4513487219810486,
+      "learning_rate": 8.387584181322233e-05,
+      "loss": 0.3038,
+      "step": 304
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5118147730827332,
+      "learning_rate": 8.376395138112754e-05,
+      "loss": 0.3482,
+      "step": 305
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.42370742559432983,
+      "learning_rate": 8.365174926712032e-05,
+      "loss": 0.3431,
+      "step": 306
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4493395984172821,
+      "learning_rate": 8.353923650696118e-05,
+      "loss": 0.3475,
+      "step": 307
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4581238329410553,
+      "learning_rate": 8.342641413927837e-05,
+      "loss": 0.3637,
+      "step": 308
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4623344838619232,
+      "learning_rate": 8.331328320555812e-05,
+      "loss": 0.3504,
+      "step": 309
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5001980066299438,
+      "learning_rate": 8.319984475013512e-05,
+      "loss": 0.4016,
+      "step": 310
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4373179078102112,
+      "learning_rate": 8.308609982018286e-05,
+      "loss": 0.3375,
+      "step": 311
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.38677600026130676,
+      "learning_rate": 8.297204946570398e-05,
+      "loss": 0.2704,
+      "step": 312
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.44620731472969055,
+      "learning_rate": 8.285769473952052e-05,
+      "loss": 0.2628,
+      "step": 313
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.49608850479125977,
+      "learning_rate": 8.274303669726426e-05,
+      "loss": 0.3184,
+      "step": 314
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4633454382419586,
+      "learning_rate": 8.262807639736692e-05,
+      "loss": 0.404,
+      "step": 315
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.47432637214660645,
+      "learning_rate": 8.251281490105045e-05,
+      "loss": 0.3733,
+      "step": 316
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.38651883602142334,
+      "learning_rate": 8.239725327231721e-05,
+      "loss": 0.303,
+      "step": 317
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4193047285079956,
+      "learning_rate": 8.228139257794012e-05,
+      "loss": 0.3084,
+      "step": 318
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4297058582305908,
+      "learning_rate": 8.216523388745287e-05,
+      "loss": 0.3503,
+      "step": 319
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4172101616859436,
+      "learning_rate": 8.204877827313997e-05,
+      "loss": 0.3307,
+      "step": 320
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.39230528473854065,
+      "learning_rate": 8.193202681002692e-05,
+      "loss": 0.339,
+      "step": 321
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2820037603378296,
+      "learning_rate": 8.181498057587027e-05,
+      "loss": 0.1672,
+      "step": 322
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3549487292766571,
+      "learning_rate": 8.169764065114764e-05,
+      "loss": 0.2718,
+      "step": 323
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.41914433240890503,
+      "learning_rate": 8.158000811904778e-05,
+      "loss": 0.2528,
+      "step": 324
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.40389931201934814,
+      "learning_rate": 8.146208406546053e-05,
+      "loss": 0.3286,
+      "step": 325
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4596593677997589,
+      "learning_rate": 8.134386957896688e-05,
+      "loss": 0.4036,
+      "step": 326
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.45193785429000854,
+      "learning_rate": 8.122536575082882e-05,
+      "loss": 0.418,
+      "step": 327
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4680340886116028,
+      "learning_rate": 8.110657367497933e-05,
+      "loss": 0.2809,
+      "step": 328
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3416012227535248,
+      "learning_rate": 8.098749444801224e-05,
+      "loss": 0.235,
+      "step": 329
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.43442729115486145,
+      "learning_rate": 8.08681291691722e-05,
+      "loss": 0.3546,
+      "step": 330
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4265907108783722,
+      "learning_rate": 8.074847894034434e-05,
+      "loss": 0.2733,
+      "step": 331
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4317761957645416,
+      "learning_rate": 8.062854486604435e-05,
+      "loss": 0.2601,
+      "step": 332
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.38275545835494995,
+      "learning_rate": 8.050832805340806e-05,
+      "loss": 0.2794,
+      "step": 333
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3954709470272064,
+      "learning_rate": 8.038782961218136e-05,
+      "loss": 0.3348,
+      "step": 334
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5072319507598877,
+      "learning_rate": 8.026705065470996e-05,
+      "loss": 0.343,
+      "step": 335
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4034208357334137,
+      "learning_rate": 8.014599229592894e-05,
+      "loss": 0.3071,
+      "step": 336
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4476311206817627,
+      "learning_rate": 8.002465565335271e-05,
+      "loss": 0.256,
+      "step": 337
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.41503357887268066,
+      "learning_rate": 7.990304184706455e-05,
+      "loss": 0.277,
+      "step": 338
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4218122363090515,
+      "learning_rate": 7.978115199970621e-05,
+      "loss": 0.3754,
+      "step": 339
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3906747102737427,
+      "learning_rate": 7.965898723646776e-05,
+      "loss": 0.304,
+      "step": 340
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5392120480537415,
+      "learning_rate": 7.953654868507699e-05,
+      "loss": 0.416,
+      "step": 341
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4278102219104767,
+      "learning_rate": 7.941383747578912e-05,
+      "loss": 0.343,
+      "step": 342
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4145180583000183,
+      "learning_rate": 7.929085474137629e-05,
+      "loss": 0.3392,
+      "step": 343
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.37485140562057495,
+      "learning_rate": 7.91676016171172e-05,
+      "loss": 0.321,
+      "step": 344
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4715069532394409,
+      "learning_rate": 7.904407924078654e-05,
+      "loss": 0.4593,
+      "step": 345
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.44009581208229065,
+      "learning_rate": 7.892028875264451e-05,
+      "loss": 0.3349,
+      "step": 346
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3941679894924164,
+      "learning_rate": 7.879623129542633e-05,
+      "loss": 0.2938,
+      "step": 347
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4242309331893921,
+      "learning_rate": 7.867190801433166e-05,
+      "loss": 0.3678,
+      "step": 348
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.40279605984687805,
+      "learning_rate": 7.854732005701402e-05,
+      "loss": 0.3434,
+      "step": 349
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.40267717838287354,
+      "learning_rate": 7.842246857357023e-05,
+      "loss": 0.29,
+      "step": 350
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3405179977416992,
+      "learning_rate": 7.829735471652978e-05,
+      "loss": 0.2241,
+      "step": 351
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.39818453788757324,
+      "learning_rate": 7.817197964084411e-05,
+      "loss": 0.2662,
+      "step": 352
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3334313929080963,
+      "learning_rate": 7.804634450387616e-05,
+      "loss": 0.1877,
+      "step": 353
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.41519755125045776,
+      "learning_rate": 7.792045046538941e-05,
+      "loss": 0.3066,
+      "step": 354
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4714450538158417,
+      "learning_rate": 7.77942986875374e-05,
+      "loss": 0.4239,
+      "step": 355
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.38951438665390015,
+      "learning_rate": 7.766789033485287e-05,
+      "loss": 0.3293,
+      "step": 356
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4261366128921509,
+      "learning_rate": 7.75412265742371e-05,
+      "loss": 0.3883,
+      "step": 357
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.43968912959098816,
+      "learning_rate": 7.741430857494904e-05,
+      "loss": 0.3126,
+      "step": 358
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4306524395942688,
+      "learning_rate": 7.728713750859458e-05,
+      "loss": 0.3449,
+      "step": 359
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.42536354064941406,
+      "learning_rate": 7.715971454911577e-05,
+      "loss": 0.3506,
+      "step": 360
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4529925286769867,
+      "learning_rate": 7.703204087277988e-05,
+      "loss": 0.3568,
+      "step": 361
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.36422109603881836,
+      "learning_rate": 7.690411765816864e-05,
+      "loss": 0.2616,
+      "step": 362
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4082551896572113,
+      "learning_rate": 7.677594608616729e-05,
+      "loss": 0.3083,
+      "step": 363
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.38873544335365295,
+      "learning_rate": 7.66475273399537e-05,
+      "loss": 0.2812,
+      "step": 364
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3795793950557709,
+      "learning_rate": 7.651886260498751e-05,
+      "loss": 0.2658,
+      "step": 365
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4415530264377594,
+      "learning_rate": 7.638995306899908e-05,
+      "loss": 0.327,
+      "step": 366
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3712081015110016,
+      "learning_rate": 7.626079992197857e-05,
+      "loss": 0.2781,
+      "step": 367
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.40912961959838867,
+      "learning_rate": 7.613140435616503e-05,
+      "loss": 0.3211,
+      "step": 368
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4007984399795532,
+      "learning_rate": 7.600176756603525e-05,
+      "loss": 0.2528,
+      "step": 369
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3977168798446655,
+      "learning_rate": 7.587189074829284e-05,
+      "loss": 0.306,
+      "step": 370
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.333519846200943,
+      "learning_rate": 7.57417751018572e-05,
+      "loss": 0.2291,
+      "step": 371
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.37693852186203003,
+      "learning_rate": 7.561142182785233e-05,
+      "loss": 0.2407,
+      "step": 372
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.40604284405708313,
+      "learning_rate": 7.548083212959588e-05,
+      "loss": 0.3062,
+      "step": 373
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4472860097885132,
+      "learning_rate": 7.535000721258791e-05,
+      "loss": 0.2598,
+      "step": 374
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4256829023361206,
+      "learning_rate": 7.521894828449994e-05,
+      "loss": 0.3627,
+      "step": 375
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4579877555370331,
+      "learning_rate": 7.508765655516358e-05,
+      "loss": 0.3436,
+      "step": 376
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.38452377915382385,
+      "learning_rate": 7.495613323655953e-05,
+      "loss": 0.3056,
+      "step": 377
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.411670446395874,
+      "learning_rate": 7.482437954280635e-05,
+      "loss": 0.2649,
+      "step": 378
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4431185722351074,
+      "learning_rate": 7.469239669014923e-05,
+      "loss": 0.3693,
+      "step": 379
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3988078534603119,
+      "learning_rate": 7.456018589694873e-05,
+      "loss": 0.2902,
+      "step": 380
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5329432487487793,
+      "learning_rate": 7.442774838366965e-05,
+      "loss": 0.3991,
+      "step": 381
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2802259624004364,
+      "learning_rate": 7.429508537286963e-05,
+      "loss": 0.1907,
+      "step": 382
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.43637794256210327,
+      "learning_rate": 7.416219808918794e-05,
+      "loss": 0.3517,
+      "step": 383
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4463917911052704,
+      "learning_rate": 7.402908775933419e-05,
+      "loss": 0.3699,
+      "step": 384
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.46155545115470886,
+      "learning_rate": 7.389575561207692e-05,
+      "loss": 0.4614,
+      "step": 385
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4998500943183899,
+      "learning_rate": 7.376220287823236e-05,
+      "loss": 0.3709,
+      "step": 386
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.45294034481048584,
+      "learning_rate": 7.3628430790653e-05,
+      "loss": 0.3206,
+      "step": 387
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.39041468501091003,
+      "learning_rate": 7.349444058421619e-05,
+      "loss": 0.2901,
+      "step": 388
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.40528184175491333,
+      "learning_rate": 7.336023349581287e-05,
+      "loss": 0.245,
+      "step": 389
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.37183213233947754,
+      "learning_rate": 7.322581076433596e-05,
+      "loss": 0.2901,
+      "step": 390
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4894709289073944,
+      "learning_rate": 7.309117363066912e-05,
+      "loss": 0.4137,
+      "step": 391
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.41468536853790283,
+      "learning_rate": 7.295632333767513e-05,
+      "loss": 0.3505,
+      "step": 392
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.356780469417572,
+      "learning_rate": 7.28212611301845e-05,
+      "loss": 0.2336,
+      "step": 393
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4367281198501587,
+      "learning_rate": 7.2685988254984e-05,
+      "loss": 0.361,
+      "step": 394
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.43233007192611694,
+      "learning_rate": 7.255050596080509e-05,
+      "loss": 0.3527,
+      "step": 395
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2684633731842041,
+      "learning_rate": 7.241481549831243e-05,
+      "loss": 0.1598,
+      "step": 396
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.41831842064857483,
+      "learning_rate": 7.22789181200923e-05,
+      "loss": 0.341,
+      "step": 397
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4490970969200134,
+      "learning_rate": 7.214281508064107e-05,
+      "loss": 0.3595,
+      "step": 398
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.36620378494262695,
+      "learning_rate": 7.200650763635366e-05,
+      "loss": 0.2399,
+      "step": 399
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.44731780886650085,
+      "learning_rate": 7.186999704551181e-05,
+      "loss": 0.3095,
+      "step": 400
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.44940871000289917,
+      "learning_rate": 7.173328456827263e-05,
+      "loss": 0.3172,
+      "step": 401
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.32171034812927246,
+      "learning_rate": 7.15963714666568e-05,
+      "loss": 0.2014,
+      "step": 402
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.43398237228393555,
+      "learning_rate": 7.145925900453709e-05,
+      "loss": 0.3541,
+      "step": 403
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3726755976676941,
+      "learning_rate": 7.132194844762654e-05,
+      "loss": 0.273,
+      "step": 404
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.43420591950416565,
+      "learning_rate": 7.118444106346687e-05,
+      "loss": 0.3587,
+      "step": 405
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.42400193214416504,
+      "learning_rate": 7.104673812141675e-05,
+      "loss": 0.3649,
+      "step": 406
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4047689139842987,
+      "learning_rate": 7.090884089264011e-05,
+      "loss": 0.2675,
+      "step": 407
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3567971885204315,
+      "learning_rate": 7.077075065009433e-05,
+      "loss": 0.2321,
+      "step": 408
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4200761616230011,
+      "learning_rate": 7.063246866851858e-05,
+      "loss": 0.2741,
+      "step": 409
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.525682806968689,
+      "learning_rate": 7.049399622442198e-05,
+      "loss": 0.3688,
+      "step": 410
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.46288007497787476,
+      "learning_rate": 7.035533459607189e-05,
+      "loss": 0.3295,
+      "step": 411
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4013391435146332,
+      "learning_rate": 7.021648506348204e-05,
+      "loss": 0.3131,
+      "step": 412
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.41562074422836304,
+      "learning_rate": 7.007744890840073e-05,
+      "loss": 0.2775,
+      "step": 413
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.43653470277786255,
+      "learning_rate": 6.993822741429907e-05,
+      "loss": 0.3803,
+      "step": 414
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3548017740249634,
+      "learning_rate": 6.979882186635897e-05,
+      "loss": 0.2141,
+      "step": 415
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4461126923561096,
+      "learning_rate": 6.965923355146147e-05,
+      "loss": 0.3511,
+      "step": 416
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3795164227485657,
+      "learning_rate": 6.951946375817474e-05,
+      "loss": 0.2828,
+      "step": 417
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.46577778458595276,
+      "learning_rate": 6.937951377674221e-05,
+      "loss": 0.2939,
+      "step": 418
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4420101046562195,
+      "learning_rate": 6.923938489907066e-05,
+      "loss": 0.3268,
+      "step": 419
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4184642732143402,
+      "learning_rate": 6.909907841871829e-05,
+      "loss": 0.2857,
+      "step": 420
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4399305582046509,
+      "learning_rate": 6.895859563088283e-05,
+      "loss": 0.3702,
+      "step": 421
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.36446893215179443,
+      "learning_rate": 6.881793783238948e-05,
+      "loss": 0.2467,
+      "step": 422
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4695288836956024,
+      "learning_rate": 6.867710632167903e-05,
+      "loss": 0.3516,
+      "step": 423
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3637704849243164,
+      "learning_rate": 6.853610239879586e-05,
+      "loss": 0.2723,
+      "step": 424
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3477213978767395,
+      "learning_rate": 6.839492736537588e-05,
+      "loss": 0.1958,
+      "step": 425
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4522038698196411,
+      "learning_rate": 6.82535825246346e-05,
+      "loss": 0.2695,
+      "step": 426
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4621041715145111,
+      "learning_rate": 6.811206918135502e-05,
+      "loss": 0.3995,
+      "step": 427
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4887210726737976,
+      "learning_rate": 6.797038864187564e-05,
+      "loss": 0.3456,
+      "step": 428
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.37454670667648315,
+      "learning_rate": 6.782854221407838e-05,
+      "loss": 0.2969,
+      "step": 429
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4346952438354492,
+      "learning_rate": 6.768653120737652e-05,
+      "loss": 0.3828,
+      "step": 430
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.404827356338501,
+      "learning_rate": 6.754435693270258e-05,
+      "loss": 0.2817,
+      "step": 431
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4594029486179352,
+      "learning_rate": 6.740202070249623e-05,
+      "loss": 0.3467,
+      "step": 432
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4898674190044403,
+      "learning_rate": 6.725952383069222e-05,
+      "loss": 0.4575,
+      "step": 433
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4370651841163635,
+      "learning_rate": 6.711686763270818e-05,
+      "loss": 0.3752,
+      "step": 434
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4255489110946655,
+      "learning_rate": 6.697405342543258e-05,
+      "loss": 0.378,
+      "step": 435
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3686978220939636,
+      "learning_rate": 6.683108252721238e-05,
+      "loss": 0.3042,
+      "step": 436
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4671451449394226,
+      "learning_rate": 6.668795625784113e-05,
+      "loss": 0.4444,
+      "step": 437
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.351331502199173,
+      "learning_rate": 6.654467593854657e-05,
+      "loss": 0.2294,
+      "step": 438
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.40956103801727295,
+      "learning_rate": 6.640124289197845e-05,
+      "loss": 0.3081,
+      "step": 439
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4217970669269562,
+      "learning_rate": 6.625765844219652e-05,
+      "loss": 0.3677,
+      "step": 440
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3987672030925751,
+      "learning_rate": 6.611392391465802e-05,
+      "loss": 0.2973,
+      "step": 441
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4282805621623993,
+      "learning_rate": 6.597004063620567e-05,
+      "loss": 0.3339,
+      "step": 442
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.29611936211586,
+      "learning_rate": 6.582600993505534e-05,
+      "loss": 0.1904,
+      "step": 443
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4696163833141327,
+      "learning_rate": 6.568183314078376e-05,
+      "loss": 0.4168,
+      "step": 444
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.33832165598869324,
+      "learning_rate": 6.553751158431627e-05,
+      "loss": 0.2301,
+      "step": 445
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.379827082157135,
+      "learning_rate": 6.539304659791456e-05,
+      "loss": 0.2657,
+      "step": 446
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.387509286403656,
+      "learning_rate": 6.524843951516434e-05,
+      "loss": 0.2709,
+      "step": 447
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4099612236022949,
+      "learning_rate": 6.510369167096308e-05,
+      "loss": 0.295,
+      "step": 448
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.38275423645973206,
+      "learning_rate": 6.495880440150756e-05,
+      "loss": 0.2644,
+      "step": 449
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3869870603084564,
+      "learning_rate": 6.481377904428171e-05,
+      "loss": 0.2606,
+      "step": 450
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.41019243001937866,
+      "learning_rate": 6.466861693804413e-05,
+      "loss": 0.3219,
+      "step": 451
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4798837900161743,
+      "learning_rate": 6.45233194228158e-05,
+      "loss": 0.4045,
+      "step": 452
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4423982501029968,
+      "learning_rate": 6.437788783986766e-05,
+      "loss": 0.2412,
+      "step": 453
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.45012742280960083,
+      "learning_rate": 6.42323235317083e-05,
+      "loss": 0.3143,
+      "step": 454
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4368098974227905,
+      "learning_rate": 6.408662784207149e-05,
+      "loss": 0.296,
+      "step": 455
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4497329890727997,
+      "learning_rate": 6.394080211590381e-05,
+      "loss": 0.2861,
+      "step": 456
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3450700342655182,
+      "learning_rate": 6.379484769935223e-05,
+      "loss": 0.2295,
+      "step": 457
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4247216284275055,
+      "learning_rate": 6.364876593975173e-05,
+      "loss": 0.2657,
+      "step": 458
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.38082388043403625,
+      "learning_rate": 6.350255818561277e-05,
+      "loss": 0.2137,
+      "step": 459
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.43894073367118835,
+      "learning_rate": 6.335622578660889e-05,
+      "loss": 0.313,
+      "step": 460
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.36916911602020264,
+      "learning_rate": 6.320977009356431e-05,
+      "loss": 0.2436,
+      "step": 461
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5032907724380493,
+      "learning_rate": 6.306319245844133e-05,
+      "loss": 0.3732,
+      "step": 462
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.40374746918678284,
+      "learning_rate": 6.291649423432799e-05,
+      "loss": 0.2411,
+      "step": 463
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.36921069025993347,
+      "learning_rate": 6.276967677542542e-05,
+      "loss": 0.266,
+      "step": 464
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.407466858625412,
+      "learning_rate": 6.262274143703554e-05,
+      "loss": 0.306,
+      "step": 465
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.42362383008003235,
+      "learning_rate": 6.24756895755484e-05,
+      "loss": 0.2955,
+      "step": 466
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3616067171096802,
+      "learning_rate": 6.232852254842962e-05,
+      "loss": 0.2596,
+      "step": 467
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4745740592479706,
+      "learning_rate": 6.218124171420806e-05,
+      "loss": 0.3476,
+      "step": 468
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4039102792739868,
+      "learning_rate": 6.203384843246307e-05,
+      "loss": 0.3447,
+      "step": 469
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3852016031742096,
+      "learning_rate": 6.188634406381207e-05,
+      "loss": 0.3348,
+      "step": 470
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4698221683502197,
+      "learning_rate": 6.173872996989793e-05,
+      "loss": 0.3371,
+      "step": 471
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3284843862056732,
+      "learning_rate": 6.159100751337642e-05,
+      "loss": 0.2062,
+      "step": 472
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3801678419113159,
+      "learning_rate": 6.144317805790361e-05,
+      "loss": 0.2848,
+      "step": 473
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.38208192586898804,
+      "learning_rate": 6.129524296812335e-05,
+      "loss": 0.2542,
+      "step": 474
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3336544334888458,
+      "learning_rate": 6.114720360965453e-05,
+      "loss": 0.187,
+      "step": 475
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.41276127099990845,
+      "learning_rate": 6.099906134907868e-05,
+      "loss": 0.2808,
+      "step": 476
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4322654604911804,
+      "learning_rate": 6.085081755392714e-05,
+      "loss": 0.35,
+      "step": 477
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.45964300632476807,
+      "learning_rate": 6.07024735926686e-05,
+      "loss": 0.3284,
+      "step": 478
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4558175802230835,
+      "learning_rate": 6.055403083469637e-05,
+      "loss": 0.3817,
+      "step": 479
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5052785873413086,
+      "learning_rate": 6.04054906503158e-05,
+      "loss": 0.4212,
+      "step": 480
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4662330746650696,
+      "learning_rate": 6.0256854410731565e-05,
+      "loss": 0.3404,
+      "step": 481
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4600191116333008,
+      "learning_rate": 6.010812348803509e-05,
+      "loss": 0.3785,
+      "step": 482
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.45039844512939453,
+      "learning_rate": 5.99592992551918e-05,
+      "loss": 0.4206,
+      "step": 483
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.42035818099975586,
+      "learning_rate": 5.9810383086028535e-05,
+      "loss": 0.3766,
+      "step": 484
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.35778695344924927,
+      "learning_rate": 5.9661376355220734e-05,
+      "loss": 0.2886,
+      "step": 485
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.48496320843696594,
+      "learning_rate": 5.9512280438279914e-05,
+      "loss": 0.4084,
+      "step": 486
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.39029964804649353,
+      "learning_rate": 5.936309671154084e-05,
+      "loss": 0.3367,
+      "step": 487
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4327787458896637,
+      "learning_rate": 5.9213826552148886e-05,
+      "loss": 0.3404,
+      "step": 488
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.41896331310272217,
+      "learning_rate": 5.906447133804731e-05,
+      "loss": 0.3544,
+      "step": 489
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.37838464975357056,
+      "learning_rate": 5.891503244796448e-05,
+      "loss": 0.2552,
+      "step": 490
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3590022027492523,
+      "learning_rate": 5.8765511261401254e-05,
+      "loss": 0.2994,
+      "step": 491
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3592306673526764,
+      "learning_rate": 5.861590915861817e-05,
+      "loss": 0.2334,
+      "step": 492
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.389812707901001,
+      "learning_rate": 5.846622752062268e-05,
+      "loss": 0.2642,
+      "step": 493
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.431536465883255,
+      "learning_rate": 5.831646772915651e-05,
+      "loss": 0.3448,
+      "step": 494
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4194413423538208,
+      "learning_rate": 5.816663116668276e-05,
+      "loss": 0.3033,
+      "step": 495
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4021466076374054,
+      "learning_rate": 5.801671921637328e-05,
+      "loss": 0.3144,
+      "step": 496
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5036161541938782,
+      "learning_rate": 5.786673326209584e-05,
+      "loss": 0.4045,
+      "step": 497
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3767940104007721,
+      "learning_rate": 5.7716674688401286e-05,
+      "loss": 0.2666,
+      "step": 498
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5003858208656311,
+      "learning_rate": 5.756654488051091e-05,
+      "loss": 0.3855,
+      "step": 499
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4110037684440613,
+      "learning_rate": 5.7416345224303524e-05,
+      "loss": 0.3001,
+      "step": 500
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.38694775104522705,
+      "learning_rate": 5.7266077106302785e-05,
+      "loss": 0.2862,
+      "step": 501
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.34114885330200195,
+      "learning_rate": 5.7115741913664264e-05,
+      "loss": 0.267,
+      "step": 502
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.37278828024864197,
+      "learning_rate": 5.696534103416276e-05,
+      "loss": 0.271,
+      "step": 503
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4475939869880676,
+      "learning_rate": 5.6814875856179414e-05,
+      "loss": 0.3599,
+      "step": 504
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5131115913391113,
+      "learning_rate": 5.666434776868895e-05,
+      "loss": 0.3666,
+      "step": 505
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4798625409603119,
+      "learning_rate": 5.651375816124679e-05,
+      "loss": 0.3203,
+      "step": 506
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.42612072825431824,
+      "learning_rate": 5.636310842397629e-05,
+      "loss": 0.3146,
+      "step": 507
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.45160189270973206,
+      "learning_rate": 5.621239994755583e-05,
+      "loss": 0.3138,
+      "step": 508
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4129311740398407,
+      "learning_rate": 5.606163412320608e-05,
+      "loss": 0.2952,
+      "step": 509
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.41268205642700195,
+      "learning_rate": 5.5910812342677065e-05,
+      "loss": 0.2656,
+      "step": 510
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.36380165815353394,
+      "learning_rate": 5.575993599823536e-05,
+      "loss": 0.2511,
+      "step": 511
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3615700900554657,
+      "learning_rate": 5.560900648265124e-05,
+      "loss": 0.1815,
+      "step": 512
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.49483224749565125,
+      "learning_rate": 5.545802518918579e-05,
+      "loss": 0.4012,
+      "step": 513
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5183067321777344,
+      "learning_rate": 5.5306993511578096e-05,
+      "loss": 0.4256,
+      "step": 514
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.48457270860671997,
+      "learning_rate": 5.515591284403234e-05,
+      "loss": 0.3114,
+      "step": 515
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.40141454339027405,
+      "learning_rate": 5.5004784581204927e-05,
+      "loss": 0.2999,
+      "step": 516
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.36214885115623474,
+      "learning_rate": 5.485361011819164e-05,
+      "loss": 0.2415,
+      "step": 517
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3985465466976166,
+      "learning_rate": 5.4702390850514726e-05,
+      "loss": 0.2639,
+      "step": 518
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4212491512298584,
+      "learning_rate": 5.455112817411006e-05,
+      "loss": 0.3155,
+      "step": 519
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.406974196434021,
+      "learning_rate": 5.4399823485314226e-05,
+      "loss": 0.2597,
+      "step": 520
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.44170641899108887,
+      "learning_rate": 5.4248478180851604e-05,
+      "loss": 0.3835,
+      "step": 521
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3631415069103241,
+      "learning_rate": 5.409709365782154e-05,
+      "loss": 0.2471,
+      "step": 522
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.464281290769577,
+      "learning_rate": 5.3945671313685386e-05,
+      "loss": 0.3869,
+      "step": 523
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.381562739610672,
+      "learning_rate": 5.379421254625366e-05,
+      "loss": 0.2764,
+      "step": 524
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4001575708389282,
+      "learning_rate": 5.364271875367311e-05,
+      "loss": 0.2963,
+      "step": 525
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.42583319544792175,
+      "learning_rate": 5.3491191334413746e-05,
+      "loss": 0.3755,
+      "step": 526
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3744286596775055,
+      "learning_rate": 5.3339631687256084e-05,
+      "loss": 0.2448,
+      "step": 527
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.39580249786376953,
+      "learning_rate": 5.318804121127807e-05,
+      "loss": 0.2644,
+      "step": 528
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.46420353651046753,
+      "learning_rate": 5.3036421305842276e-05,
+      "loss": 0.3886,
+      "step": 529
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3602772355079651,
+      "learning_rate": 5.288477337058293e-05,
+      "loss": 0.2409,
+      "step": 530
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4621635973453522,
+      "learning_rate": 5.273309880539301e-05,
+      "loss": 0.368,
+      "step": 531
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.534472644329071,
+      "learning_rate": 5.258139901041132e-05,
+      "loss": 0.3049,
+      "step": 532
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3050483465194702,
+      "learning_rate": 5.242967538600957e-05,
+      "loss": 0.1831,
+      "step": 533
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.427960067987442,
+      "learning_rate": 5.227792933277943e-05,
+      "loss": 0.3298,
+      "step": 534
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.31485849618911743,
+      "learning_rate": 5.212616225151965e-05,
+      "loss": 0.2202,
+      "step": 535
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.44149473309516907,
+      "learning_rate": 5.197437554322304e-05,
+      "loss": 0.4113,
+      "step": 536
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4465602934360504,
+      "learning_rate": 5.182257060906365e-05,
+      "loss": 0.3018,
+      "step": 537
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4594295620918274,
+      "learning_rate": 5.167074885038373e-05,
+      "loss": 0.3268,
+      "step": 538
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3535444140434265,
+      "learning_rate": 5.151891166868086e-05,
+      "loss": 0.2448,
+      "step": 539
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5269205570220947,
+      "learning_rate": 5.1367060465595006e-05,
+      "loss": 0.4655,
+      "step": 540
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3851945698261261,
+      "learning_rate": 5.121519664289553e-05,
+      "loss": 0.2964,
+      "step": 541
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4726141393184662,
+      "learning_rate": 5.106332160246834e-05,
+      "loss": 0.4122,
+      "step": 542
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.37341347336769104,
+      "learning_rate": 5.0911436746302834e-05,
+      "loss": 0.2887,
+      "step": 543
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3809630274772644,
+      "learning_rate": 5.075954347647909e-05,
+      "loss": 0.2476,
+      "step": 544
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3767317235469818,
+      "learning_rate": 5.0607643195154796e-05,
+      "loss": 0.2833,
+      "step": 545
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3544347584247589,
+      "learning_rate": 5.045573730455241e-05,
+      "loss": 0.286,
+      "step": 546
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.34240320324897766,
+      "learning_rate": 5.030382720694612e-05,
+      "loss": 0.2543,
+      "step": 547
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4301561117172241,
+      "learning_rate": 5.0151914304649015e-05,
+      "loss": 0.35,
+      "step": 548
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.34088242053985596,
+      "learning_rate": 5e-05,
+      "loss": 0.228,
+      "step": 549
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4396790564060211,
+      "learning_rate": 4.984808569535101e-05,
+      "loss": 0.3361,
+      "step": 550
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4188593029975891,
+      "learning_rate": 4.969617279305388e-05,
+      "loss": 0.3144,
+      "step": 551
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3811877965927124,
+      "learning_rate": 4.954426269544761e-05,
+      "loss": 0.2667,
+      "step": 552
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.40995457768440247,
+      "learning_rate": 4.939235680484522e-05,
+      "loss": 0.3144,
+      "step": 553
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3613314926624298,
+      "learning_rate": 4.924045652352092e-05,
+      "loss": 0.2425,
+      "step": 554
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.38490086793899536,
+      "learning_rate": 4.908856325369718e-05,
+      "loss": 0.2996,
+      "step": 555
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3282059133052826,
+      "learning_rate": 4.893667839753168e-05,
+      "loss": 0.1906,
+      "step": 556
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4107285737991333,
+      "learning_rate": 4.878480335710448e-05,
+      "loss": 0.3535,
+      "step": 557
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.2970867156982422,
+      "learning_rate": 4.8632939534405006e-05,
+      "loss": 0.1745,
+      "step": 558
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.39580798149108887,
+      "learning_rate": 4.8481088331319146e-05,
+      "loss": 0.3176,
+      "step": 559
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3547394573688507,
+      "learning_rate": 4.832925114961629e-05,
+      "loss": 0.2462,
+      "step": 560
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4645850956439972,
+      "learning_rate": 4.817742939093635e-05,
+      "loss": 0.4052,
+      "step": 561
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4204746186733246,
+      "learning_rate": 4.8025624456776966e-05,
+      "loss": 0.268,
+      "step": 562
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3667820394039154,
+      "learning_rate": 4.787383774848037e-05,
+      "loss": 0.2983,
+      "step": 563
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.479818731546402,
+      "learning_rate": 4.772207066722057e-05,
+      "loss": 0.3442,
+      "step": 564
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4877625107765198,
+      "learning_rate": 4.757032461399044e-05,
+      "loss": 0.3294,
+      "step": 565
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3830382227897644,
+      "learning_rate": 4.7418600989588694e-05,
+      "loss": 0.2375,
+      "step": 566
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3333114981651306,
+      "learning_rate": 4.726690119460701e-05,
+      "loss": 0.1962,
+      "step": 567
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4095388352870941,
+      "learning_rate": 4.7115226629417075e-05,
+      "loss": 0.3104,
+      "step": 568
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4421548843383789,
+      "learning_rate": 4.6963578694157736e-05,
+      "loss": 0.2638,
+      "step": 569
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4537515640258789,
+      "learning_rate": 4.681195878872194e-05,
+      "loss": 0.3362,
+      "step": 570
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3993587791919708,
+      "learning_rate": 4.666036831274392e-05,
+      "loss": 0.2565,
+      "step": 571
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4283117651939392,
+      "learning_rate": 4.6508808665586265e-05,
+      "loss": 0.3057,
+      "step": 572
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.33654728531837463,
+      "learning_rate": 4.635728124632692e-05,
+      "loss": 0.2612,
+      "step": 573
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4272298514842987,
+      "learning_rate": 4.6205787453746336e-05,
+      "loss": 0.282,
+      "step": 574
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3670824468135834,
+      "learning_rate": 4.605432868631462e-05,
+      "loss": 0.2354,
+      "step": 575
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5032891631126404,
+      "learning_rate": 4.590290634217848e-05,
+      "loss": 0.3752,
+      "step": 576
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4507879316806793,
+      "learning_rate": 4.57515218191484e-05,
+      "loss": 0.3363,
+      "step": 577
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.37974223494529724,
+      "learning_rate": 4.5600176514685786e-05,
+      "loss": 0.2745,
+      "step": 578
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.42613649368286133,
+      "learning_rate": 4.5448871825889946e-05,
+      "loss": 0.3073,
+      "step": 579
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4176366627216339,
+      "learning_rate": 4.52976091494853e-05,
+      "loss": 0.2573,
+      "step": 580
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3846084475517273,
+      "learning_rate": 4.514638988180837e-05,
+      "loss": 0.2727,
+      "step": 581
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4902336001396179,
+      "learning_rate": 4.4995215418795085e-05,
+      "loss": 0.3941,
+      "step": 582
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.41963398456573486,
+      "learning_rate": 4.484408715596768e-05,
+      "loss": 0.2968,
+      "step": 583
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4140608012676239,
+      "learning_rate": 4.4693006488421915e-05,
+      "loss": 0.2776,
+      "step": 584
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4394283592700958,
+      "learning_rate": 4.454197481081422e-05,
+      "loss": 0.2955,
+      "step": 585
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4049432575702667,
+      "learning_rate": 4.439099351734878e-05,
+      "loss": 0.2811,
+      "step": 586
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.40243345499038696,
+      "learning_rate": 4.4240064001764646e-05,
+      "loss": 0.265,
+      "step": 587
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4278285503387451,
+      "learning_rate": 4.4089187657322953e-05,
+      "loss": 0.2675,
+      "step": 588
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3500414192676544,
+      "learning_rate": 4.393836587679394e-05,
+      "loss": 0.2449,
+      "step": 589
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4892849326133728,
+      "learning_rate": 4.3787600052444174e-05,
+      "loss": 0.3522,
+      "step": 590
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4855495095252991,
+      "learning_rate": 4.363689157602373e-05,
+      "loss": 0.4272,
+      "step": 591
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.37122058868408203,
+      "learning_rate": 4.348624183875322e-05,
+      "loss": 0.2937,
+      "step": 592
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.356032133102417,
+      "learning_rate": 4.333565223131107e-05,
+      "loss": 0.2191,
+      "step": 593
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4197503924369812,
+      "learning_rate": 4.318512414382059e-05,
+      "loss": 0.3349,
+      "step": 594
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.39776870608329773,
+      "learning_rate": 4.3034658965837255e-05,
+      "loss": 0.1818,
+      "step": 595
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4147531986236572,
+      "learning_rate": 4.288425808633575e-05,
+      "loss": 0.336,
+      "step": 596
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3913092017173767,
+      "learning_rate": 4.273392289369723e-05,
+      "loss": 0.2746,
+      "step": 597
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.36486339569091797,
+      "learning_rate": 4.258365477569648e-05,
+      "loss": 0.227,
+      "step": 598
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4407197833061218,
+      "learning_rate": 4.2433455119489105e-05,
+      "loss": 0.3245,
+      "step": 599
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4557214677333832,
+      "learning_rate": 4.228332531159871e-05,
+      "loss": 0.3608,
+      "step": 600
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4273921251296997,
+      "learning_rate": 4.2133266737904176e-05,
+      "loss": 0.2968,
+      "step": 601
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4376170039176941,
+      "learning_rate": 4.1983280783626724e-05,
+      "loss": 0.3633,
+      "step": 602
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.444912850856781,
+      "learning_rate": 4.183336883331723e-05,
+      "loss": 0.3685,
+      "step": 603
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4346312880516052,
+      "learning_rate": 4.1683532270843504e-05,
+      "loss": 0.3486,
+      "step": 604
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4119686484336853,
+      "learning_rate": 4.153377247937732e-05,
+      "loss": 0.3092,
+      "step": 605
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4005814492702484,
+      "learning_rate": 4.138409084138185e-05,
+      "loss": 0.3097,
+      "step": 606
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4748894274234772,
+      "learning_rate": 4.1234488738598744e-05,
+      "loss": 0.2733,
+      "step": 607
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.33364054560661316,
+      "learning_rate": 4.108496755203553e-05,
+      "loss": 0.2098,
+      "step": 608
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.41625064611434937,
+      "learning_rate": 4.0935528661952716e-05,
+      "loss": 0.3067,
+      "step": 609
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.40368083119392395,
+      "learning_rate": 4.0786173447851126e-05,
+      "loss": 0.3118,
+      "step": 610
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.392261266708374,
+      "learning_rate": 4.063690328845916e-05,
+      "loss": 0.269,
+      "step": 611
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.34611016511917114,
+      "learning_rate": 4.04877195617201e-05,
+      "loss": 0.2569,
+      "step": 612
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4573889672756195,
+      "learning_rate": 4.033862364477927e-05,
+      "loss": 0.355,
+      "step": 613
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3446807563304901,
+      "learning_rate": 4.0189616913971484e-05,
+      "loss": 0.2251,
+      "step": 614
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.41673514246940613,
+      "learning_rate": 4.0040700744808204e-05,
+      "loss": 0.385,
+      "step": 615
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3703855872154236,
+      "learning_rate": 3.989187651196493e-05,
+      "loss": 0.2316,
+      "step": 616
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.41116270422935486,
+      "learning_rate": 3.974314558926844e-05,
+      "loss": 0.3043,
+      "step": 617
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.36566266417503357,
+      "learning_rate": 3.9594509349684216e-05,
+      "loss": 0.3039,
+      "step": 618
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.40920472145080566,
+      "learning_rate": 3.9445969165303647e-05,
+      "loss": 0.2838,
+      "step": 619
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4240154027938843,
+      "learning_rate": 3.929752640733141e-05,
+      "loss": 0.3118,
+      "step": 620
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.42745116353034973,
+      "learning_rate": 3.914918244607287e-05,
+      "loss": 0.2864,
+      "step": 621
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3916018009185791,
+      "learning_rate": 3.900093865092134e-05,
+      "loss": 0.2405,
+      "step": 622
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.38008588552474976,
+      "learning_rate": 3.885279639034546e-05,
+      "loss": 0.2444,
+      "step": 623
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4288583993911743,
+      "learning_rate": 3.870475703187667e-05,
+      "loss": 0.2701,
+      "step": 624
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4599088728427887,
+      "learning_rate": 3.855682194209639e-05,
+      "loss": 0.292,
+      "step": 625
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3144804537296295,
+      "learning_rate": 3.840899248662358e-05,
+      "loss": 0.2212,
+      "step": 626
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.39628955721855164,
+      "learning_rate": 3.8261270030102084e-05,
+      "loss": 0.2823,
+      "step": 627
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.42753270268440247,
+      "learning_rate": 3.8113655936187947e-05,
+      "loss": 0.3056,
+      "step": 628
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3899834454059601,
+      "learning_rate": 3.796615156753696e-05,
+      "loss": 0.3375,
+      "step": 629
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4072708785533905,
+      "learning_rate": 3.7818758285791955e-05,
+      "loss": 0.3034,
+      "step": 630
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.37316933274269104,
+      "learning_rate": 3.767147745157039e-05,
+      "loss": 0.3322,
+      "step": 631
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5441315770149231,
+      "learning_rate": 3.7524310424451635e-05,
+      "loss": 0.3682,
+      "step": 632
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4088776111602783,
+      "learning_rate": 3.7377258562964454e-05,
+      "loss": 0.3208,
+      "step": 633
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.42660993337631226,
+      "learning_rate": 3.723032322457458e-05,
+      "loss": 0.3428,
+      "step": 634
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.49614468216896057,
+      "learning_rate": 3.708350576567204e-05,
+      "loss": 0.4129,
+      "step": 635
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3900456428527832,
+      "learning_rate": 3.6936807541558674e-05,
+      "loss": 0.2697,
+      "step": 636
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.47792574763298035,
+      "learning_rate": 3.6790229906435705e-05,
+      "loss": 0.4538,
+      "step": 637
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.39278462529182434,
+      "learning_rate": 3.664377421339111e-05,
+      "loss": 0.2113,
+      "step": 638
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3987380862236023,
+      "learning_rate": 3.6497441814387247e-05,
+      "loss": 0.2929,
+      "step": 639
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3924303948879242,
+      "learning_rate": 3.6351234060248286e-05,
+      "loss": 0.2891,
+      "step": 640
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4462815225124359,
+      "learning_rate": 3.6205152300647784e-05,
+      "loss": 0.3793,
+      "step": 641
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.361194372177124,
+      "learning_rate": 3.605919788409621e-05,
+      "loss": 0.304,
+      "step": 642
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3908407390117645,
+      "learning_rate": 3.591337215792852e-05,
+      "loss": 0.3189,
+      "step": 643
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.37534841895103455,
+      "learning_rate": 3.5767676468291713e-05,
+      "loss": 0.237,
+      "step": 644
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4761614203453064,
+      "learning_rate": 3.562211216013235e-05,
+      "loss": 0.4667,
+      "step": 645
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.40542420744895935,
+      "learning_rate": 3.5476680577184206e-05,
+      "loss": 0.33,
+      "step": 646
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.33448925614356995,
+      "learning_rate": 3.533138306195588e-05,
+      "loss": 0.2378,
+      "step": 647
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4175407290458679,
+      "learning_rate": 3.5186220955718306e-05,
+      "loss": 0.2922,
+      "step": 648
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3705791234970093,
+      "learning_rate": 3.5041195598492446e-05,
+      "loss": 0.3063,
+      "step": 649
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3703921139240265,
+      "learning_rate": 3.489630832903694e-05,
+      "loss": 0.2685,
+      "step": 650
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.44679349660873413,
+      "learning_rate": 3.475156048483567e-05,
+      "loss": 0.325,
+      "step": 651
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4002423882484436,
+      "learning_rate": 3.460695340208546e-05,
+      "loss": 0.3584,
+      "step": 652
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.38548558950424194,
+      "learning_rate": 3.446248841568375e-05,
+      "loss": 0.2725,
+      "step": 653
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.41728684306144714,
+      "learning_rate": 3.431816685921625e-05,
+      "loss": 0.3218,
+      "step": 654
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3534894585609436,
+      "learning_rate": 3.417399006494466e-05,
+      "loss": 0.2367,
+      "step": 655
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3368327021598816,
+      "learning_rate": 3.402995936379433e-05,
+      "loss": 0.2411,
+      "step": 656
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.36815300583839417,
+      "learning_rate": 3.3886076085341986e-05,
+      "loss": 0.2482,
+      "step": 657
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3257908821105957,
+      "learning_rate": 3.37423415578035e-05,
+      "loss": 0.2375,
+      "step": 658
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.36529654264450073,
+      "learning_rate": 3.3598757108021546e-05,
+      "loss": 0.2368,
+      "step": 659
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3795417845249176,
+      "learning_rate": 3.345532406145345e-05,
+      "loss": 0.2878,
+      "step": 660
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.34727954864501953,
+      "learning_rate": 3.331204374215888e-05,
+      "loss": 0.2283,
+      "step": 661
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3729381561279297,
+      "learning_rate": 3.316891747278761e-05,
+      "loss": 0.2449,
+      "step": 662
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.4766552746295929,
+      "learning_rate": 3.302594657456744e-05,
+      "loss": 0.3729,
+      "step": 663
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.45169883966445923,
+      "learning_rate": 3.288313236729183e-05,
+      "loss": 0.3373,
+      "step": 664
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.4404531717300415,
+      "learning_rate": 3.274047616930781e-05,
+      "loss": 0.3399,
+      "step": 665
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.4247913062572479,
+      "learning_rate": 3.259797929750378e-05,
+      "loss": 0.2615,
+      "step": 666
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3882124722003937,
+      "learning_rate": 3.245564306729744e-05,
+      "loss": 0.2928,
+      "step": 667
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3825494945049286,
+      "learning_rate": 3.231346879262349e-05,
+      "loss": 0.2818,
+      "step": 668
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.38851484656333923,
+      "learning_rate": 3.217145778592162e-05,
+      "loss": 0.3119,
+      "step": 669
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4527076184749603,
+      "learning_rate": 3.202961135812437e-05,
+      "loss": 0.3445,
+      "step": 670
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.41588783264160156,
+      "learning_rate": 3.1887930818644996e-05,
+      "loss": 0.3486,
+      "step": 671
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4113435447216034,
+      "learning_rate": 3.1746417475365405e-05,
+      "loss": 0.2806,
+      "step": 672
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.45771273970603943,
+      "learning_rate": 3.1605072634624125e-05,
+      "loss": 0.4103,
+      "step": 673
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4430786073207855,
+      "learning_rate": 3.146389760120416e-05,
+      "loss": 0.2651,
+      "step": 674
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4159556031227112,
+      "learning_rate": 3.132289367832097e-05,
+      "loss": 0.292,
+      "step": 675
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4217040240764618,
+      "learning_rate": 3.118206216761053e-05,
+      "loss": 0.3106,
+      "step": 676
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3670906126499176,
+      "learning_rate": 3.104140436911719e-05,
+      "loss": 0.189,
+      "step": 677
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3929611146450043,
+      "learning_rate": 3.0900921581281725e-05,
+      "loss": 0.3106,
+      "step": 678
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4370023012161255,
+      "learning_rate": 3.076061510092935e-05,
+      "loss": 0.3362,
+      "step": 679
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.39130982756614685,
+      "learning_rate": 3.062048622325779e-05,
+      "loss": 0.3072,
+      "step": 680
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4033917188644409,
+      "learning_rate": 3.0480536241825263e-05,
+      "loss": 0.2458,
+      "step": 681
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.43826785683631897,
+      "learning_rate": 3.034076644853853e-05,
+      "loss": 0.2895,
+      "step": 682
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.46332448720932007,
+      "learning_rate": 3.0201178133641038e-05,
+      "loss": 0.3829,
+      "step": 683
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3498688340187073,
+      "learning_rate": 3.0061772585700953e-05,
+      "loss": 0.2358,
+      "step": 684
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3833886981010437,
+      "learning_rate": 2.992255109159926e-05,
+      "loss": 0.2501,
+      "step": 685
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3906776010990143,
+      "learning_rate": 2.9783514936517965e-05,
+      "loss": 0.309,
+      "step": 686
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.40844976902008057,
+      "learning_rate": 2.9644665403928117e-05,
+      "loss": 0.2933,
+      "step": 687
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4485379159450531,
+      "learning_rate": 2.950600377557804e-05,
+      "loss": 0.4246,
+      "step": 688
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.398406982421875,
+      "learning_rate": 2.9367531331481436e-05,
+      "loss": 0.2596,
+      "step": 689
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.40626242756843567,
+      "learning_rate": 2.9229249349905684e-05,
+      "loss": 0.3252,
+      "step": 690
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.39439576864242554,
+      "learning_rate": 2.909115910735991e-05,
+      "loss": 0.2669,
+      "step": 691
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.35952499508857727,
+      "learning_rate": 2.895326187858326e-05,
+      "loss": 0.2393,
+      "step": 692
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.476965993642807,
+      "learning_rate": 2.881555893653314e-05,
+      "loss": 0.4749,
+      "step": 693
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.40909814834594727,
+      "learning_rate": 2.8678051552373487e-05,
+      "loss": 0.3012,
+      "step": 694
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.34073591232299805,
+      "learning_rate": 2.854074099546291e-05,
+      "loss": 0.2452,
+      "step": 695
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.40000537037849426,
+      "learning_rate": 2.8403628533343206e-05,
+      "loss": 0.2009,
+      "step": 696
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.38767191767692566,
+      "learning_rate": 2.826671543172738e-05,
+      "loss": 0.3107,
+      "step": 697
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.35624590516090393,
+      "learning_rate": 2.8130002954488183e-05,
+      "loss": 0.2552,
+      "step": 698
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3999212682247162,
+      "learning_rate": 2.799349236364634e-05,
+      "loss": 0.3384,
+      "step": 699
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.41786062717437744,
+      "learning_rate": 2.7857184919358937e-05,
+      "loss": 0.3188,
+      "step": 700
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3465578258037567,
+      "learning_rate": 2.7721081879907718e-05,
+      "loss": 0.2404,
+      "step": 701
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4516163766384125,
+      "learning_rate": 2.7585184501687577e-05,
+      "loss": 0.3522,
+      "step": 702
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4111158549785614,
+      "learning_rate": 2.74494940391949e-05,
+      "loss": 0.357,
+      "step": 703
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4369189143180847,
+      "learning_rate": 2.731401174501601e-05,
+      "loss": 0.3152,
+      "step": 704
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.37029463052749634,
+      "learning_rate": 2.7178738869815506e-05,
+      "loss": 0.2604,
+      "step": 705
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3447953462600708,
+      "learning_rate": 2.7043676662324878e-05,
+      "loss": 0.226,
+      "step": 706
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.41446515917778015,
+      "learning_rate": 2.69088263693309e-05,
+      "loss": 0.3036,
+      "step": 707
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.37902921438217163,
+      "learning_rate": 2.6774189235664026e-05,
+      "loss": 0.2527,
+      "step": 708
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.4798141419887543,
+      "learning_rate": 2.663976650418715e-05,
+      "loss": 0.4065,
+      "step": 709
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.45657438039779663,
+      "learning_rate": 2.650555941578381e-05,
+      "loss": 0.4167,
+      "step": 710
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.4082154929637909,
+      "learning_rate": 2.6371569209347014e-05,
+      "loss": 0.2944,
+      "step": 711
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.41583922505378723,
+      "learning_rate": 2.6237797121767634e-05,
+      "loss": 0.3708,
+      "step": 712
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3839868903160095,
+      "learning_rate": 2.6104244387923082e-05,
+      "loss": 0.2698,
+      "step": 713
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3401528596878052,
+      "learning_rate": 2.5970912240665813e-05,
+      "loss": 0.2001,
+      "step": 714
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.38874009251594543,
+      "learning_rate": 2.5837801910812053e-05,
+      "loss": 0.2999,
+      "step": 715
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.33934372663497925,
+      "learning_rate": 2.5704914627130374e-05,
+      "loss": 0.2624,
+      "step": 716
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.4392395317554474,
+      "learning_rate": 2.5572251616330373e-05,
+      "loss": 0.3792,
+      "step": 717
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.386403352022171,
+      "learning_rate": 2.5439814103051284e-05,
+      "loss": 0.2239,
+      "step": 718
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.31485530734062195,
+      "learning_rate": 2.530760330985079e-05,
+      "loss": 0.183,
+      "step": 719
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.45125994086265564,
+      "learning_rate": 2.517562045719367e-05,
+      "loss": 0.3995,
+      "step": 720
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3921152353286743,
+      "learning_rate": 2.504386676344047e-05,
+      "loss": 0.3222,
+      "step": 721
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4134279489517212,
+      "learning_rate": 2.4912343444836445e-05,
+      "loss": 0.3354,
+      "step": 722
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4451929032802582,
+      "learning_rate": 2.4781051715500076e-05,
+      "loss": 0.3869,
+      "step": 723
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3529045879840851,
+      "learning_rate": 2.46499927874121e-05,
+      "loss": 0.2156,
+      "step": 724
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3834676444530487,
+      "learning_rate": 2.4519167870404125e-05,
+      "loss": 0.2475,
+      "step": 725
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.43510353565216064,
+      "learning_rate": 2.4388578172147675e-05,
+      "loss": 0.3217,
+      "step": 726
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.387541800737381,
+      "learning_rate": 2.4258224898142807e-05,
+      "loss": 0.2327,
+      "step": 727
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4362310767173767,
+      "learning_rate": 2.4128109251707155e-05,
+      "loss": 0.3878,
+      "step": 728
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3979613482952118,
+      "learning_rate": 2.399823243396476e-05,
+      "loss": 0.2884,
+      "step": 729
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3934382498264313,
+      "learning_rate": 2.3868595643834995e-05,
+      "loss": 0.255,
+      "step": 730
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4257848262786865,
+      "learning_rate": 2.373920007802144e-05,
+      "loss": 0.3247,
+      "step": 731
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3740795850753784,
+      "learning_rate": 2.361004693100094e-05,
+      "loss": 0.2647,
+      "step": 732
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4017227590084076,
+      "learning_rate": 2.3481137395012513e-05,
+      "loss": 0.3015,
+      "step": 733
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3802047073841095,
+      "learning_rate": 2.3352472660046295e-05,
+      "loss": 0.2571,
+      "step": 734
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3709452450275421,
+      "learning_rate": 2.322405391383273e-05,
+      "loss": 0.3034,
+      "step": 735
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4259112775325775,
+      "learning_rate": 2.3095882341831372e-05,
+      "loss": 0.401,
+      "step": 736
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3781328797340393,
+      "learning_rate": 2.296795912722014e-05,
+      "loss": 0.3018,
+      "step": 737
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3574370741844177,
+      "learning_rate": 2.284028545088423e-05,
+      "loss": 0.2579,
+      "step": 738
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3482043743133545,
+      "learning_rate": 2.2712862491405436e-05,
+      "loss": 0.2644,
+      "step": 739
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3467924892902374,
+      "learning_rate": 2.258569142505098e-05,
+      "loss": 0.2472,
+      "step": 740
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.44436389207839966,
+      "learning_rate": 2.2458773425762912e-05,
+      "loss": 0.2769,
+      "step": 741
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4540431797504425,
+      "learning_rate": 2.2332109665147127e-05,
+      "loss": 0.3808,
+      "step": 742
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4173644185066223,
+      "learning_rate": 2.2205701312462617e-05,
+      "loss": 0.3577,
+      "step": 743
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.39266419410705566,
+      "learning_rate": 2.2079549534610606e-05,
+      "loss": 0.2936,
+      "step": 744
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4002622365951538,
+      "learning_rate": 2.1953655496123853e-05,
+      "loss": 0.2844,
+      "step": 745
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4227045774459839,
+      "learning_rate": 2.1828020359155905e-05,
+      "loss": 0.2866,
+      "step": 746
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.37462419271469116,
+      "learning_rate": 2.1702645283470236e-05,
+      "loss": 0.2438,
+      "step": 747
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.46883416175842285,
+      "learning_rate": 2.1577531426429782e-05,
+      "loss": 0.3699,
+      "step": 748
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3425189256668091,
+      "learning_rate": 2.1452679942985993e-05,
+      "loss": 0.2275,
+      "step": 749
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4499514698982239,
+      "learning_rate": 2.132809198566837e-05,
+      "loss": 0.3443,
+      "step": 750
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.46920251846313477,
+      "learning_rate": 2.1203768704573672e-05,
+      "loss": 0.3687,
+      "step": 751
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3865799903869629,
+      "learning_rate": 2.1079711247355505e-05,
+      "loss": 0.2733,
+      "step": 752
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.41232362389564514,
+      "learning_rate": 2.095592075921347e-05,
+      "loss": 0.246,
+      "step": 753
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.32365286350250244,
+      "learning_rate": 2.08323983828828e-05,
+      "loss": 0.1873,
+      "step": 754
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4555100202560425,
+      "learning_rate": 2.0709145258623704e-05,
+      "loss": 0.3822,
+      "step": 755
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.41039618849754333,
+      "learning_rate": 2.0586162524210895e-05,
+      "loss": 0.283,
+      "step": 756
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.43508920073509216,
+      "learning_rate": 2.0463451314923015e-05,
+      "loss": 0.3266,
+      "step": 757
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.43862390518188477,
+      "learning_rate": 2.0341012763532243e-05,
+      "loss": 0.3413,
+      "step": 758
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3817911446094513,
+      "learning_rate": 2.021884800029379e-05,
+      "loss": 0.2646,
+      "step": 759
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3952636122703552,
+      "learning_rate": 2.009695815293548e-05,
+      "loss": 0.2985,
+      "step": 760
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4302140474319458,
+      "learning_rate": 1.9975344346647297e-05,
+      "loss": 0.3136,
+      "step": 761
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.38452428579330444,
+      "learning_rate": 1.9854007704071064e-05,
+      "loss": 0.2897,
+      "step": 762
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3167802095413208,
+      "learning_rate": 1.973294934529007e-05,
+      "loss": 0.2055,
+      "step": 763
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3533049523830414,
+      "learning_rate": 1.961217038781863e-05,
+      "loss": 0.2597,
+      "step": 764
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.4575469195842743,
+      "learning_rate": 1.9491671946591962e-05,
+      "loss": 0.3197,
+      "step": 765
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.4529496729373932,
+      "learning_rate": 1.9371455133955675e-05,
+      "loss": 0.3478,
+      "step": 766
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3405405580997467,
+      "learning_rate": 1.925152105965567e-05,
+      "loss": 0.268,
+      "step": 767
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3697284758090973,
+      "learning_rate": 1.9131870830827818e-05,
+      "loss": 0.2315,
+      "step": 768
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.41890275478363037,
+      "learning_rate": 1.9012505551987765e-05,
+      "loss": 0.2721,
+      "step": 769
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.37630268931388855,
+      "learning_rate": 1.8893426325020686e-05,
+      "loss": 0.2562,
+      "step": 770
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.4262970983982086,
+      "learning_rate": 1.8774634249171185e-05,
+      "loss": 0.4016,
+      "step": 771
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3830007016658783,
+      "learning_rate": 1.8656130421033123e-05,
+      "loss": 0.2976,
+      "step": 772
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.34678715467453003,
+      "learning_rate": 1.8537915934539486e-05,
+      "loss": 0.1675,
+      "step": 773
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.38951581716537476,
+      "learning_rate": 1.841999188095224e-05,
+      "loss": 0.3078,
+      "step": 774
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.41767624020576477,
+      "learning_rate": 1.830235934885237e-05,
+      "loss": 0.3639,
+      "step": 775
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.35205718874931335,
+      "learning_rate": 1.818501942412975e-05,
+      "loss": 0.2867,
+      "step": 776
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3999721109867096,
+      "learning_rate": 1.8067973189973075e-05,
+      "loss": 0.3284,
+      "step": 777
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.4182070195674896,
+      "learning_rate": 1.7951221726860045e-05,
+      "loss": 0.337,
+      "step": 778
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.41383877396583557,
+      "learning_rate": 1.7834766112547142e-05,
+      "loss": 0.3111,
+      "step": 779
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.381978303194046,
+      "learning_rate": 1.771860742205988e-05,
+      "loss": 0.3027,
+      "step": 780
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.41920870542526245,
+      "learning_rate": 1.7602746727682796e-05,
+      "loss": 0.3606,
+      "step": 781
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3676052391529083,
+      "learning_rate": 1.7487185098949565e-05,
+      "loss": 0.2786,
+      "step": 782
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.43040353059768677,
+      "learning_rate": 1.7371923602633078e-05,
+      "loss": 0.3621,
+      "step": 783
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.4287657141685486,
+      "learning_rate": 1.725696330273575e-05,
+      "loss": 0.3691,
+      "step": 784
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.40313223004341125,
+      "learning_rate": 1.7142305260479474e-05,
+      "loss": 0.362,
+      "step": 785
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3955127000808716,
+      "learning_rate": 1.7027950534296027e-05,
+      "loss": 0.3765,
+      "step": 786
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.34894323348999023,
+      "learning_rate": 1.6913900179817144e-05,
+      "loss": 0.2277,
+      "step": 787
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.39063969254493713,
+      "learning_rate": 1.6800155249864896e-05,
+      "loss": 0.3301,
+      "step": 788
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3707781136035919,
+      "learning_rate": 1.668671679444192e-05,
+      "loss": 0.286,
+      "step": 789
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.4262647330760956,
+      "learning_rate": 1.6573585860721646e-05,
+      "loss": 0.3549,
+      "step": 790
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3833974003791809,
+      "learning_rate": 1.646076349303884e-05,
+      "loss": 0.2805,
+      "step": 791
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3846880793571472,
+      "learning_rate": 1.63482507328797e-05,
+      "loss": 0.2934,
+      "step": 792
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.30670422315597534,
+      "learning_rate": 1.6236048618872456e-05,
+      "loss": 0.2128,
+      "step": 793
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.502191424369812,
+      "learning_rate": 1.6124158186777676e-05,
+      "loss": 0.3166,
+      "step": 794
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4251638650894165,
+      "learning_rate": 1.6012580469478743e-05,
+      "loss": 0.3323,
+      "step": 795
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3232210576534271,
+      "learning_rate": 1.5901316496972262e-05,
+      "loss": 0.1939,
+      "step": 796
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3785446286201477,
+      "learning_rate": 1.5790367296358644e-05,
+      "loss": 0.2983,
+      "step": 797
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.31697627902030945,
+      "learning_rate": 1.5679733891832556e-05,
+      "loss": 0.2113,
+      "step": 798
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.38774025440216064,
+      "learning_rate": 1.55694173046735e-05,
+      "loss": 0.2785,
+      "step": 799
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3695196807384491,
+      "learning_rate": 1.5459418553236343e-05,
+      "loss": 0.2784,
+      "step": 800
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4070797264575958,
+      "learning_rate": 1.5349738652941968e-05,
+      "loss": 0.3233,
+      "step": 801
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.37904244661331177,
+      "learning_rate": 1.5240378616267886e-05,
+      "loss": 0.2564,
+      "step": 802
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3779933750629425,
+      "learning_rate": 1.5131339452738863e-05,
+      "loss": 0.3678,
+      "step": 803
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3886728584766388,
+      "learning_rate": 1.5022622168917649e-05,
+      "loss": 0.2751,
+      "step": 804
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3595404624938965,
+      "learning_rate": 1.4914227768395595e-05,
+      "loss": 0.2592,
+      "step": 805
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.4031426012516022,
+      "learning_rate": 1.4806157251783515e-05,
+      "loss": 0.2972,
+      "step": 806
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.36621609330177307,
+      "learning_rate": 1.4698411616702356e-05,
+      "loss": 0.253,
+      "step": 807
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.38698112964630127,
+      "learning_rate": 1.4590991857774038e-05,
+      "loss": 0.3141,
+      "step": 808
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.40530022978782654,
+      "learning_rate": 1.4483898966612209e-05,
+      "loss": 0.3769,
+      "step": 809
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3705315887928009,
+      "learning_rate": 1.437713393181317e-05,
+      "loss": 0.2756,
+      "step": 810
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.40540188550949097,
+      "learning_rate": 1.4270697738946704e-05,
+      "loss": 0.2587,
+      "step": 811
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3355952203273773,
+      "learning_rate": 1.4164591370547004e-05,
+      "loss": 0.2367,
+      "step": 812
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.39372363686561584,
+      "learning_rate": 1.4058815806103542e-05,
+      "loss": 0.2863,
+      "step": 813
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.41405317187309265,
+      "learning_rate": 1.3953372022052107e-05,
+      "loss": 0.3338,
+      "step": 814
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.34528371691703796,
+      "learning_rate": 1.3848260991765755e-05,
+      "loss": 0.2889,
+      "step": 815
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.4991545081138611,
+      "learning_rate": 1.3743483685545811e-05,
+      "loss": 0.41,
+      "step": 816
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.41992202401161194,
+      "learning_rate": 1.363904107061294e-05,
+      "loss": 0.3185,
+      "step": 817
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.4431324601173401,
+      "learning_rate": 1.3534934111098179e-05,
+      "loss": 0.3361,
+      "step": 818
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3921603262424469,
+      "learning_rate": 1.3431163768034077e-05,
+      "loss": 0.3133,
+      "step": 819
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3576185703277588,
+      "learning_rate": 1.3327730999345817e-05,
+      "loss": 0.2269,
+      "step": 820
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.35856878757476807,
+      "learning_rate": 1.3224636759842363e-05,
+      "loss": 0.2933,
+      "step": 821
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.33871906995773315,
+      "learning_rate": 1.3121882001207614e-05,
+      "loss": 0.2214,
+      "step": 822
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.39352303743362427,
+      "learning_rate": 1.3019467671991692e-05,
+      "loss": 0.2456,
+      "step": 823
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.4099547266960144,
+      "learning_rate": 1.2917394717602121e-05,
+      "loss": 0.3549,
+      "step": 824
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3432406485080719,
+      "learning_rate": 1.2815664080295159e-05,
+      "loss": 0.2423,
+      "step": 825
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3456730842590332,
+      "learning_rate": 1.2714276699166994e-05,
+      "loss": 0.2323,
+      "step": 826
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3806013762950897,
+      "learning_rate": 1.261323351014525e-05,
+      "loss": 0.257,
+      "step": 827
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.32994189858436584,
+      "learning_rate": 1.251253544598014e-05,
+      "loss": 0.2059,
+      "step": 828
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3641059696674347,
+      "learning_rate": 1.241218343623602e-05,
+      "loss": 0.3046,
+      "step": 829
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.39203181862831116,
+      "learning_rate": 1.2312178407282749e-05,
+      "loss": 0.3047,
+      "step": 830
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3873448669910431,
+      "learning_rate": 1.2212521282287092e-05,
+      "loss": 0.2999,
+      "step": 831
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.43280982971191406,
+      "learning_rate": 1.2113212981204292e-05,
+      "loss": 0.3629,
+      "step": 832
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.42592155933380127,
+      "learning_rate": 1.2014254420769466e-05,
+      "loss": 0.3262,
+      "step": 833
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3780333995819092,
+      "learning_rate": 1.1915646514489292e-05,
+      "loss": 0.29,
+      "step": 834
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.38742387294769287,
+      "learning_rate": 1.1817390172633403e-05,
+      "loss": 0.3148,
+      "step": 835
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.34696269035339355,
+      "learning_rate": 1.1719486302226118e-05,
+      "loss": 0.2396,
+      "step": 836
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5252797603607178,
+      "learning_rate": 1.1621935807038003e-05,
+      "loss": 0.483,
+      "step": 837
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.4171554744243622,
+      "learning_rate": 1.152473958757756e-05,
+      "loss": 0.3367,
+      "step": 838
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.35319188237190247,
+      "learning_rate": 1.1427898541082855e-05,
+      "loss": 0.2354,
+      "step": 839
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.43658795952796936,
+      "learning_rate": 1.133141356151336e-05,
+      "loss": 0.3273,
+      "step": 840
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.4028431177139282,
+      "learning_rate": 1.123528553954154e-05,
+      "loss": 0.3028,
+      "step": 841
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.30417734384536743,
+      "learning_rate": 1.1139515362544755e-05,
+      "loss": 0.2006,
+      "step": 842
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.4621390700340271,
+      "learning_rate": 1.1044103914597031e-05,
+      "loss": 0.387,
+      "step": 843
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5159476399421692,
+      "learning_rate": 1.0949052076460853e-05,
+      "loss": 0.4158,
+      "step": 844
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.44435352087020874,
+      "learning_rate": 1.085436072557911e-05,
+      "loss": 0.3199,
+      "step": 845
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3563813269138336,
+      "learning_rate": 1.0760030736066951e-05,
+      "loss": 0.2455,
+      "step": 846
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3790179491043091,
+      "learning_rate": 1.0666062978703733e-05,
+      "loss": 0.2853,
+      "step": 847
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3730546236038208,
+      "learning_rate": 1.0572458320924943e-05,
+      "loss": 0.264,
+      "step": 848
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.43513819575309753,
+      "learning_rate": 1.0479217626814253e-05,
+      "loss": 0.3131,
+      "step": 849
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3191884160041809,
+      "learning_rate": 1.0386341757095502e-05,
+      "loss": 0.1952,
+      "step": 850
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3954907953739166,
+      "learning_rate": 1.0293831569124774e-05,
+      "loss": 0.3285,
+      "step": 851
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4020962417125702,
+      "learning_rate": 1.0201687916882418e-05,
+      "loss": 0.3469,
+      "step": 852
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3509984612464905,
+      "learning_rate": 1.0109911650965314e-05,
+      "loss": 0.2091,
+      "step": 853
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.558068573474884,
+      "learning_rate": 1.0018503618578818e-05,
+      "loss": 0.5226,
+      "step": 854
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.39664986729621887,
+      "learning_rate": 9.927464663529118e-06,
+      "loss": 0.2754,
+      "step": 855
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.44442465901374817,
+      "learning_rate": 9.836795626215356e-06,
+      "loss": 0.3699,
+      "step": 856
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.365354984998703,
+      "learning_rate": 9.746497343621857e-06,
+      "loss": 0.2974,
+      "step": 857
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.33468684554100037,
+      "learning_rate": 9.656570649310481e-06,
+      "loss": 0.2339,
+      "step": 858
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.36793625354766846,
+      "learning_rate": 9.567016373412857e-06,
+      "loss": 0.3041,
+      "step": 859
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.46790769696235657,
+      "learning_rate": 9.477835342622759e-06,
+      "loss": 0.3799,
+      "step": 860
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.4542980492115021,
+      "learning_rate": 9.389028380188419e-06,
+      "loss": 0.4007,
+      "step": 861
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.37855392694473267,
+      "learning_rate": 9.300596305905013e-06,
+      "loss": 0.293,
+      "step": 862
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3878996670246124,
+      "learning_rate": 9.212539936107029e-06,
+      "loss": 0.2728,
+      "step": 863
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.34039977192878723,
+      "learning_rate": 9.124860083660769e-06,
+      "loss": 0.2652,
+      "step": 864
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3311023414134979,
+      "learning_rate": 9.037557557956766e-06,
+      "loss": 0.1982,
+      "step": 865
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.4137350916862488,
+      "learning_rate": 8.950633164902467e-06,
+      "loss": 0.3247,
+      "step": 866
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.4069403111934662,
+      "learning_rate": 8.86408770691462e-06,
+      "loss": 0.2845,
+      "step": 867
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.33959850668907166,
+      "learning_rate": 8.777921982911996e-06,
+      "loss": 0.2373,
+      "step": 868
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.387346476316452,
+      "learning_rate": 8.692136788307903e-06,
+      "loss": 0.2683,
+      "step": 869
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.33370476961135864,
+      "learning_rate": 8.606732915003002e-06,
+      "loss": 0.2477,
+      "step": 870
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3765987455844879,
+      "learning_rate": 8.521711151377803e-06,
+      "loss": 0.262,
+      "step": 871
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.41339150071144104,
+      "learning_rate": 8.437072282285535e-06,
+      "loss": 0.3251,
+      "step": 872
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3473236858844757,
+      "learning_rate": 8.35281708904485e-06,
+      "loss": 0.2351,
+      "step": 873
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.38972267508506775,
+      "learning_rate": 8.268946349432582e-06,
+      "loss": 0.3143,
+      "step": 874
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.4202972948551178,
+      "learning_rate": 8.185460837676612e-06,
+      "loss": 0.2975,
+      "step": 875
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.4606451094150543,
+      "learning_rate": 8.102361324448715e-06,
+      "loss": 0.2833,
+      "step": 876
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.42644232511520386,
+      "learning_rate": 8.019648576857425e-06,
+      "loss": 0.3081,
+      "step": 877
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.4242898225784302,
+      "learning_rate": 7.937323358440935e-06,
+      "loss": 0.3451,
+      "step": 878
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3542291522026062,
+      "learning_rate": 7.85538642916015e-06,
+      "loss": 0.2743,
+      "step": 879
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.33673974871635437,
+      "learning_rate": 7.773838545391515e-06,
+      "loss": 0.2067,
+      "step": 880
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3924747109413147,
+      "learning_rate": 7.692680459920188e-06,
+      "loss": 0.2897,
+      "step": 881
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.33239954710006714,
+      "learning_rate": 7.6119129219329395e-06,
+      "loss": 0.2183,
+      "step": 882
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.43247902393341064,
+      "learning_rate": 7.5315366770114195e-06,
+      "loss": 0.3292,
+      "step": 883
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.4543492794036865,
+      "learning_rate": 7.4515524671250725e-06,
+      "loss": 0.2968,
+      "step": 884
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.39136525988578796,
+      "learning_rate": 7.371961030624452e-06,
+      "loss": 0.2538,
+      "step": 885
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3515141010284424,
+      "learning_rate": 7.292763102234329e-06,
+      "loss": 0.269,
+      "step": 886
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.36654534935951233,
+      "learning_rate": 7.213959413046894e-06,
+      "loss": 0.2089,
+      "step": 887
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3514954447746277,
+      "learning_rate": 7.135550690515052e-06,
+      "loss": 0.2764,
+      "step": 888
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.36854228377342224,
+      "learning_rate": 7.057537658445701e-06,
+      "loss": 0.238,
+      "step": 889
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3011555075645447,
+      "learning_rate": 6.979921036993042e-06,
+      "loss": 0.1986,
+      "step": 890
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4285282492637634,
+      "learning_rate": 6.902701542651874e-06,
+      "loss": 0.285,
+      "step": 891
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3245009481906891,
+      "learning_rate": 6.825879888251135e-06,
+      "loss": 0.1793,
+      "step": 892
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.38561704754829407,
+      "learning_rate": 6.749456782947122e-06,
+      "loss": 0.2576,
+      "step": 893
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4220375120639801,
+      "learning_rate": 6.6734329322171165e-06,
+      "loss": 0.3059,
+      "step": 894
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.391051322221756,
+      "learning_rate": 6.597809037852726e-06,
+      "loss": 0.3383,
+      "step": 895
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.38187360763549805,
+      "learning_rate": 6.522585797953579e-06,
+      "loss": 0.238,
+      "step": 896
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4633513391017914,
+      "learning_rate": 6.447763906920679e-06,
+      "loss": 0.3882,
+      "step": 897
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.48090532422065735,
+      "learning_rate": 6.373344055450165e-06,
+      "loss": 0.3395,
+      "step": 898
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3448203206062317,
+      "learning_rate": 6.2993269305268495e-06,
+      "loss": 0.2316,
+      "step": 899
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.41201066970825195,
+      "learning_rate": 6.2257132154178665e-06,
+      "loss": 0.3416,
+      "step": 900
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.4267149567604065,
+      "learning_rate": 6.152503589666425e-06,
+      "loss": 0.3322,
+      "step": 901
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.413591206073761,
+      "learning_rate": 6.079698729085498e-06,
+      "loss": 0.3095,
+      "step": 902
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3144916296005249,
+      "learning_rate": 6.007299305751585e-06,
+      "loss": 0.1855,
+      "step": 903
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.34966927766799927,
+      "learning_rate": 5.935305987998496e-06,
+      "loss": 0.2445,
+      "step": 904
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.36850506067276,
+      "learning_rate": 5.863719440411214e-06,
+      "loss": 0.2391,
+      "step": 905
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.37980982661247253,
+      "learning_rate": 5.792540323819751e-06,
+      "loss": 0.3392,
+      "step": 906
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.36540210247039795,
+      "learning_rate": 5.721769295293034e-06,
+      "loss": 0.2487,
+      "step": 907
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.4353797733783722,
+      "learning_rate": 5.651407008132809e-06,
+      "loss": 0.3582,
+      "step": 908
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3671395480632782,
+      "learning_rate": 5.5814541118677284e-06,
+      "loss": 0.2116,
+      "step": 909
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.4240861237049103,
+      "learning_rate": 5.5119112522471924e-06,
+      "loss": 0.3495,
+      "step": 910
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3780825138092041,
+      "learning_rate": 5.442779071235516e-06,
+      "loss": 0.2536,
+      "step": 911
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3448036313056946,
+      "learning_rate": 5.374058207005944e-06,
+      "loss": 0.2392,
+      "step": 912
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3841763138771057,
+      "learning_rate": 5.305749293934764e-06,
+      "loss": 0.2543,
+      "step": 913
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.4242177903652191,
+      "learning_rate": 5.237852962595469e-06,
+      "loss": 0.2981,
+      "step": 914
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3546169400215149,
+      "learning_rate": 5.170369839752925e-06,
+      "loss": 0.2047,
+      "step": 915
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.380683571100235,
+      "learning_rate": 5.1033005483575925e-06,
+      "loss": 0.2585,
+      "step": 916
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.4152151942253113,
+      "learning_rate": 5.036645707539745e-06,
+      "loss": 0.2585,
+      "step": 917
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3874864876270294,
+      "learning_rate": 4.9704059326038055e-06,
+      "loss": 0.2672,
+      "step": 918
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.4094529151916504,
+      "learning_rate": 4.90458183502262e-06,
+      "loss": 0.2893,
+      "step": 919
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.4090777337551117,
+      "learning_rate": 4.839174022431858e-06,
+      "loss": 0.3671,
+      "step": 920
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.4188554286956787,
+      "learning_rate": 4.7741830986243356e-06,
+      "loss": 0.3102,
+      "step": 921
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3730091154575348,
+      "learning_rate": 4.709609663544534e-06,
+      "loss": 0.2852,
+      "step": 922
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.38047468662261963,
+      "learning_rate": 4.645454313282965e-06,
+      "loss": 0.2954,
+      "step": 923
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4720805585384369,
+      "learning_rate": 4.581717640070743e-06,
+      "loss": 0.3471,
+      "step": 924
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4264400899410248,
+      "learning_rate": 4.5184002322740785e-06,
+      "loss": 0.346,
+      "step": 925
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.31896501779556274,
+      "learning_rate": 4.455502674388873e-06,
+      "loss": 0.2315,
+      "step": 926
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4741848111152649,
+      "learning_rate": 4.3930255470352736e-06,
+      "loss": 0.3508,
+      "step": 927
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.44249042868614197,
+      "learning_rate": 4.330969426952375e-06,
+      "loss": 0.3253,
+      "step": 928
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3980604410171509,
+      "learning_rate": 4.269334886992876e-06,
+      "loss": 0.2856,
+      "step": 929
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.33377861976623535,
+      "learning_rate": 4.208122496117744e-06,
+      "loss": 0.2117,
+      "step": 930
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.394549161195755,
+      "learning_rate": 4.147332819391048e-06,
+      "loss": 0.2755,
+      "step": 931
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4144418239593506,
+      "learning_rate": 4.0869664179746694e-06,
+      "loss": 0.2635,
+      "step": 932
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.35250324010849,
+      "learning_rate": 4.027023849123157e-06,
+      "loss": 0.2294,
+      "step": 933
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.38733455538749695,
+      "learning_rate": 3.967505666178556e-06,
+      "loss": 0.2676,
+      "step": 934
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.39043229818344116,
+      "learning_rate": 3.908412418565371e-06,
+      "loss": 0.2569,
+      "step": 935
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3419300317764282,
+      "learning_rate": 3.849744651785381e-06,
+      "loss": 0.2464,
+      "step": 936
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3910216689109802,
+      "learning_rate": 3.7915029074126974e-06,
+      "loss": 0.355,
+      "step": 937
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.50761479139328,
+      "learning_rate": 3.7336877230887246e-06,
+      "loss": 0.422,
+      "step": 938
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.4381940960884094,
+      "learning_rate": 3.676299632517216e-06,
+      "loss": 0.4159,
+      "step": 939
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.4971286952495575,
+      "learning_rate": 3.619339165459307e-06,
+      "loss": 0.3299,
+      "step": 940
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.35524582862854004,
+      "learning_rate": 3.562806847728678e-06,
+      "loss": 0.2657,
+      "step": 941
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.4120367765426636,
+      "learning_rate": 3.5067032011866783e-06,
+      "loss": 0.3563,
+      "step": 942
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3152506649494171,
+      "learning_rate": 3.4510287437374835e-06,
+      "loss": 0.1662,
+      "step": 943
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3767205476760864,
+      "learning_rate": 3.3957839893233536e-06,
+      "loss": 0.2408,
+      "step": 944
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3337627053260803,
+      "learning_rate": 3.340969447919873e-06,
+      "loss": 0.2414,
+      "step": 945
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.32742202281951904,
+      "learning_rate": 3.286585625531241e-06,
+      "loss": 0.2095,
+      "step": 946
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.34052425622940063,
+      "learning_rate": 3.232633024185583e-06,
+      "loss": 0.2212,
+      "step": 947
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.33039572834968567,
+      "learning_rate": 3.1791121419303794e-06,
+      "loss": 0.2211,
+      "step": 948
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.42443013191223145,
+      "learning_rate": 3.1260234728277717e-06,
+      "loss": 0.278,
+      "step": 949
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.4003521203994751,
+      "learning_rate": 3.0733675069500865e-06,
+      "loss": 0.2895,
+      "step": 950
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.4332895874977112,
+      "learning_rate": 3.0211447303752695e-06,
+      "loss": 0.2965,
+      "step": 951
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3819825351238251,
+      "learning_rate": 2.9693556251824185e-06,
+      "loss": 0.2882,
+      "step": 952
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.47884687781333923,
+      "learning_rate": 2.9180006694472906e-06,
+      "loss": 0.4212,
+      "step": 953
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.4191427528858185,
+      "learning_rate": 2.867080337237954e-06,
+      "loss": 0.3637,
+      "step": 954
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4317600727081299,
+      "learning_rate": 2.8165950986103805e-06,
+      "loss": 0.368,
+      "step": 955
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4285089671611786,
+      "learning_rate": 2.7665454196040664e-06,
+      "loss": 0.3267,
+      "step": 956
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.40216952562332153,
+      "learning_rate": 2.716931762237801e-06,
+      "loss": 0.279,
+      "step": 957
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4264661967754364,
+      "learning_rate": 2.667754584505372e-06,
+      "loss": 0.3342,
+      "step": 958
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5141720771789551,
+      "learning_rate": 2.6190143403713174e-06,
+      "loss": 0.3507,
+      "step": 959
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4230804145336151,
+      "learning_rate": 2.5707114797667465e-06,
+      "loss": 0.2819,
+      "step": 960
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.40919217467308044,
+      "learning_rate": 2.522846448585231e-06,
+      "loss": 0.349,
+      "step": 961
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4233115017414093,
+      "learning_rate": 2.4754196886785986e-06,
+      "loss": 0.2615,
+      "step": 962
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.32790645956993103,
+      "learning_rate": 2.4284316378529404e-06,
+      "loss": 0.2638,
+      "step": 963
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3933560848236084,
+      "learning_rate": 2.3818827298645207e-06,
+      "loss": 0.2482,
+      "step": 964
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3430630564689636,
+      "learning_rate": 2.335773394415802e-06,
+      "loss": 0.2201,
+      "step": 965
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.39491233229637146,
+      "learning_rate": 2.2901040571514322e-06,
+      "loss": 0.2974,
+      "step": 966
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.42395198345184326,
+      "learning_rate": 2.2448751396543787e-06,
+      "loss": 0.2932,
+      "step": 967
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.33768317103385925,
+      "learning_rate": 2.2000870594419908e-06,
+      "loss": 0.1831,
+      "step": 968
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3686957061290741,
+      "learning_rate": 2.155740229962161e-06,
+      "loss": 0.2596,
+      "step": 969
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3499439060688019,
+      "learning_rate": 2.1118350605894955e-06,
+      "loss": 0.3024,
+      "step": 970
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.35845717787742615,
+      "learning_rate": 2.068371956621562e-06,
+      "loss": 0.258,
+      "step": 971
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.4016646146774292,
+      "learning_rate": 2.0253513192751373e-06,
+      "loss": 0.3204,
+      "step": 972
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.38734307885169983,
+      "learning_rate": 1.982773545682459e-06,
+      "loss": 0.2911,
+      "step": 973
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3539869487285614,
+      "learning_rate": 1.9406390288876586e-06,
+      "loss": 0.223,
+      "step": 974
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3853650987148285,
+      "learning_rate": 1.8989481578430223e-06,
+      "loss": 0.2413,
+      "step": 975
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4223552644252777,
+      "learning_rate": 1.8577013174054857e-06,
+      "loss": 0.3261,
+      "step": 976
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4064064025878906,
+      "learning_rate": 1.8168988883330185e-06,
+      "loss": 0.3506,
+      "step": 977
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.47123757004737854,
+      "learning_rate": 1.7765412472811771e-06,
+      "loss": 0.4422,
+      "step": 978
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.398388534784317,
+      "learning_rate": 1.7366287667995417e-06,
+      "loss": 0.2235,
+      "step": 979
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3645593225955963,
+      "learning_rate": 1.697161815328363e-06,
+      "loss": 0.2561,
+      "step": 980
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.33609190583229065,
+      "learning_rate": 1.6581407571951092e-06,
+      "loss": 0.2756,
+      "step": 981
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3676416873931885,
+      "learning_rate": 1.6195659526111185e-06,
+      "loss": 0.2628,
+      "step": 982
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4234568476676941,
+      "learning_rate": 1.5814377576682527e-06,
+      "loss": 0.303,
+      "step": 983
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3596719801425934,
+      "learning_rate": 1.5437565243356656e-06,
+      "loss": 0.1933,
+      "step": 984
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.42232537269592285,
+      "learning_rate": 1.5065226004564893e-06,
+      "loss": 0.3734,
+      "step": 985
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.38415658473968506,
+      "learning_rate": 1.4697363297446477e-06,
+      "loss": 0.2759,
+      "step": 986
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.4657137393951416,
+      "learning_rate": 1.4333980517817203e-06,
+      "loss": 0.3853,
+      "step": 987
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3416013717651367,
+      "learning_rate": 1.3975081020137392e-06,
+      "loss": 0.2035,
+      "step": 988
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5011183023452759,
+      "learning_rate": 1.3620668117481472e-06,
+      "loss": 0.3493,
+      "step": 989
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3816926181316376,
+      "learning_rate": 1.3270745081506997e-06,
+      "loss": 0.3283,
+      "step": 990
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5070998072624207,
+      "learning_rate": 1.292531514242501e-06,
+      "loss": 0.4696,
+      "step": 991
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.40773579478263855,
+      "learning_rate": 1.2584381488969454e-06,
+      "loss": 0.2974,
+      "step": 992
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3139830529689789,
+      "learning_rate": 1.2247947268368364e-06,
+      "loss": 0.2197,
+      "step": 993
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3780119717121124,
+      "learning_rate": 1.191601558631461e-06,
+      "loss": 0.2619,
+      "step": 994
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3527129292488098,
+      "learning_rate": 1.1588589506937198e-06,
+      "loss": 0.2688,
+      "step": 995
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5044612288475037,
+      "learning_rate": 1.126567205277279e-06,
+      "loss": 0.3573,
+      "step": 996
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3708314597606659,
+      "learning_rate": 1.094726620473835e-06,
+      "loss": 0.2795,
+      "step": 997
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.4164911210536957,
+      "learning_rate": 1.0633374902103088e-06,
+      "loss": 0.3088,
+      "step": 998
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.38613763451576233,
+      "learning_rate": 1.0324001042461395e-06,
+      "loss": 0.2573,
+      "step": 999
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.416906476020813,
+      "learning_rate": 1.0019147481706625e-06,
+      "loss": 0.3277,
+      "step": 1000
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.4968281686306,
+      "learning_rate": 9.718817034003901e-07,
+      "loss": 0.4193,
+      "step": 1001
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3328755497932434,
+      "learning_rate": 9.423012471764914e-07,
+      "loss": 0.198,
+      "step": 1002
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3636288046836853,
+      "learning_rate": 9.131736525621603e-07,
+      "loss": 0.2204,
+      "step": 1003
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.42198479175567627,
+      "learning_rate": 8.844991884401854e-07,
+      "loss": 0.2807,
+      "step": 1004
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.39438503980636597,
+      "learning_rate": 8.56278119510362e-07,
+      "loss": 0.2664,
+      "step": 1005
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.44490867853164673,
+      "learning_rate": 8.285107062871333e-07,
+      "loss": 0.3478,
+      "step": 1006
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3523523807525635,
+      "learning_rate": 8.011972050971483e-07,
+      "loss": 0.2738,
+      "step": 1007
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.4044662117958069,
+      "learning_rate": 7.74337868076902e-07,
+      "loss": 0.3216,
+      "step": 1008
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.45560404658317566,
+      "learning_rate": 7.479329431703985e-07,
+      "loss": 0.4045,
+      "step": 1009
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.46285828948020935,
+      "learning_rate": 7.21982674126881e-07,
+      "loss": 0.3443,
+      "step": 1010
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.36943697929382324,
+      "learning_rate": 6.964873004985717e-07,
+      "loss": 0.2318,
+      "step": 1011
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.47358638048171997,
+      "learning_rate": 6.714470576384579e-07,
+      "loss": 0.3709,
+      "step": 1012
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3758651614189148,
+      "learning_rate": 6.468621766981154e-07,
+      "loss": 0.2717,
+      "step": 1013
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3843158781528473,
+      "learning_rate": 6.227328846255931e-07,
+      "loss": 0.2687,
+      "step": 1014
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3527068793773651,
+      "learning_rate": 5.990594041632991e-07,
+      "loss": 0.2499,
+      "step": 1015
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3666359782218933,
+      "learning_rate": 5.758419538459459e-07,
+      "loss": 0.269,
+      "step": 1016
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.45985522866249084,
+      "learning_rate": 5.530807479985633e-07,
+      "loss": 0.3482,
+      "step": 1017
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3351965546607971,
+      "learning_rate": 5.307759967344672e-07,
+      "loss": 0.2629,
+      "step": 1018
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4313323199748993,
+      "learning_rate": 5.089279059533658e-07,
+      "loss": 0.3296,
+      "step": 1019
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.44096696376800537,
+      "learning_rate": 4.87536677339423e-07,
+      "loss": 0.3566,
+      "step": 1020
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3699873983860016,
+      "learning_rate": 4.666025083594483e-07,
+      "loss": 0.2561,
+      "step": 1021
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3967258036136627,
+      "learning_rate": 4.461255922609986e-07,
+      "loss": 0.2996,
+      "step": 1022
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.44547387957572937,
+      "learning_rate": 4.261061180706627e-07,
+      "loss": 0.3546,
+      "step": 1023
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4082671105861664,
+      "learning_rate": 4.065442705922906e-07,
+      "loss": 0.3672,
+      "step": 1024
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4149584174156189,
+      "learning_rate": 3.8744023040528374e-07,
+      "loss": 0.2984,
+      "step": 1025
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.41881927847862244,
+      "learning_rate": 3.687941738629186e-07,
+      "loss": 0.3459,
+      "step": 1026
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.39721134305000305,
+      "learning_rate": 3.5060627309074224e-07,
+      "loss": 0.2751,
+      "step": 1027
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.40128853917121887,
+      "learning_rate": 3.3287669598497383e-07,
+      "loss": 0.2707,
+      "step": 1028
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4406619369983673,
+      "learning_rate": 3.156056062109503e-07,
+      "loss": 0.3462,
+      "step": 1029
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3375759422779083,
+      "learning_rate": 2.987931632016272e-07,
+      "loss": 0.2176,
+      "step": 1030
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4124433696269989,
+      "learning_rate": 2.824395221560805e-07,
+      "loss": 0.2863,
+      "step": 1031
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.38367995619773865,
+      "learning_rate": 2.665448340381016e-07,
+      "loss": 0.2667,
+      "step": 1032
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.37265920639038086,
+      "learning_rate": 2.511092455747932e-07,
+      "loss": 0.2871,
+      "step": 1033
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3789976239204407,
+      "learning_rate": 2.361328992552314e-07,
+      "loss": 0.2494,
+      "step": 1034
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.48114919662475586,
+      "learning_rate": 2.2161593332910013e-07,
+      "loss": 0.3579,
+      "step": 1035
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4425215423107147,
+      "learning_rate": 2.0755848180547543e-07,
+      "loss": 0.3963,
+      "step": 1036
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.32901331782341003,
+      "learning_rate": 1.9396067445155986e-07,
+      "loss": 0.2204,
+      "step": 1037
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4334326684474945,
+      "learning_rate": 1.8082263679148337e-07,
+      "loss": 0.3967,
+      "step": 1038
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4191541373729706,
+      "learning_rate": 1.681444901051432e-07,
+      "loss": 0.3374,
+      "step": 1039
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.39676961302757263,
+      "learning_rate": 1.5592635142709367e-07,
+      "loss": 0.2838,
+      "step": 1040
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.38776692748069763,
+      "learning_rate": 1.4416833354546356e-07,
+      "loss": 0.269,
+      "step": 1041
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.40096643567085266,
+      "learning_rate": 1.328705450009071e-07,
+      "loss": 0.2566,
+      "step": 1042
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.43819060921669006,
+      "learning_rate": 1.2203309008561592e-07,
+      "loss": 0.3519,
+      "step": 1043
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.4747885763645172,
+      "learning_rate": 1.1165606884234181e-07,
+      "loss": 0.3204,
+      "step": 1044
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.4276343882083893,
+      "learning_rate": 1.0173957706348659e-07,
+      "loss": 0.3291,
+      "step": 1045
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.4158606231212616,
+      "learning_rate": 9.228370629019711e-08,
+      "loss": 0.2904,
+      "step": 1046
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.44644296169281006,
+      "learning_rate": 8.328854381154938e-08,
+      "loss": 0.3455,
+      "step": 1047
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.34873446822166443,
+      "learning_rate": 7.475417266371576e-08,
+      "loss": 0.2699,
+      "step": 1048
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.36457881331443787,
+      "learning_rate": 6.668067162921566e-08,
+      "loss": 0.2634,
+      "step": 1049
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.39075329899787903,
+      "learning_rate": 5.906811523618272e-08,
+      "loss": 0.2967,
+      "step": 1050
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.4555211067199707,
+      "learning_rate": 5.191657375767656e-08,
+      "loss": 0.3807,
+      "step": 1051
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.39703041315078735,
+      "learning_rate": 4.522611321103321e-08,
+      "loss": 0.2763,
+      "step": 1052
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3615953326225281,
+      "learning_rate": 3.8996795357254535e-08,
+      "loss": 0.2181,
+      "step": 1053
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5408671498298645,
+      "learning_rate": 3.322867770044202e-08,
+      "loss": 0.374,
+      "step": 1054
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.463867723941803,
+      "learning_rate": 2.792181348726941e-08,
+      "loss": 0.3265,
+      "step": 1055
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.46088552474975586,
+      "learning_rate": 2.3076251706477536e-08,
+      "loss": 0.3334,
+      "step": 1056
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.4711132347583771,
+      "learning_rate": 1.869203708843581e-08,
+      "loss": 0.3801,
+      "step": 1057
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.39382699131965637,
+      "learning_rate": 1.476921010471477e-08,
+      "loss": 0.3173,
+      "step": 1058
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.4106517434120178,
+      "learning_rate": 1.1307806967741919e-08,
+      "loss": 0.2978,
+      "step": 1059
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.32475370168685913,
+      "learning_rate": 8.307859630429793e-09,
+      "loss": 0.2326,
+      "step": 1060
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.45648568868637085,
+      "learning_rate": 5.7693957858984125e-09,
+      "loss": 0.3561,
+      "step": 1061
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.41961532831192017,
+      "learning_rate": 3.6924388672254785e-09,
+      "loss": 0.3269,
+      "step": 1062
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.3751271069049835,
+      "learning_rate": 2.0770080472298783e-09,
+      "loss": 0.2859,
+      "step": 1063
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.4922095239162445,
+      "learning_rate": 9.231182382773984e-10,
+      "loss": 0.3797,
+      "step": 1064
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.38687193393707275,
+      "learning_rate": 2.307800921641512e-10,
+      "loss": 0.2846,
+      "step": 1065
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.390020489692688,
+      "learning_rate": 0.0,
+      "loss": 0.2777,
+      "step": 1066
+    },
+    {
+      "epoch": 1.0,
+      "step": 1066,
+      "total_flos": 1.171159722670162e+17,
+      "train_loss": 0.31241298194338635,
+      "train_runtime": 51725.3719,
+      "train_samples_per_second": 0.66,
+      "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
+}