Upload folder using huggingface_hub

.gitattributes

@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+tokenizer.json filter=lfs diff=lfs merge=lfs -text

added_tokens.json

@@ -0,0 +1,28 @@
+{
+  "</think>": 151668,
+  "</tool_call>": 151658,
+  "</tool_response>": 151666,
+  "<think>": 151667,
+  "<tool_call>": 151657,
+  "<tool_response>": 151665,
+  "<|box_end|>": 151649,
+  "<|box_start|>": 151648,
+  "<|endoftext|>": 151643,
+  "<|file_sep|>": 151664,
+  "<|fim_middle|>": 151660,
+  "<|fim_pad|>": 151662,
+  "<|fim_prefix|>": 151659,
+  "<|fim_suffix|>": 151661,
+  "<|im_end|>": 151645,
+  "<|im_start|>": 151644,
+  "<|image_pad|>": 151655,
+  "<|object_ref_end|>": 151647,
+  "<|object_ref_start|>": 151646,
+  "<|quad_end|>": 151651,
+  "<|quad_start|>": 151650,
+  "<|repo_name|>": 151663,
+  "<|video_pad|>": 151656,
+  "<|vision_end|>": 151653,
+  "<|vision_pad|>": 151654,
+  "<|vision_start|>": 151652
+}

chat_template.jinja

@@ -0,0 +1,85 @@
+{%- if tools %}
+    {{- '<|im_start|>system\n' }}
+    {%- if messages[0].role == 'system' %}
+        {{- messages[0].content + '\n\n' }}
+    {%- endif %}
+    {{- "# Tools\n\nYou may call one or more functions to assist with the user query.\n\nYou are provided with function signatures within <tools></tools> XML tags:\n<tools>" }}
+    {%- for tool in tools %}
+        {{- "\n" }}
+        {{- tool | tojson }}
+    {%- endfor %}
+    {{- "\n</tools>\n\nFor each function call, return a json object with function name and arguments within <tool_call></tool_call> XML tags:\n<tool_call>\n{\"name\": <function-name>, \"arguments\": <args-json-object>}\n</tool_call><|im_end|>\n" }}
+{%- else %}
+    {%- if messages[0].role == 'system' %}
+        {{- '<|im_start|>system\n' + messages[0].content + '<|im_end|>\n' }}
+    {%- endif %}
+{%- endif %}
+{%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %}
+{%- for message in messages[::-1] %}
+    {%- set index = (messages|length - 1) - loop.index0 %}
+    {%- if ns.multi_step_tool and message.role == "user" and not(message.content.startswith('<tool_response>') and message.content.endswith('</tool_response>')) %}
+        {%- set ns.multi_step_tool = false %}
+        {%- set ns.last_query_index = index %}
+    {%- endif %}
+{%- endfor %}
+{%- for message in messages %}
+    {%- if (message.role == "user") or (message.role == "system" and not loop.first) %}
+        {{- '<|im_start|>' + message.role + '\n' + message.content + '<|im_end|>' + '\n' }}
+    {%- elif message.role == "assistant" %}
+        {%- set content = message.content %}
+        {%- set reasoning_content = '' %}
+        {%- if message.reasoning_content is defined and message.reasoning_content is not none %}
+            {%- set reasoning_content = message.reasoning_content %}
+        {%- else %}
+            {%- if '</think>' in message.content %}
+                {%- set content = message.content.split('</think>')[-1].lstrip('\n') %}
+                {%- set reasoning_content = message.content.split('</think>')[0].rstrip('\n').split('<think>')[-1].lstrip('\n') %}
+            {%- endif %}
+        {%- endif %}
+        {%- if loop.index0 > ns.last_query_index %}
+            {%- if loop.last or (not loop.last and reasoning_content) %}
+                {{- '<|im_start|>' + message.role + '\n<think>\n' + reasoning_content.strip('\n') + '\n</think>\n\n' + content.lstrip('\n') }}
+            {%- else %}
+                {{- '<|im_start|>' + message.role + '\n' + content }}
+            {%- endif %}
+        {%- else %}
+            {{- '<|im_start|>' + message.role + '\n' + content }}
+        {%- endif %}
+        {%- if message.tool_calls %}
+            {%- for tool_call in message.tool_calls %}
+                {%- if (loop.first and content) or (not loop.first) %}
+                    {{- '\n' }}
+                {%- endif %}
+                {%- if tool_call.function %}
+                    {%- set tool_call = tool_call.function %}
+                {%- endif %}
+                {{- '<tool_call>\n{"name": "' }}
+                {{- tool_call.name }}
+                {{- '", "arguments": ' }}
+                {%- if tool_call.arguments is string %}
+                    {{- tool_call.arguments }}
+                {%- else %}
+                    {{- tool_call.arguments | tojson }}
+                {%- endif %}
+                {{- '}\n</tool_call>' }}
+            {%- endfor %}
+        {%- endif %}
+        {{- '<|im_end|>\n' }}
+    {%- elif message.role == "tool" %}
+        {%- if loop.first or (messages[loop.index0 - 1].role != "tool") %}
+            {{- '<|im_start|>user' }}
+        {%- endif %}
+        {{- '\n<tool_response>\n' }}
+        {{- message.content }}
+        {{- '\n</tool_response>' }}
+        {%- if loop.last or (messages[loop.index0 + 1].role != "tool") %}
+            {{- '<|im_end|>\n' }}
+        {%- endif %}
+    {%- endif %}
+{%- endfor %}
+{%- if add_generation_prompt %}
+    {{- '<|im_start|>assistant\n' }}
+    {%- if enable_thinking is defined and enable_thinking is false %}
+        {{- '<think>\n\n</think>\n\n' }}
+    {%- endif %}
+{%- endif %}

config.json

@@ -0,0 +1,38 @@
+{
+  "architectures": [
+    "Qwen3MoeForCausalLM"
+  ],
+  "attention_bias": false,
+  "attention_dropout": 0.0,
+  "bos_token_id": 151643,
+  "decoder_sparse_step": 1,
+  "eos_token_id": 151643,
+  "head_dim": 128,
+  "hidden_act": "silu",
+  "hidden_size": 2048,
+  "initializer_range": 0.02,
+  "intermediate_size": 6144,
+  "max_position_embeddings": 32768,
+  "max_window_layers": 48,
+  "mlp_only_layers": [],
+  "model_type": "qwen3_moe",
+  "moe_intermediate_size": 768,
+  "norm_topk_prob": true,
+  "num_attention_heads": 32,
+  "num_experts": 128,
+  "num_experts_per_tok": 8,
+  "num_hidden_layers": 48,
+  "num_key_value_heads": 4,
+  "output_router_logits": false,
+  "rms_norm_eps": 1e-06,
+  "rope_scaling": null,
+  "rope_theta": 1000000.0,
+  "router_aux_loss_coef": 0.001,
+  "sliding_window": null,
+  "tie_word_embeddings": false,
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.52.0.dev0",
+  "use_cache": false,
+  "use_sliding_window": false,
+  "vocab_size": 151936
+}

generation_config.json

@@ -0,0 +1,6 @@
+{
+  "bos_token_id": 151643,
+  "eos_token_id": 151643,
+  "max_new_tokens": 2048,
+  "transformers_version": "4.52.0.dev0"
+}

model-00001-of-00013.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cc894ba45d367c21dc5047cd229ab0fa2c6dbdf0fc5809b5e5125756e8e40d75
+size 4997184968

model-00002-of-00013.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d2238a110b9b60f223126cbc7066f6b480a226d6059126fed11d033358ccee1e
+size 4997741608

model-00003-of-00013.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:346d5ac9f942a96885bc3605b1bd87ce745483d5f587dab7666098065543e13e
+size 4997742208

model-00004-of-00013.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4784b1d8461eb6bd65a7e03c7a7f3359b05ba2076eacc3ca59f79368d7a6e33f
+size 4997743184

model-00005-of-00013.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:464fd04789559e854889ebefa7bc9a9c1dd557c4a4c45c6c3eab7c75f67ca8d5
+size 4997743184

model-00006-of-00013.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3e2d8b3053937fa59a6ed32a9fd307619ec8efd0db9cc0e055de828c50dabb91
+size 4997743184

model-00007-of-00013.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:40041164fb7b5698db6d82beb90a95b607fc4b3b3ca7c34aa527ff3232f501a0
+size 4997743184

model-00008-of-00013.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0096bacda5b37df99471dca586677cdd65d9e41d6097d100922fd4d0e408dd63
+size 4997743184

model-00009-of-00013.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:24ed0468994d6e61696f4ebbb37547c3d868bd54ae736daa0cd4a568bc4d77a8
+size 4997743184

model-00010-of-00013.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d068c5da1e8ffe41fbe4d95dead1cc0fec2d7fb4ccb29bc7d177a472d8958f15
+size 4997743184

model-00011-of-00013.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:24bd0f33a58da18a89e5de622fff18e22a0ae3812c7c503fb42e85f7d72067ab
+size 4997743184

model-00012-of-00013.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d56e8f8f943406d316004dbd3ab1eb929c0bab08f03c871b23ea1c59a5fc5ca8
+size 4997743184

model-00013-of-00013.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:510870998d1334f1ea40a86cb3f56b93cd153b5a29496a479d6d6f67eaa24254
+size 1094220288

special_tokens_map.json

@@ -0,0 +1,31 @@
+{
+  "additional_special_tokens": [
+    "<|im_start|>",
+    "<|im_end|>",
+    "<|object_ref_start|>",
+    "<|object_ref_end|>",
+    "<|box_start|>",
+    "<|box_end|>",
+    "<|quad_start|>",
+    "<|quad_end|>",
+    "<|vision_start|>",
+    "<|vision_end|>",
+    "<|vision_pad|>",
+    "<|image_pad|>",
+    "<|video_pad|>"
+  ],
+  "eos_token": {
+    "content": "<|im_end|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": {
+    "content": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
+}

tokenizer.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:aeb13307a71acd8fe81861d94ad54ab689df773318809eed3cbe794b4492dae4
+size 11422654

tokenizer_config.json

@@ -0,0 +1,240 @@
+{
+  "add_bos_token": false,
+  "add_prefix_space": false,
+  "added_tokens_decoder": {
+    "151643": {
+      "content": "<|endoftext|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151644": {
+      "content": "<|im_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151645": {
+      "content": "<|im_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151646": {
+      "content": "<|object_ref_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151647": {
+      "content": "<|object_ref_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151648": {
+      "content": "<|box_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151649": {
+      "content": "<|box_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151650": {
+      "content": "<|quad_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151651": {
+      "content": "<|quad_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151652": {
+      "content": "<|vision_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151653": {
+      "content": "<|vision_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151654": {
+      "content": "<|vision_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151655": {
+      "content": "<|image_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151656": {
+      "content": "<|video_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151657": {
+      "content": "<tool_call>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151658": {
+      "content": "</tool_call>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151659": {
+      "content": "<|fim_prefix|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151660": {
+      "content": "<|fim_middle|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151661": {
+      "content": "<|fim_suffix|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151662": {
+      "content": "<|fim_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151663": {
+      "content": "<|repo_name|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151664": {
+      "content": "<|file_sep|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151665": {
+      "content": "<tool_response>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151666": {
+      "content": "</tool_response>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151667": {
+      "content": "<think>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151668": {
+      "content": "</think>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    }
+  },
+  "additional_special_tokens": [
+    "<|im_start|>",
+    "<|im_end|>",
+    "<|object_ref_start|>",
+    "<|object_ref_end|>",
+    "<|box_start|>",
+    "<|box_end|>",
+    "<|quad_start|>",
+    "<|quad_end|>",
+    "<|vision_start|>",
+    "<|vision_end|>",
+    "<|vision_pad|>",
+    "<|image_pad|>",
+    "<|video_pad|>"
+  ],
+  "bos_token": null,
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "<|im_end|>",
+  "errors": "replace",
+  "extra_special_tokens": {},
+  "model_max_length": 131072,
+  "pad_token": "<|endoftext|>",
+  "padding_side": "right",
+  "split_special_tokens": false,
+  "tokenizer_class": "Qwen2Tokenizer",
+  "unk_token": null
+}

trainer_state.json

@@ -0,0 +1,3331 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 471,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0021231422505307855,
+      "grad_norm": 6.233692311689662,
+      "learning_rate": 0.0,
+      "loss": 1.3677,
+      "step": 1
+    },
+    {
+      "epoch": 0.004246284501061571,
+      "grad_norm": 6.036329344230565,
+      "learning_rate": 1.2499999999999999e-06,
+      "loss": 1.4092,
+      "step": 2
+    },
+    {
+      "epoch": 0.006369426751592357,
+      "grad_norm": 6.056066195237692,
+      "learning_rate": 2.4999999999999998e-06,
+      "loss": 1.3995,
+      "step": 3
+    },
+    {
+      "epoch": 0.008492569002123142,
+      "grad_norm": 5.922965787004136,
+      "learning_rate": 3.75e-06,
+      "loss": 1.3975,
+      "step": 4
+    },
+    {
+      "epoch": 0.010615711252653927,
+      "grad_norm": 5.286633700836792,
+      "learning_rate": 4.9999999999999996e-06,
+      "loss": 1.3654,
+      "step": 5
+    },
+    {
+      "epoch": 0.012738853503184714,
+      "grad_norm": 4.272128958690016,
+      "learning_rate": 6.25e-06,
+      "loss": 1.357,
+      "step": 6
+    },
+    {
+      "epoch": 0.014861995753715499,
+      "grad_norm": 3.3415834474393087,
+      "learning_rate": 7.5e-06,
+      "loss": 1.2136,
+      "step": 7
+    },
+    {
+      "epoch": 0.016985138004246284,
+      "grad_norm": 2.97877807094389,
+      "learning_rate": 8.750000000000001e-06,
+      "loss": 1.2796,
+      "step": 8
+    },
+    {
+      "epoch": 0.01910828025477707,
+      "grad_norm": 2.871670675310436,
+      "learning_rate": 9.999999999999999e-06,
+      "loss": 1.2221,
+      "step": 9
+    },
+    {
+      "epoch": 0.021231422505307854,
+      "grad_norm": 3.328688346998038,
+      "learning_rate": 1.125e-05,
+      "loss": 1.2234,
+      "step": 10
+    },
+    {
+      "epoch": 0.02335456475583864,
+      "grad_norm": 2.803564933108231,
+      "learning_rate": 1.25e-05,
+      "loss": 1.1813,
+      "step": 11
+    },
+    {
+      "epoch": 0.025477707006369428,
+      "grad_norm": 2.6965733342170646,
+      "learning_rate": 1.375e-05,
+      "loss": 1.1396,
+      "step": 12
+    },
+    {
+      "epoch": 0.027600849256900213,
+      "grad_norm": 3.096481749634196,
+      "learning_rate": 1.5e-05,
+      "loss": 1.2022,
+      "step": 13
+    },
+    {
+      "epoch": 0.029723991507430998,
+      "grad_norm": 2.274249270127275,
+      "learning_rate": 1.625e-05,
+      "loss": 1.1202,
+      "step": 14
+    },
+    {
+      "epoch": 0.03184713375796178,
+      "grad_norm": 3.884266388124913,
+      "learning_rate": 1.7500000000000002e-05,
+      "loss": 1.1371,
+      "step": 15
+    },
+    {
+      "epoch": 0.03397027600849257,
+      "grad_norm": 1.6993184946564164,
+      "learning_rate": 1.8750000000000002e-05,
+      "loss": 1.1044,
+      "step": 16
+    },
+    {
+      "epoch": 0.036093418259023353,
+      "grad_norm": 1.3282225362466542,
+      "learning_rate": 1.9999999999999998e-05,
+      "loss": 1.0696,
+      "step": 17
+    },
+    {
+      "epoch": 0.03821656050955414,
+      "grad_norm": 1.2132093388653153,
+      "learning_rate": 2.125e-05,
+      "loss": 1.0457,
+      "step": 18
+    },
+    {
+      "epoch": 0.040339702760084924,
+      "grad_norm": 1.2631509255285398,
+      "learning_rate": 2.25e-05,
+      "loss": 1.0374,
+      "step": 19
+    },
+    {
+      "epoch": 0.04246284501061571,
+      "grad_norm": 1.194394279683961,
+      "learning_rate": 2.3749999999999998e-05,
+      "loss": 0.9467,
+      "step": 20
+    },
+    {
+      "epoch": 0.044585987261146494,
+      "grad_norm": 1.098708611169601,
+      "learning_rate": 2.5e-05,
+      "loss": 1.0645,
+      "step": 21
+    },
+    {
+      "epoch": 0.04670912951167728,
+      "grad_norm": 1.0283648174128268,
+      "learning_rate": 2.625e-05,
+      "loss": 1.0394,
+      "step": 22
+    },
+    {
+      "epoch": 0.04883227176220807,
+      "grad_norm": 0.9732861680373749,
+      "learning_rate": 2.75e-05,
+      "loss": 0.9515,
+      "step": 23
+    },
+    {
+      "epoch": 0.050955414012738856,
+      "grad_norm": 1.017693167952402,
+      "learning_rate": 2.875e-05,
+      "loss": 0.9749,
+      "step": 24
+    },
+    {
+      "epoch": 0.05307855626326964,
+      "grad_norm": 1.1330379055126547,
+      "learning_rate": 3e-05,
+      "loss": 1.0073,
+      "step": 25
+    },
+    {
+      "epoch": 0.055201698513800426,
+      "grad_norm": 1.0745948466977544,
+      "learning_rate": 2.9999629537566803e-05,
+      "loss": 1.0051,
+      "step": 26
+    },
+    {
+      "epoch": 0.05732484076433121,
+      "grad_norm": 0.8795447467848742,
+      "learning_rate": 2.9998518168566207e-05,
+      "loss": 1.0527,
+      "step": 27
+    },
+    {
+      "epoch": 0.059447983014861996,
+      "grad_norm": 1.1167437998178333,
+      "learning_rate": 2.999666594789427e-05,
+      "loss": 1.0584,
+      "step": 28
+    },
+    {
+      "epoch": 0.06157112526539278,
+      "grad_norm": 1.0126995680355033,
+      "learning_rate": 2.999407296704142e-05,
+      "loss": 1.0291,
+      "step": 29
+    },
+    {
+      "epoch": 0.06369426751592357,
+      "grad_norm": 0.8834749944368594,
+      "learning_rate": 2.9990739354087918e-05,
+      "loss": 1.0342,
+      "step": 30
+    },
+    {
+      "epoch": 0.06581740976645435,
+      "grad_norm": 0.980242387134944,
+      "learning_rate": 2.9986665273697548e-05,
+      "loss": 1.0209,
+      "step": 31
+    },
+    {
+      "epoch": 0.06794055201698514,
+      "grad_norm": 0.8415998788153699,
+      "learning_rate": 2.9981850927109472e-05,
+      "loss": 1.0274,
+      "step": 32
+    },
+    {
+      "epoch": 0.07006369426751592,
+      "grad_norm": 0.9010169241333265,
+      "learning_rate": 2.9976296552128305e-05,
+      "loss": 0.9844,
+      "step": 33
+    },
+    {
+      "epoch": 0.07218683651804671,
+      "grad_norm": 0.8466313083198588,
+      "learning_rate": 2.9970002423112342e-05,
+      "loss": 0.9862,
+      "step": 34
+    },
+    {
+      "epoch": 0.07430997876857749,
+      "grad_norm": 3.0775818765498997,
+      "learning_rate": 2.996296885096003e-05,
+      "loss": 0.9921,
+      "step": 35
+    },
+    {
+      "epoch": 0.07643312101910828,
+      "grad_norm": 0.9668150868467069,
+      "learning_rate": 2.9955196183094604e-05,
+      "loss": 1.0312,
+      "step": 36
+    },
+    {
+      "epoch": 0.07855626326963906,
+      "grad_norm": 0.8491853067637553,
+      "learning_rate": 2.9946684803446928e-05,
+      "loss": 0.9704,
+      "step": 37
+    },
+    {
+      "epoch": 0.08067940552016985,
+      "grad_norm": 1.4205641764205312,
+      "learning_rate": 2.9937435132436517e-05,
+      "loss": 0.9659,
+      "step": 38
+    },
+    {
+      "epoch": 0.08280254777070063,
+      "grad_norm": 0.9441030271305156,
+      "learning_rate": 2.9927447626950795e-05,
+      "loss": 0.9507,
+      "step": 39
+    },
+    {
+      "epoch": 0.08492569002123142,
+      "grad_norm": 1.2675209983159783,
+      "learning_rate": 2.9916722780322504e-05,
+      "loss": 1.002,
+      "step": 40
+    },
+    {
+      "epoch": 0.0870488322717622,
+      "grad_norm": 1.6463034987153113,
+      "learning_rate": 2.9905261122305344e-05,
+      "loss": 0.9619,
+      "step": 41
+    },
+    {
+      "epoch": 0.08917197452229299,
+      "grad_norm": 0.9187757905390965,
+      "learning_rate": 2.9893063219047815e-05,
+      "loss": 0.9691,
+      "step": 42
+    },
+    {
+      "epoch": 0.09129511677282377,
+      "grad_norm": 0.9018029010217173,
+      "learning_rate": 2.988012967306524e-05,
+      "loss": 0.9658,
+      "step": 43
+    },
+    {
+      "epoch": 0.09341825902335456,
+      "grad_norm": 0.7942951837701693,
+      "learning_rate": 2.9866461123210005e-05,
+      "loss": 1.036,
+      "step": 44
+    },
+    {
+      "epoch": 0.09554140127388536,
+      "grad_norm": 0.7920942774535104,
+      "learning_rate": 2.985205824464001e-05,
+      "loss": 1.0362,
+      "step": 45
+    },
+    {
+      "epoch": 0.09766454352441614,
+      "grad_norm": 1.1538568110872267,
+      "learning_rate": 2.983692174878531e-05,
+      "loss": 1.0226,
+      "step": 46
+    },
+    {
+      "epoch": 0.09978768577494693,
+      "grad_norm": 0.8728667083119837,
+      "learning_rate": 2.9821052383312987e-05,
+      "loss": 0.9848,
+      "step": 47
+    },
+    {
+      "epoch": 0.10191082802547771,
+      "grad_norm": 0.9041523514230781,
+      "learning_rate": 2.980445093209021e-05,
+      "loss": 0.9875,
+      "step": 48
+    },
+    {
+      "epoch": 0.1040339702760085,
+      "grad_norm": 1.0596951962405752,
+      "learning_rate": 2.9787118215145502e-05,
+      "loss": 0.9708,
+      "step": 49
+    },
+    {
+      "epoch": 0.10615711252653928,
+      "grad_norm": 0.9427913699014273,
+      "learning_rate": 2.976905508862828e-05,
+      "loss": 1.009,
+      "step": 50
+    },
+    {
+      "epoch": 0.10828025477707007,
+      "grad_norm": 0.8275899758352401,
+      "learning_rate": 2.9750262444766502e-05,
+      "loss": 0.9985,
+      "step": 51
+    },
+    {
+      "epoch": 0.11040339702760085,
+      "grad_norm": 0.975852608473361,
+      "learning_rate": 2.9730741211822654e-05,
+      "loss": 1.0062,
+      "step": 52
+    },
+    {
+      "epoch": 0.11252653927813164,
+      "grad_norm": 0.8051301886044567,
+      "learning_rate": 2.9710492354047857e-05,
+      "loss": 0.9339,
+      "step": 53
+    },
+    {
+      "epoch": 0.11464968152866242,
+      "grad_norm": 0.8707635236666643,
+      "learning_rate": 2.968951687163426e-05,
+      "loss": 0.9955,
+      "step": 54
+    },
+    {
+      "epoch": 0.11677282377919321,
+      "grad_norm": 0.8254929493091868,
+      "learning_rate": 2.9667815800665637e-05,
+      "loss": 1.0298,
+      "step": 55
+    },
+    {
+      "epoch": 0.11889596602972399,
+      "grad_norm": 0.7562926195663171,
+      "learning_rate": 2.9645390213066193e-05,
+      "loss": 1.0611,
+      "step": 56
+    },
+    {
+      "epoch": 0.12101910828025478,
+      "grad_norm": 0.8226232019747459,
+      "learning_rate": 2.9622241216547623e-05,
+      "loss": 0.9991,
+      "step": 57
+    },
+    {
+      "epoch": 0.12314225053078556,
+      "grad_norm": 0.7982200631080218,
+      "learning_rate": 2.9598369954554405e-05,
+      "loss": 0.9991,
+      "step": 58
+    },
+    {
+      "epoch": 0.12526539278131635,
+      "grad_norm": 0.6963466420104463,
+      "learning_rate": 2.957377760620732e-05,
+      "loss": 0.9941,
+      "step": 59
+    },
+    {
+      "epoch": 0.12738853503184713,
+      "grad_norm": 0.7408513127997496,
+      "learning_rate": 2.9548465386245185e-05,
+      "loss": 0.9874,
+      "step": 60
+    },
+    {
+      "epoch": 0.12951167728237792,
+      "grad_norm": 0.8271991090411744,
+      "learning_rate": 2.952243454496488e-05,
+      "loss": 0.9139,
+      "step": 61
+    },
+    {
+      "epoch": 0.1316348195329087,
+      "grad_norm": 1.0499980155535253,
+      "learning_rate": 2.9495686368159592e-05,
+      "loss": 0.9768,
+      "step": 62
+    },
+    {
+      "epoch": 0.1337579617834395,
+      "grad_norm": 0.7840493370566994,
+      "learning_rate": 2.946822217705526e-05,
+      "loss": 0.999,
+      "step": 63
+    },
+    {
+      "epoch": 0.13588110403397027,
+      "grad_norm": 0.7806728112401357,
+      "learning_rate": 2.9440043328245366e-05,
+      "loss": 1.0215,
+      "step": 64
+    },
+    {
+      "epoch": 0.13800424628450106,
+      "grad_norm": 0.7631865352856676,
+      "learning_rate": 2.9411151213623894e-05,
+      "loss": 1.0313,
+      "step": 65
+    },
+    {
+      "epoch": 0.14012738853503184,
+      "grad_norm": 0.7125983390885618,
+      "learning_rate": 2.938154726031659e-05,
+      "loss": 0.9669,
+      "step": 66
+    },
+    {
+      "epoch": 0.14225053078556263,
+      "grad_norm": 0.7130914703890735,
+      "learning_rate": 2.9351232930610473e-05,
+      "loss": 0.9745,
+      "step": 67
+    },
+    {
+      "epoch": 0.14437367303609341,
+      "grad_norm": 0.7946605985222294,
+      "learning_rate": 2.932020972188157e-05,
+      "loss": 0.9708,
+      "step": 68
+    },
+    {
+      "epoch": 0.1464968152866242,
+      "grad_norm": 0.6811138210740744,
+      "learning_rate": 2.9288479166521014e-05,
+      "loss": 0.9494,
+      "step": 69
+    },
+    {
+      "epoch": 0.14861995753715498,
+      "grad_norm": 0.772893213577846,
+      "learning_rate": 2.9256042831859295e-05,
+      "loss": 1.0612,
+      "step": 70
+    },
+    {
+      "epoch": 0.15074309978768577,
+      "grad_norm": 0.7592068515404989,
+      "learning_rate": 2.9222902320088882e-05,
+      "loss": 1.0019,
+      "step": 71
+    },
+    {
+      "epoch": 0.15286624203821655,
+      "grad_norm": 0.7696389426407971,
+      "learning_rate": 2.9189059268185057e-05,
+      "loss": 1.0222,
+      "step": 72
+    },
+    {
+      "epoch": 0.15498938428874734,
+      "grad_norm": 0.7196168653178467,
+      "learning_rate": 2.9154515347825065e-05,
+      "loss": 0.9702,
+      "step": 73
+    },
+    {
+      "epoch": 0.15711252653927812,
+      "grad_norm": 0.8084046623013887,
+      "learning_rate": 2.9119272265305546e-05,
+      "loss": 0.9476,
+      "step": 74
+    },
+    {
+      "epoch": 0.1592356687898089,
+      "grad_norm": 0.7666084490813502,
+      "learning_rate": 2.9083331761458247e-05,
+      "loss": 0.9927,
+      "step": 75
+    },
+    {
+      "epoch": 0.1613588110403397,
+      "grad_norm": 0.8464390116357674,
+      "learning_rate": 2.904669561156404e-05,
+      "loss": 0.909,
+      "step": 76
+    },
+    {
+      "epoch": 0.16348195329087048,
+      "grad_norm": 0.6443641909940253,
+      "learning_rate": 2.9009365625265215e-05,
+      "loss": 0.9603,
+      "step": 77
+    },
+    {
+      "epoch": 0.16560509554140126,
+      "grad_norm": 0.7599832914924824,
+      "learning_rate": 2.8971343646476114e-05,
+      "loss": 0.9786,
+      "step": 78
+    },
+    {
+      "epoch": 0.16772823779193205,
+      "grad_norm": 0.806592225804777,
+      "learning_rate": 2.893263155329204e-05,
+      "loss": 0.9774,
+      "step": 79
+    },
+    {
+      "epoch": 0.16985138004246284,
+      "grad_norm": 0.7338345781204999,
+      "learning_rate": 2.8893231257896502e-05,
+      "loss": 1.076,
+      "step": 80
+    },
+    {
+      "epoch": 0.17197452229299362,
+      "grad_norm": 0.7256977111670229,
+      "learning_rate": 2.8853144706466725e-05,
+      "loss": 1.0003,
+      "step": 81
+    },
+    {
+      "epoch": 0.1740976645435244,
+      "grad_norm": 0.9486924335177603,
+      "learning_rate": 2.881237387907757e-05,
+      "loss": 1.061,
+      "step": 82
+    },
+    {
+      "epoch": 0.1762208067940552,
+      "grad_norm": 0.7245363724222085,
+      "learning_rate": 2.8770920789603687e-05,
+      "loss": 0.9772,
+      "step": 83
+    },
+    {
+      "epoch": 0.17834394904458598,
+      "grad_norm": 0.8462340519151804,
+      "learning_rate": 2.8728787485620068e-05,
+      "loss": 1.0097,
+      "step": 84
+    },
+    {
+      "epoch": 0.18046709129511676,
+      "grad_norm": 0.7351160195293913,
+      "learning_rate": 2.868597604830088e-05,
+      "loss": 0.99,
+      "step": 85
+    },
+    {
+      "epoch": 0.18259023354564755,
+      "grad_norm": 0.7346962642071149,
+      "learning_rate": 2.864248859231669e-05,
+      "loss": 1.0185,
+      "step": 86
+    },
+    {
+      "epoch": 0.18471337579617833,
+      "grad_norm": 0.7264358663319485,
+      "learning_rate": 2.859832726573e-05,
+      "loss": 1.018,
+      "step": 87
+    },
+    {
+      "epoch": 0.18683651804670912,
+      "grad_norm": 0.6502521161134233,
+      "learning_rate": 2.855349424988915e-05,
+      "loss": 0.9539,
+      "step": 88
+    },
+    {
+      "epoch": 0.18895966029723993,
+      "grad_norm": 0.7429051973625352,
+      "learning_rate": 2.8507991759320545e-05,
+      "loss": 0.932,
+      "step": 89
+    },
+    {
+      "epoch": 0.1910828025477707,
+      "grad_norm": 0.7337271359665005,
+      "learning_rate": 2.8461822041619312e-05,
+      "loss": 1.0024,
+      "step": 90
+    },
+    {
+      "epoch": 0.1932059447983015,
+      "grad_norm": 0.696436735871088,
+      "learning_rate": 2.841498737733824e-05,
+      "loss": 0.9241,
+      "step": 91
+    },
+    {
+      "epoch": 0.19532908704883228,
+      "grad_norm": 0.8367022783932109,
+      "learning_rate": 2.8367490079875154e-05,
+      "loss": 0.9643,
+      "step": 92
+    },
+    {
+      "epoch": 0.19745222929936307,
+      "grad_norm": 0.7508090587584874,
+      "learning_rate": 2.831933249535865e-05,
+      "loss": 0.9753,
+      "step": 93
+    },
+    {
+      "epoch": 0.19957537154989385,
+      "grad_norm": 0.7462348768378886,
+      "learning_rate": 2.827051700253217e-05,
+      "loss": 0.958,
+      "step": 94
+    },
+    {
+      "epoch": 0.20169851380042464,
+      "grad_norm": 0.7525272193363061,
+      "learning_rate": 2.8221046012636558e-05,
+      "loss": 1.0095,
+      "step": 95
+    },
+    {
+      "epoch": 0.20382165605095542,
+      "grad_norm": 0.6809918403435731,
+      "learning_rate": 2.817092196929091e-05,
+      "loss": 0.9593,
+      "step": 96
+    },
+    {
+      "epoch": 0.2059447983014862,
+      "grad_norm": 0.6932971348903458,
+      "learning_rate": 2.8120147348371912e-05,
+      "loss": 0.9581,
+      "step": 97
+    },
+    {
+      "epoch": 0.208067940552017,
+      "grad_norm": 0.7197887059084126,
+      "learning_rate": 2.8068724657891507e-05,
+      "loss": 0.9432,
+      "step": 98
+    },
+    {
+      "epoch": 0.21019108280254778,
+      "grad_norm": 0.7347184474062712,
+      "learning_rate": 2.801665643787303e-05,
+      "loss": 1.0137,
+      "step": 99
+    },
+    {
+      "epoch": 0.21231422505307856,
+      "grad_norm": 0.673328010924802,
+      "learning_rate": 2.7963945260225748e-05,
+      "loss": 1.0148,
+      "step": 100
+    },
+    {
+      "epoch": 0.21443736730360935,
+      "grad_norm": 0.7214998628963789,
+      "learning_rate": 2.7910593728617813e-05,
+      "loss": 1.0484,
+      "step": 101
+    },
+    {
+      "epoch": 0.21656050955414013,
+      "grad_norm": 0.6722704514414892,
+      "learning_rate": 2.7856604478347655e-05,
+      "loss": 0.9654,
+      "step": 102
+    },
+    {
+      "epoch": 0.21868365180467092,
+      "grad_norm": 0.6556980159522945,
+      "learning_rate": 2.7801980176213798e-05,
+      "loss": 0.9148,
+      "step": 103
+    },
+    {
+      "epoch": 0.2208067940552017,
+      "grad_norm": 0.7796844299678404,
+      "learning_rate": 2.7746723520383174e-05,
+      "loss": 0.9803,
+      "step": 104
+    },
+    {
+      "epoch": 0.2229299363057325,
+      "grad_norm": 0.69615473393741,
+      "learning_rate": 2.76908372402578e-05,
+      "loss": 0.9389,
+      "step": 105
+    },
+    {
+      "epoch": 0.22505307855626328,
+      "grad_norm": 0.665149808571595,
+      "learning_rate": 2.763432409633998e-05,
+      "loss": 0.9407,
+      "step": 106
+    },
+    {
+      "epoch": 0.22717622080679406,
+      "grad_norm": 0.6679067532269544,
+      "learning_rate": 2.7577186880095966e-05,
+      "loss": 0.9864,
+      "step": 107
+    },
+    {
+      "epoch": 0.22929936305732485,
+      "grad_norm": 0.717736772052187,
+      "learning_rate": 2.7519428413818034e-05,
+      "loss": 0.9568,
+      "step": 108
+    },
+    {
+      "epoch": 0.23142250530785563,
+      "grad_norm": 0.7365137124208776,
+      "learning_rate": 2.7461051550485116e-05,
+      "loss": 1.0027,
+      "step": 109
+    },
+    {
+      "epoch": 0.23354564755838642,
+      "grad_norm": 0.7060692871299123,
+      "learning_rate": 2.740205917362186e-05,
+      "loss": 1.0176,
+      "step": 110
+    },
+    {
+      "epoch": 0.2356687898089172,
+      "grad_norm": 0.6086155886816632,
+      "learning_rate": 2.7342454197156194e-05,
+      "loss": 0.9614,
+      "step": 111
+    },
+    {
+      "epoch": 0.23779193205944799,
+      "grad_norm": 0.6391029724554078,
+      "learning_rate": 2.728223956527539e-05,
+      "loss": 0.979,
+      "step": 112
+    },
+    {
+      "epoch": 0.23991507430997877,
+      "grad_norm": 0.7225858742160614,
+      "learning_rate": 2.722141825228066e-05,
+      "loss": 0.9522,
+      "step": 113
+    },
+    {
+      "epoch": 0.24203821656050956,
+      "grad_norm": 0.6685886845969267,
+      "learning_rate": 2.7159993262440228e-05,
+      "loss": 0.9574,
+      "step": 114
+    },
+    {
+      "epoch": 0.24416135881104034,
+      "grad_norm": 0.6648088183213471,
+      "learning_rate": 2.7097967629840906e-05,
+      "loss": 1.008,
+      "step": 115
+    },
+    {
+      "epoch": 0.24628450106157113,
+      "grad_norm": 0.7204962333279492,
+      "learning_rate": 2.703534441823827e-05,
+      "loss": 0.9785,
+      "step": 116
+    },
+    {
+      "epoch": 0.2484076433121019,
+      "grad_norm": 0.6804549324723432,
+      "learning_rate": 2.6972126720905293e-05,
+      "loss": 0.9816,
+      "step": 117
+    },
+    {
+      "epoch": 0.2505307855626327,
+      "grad_norm": 0.6837624060695675,
+      "learning_rate": 2.6908317660479583e-05,
+      "loss": 0.9813,
+      "step": 118
+    },
+    {
+      "epoch": 0.2526539278131635,
+      "grad_norm": 0.6706972759019354,
+      "learning_rate": 2.68439203888091e-05,
+      "loss": 1.0129,
+      "step": 119
+    },
+    {
+      "epoch": 0.25477707006369427,
+      "grad_norm": 0.7603607267519649,
+      "learning_rate": 2.6778938086796512e-05,
+      "loss": 1.0785,
+      "step": 120
+    },
+    {
+      "epoch": 0.25690021231422505,
+      "grad_norm": 0.701884945821394,
+      "learning_rate": 2.6713373964242043e-05,
+      "loss": 0.9629,
+      "step": 121
+    },
+    {
+      "epoch": 0.25902335456475584,
+      "grad_norm": 0.7096804682790832,
+      "learning_rate": 2.6647231259684946e-05,
+      "loss": 0.9646,
+      "step": 122
+    },
+    {
+      "epoch": 0.2611464968152866,
+      "grad_norm": 0.6445523205552592,
+      "learning_rate": 2.6580513240243524e-05,
+      "loss": 1.0163,
+      "step": 123
+    },
+    {
+      "epoch": 0.2632696390658174,
+      "grad_norm": 0.6863246456533727,
+      "learning_rate": 2.651322320145375e-05,
+      "loss": 0.9842,
+      "step": 124
+    },
+    {
+      "epoch": 0.2653927813163482,
+      "grad_norm": 0.691665965346161,
+      "learning_rate": 2.6445364467106494e-05,
+      "loss": 0.9421,
+      "step": 125
+    },
+    {
+      "epoch": 0.267515923566879,
+      "grad_norm": 0.7584893620552549,
+      "learning_rate": 2.637694038908333e-05,
+      "loss": 1.0348,
+      "step": 126
+    },
+    {
+      "epoch": 0.26963906581740976,
+      "grad_norm": 0.6613598336840398,
+      "learning_rate": 2.630795434719099e-05,
+      "loss": 0.9464,
+      "step": 127
+    },
+    {
+      "epoch": 0.27176220806794055,
+      "grad_norm": 0.6713699116927843,
+      "learning_rate": 2.623840974899439e-05,
+      "loss": 0.8904,
+      "step": 128
+    },
+    {
+      "epoch": 0.27388535031847133,
+      "grad_norm": 0.816037128777881,
+      "learning_rate": 2.616831002964834e-05,
+      "loss": 0.9743,
+      "step": 129
+    },
+    {
+      "epoch": 0.2760084925690021,
+      "grad_norm": 0.7338762748040034,
+      "learning_rate": 2.609765865172786e-05,
+      "loss": 0.9847,
+      "step": 130
+    },
+    {
+      "epoch": 0.2781316348195329,
+      "grad_norm": 0.6161573010385878,
+      "learning_rate": 2.6026459105057127e-05,
+      "loss": 0.9607,
+      "step": 131
+    },
+    {
+      "epoch": 0.2802547770700637,
+      "grad_norm": 0.6784919139990105,
+      "learning_rate": 2.5954714906537116e-05,
+      "loss": 0.9833,
+      "step": 132
+    },
+    {
+      "epoch": 0.2823779193205945,
+      "grad_norm": 0.7455868873904333,
+      "learning_rate": 2.5882429599971872e-05,
+      "loss": 0.9333,
+      "step": 133
+    },
+    {
+      "epoch": 0.28450106157112526,
+      "grad_norm": 0.6606022069654393,
+      "learning_rate": 2.580960675589347e-05,
+      "loss": 1.0049,
+      "step": 134
+    },
+    {
+      "epoch": 0.28662420382165604,
+      "grad_norm": 0.6870029740052253,
+      "learning_rate": 2.573624997138565e-05,
+      "loss": 1.0137,
+      "step": 135
+    },
+    {
+      "epoch": 0.28874734607218683,
+      "grad_norm": 0.6822500856445419,
+      "learning_rate": 2.5662362869906123e-05,
+      "loss": 1.0007,
+      "step": 136
+    },
+    {
+      "epoch": 0.2908704883227176,
+      "grad_norm": 0.6035427107718914,
+      "learning_rate": 2.558794910110761e-05,
+      "loss": 0.937,
+      "step": 137
+    },
+    {
+      "epoch": 0.2929936305732484,
+      "grad_norm": 0.6041202314277974,
+      "learning_rate": 2.5513012340657557e-05,
+      "loss": 0.9461,
+      "step": 138
+    },
+    {
+      "epoch": 0.2951167728237792,
+      "grad_norm": 0.6390470706425474,
+      "learning_rate": 2.5437556290056575e-05,
+      "loss": 0.9358,
+      "step": 139
+    },
+    {
+      "epoch": 0.29723991507430997,
+      "grad_norm": 0.7195254765489666,
+      "learning_rate": 2.5361584676455627e-05,
+      "loss": 0.9976,
+      "step": 140
+    },
+    {
+      "epoch": 0.29936305732484075,
+      "grad_norm": 0.6521311999355591,
+      "learning_rate": 2.5285101252471874e-05,
+      "loss": 0.9377,
+      "step": 141
+    },
+    {
+      "epoch": 0.30148619957537154,
+      "grad_norm": 0.6409749095351458,
+      "learning_rate": 2.5208109796003364e-05,
+      "loss": 1.0104,
+      "step": 142
+    },
+    {
+      "epoch": 0.3036093418259023,
+      "grad_norm": 0.6315495351185857,
+      "learning_rate": 2.5130614110042415e-05,
+      "loss": 1.0316,
+      "step": 143
+    },
+    {
+      "epoch": 0.3057324840764331,
+      "grad_norm": 0.6737333438485361,
+      "learning_rate": 2.5052618022487733e-05,
+      "loss": 0.9078,
+      "step": 144
+    },
+    {
+      "epoch": 0.3078556263269639,
+      "grad_norm": 0.6808606797582761,
+      "learning_rate": 2.4974125385955374e-05,
+      "loss": 0.9745,
+      "step": 145
+    },
+    {
+      "epoch": 0.3099787685774947,
+      "grad_norm": 0.6728298363507792,
+      "learning_rate": 2.4895140077588412e-05,
+      "loss": 0.9231,
+      "step": 146
+    },
+    {
+      "epoch": 0.31210191082802546,
+      "grad_norm": 0.737270504862064,
+      "learning_rate": 2.481566599886546e-05,
+      "loss": 0.9499,
+      "step": 147
+    },
+    {
+      "epoch": 0.31422505307855625,
+      "grad_norm": 0.6659360676868527,
+      "learning_rate": 2.473570707540793e-05,
+      "loss": 1.036,
+      "step": 148
+    },
+    {
+      "epoch": 0.31634819532908703,
+      "grad_norm": 0.6279327692414004,
+      "learning_rate": 2.4655267256786126e-05,
+      "loss": 1.0277,
+      "step": 149
+    },
+    {
+      "epoch": 0.3184713375796178,
+      "grad_norm": 0.6827718930790617,
+      "learning_rate": 2.4574350516324176e-05,
+      "loss": 0.9544,
+      "step": 150
+    },
+    {
+      "epoch": 0.3205944798301486,
+      "grad_norm": 0.6764712245297005,
+      "learning_rate": 2.4492960850903757e-05,
+      "loss": 0.9915,
+      "step": 151
+    },
+    {
+      "epoch": 0.3227176220806794,
+      "grad_norm": 0.6118211254231126,
+      "learning_rate": 2.4411102280766658e-05,
+      "loss": 0.9918,
+      "step": 152
+    },
+    {
+      "epoch": 0.3248407643312102,
+      "grad_norm": 0.6445432947438029,
+      "learning_rate": 2.432877884931623e-05,
+      "loss": 1.0199,
+      "step": 153
+    },
+    {
+      "epoch": 0.32696390658174096,
+      "grad_norm": 0.6555286244268101,
+      "learning_rate": 2.4245994622917636e-05,
+      "loss": 1.0144,
+      "step": 154
+    },
+    {
+      "epoch": 0.32908704883227174,
+      "grad_norm": 0.5747726408332037,
+      "learning_rate": 2.4162753690696998e-05,
+      "loss": 1.0059,
+      "step": 155
+    },
+    {
+      "epoch": 0.33121019108280253,
+      "grad_norm": 0.6633624122819982,
+      "learning_rate": 2.4079060164339427e-05,
+      "loss": 0.9892,
+      "step": 156
+    },
+    {
+      "epoch": 0.3333333333333333,
+      "grad_norm": 0.624099565922787,
+      "learning_rate": 2.3994918177885906e-05,
+      "loss": 1.0363,
+      "step": 157
+    },
+    {
+      "epoch": 0.3354564755838641,
+      "grad_norm": 0.5630337003726085,
+      "learning_rate": 2.391033188752911e-05,
+      "loss": 0.9432,
+      "step": 158
+    },
+    {
+      "epoch": 0.3375796178343949,
+      "grad_norm": 0.6572321757159671,
+      "learning_rate": 2.38253054714081e-05,
+      "loss": 1.0231,
+      "step": 159
+    },
+    {
+      "epoch": 0.33970276008492567,
+      "grad_norm": 0.6305040778279057,
+      "learning_rate": 2.3739843129401932e-05,
+      "loss": 0.9562,
+      "step": 160
+    },
+    {
+      "epoch": 0.34182590233545646,
+      "grad_norm": 0.6727626811450059,
+      "learning_rate": 2.365394908292224e-05,
+      "loss": 0.9246,
+      "step": 161
+    },
+    {
+      "epoch": 0.34394904458598724,
+      "grad_norm": 0.6145568336152458,
+      "learning_rate": 2.356762757470468e-05,
+      "loss": 0.965,
+      "step": 162
+    },
+    {
+      "epoch": 0.346072186836518,
+      "grad_norm": 0.6416356795188216,
+      "learning_rate": 2.3480882868599383e-05,
+      "loss": 0.9732,
+      "step": 163
+    },
+    {
+      "epoch": 0.3481953290870488,
+      "grad_norm": 0.6398547486961889,
+      "learning_rate": 2.3393719249360335e-05,
+      "loss": 1.0111,
+      "step": 164
+    },
+    {
+      "epoch": 0.3503184713375796,
+      "grad_norm": 0.6143238222438724,
+      "learning_rate": 2.3306141022433728e-05,
+      "loss": 1.0175,
+      "step": 165
+    },
+    {
+      "epoch": 0.3524416135881104,
+      "grad_norm": 0.6361075276958069,
+      "learning_rate": 2.3218152513745306e-05,
+      "loss": 0.9302,
+      "step": 166
+    },
+    {
+      "epoch": 0.35456475583864117,
+      "grad_norm": 0.6901788737924308,
+      "learning_rate": 2.3129758069486665e-05,
+      "loss": 0.9629,
+      "step": 167
+    },
+    {
+      "epoch": 0.35668789808917195,
+      "grad_norm": 0.5744122476891506,
+      "learning_rate": 2.3040962055900598e-05,
+      "loss": 1.0089,
+      "step": 168
+    },
+    {
+      "epoch": 0.35881104033970274,
+      "grad_norm": 0.6397560453387058,
+      "learning_rate": 2.2951768859065405e-05,
+      "loss": 0.9834,
+      "step": 169
+    },
+    {
+      "epoch": 0.3609341825902335,
+      "grad_norm": 0.6171130142316281,
+      "learning_rate": 2.2862182884678243e-05,
+      "loss": 0.9143,
+      "step": 170
+    },
+    {
+      "epoch": 0.3630573248407643,
+      "grad_norm": 0.6105813496358233,
+      "learning_rate": 2.2772208557837526e-05,
+      "loss": 0.953,
+      "step": 171
+    },
+    {
+      "epoch": 0.3651804670912951,
+      "grad_norm": 0.6094186393268805,
+      "learning_rate": 2.2681850322824324e-05,
+      "loss": 0.9423,
+      "step": 172
+    },
+    {
+      "epoch": 0.3673036093418259,
+      "grad_norm": 0.6168713424561586,
+      "learning_rate": 2.2591112642882853e-05,
+      "loss": 0.9377,
+      "step": 173
+    },
+    {
+      "epoch": 0.36942675159235666,
+      "grad_norm": 0.6038856999882359,
+      "learning_rate": 2.25e-05,
+      "loss": 0.9765,
+      "step": 174
+    },
+    {
+      "epoch": 0.37154989384288745,
+      "grad_norm": 0.6042658108338912,
+      "learning_rate": 2.2408516894683952e-05,
+      "loss": 0.9511,
+      "step": 175
+    },
+    {
+      "epoch": 0.37367303609341823,
+      "grad_norm": 0.6502768700665983,
+      "learning_rate": 2.2316667845741885e-05,
+      "loss": 0.989,
+      "step": 176
+    },
+    {
+      "epoch": 0.37579617834394907,
+      "grad_norm": 0.6428706495908844,
+      "learning_rate": 2.2224457390056747e-05,
+      "loss": 0.9461,
+      "step": 177
+    },
+    {
+      "epoch": 0.37791932059447986,
+      "grad_norm": 0.6077829967160175,
+      "learning_rate": 2.2131890082363176e-05,
+      "loss": 0.9845,
+      "step": 178
+    },
+    {
+      "epoch": 0.38004246284501064,
+      "grad_norm": 0.6332362664782106,
+      "learning_rate": 2.2038970495022515e-05,
+      "loss": 0.9243,
+      "step": 179
+    },
+    {
+      "epoch": 0.3821656050955414,
+      "grad_norm": 0.6871006568383716,
+      "learning_rate": 2.194570321779695e-05,
+      "loss": 1.0147,
+      "step": 180
+    },
+    {
+      "epoch": 0.3842887473460722,
+      "grad_norm": 0.6463380640823357,
+      "learning_rate": 2.185209285762281e-05,
+      "loss": 0.9812,
+      "step": 181
+    },
+    {
+      "epoch": 0.386411889596603,
+      "grad_norm": 0.5883115803971956,
+      "learning_rate": 2.1758144038383e-05,
+      "loss": 0.9982,
+      "step": 182
+    },
+    {
+      "epoch": 0.3885350318471338,
+      "grad_norm": 0.6375181055091449,
+      "learning_rate": 2.166386140067861e-05,
+      "loss": 0.9368,
+      "step": 183
+    },
+    {
+      "epoch": 0.39065817409766457,
+      "grad_norm": 0.6371811007750889,
+      "learning_rate": 2.1569249601599697e-05,
+      "loss": 0.9167,
+      "step": 184
+    },
+    {
+      "epoch": 0.39278131634819535,
+      "grad_norm": 0.6062348447482547,
+      "learning_rate": 2.147431331449522e-05,
+      "loss": 0.9875,
+      "step": 185
+    },
+    {
+      "epoch": 0.39490445859872614,
+      "grad_norm": 0.6671182861464395,
+      "learning_rate": 2.1379057228742246e-05,
+      "loss": 1.0217,
+      "step": 186
+    },
+    {
+      "epoch": 0.3970276008492569,
+      "grad_norm": 0.6028813716348284,
+      "learning_rate": 2.128348604951428e-05,
+      "loss": 1.0129,
+      "step": 187
+    },
+    {
+      "epoch": 0.3991507430997877,
+      "grad_norm": 0.5778363523300273,
+      "learning_rate": 2.1187604497548862e-05,
+      "loss": 1.0171,
+      "step": 188
+    },
+    {
+      "epoch": 0.4012738853503185,
+      "grad_norm": 0.6855476054674353,
+      "learning_rate": 2.1091417308914406e-05,
+      "loss": 0.9339,
+      "step": 189
+    },
+    {
+      "epoch": 0.4033970276008493,
+      "grad_norm": 0.55982548537731,
+      "learning_rate": 2.0994929234776232e-05,
+      "loss": 0.8762,
+      "step": 190
+    },
+    {
+      "epoch": 0.40552016985138006,
+      "grad_norm": 0.6258582771861254,
+      "learning_rate": 2.089814504116191e-05,
+      "loss": 0.9444,
+      "step": 191
+    },
+    {
+      "epoch": 0.40764331210191085,
+      "grad_norm": 0.6562202233796375,
+      "learning_rate": 2.0801069508725815e-05,
+      "loss": 0.9461,
+      "step": 192
+    },
+    {
+      "epoch": 0.40976645435244163,
+      "grad_norm": 0.579424184880784,
+      "learning_rate": 2.0703707432513006e-05,
+      "loss": 0.9793,
+      "step": 193
+    },
+    {
+      "epoch": 0.4118895966029724,
+      "grad_norm": 0.6330424612203424,
+      "learning_rate": 2.0606063621722374e-05,
+      "loss": 0.9634,
+      "step": 194
+    },
+    {
+      "epoch": 0.4140127388535032,
+      "grad_norm": 0.6390702566723381,
+      "learning_rate": 2.050814289946908e-05,
+      "loss": 0.9219,
+      "step": 195
+    },
+    {
+      "epoch": 0.416135881104034,
+      "grad_norm": 0.6275024889619232,
+      "learning_rate": 2.0409950102546334e-05,
+      "loss": 0.9155,
+      "step": 196
+    },
+    {
+      "epoch": 0.4182590233545648,
+      "grad_norm": 0.6029024758044792,
+      "learning_rate": 2.0311490081186458e-05,
+      "loss": 0.9711,
+      "step": 197
+    },
+    {
+      "epoch": 0.42038216560509556,
+      "grad_norm": 0.6725742996333739,
+      "learning_rate": 2.021276769882133e-05,
+      "loss": 1.0026,
+      "step": 198
+    },
+    {
+      "epoch": 0.42250530785562634,
+      "grad_norm": 0.6705704166609441,
+      "learning_rate": 2.0113787831842155e-05,
+      "loss": 0.9382,
+      "step": 199
+    },
+    {
+      "epoch": 0.42462845010615713,
+      "grad_norm": 0.6133264395984722,
+      "learning_rate": 2.0014555369358576e-05,
+      "loss": 0.9553,
+      "step": 200
+    },
+    {
+      "epoch": 0.4267515923566879,
+      "grad_norm": 0.6353000261637869,
+      "learning_rate": 1.9915075212957192e-05,
+      "loss": 0.9745,
+      "step": 201
+    },
+    {
+      "epoch": 0.4288747346072187,
+      "grad_norm": 0.6490705651396219,
+      "learning_rate": 1.9815352276459447e-05,
+      "loss": 0.9682,
+      "step": 202
+    },
+    {
+      "epoch": 0.4309978768577495,
+      "grad_norm": 0.7067729195244077,
+      "learning_rate": 1.9715391485678893e-05,
+      "loss": 0.9225,
+      "step": 203
+    },
+    {
+      "epoch": 0.43312101910828027,
+      "grad_norm": 0.6105521305677857,
+      "learning_rate": 1.961519777817791e-05,
+      "loss": 0.9571,
+      "step": 204
+    },
+    {
+      "epoch": 0.43524416135881105,
+      "grad_norm": 0.5986113568625525,
+      "learning_rate": 1.9514776103023783e-05,
+      "loss": 1.0375,
+      "step": 205
+    },
+    {
+      "epoch": 0.43736730360934184,
+      "grad_norm": 0.6239888084051822,
+      "learning_rate": 1.9414131420544258e-05,
+      "loss": 0.9442,
+      "step": 206
+    },
+    {
+      "epoch": 0.4394904458598726,
+      "grad_norm": 0.6018088924811278,
+      "learning_rate": 1.9313268702082536e-05,
+      "loss": 0.9484,
+      "step": 207
+    },
+    {
+      "epoch": 0.4416135881104034,
+      "grad_norm": 0.6351267682362137,
+      "learning_rate": 1.921219292975169e-05,
+      "loss": 0.9109,
+      "step": 208
+    },
+    {
+      "epoch": 0.4437367303609342,
+      "grad_norm": 0.5797747650598362,
+      "learning_rate": 1.9110909096188606e-05,
+      "loss": 0.9167,
+      "step": 209
+    },
+    {
+      "epoch": 0.445859872611465,
+      "grad_norm": 0.6326398300687381,
+      "learning_rate": 1.900942220430733e-05,
+      "loss": 0.9775,
+      "step": 210
+    },
+    {
+      "epoch": 0.44798301486199577,
+      "grad_norm": 0.5993446508985663,
+      "learning_rate": 1.890773726705198e-05,
+      "loss": 0.9524,
+      "step": 211
+    },
+    {
+      "epoch": 0.45010615711252655,
+      "grad_norm": 0.6109185912050061,
+      "learning_rate": 1.880585930714914e-05,
+      "loss": 0.9273,
+      "step": 212
+    },
+    {
+      "epoch": 0.45222929936305734,
+      "grad_norm": 0.6385822979906733,
+      "learning_rate": 1.8703793356859717e-05,
+      "loss": 0.9269,
+      "step": 213
+    },
+    {
+      "epoch": 0.4543524416135881,
+      "grad_norm": 0.607410319201568,
+      "learning_rate": 1.8601544457730426e-05,
+      "loss": 0.9085,
+      "step": 214
+    },
+    {
+      "epoch": 0.4564755838641189,
+      "grad_norm": 0.5992615563122552,
+      "learning_rate": 1.849911766034472e-05,
+      "loss": 0.9203,
+      "step": 215
+    },
+    {
+      "epoch": 0.4585987261146497,
+      "grad_norm": 0.6500875849339258,
+      "learning_rate": 1.8396518024073356e-05,
+      "loss": 0.968,
+      "step": 216
+    },
+    {
+      "epoch": 0.4607218683651805,
+      "grad_norm": 0.6219814557801068,
+      "learning_rate": 1.8293750616824443e-05,
+      "loss": 1.0103,
+      "step": 217
+    },
+    {
+      "epoch": 0.46284501061571126,
+      "grad_norm": 0.5719739307553183,
+      "learning_rate": 1.819082051479315e-05,
+      "loss": 0.9416,
+      "step": 218
+    },
+    {
+      "epoch": 0.46496815286624205,
+      "grad_norm": 0.5794827261117855,
+      "learning_rate": 1.8087732802210956e-05,
+      "loss": 0.9242,
+      "step": 219
+    },
+    {
+      "epoch": 0.46709129511677283,
+      "grad_norm": 0.5984472139940352,
+      "learning_rate": 1.7984492571094497e-05,
+      "loss": 0.9447,
+      "step": 220
+    },
+    {
+      "epoch": 0.4692144373673036,
+      "grad_norm": 0.616893517178056,
+      "learning_rate": 1.788110492099407e-05,
+      "loss": 1.0119,
+      "step": 221
+    },
+    {
+      "epoch": 0.4713375796178344,
+      "grad_norm": 0.5598209393737393,
+      "learning_rate": 1.777757495874174e-05,
+      "loss": 0.8852,
+      "step": 222
+    },
+    {
+      "epoch": 0.4734607218683652,
+      "grad_norm": 0.6397664463775209,
+      "learning_rate": 1.7673907798199054e-05,
+      "loss": 0.9221,
+      "step": 223
+    },
+    {
+      "epoch": 0.47558386411889597,
+      "grad_norm": 0.5813852755173831,
+      "learning_rate": 1.75701085600045e-05,
+      "loss": 0.9963,
+      "step": 224
+    },
+    {
+      "epoch": 0.47770700636942676,
+      "grad_norm": 0.665489852109793,
+      "learning_rate": 1.7466182371320518e-05,
+      "loss": 0.926,
+      "step": 225
+    },
+    {
+      "epoch": 0.47983014861995754,
+      "grad_norm": 0.5821224413128452,
+      "learning_rate": 1.7362134365580268e-05,
+      "loss": 0.922,
+      "step": 226
+    },
+    {
+      "epoch": 0.4819532908704883,
+      "grad_norm": 0.5356404406560891,
+      "learning_rate": 1.7257969682234073e-05,
+      "loss": 0.9589,
+      "step": 227
+    },
+    {
+      "epoch": 0.4840764331210191,
+      "grad_norm": 0.6108134895245437,
+      "learning_rate": 1.7153693466495538e-05,
+      "loss": 0.9597,
+      "step": 228
+    },
+    {
+      "epoch": 0.4861995753715499,
+      "grad_norm": 0.6137774258222183,
+      "learning_rate": 1.7049310869087402e-05,
+      "loss": 0.9799,
+      "step": 229
+    },
+    {
+      "epoch": 0.4883227176220807,
+      "grad_norm": 0.5726381735894963,
+      "learning_rate": 1.6944827045987148e-05,
+      "loss": 0.9488,
+      "step": 230
+    },
+    {
+      "epoch": 0.49044585987261147,
+      "grad_norm": 0.6215657856126448,
+      "learning_rate": 1.6840247158172277e-05,
+      "loss": 0.9082,
+      "step": 231
+    },
+    {
+      "epoch": 0.49256900212314225,
+      "grad_norm": 0.6328150266250187,
+      "learning_rate": 1.673557637136542e-05,
+      "loss": 1.0402,
+      "step": 232
+    },
+    {
+      "epoch": 0.49469214437367304,
+      "grad_norm": 0.6045652853253233,
+      "learning_rate": 1.663081985577916e-05,
+      "loss": 0.9112,
+      "step": 233
+    },
+    {
+      "epoch": 0.4968152866242038,
+      "grad_norm": 0.5785695616674981,
+      "learning_rate": 1.652598278586065e-05,
+      "loss": 0.8967,
+      "step": 234
+    },
+    {
+      "epoch": 0.4989384288747346,
+      "grad_norm": 0.6276478280684502,
+      "learning_rate": 1.6421070340036026e-05,
+      "loss": 0.9626,
+      "step": 235
+    },
+    {
+      "epoch": 0.5010615711252654,
+      "grad_norm": 0.5887038659784888,
+      "learning_rate": 1.631608770045461e-05,
+      "loss": 0.8641,
+      "step": 236
+    },
+    {
+      "epoch": 0.5031847133757962,
+      "grad_norm": 0.6927396901880972,
+      "learning_rate": 1.6211040052732958e-05,
+      "loss": 0.9537,
+      "step": 237
+    },
+    {
+      "epoch": 0.505307855626327,
+      "grad_norm": 0.5744893328350608,
+      "learning_rate": 1.6105932585698687e-05,
+      "loss": 0.9803,
+      "step": 238
+    },
+    {
+      "epoch": 0.5074309978768577,
+      "grad_norm": 0.5870783428723387,
+      "learning_rate": 1.6000770491134195e-05,
+      "loss": 0.9281,
+      "step": 239
+    },
+    {
+      "epoch": 0.5095541401273885,
+      "grad_norm": 0.5929932667698478,
+      "learning_rate": 1.5895558963520207e-05,
+      "loss": 1.0013,
+      "step": 240
+    },
+    {
+      "epoch": 0.5116772823779193,
+      "grad_norm": 0.5725128495696129,
+      "learning_rate": 1.5790303199779194e-05,
+      "loss": 0.9386,
+      "step": 241
+    },
+    {
+      "epoch": 0.5138004246284501,
+      "grad_norm": 0.6107066072742894,
+      "learning_rate": 1.568500839901867e-05,
+      "loss": 0.8998,
+      "step": 242
+    },
+    {
+      "epoch": 0.5159235668789809,
+      "grad_norm": 0.5913797577531015,
+      "learning_rate": 1.5579679762274377e-05,
+      "loss": 0.9059,
+      "step": 243
+    },
+    {
+      "epoch": 0.5180467091295117,
+      "grad_norm": 0.6181475328183207,
+      "learning_rate": 1.5474322492253382e-05,
+      "loss": 0.9877,
+      "step": 244
+    },
+    {
+      "epoch": 0.5201698513800425,
+      "grad_norm": 0.5833190722915838,
+      "learning_rate": 1.5368941793077115e-05,
+      "loss": 0.9947,
+      "step": 245
+    },
+    {
+      "epoch": 0.5222929936305732,
+      "grad_norm": 0.5771700964494837,
+      "learning_rate": 1.5263542870024263e-05,
+      "loss": 0.9787,
+      "step": 246
+    },
+    {
+      "epoch": 0.524416135881104,
+      "grad_norm": 0.6069660458860243,
+      "learning_rate": 1.5158130929273695e-05,
+      "loss": 0.9581,
+      "step": 247
+    },
+    {
+      "epoch": 0.5265392781316348,
+      "grad_norm": 0.5709532636436643,
+      "learning_rate": 1.505271117764728e-05,
+      "loss": 0.9278,
+      "step": 248
+    },
+    {
+      "epoch": 0.5286624203821656,
+      "grad_norm": 0.6433938701377838,
+      "learning_rate": 1.494728882235272e-05,
+      "loss": 0.9655,
+      "step": 249
+    },
+    {
+      "epoch": 0.5307855626326964,
+      "grad_norm": 0.5822393752360885,
+      "learning_rate": 1.4841869070726307e-05,
+      "loss": 0.9787,
+      "step": 250
+    },
+    {
+      "epoch": 0.5329087048832272,
+      "grad_norm": 0.6267023147379772,
+      "learning_rate": 1.473645712997574e-05,
+      "loss": 0.9342,
+      "step": 251
+    },
+    {
+      "epoch": 0.535031847133758,
+      "grad_norm": 0.6375126071613487,
+      "learning_rate": 1.4631058206922886e-05,
+      "loss": 0.9348,
+      "step": 252
+    },
+    {
+      "epoch": 0.5371549893842887,
+      "grad_norm": 0.6127085617357599,
+      "learning_rate": 1.4525677507746615e-05,
+      "loss": 0.9611,
+      "step": 253
+    },
+    {
+      "epoch": 0.5392781316348195,
+      "grad_norm": 0.5751592999711028,
+      "learning_rate": 1.4420320237725632e-05,
+      "loss": 1.0019,
+      "step": 254
+    },
+    {
+      "epoch": 0.5414012738853503,
+      "grad_norm": 0.6048766878698145,
+      "learning_rate": 1.4314991600981336e-05,
+      "loss": 0.9246,
+      "step": 255
+    },
+    {
+      "epoch": 0.5435244161358811,
+      "grad_norm": 0.585394770621169,
+      "learning_rate": 1.4209696800220807e-05,
+      "loss": 0.9185,
+      "step": 256
+    },
+    {
+      "epoch": 0.5456475583864119,
+      "grad_norm": 0.6401236988459983,
+      "learning_rate": 1.4104441036479797e-05,
+      "loss": 0.9277,
+      "step": 257
+    },
+    {
+      "epoch": 0.5477707006369427,
+      "grad_norm": 0.5987058316881022,
+      "learning_rate": 1.399922950886581e-05,
+      "loss": 0.9692,
+      "step": 258
+    },
+    {
+      "epoch": 0.5498938428874734,
+      "grad_norm": 0.5572446710237258,
+      "learning_rate": 1.3894067414301315e-05,
+      "loss": 0.9375,
+      "step": 259
+    },
+    {
+      "epoch": 0.5520169851380042,
+      "grad_norm": 0.5955717580516604,
+      "learning_rate": 1.3788959947267042e-05,
+      "loss": 0.9098,
+      "step": 260
+    },
+    {
+      "epoch": 0.554140127388535,
+      "grad_norm": 0.6834272471980338,
+      "learning_rate": 1.3683912299545391e-05,
+      "loss": 0.9523,
+      "step": 261
+    },
+    {
+      "epoch": 0.5562632696390658,
+      "grad_norm": 0.6256533610035157,
+      "learning_rate": 1.3578929659963977e-05,
+      "loss": 0.9969,
+      "step": 262
+    },
+    {
+      "epoch": 0.5583864118895966,
+      "grad_norm": 0.5436695647032473,
+      "learning_rate": 1.347401721413935e-05,
+      "loss": 0.8784,
+      "step": 263
+    },
+    {
+      "epoch": 0.5605095541401274,
+      "grad_norm": 0.5704479762497785,
+      "learning_rate": 1.336918014422084e-05,
+      "loss": 0.9338,
+      "step": 264
+    },
+    {
+      "epoch": 0.5626326963906582,
+      "grad_norm": 0.6086202857937474,
+      "learning_rate": 1.3264423628634583e-05,
+      "loss": 0.9847,
+      "step": 265
+    },
+    {
+      "epoch": 0.564755838641189,
+      "grad_norm": 0.6354163585809983,
+      "learning_rate": 1.3159752841827724e-05,
+      "loss": 0.9815,
+      "step": 266
+    },
+    {
+      "epoch": 0.5668789808917197,
+      "grad_norm": 0.5785366526905389,
+      "learning_rate": 1.3055172954012856e-05,
+      "loss": 0.9393,
+      "step": 267
+    },
+    {
+      "epoch": 0.5690021231422505,
+      "grad_norm": 0.5777363950316821,
+      "learning_rate": 1.2950689130912599e-05,
+      "loss": 1.0318,
+      "step": 268
+    },
+    {
+      "epoch": 0.5711252653927813,
+      "grad_norm": 0.5400938772826588,
+      "learning_rate": 1.2846306533504465e-05,
+      "loss": 0.9302,
+      "step": 269
+    },
+    {
+      "epoch": 0.5732484076433121,
+      "grad_norm": 0.5957886887990331,
+      "learning_rate": 1.2742030317765933e-05,
+      "loss": 0.947,
+      "step": 270
+    },
+    {
+      "epoch": 0.5753715498938429,
+      "grad_norm": 0.5685633933725828,
+      "learning_rate": 1.2637865634419735e-05,
+      "loss": 0.934,
+      "step": 271
+    },
+    {
+      "epoch": 0.5774946921443737,
+      "grad_norm": 0.6038992480186266,
+      "learning_rate": 1.2533817628679485e-05,
+      "loss": 0.9871,
+      "step": 272
+    },
+    {
+      "epoch": 0.5796178343949044,
+      "grad_norm": 0.5531455689022556,
+      "learning_rate": 1.24298914399955e-05,
+      "loss": 0.909,
+      "step": 273
+    },
+    {
+      "epoch": 0.5817409766454352,
+      "grad_norm": 0.5562580177389622,
+      "learning_rate": 1.2326092201800948e-05,
+      "loss": 0.9864,
+      "step": 274
+    },
+    {
+      "epoch": 0.583864118895966,
+      "grad_norm": 0.5650599912328349,
+      "learning_rate": 1.2222425041258267e-05,
+      "loss": 0.9932,
+      "step": 275
+    },
+    {
+      "epoch": 0.5859872611464968,
+      "grad_norm": 0.6128698371500075,
+      "learning_rate": 1.2118895079005929e-05,
+      "loss": 0.9663,
+      "step": 276
+    },
+    {
+      "epoch": 0.5881104033970276,
+      "grad_norm": 0.6402876844624192,
+      "learning_rate": 1.2015507428905509e-05,
+      "loss": 0.9367,
+      "step": 277
+    },
+    {
+      "epoch": 0.5902335456475584,
+      "grad_norm": 0.5802127210062712,
+      "learning_rate": 1.1912267197789047e-05,
+      "loss": 0.9369,
+      "step": 278
+    },
+    {
+      "epoch": 0.5923566878980892,
+      "grad_norm": 0.6212326138735739,
+      "learning_rate": 1.1809179485206847e-05,
+      "loss": 0.9742,
+      "step": 279
+    },
+    {
+      "epoch": 0.5944798301486199,
+      "grad_norm": 0.548301604024312,
+      "learning_rate": 1.1706249383175558e-05,
+      "loss": 0.9866,
+      "step": 280
+    },
+    {
+      "epoch": 0.5966029723991507,
+      "grad_norm": 0.6101785900734475,
+      "learning_rate": 1.1603481975926643e-05,
+      "loss": 0.9784,
+      "step": 281
+    },
+    {
+      "epoch": 0.5987261146496815,
+      "grad_norm": 0.6052900059845123,
+      "learning_rate": 1.1500882339655278e-05,
+      "loss": 0.9075,
+      "step": 282
+    },
+    {
+      "epoch": 0.6008492569002123,
+      "grad_norm": 0.6185657143466954,
+      "learning_rate": 1.1398455542269578e-05,
+      "loss": 0.9706,
+      "step": 283
+    },
+    {
+      "epoch": 0.6029723991507431,
+      "grad_norm": 0.5751738279124855,
+      "learning_rate": 1.1296206643140284e-05,
+      "loss": 0.9598,
+      "step": 284
+    },
+    {
+      "epoch": 0.6050955414012739,
+      "grad_norm": 0.6416444090869639,
+      "learning_rate": 1.1194140692850863e-05,
+      "loss": 0.9788,
+      "step": 285
+    },
+    {
+      "epoch": 0.6072186836518046,
+      "grad_norm": 0.5714386078956842,
+      "learning_rate": 1.1092262732948017e-05,
+      "loss": 0.919,
+      "step": 286
+    },
+    {
+      "epoch": 0.6093418259023354,
+      "grad_norm": 0.545987242527638,
+      "learning_rate": 1.0990577795692676e-05,
+      "loss": 0.9136,
+      "step": 287
+    },
+    {
+      "epoch": 0.6114649681528662,
+      "grad_norm": 0.5656751198434353,
+      "learning_rate": 1.0889090903811397e-05,
+      "loss": 0.9587,
+      "step": 288
+    },
+    {
+      "epoch": 0.613588110403397,
+      "grad_norm": 0.5636926547779225,
+      "learning_rate": 1.0787807070248306e-05,
+      "loss": 0.8874,
+      "step": 289
+    },
+    {
+      "epoch": 0.6157112526539278,
+      "grad_norm": 0.5700481388735775,
+      "learning_rate": 1.068673129791747e-05,
+      "loss": 0.9478,
+      "step": 290
+    },
+    {
+      "epoch": 0.6178343949044586,
+      "grad_norm": 0.5614742249715488,
+      "learning_rate": 1.0585868579455745e-05,
+      "loss": 0.955,
+      "step": 291
+    },
+    {
+      "epoch": 0.6199575371549894,
+      "grad_norm": 0.581521611639187,
+      "learning_rate": 1.0485223896976221e-05,
+      "loss": 0.9676,
+      "step": 292
+    },
+    {
+      "epoch": 0.6220806794055201,
+      "grad_norm": 0.5903813523116045,
+      "learning_rate": 1.0384802221822098e-05,
+      "loss": 0.9815,
+      "step": 293
+    },
+    {
+      "epoch": 0.6242038216560509,
+      "grad_norm": 0.5837585878828937,
+      "learning_rate": 1.028460851432111e-05,
+      "loss": 0.8784,
+      "step": 294
+    },
+    {
+      "epoch": 0.6263269639065817,
+      "grad_norm": 0.5568116286168524,
+      "learning_rate": 1.0184647723540557e-05,
+      "loss": 0.9364,
+      "step": 295
+    },
+    {
+      "epoch": 0.6284501061571125,
+      "grad_norm": 0.5990779103796753,
+      "learning_rate": 1.0084924787042809e-05,
+      "loss": 0.9763,
+      "step": 296
+    },
+    {
+      "epoch": 0.6305732484076433,
+      "grad_norm": 0.5866020091761237,
+      "learning_rate": 9.985444630641426e-06,
+      "loss": 0.9582,
+      "step": 297
+    },
+    {
+      "epoch": 0.6326963906581741,
+      "grad_norm": 0.5323296010758973,
+      "learning_rate": 9.886212168157848e-06,
+      "loss": 0.9254,
+      "step": 298
+    },
+    {
+      "epoch": 0.6348195329087049,
+      "grad_norm": 0.5522784198698178,
+      "learning_rate": 9.787232301178669e-06,
+      "loss": 0.9657,
+      "step": 299
+    },
+    {
+      "epoch": 0.6369426751592356,
+      "grad_norm": 0.575212640738695,
+      "learning_rate": 9.688509918813547e-06,
+      "loss": 0.9051,
+      "step": 300
+    },
+    {
+      "epoch": 0.6390658174097664,
+      "grad_norm": 0.5674507026773249,
+      "learning_rate": 9.590049897453668e-06,
+      "loss": 0.984,
+      "step": 301
+    },
+    {
+      "epoch": 0.6411889596602972,
+      "grad_norm": 0.55515175444595,
+      "learning_rate": 9.491857100530919e-06,
+      "loss": 0.9484,
+      "step": 302
+    },
+    {
+      "epoch": 0.643312101910828,
+      "grad_norm": 0.5292179686500557,
+      "learning_rate": 9.393936378277632e-06,
+      "loss": 0.9365,
+      "step": 303
+    },
+    {
+      "epoch": 0.6454352441613588,
+      "grad_norm": 0.5949560633293998,
+      "learning_rate": 9.296292567486997e-06,
+      "loss": 0.9013,
+      "step": 304
+    },
+    {
+      "epoch": 0.6475583864118896,
+      "grad_norm": 0.5640508209822152,
+      "learning_rate": 9.198930491274188e-06,
+      "loss": 0.9224,
+      "step": 305
+    },
+    {
+      "epoch": 0.6496815286624203,
+      "grad_norm": 0.564377192493895,
+      "learning_rate": 9.10185495883809e-06,
+      "loss": 0.9493,
+      "step": 306
+    },
+    {
+      "epoch": 0.6518046709129511,
+      "grad_norm": 0.5436911185747707,
+      "learning_rate": 9.005070765223768e-06,
+      "loss": 0.9504,
+      "step": 307
+    },
+    {
+      "epoch": 0.6539278131634819,
+      "grad_norm": 0.5364497163093522,
+      "learning_rate": 8.908582691085593e-06,
+      "loss": 0.9543,
+      "step": 308
+    },
+    {
+      "epoch": 0.6560509554140127,
+      "grad_norm": 0.6063724960512262,
+      "learning_rate": 8.812395502451139e-06,
+      "loss": 1.0194,
+      "step": 309
+    },
+    {
+      "epoch": 0.6581740976645435,
+      "grad_norm": 0.572682762237652,
+      "learning_rate": 8.716513950485725e-06,
+      "loss": 0.9761,
+      "step": 310
+    },
+    {
+      "epoch": 0.6602972399150743,
+      "grad_norm": 0.6237775922204165,
+      "learning_rate": 8.620942771257755e-06,
+      "loss": 0.9262,
+      "step": 311
+    },
+    {
+      "epoch": 0.6624203821656051,
+      "grad_norm": 0.5528221980335309,
+      "learning_rate": 8.525686685504781e-06,
+      "loss": 0.8953,
+      "step": 312
+    },
+    {
+      "epoch": 0.6645435244161358,
+      "grad_norm": 0.5889044724013025,
+      "learning_rate": 8.430750398400309e-06,
+      "loss": 0.943,
+      "step": 313
+    },
+    {
+      "epoch": 0.6666666666666666,
+      "grad_norm": 0.5125143671796547,
+      "learning_rate": 8.336138599321391e-06,
+      "loss": 0.9601,
+      "step": 314
+    },
+    {
+      "epoch": 0.6687898089171974,
+      "grad_norm": 0.5476667961262259,
+      "learning_rate": 8.241855961617e-06,
+      "loss": 0.901,
+      "step": 315
+    },
+    {
+      "epoch": 0.6709129511677282,
+      "grad_norm": 0.49845108048267994,
+      "learning_rate": 8.147907142377198e-06,
+      "loss": 0.9787,
+      "step": 316
+    },
+    {
+      "epoch": 0.673036093418259,
+      "grad_norm": 0.5404151085945639,
+      "learning_rate": 8.054296782203052e-06,
+      "loss": 0.9426,
+      "step": 317
+    },
+    {
+      "epoch": 0.6751592356687898,
+      "grad_norm": 0.5347124147295028,
+      "learning_rate": 7.961029504977487e-06,
+      "loss": 0.9228,
+      "step": 318
+    },
+    {
+      "epoch": 0.6772823779193206,
+      "grad_norm": 0.5509814795088901,
+      "learning_rate": 7.868109917636823e-06,
+      "loss": 0.9289,
+      "step": 319
+    },
+    {
+      "epoch": 0.6794055201698513,
+      "grad_norm": 0.5316099445221201,
+      "learning_rate": 7.77554260994326e-06,
+      "loss": 0.8699,
+      "step": 320
+    },
+    {
+      "epoch": 0.6815286624203821,
+      "grad_norm": 0.49836894602961657,
+      "learning_rate": 7.683332154258117e-06,
+      "loss": 0.9747,
+      "step": 321
+    },
+    {
+      "epoch": 0.6836518046709129,
+      "grad_norm": 0.5546979465633,
+      "learning_rate": 7.591483105316048e-06,
+      "loss": 0.8788,
+      "step": 322
+    },
+    {
+      "epoch": 0.6857749469214437,
+      "grad_norm": 0.6054955853270848,
+      "learning_rate": 7.500000000000004e-06,
+      "loss": 0.9601,
+      "step": 323
+    },
+    {
+      "epoch": 0.6878980891719745,
+      "grad_norm": 0.5434420921617582,
+      "learning_rate": 7.408887357117147e-06,
+      "loss": 0.9726,
+      "step": 324
+    },
+    {
+      "epoch": 0.6900212314225053,
+      "grad_norm": 0.5502097893793918,
+      "learning_rate": 7.3181496771756754e-06,
+      "loss": 0.966,
+      "step": 325
+    },
+    {
+      "epoch": 0.692144373673036,
+      "grad_norm": 0.5336428055938336,
+      "learning_rate": 7.227791442162475e-06,
+      "loss": 0.9168,
+      "step": 326
+    },
+    {
+      "epoch": 0.6942675159235668,
+      "grad_norm": 0.5567158996740152,
+      "learning_rate": 7.137817115321759e-06,
+      "loss": 0.9536,
+      "step": 327
+    },
+    {
+      "epoch": 0.6963906581740976,
+      "grad_norm": 0.5246357810088409,
+      "learning_rate": 7.048231140934595e-06,
+      "loss": 0.8912,
+      "step": 328
+    },
+    {
+      "epoch": 0.6985138004246284,
+      "grad_norm": 0.564409375685434,
+      "learning_rate": 6.959037944099402e-06,
+      "loss": 0.9119,
+      "step": 329
+    },
+    {
+      "epoch": 0.7006369426751592,
+      "grad_norm": 0.5533102306545692,
+      "learning_rate": 6.870241930513338e-06,
+      "loss": 0.9712,
+      "step": 330
+    },
+    {
+      "epoch": 0.70276008492569,
+      "grad_norm": 0.539122288480235,
+      "learning_rate": 6.781847486254698e-06,
+      "loss": 0.9227,
+      "step": 331
+    },
+    {
+      "epoch": 0.7048832271762208,
+      "grad_norm": 0.5382383698751078,
+      "learning_rate": 6.6938589775662705e-06,
+      "loss": 0.9448,
+      "step": 332
+    },
+    {
+      "epoch": 0.7070063694267515,
+      "grad_norm": 0.5368990830270315,
+      "learning_rate": 6.6062807506396696e-06,
+      "loss": 0.9403,
+      "step": 333
+    },
+    {
+      "epoch": 0.7091295116772823,
+      "grad_norm": 0.520929253144545,
+      "learning_rate": 6.519117131400617e-06,
+      "loss": 0.932,
+      "step": 334
+    },
+    {
+      "epoch": 0.7112526539278131,
+      "grad_norm": 0.5251850810798755,
+      "learning_rate": 6.432372425295321e-06,
+      "loss": 0.9192,
+      "step": 335
+    },
+    {
+      "epoch": 0.7133757961783439,
+      "grad_norm": 0.5421931278275105,
+      "learning_rate": 6.346050917077762e-06,
+      "loss": 0.9541,
+      "step": 336
+    },
+    {
+      "epoch": 0.7154989384288747,
+      "grad_norm": 0.5229472361556049,
+      "learning_rate": 6.260156870598072e-06,
+      "loss": 0.9779,
+      "step": 337
+    },
+    {
+      "epoch": 0.7176220806794055,
+      "grad_norm": 0.5400736058203549,
+      "learning_rate": 6.174694528591902e-06,
+      "loss": 0.8923,
+      "step": 338
+    },
+    {
+      "epoch": 0.7197452229299363,
+      "grad_norm": 0.5085373634938479,
+      "learning_rate": 6.089668112470892e-06,
+      "loss": 0.9448,
+      "step": 339
+    },
+    {
+      "epoch": 0.721868365180467,
+      "grad_norm": 0.5298949149149584,
+      "learning_rate": 6.005081822114096e-06,
+      "loss": 0.9276,
+      "step": 340
+    },
+    {
+      "epoch": 0.7239915074309978,
+      "grad_norm": 0.5432948704397309,
+      "learning_rate": 5.920939835660572e-06,
+      "loss": 0.959,
+      "step": 341
+    },
+    {
+      "epoch": 0.7261146496815286,
+      "grad_norm": 0.507224891568453,
+      "learning_rate": 5.8372463093030006e-06,
+      "loss": 0.9318,
+      "step": 342
+    },
+    {
+      "epoch": 0.7282377919320594,
+      "grad_norm": 0.5063069044994013,
+      "learning_rate": 5.754005377082365e-06,
+      "loss": 0.9895,
+      "step": 343
+    },
+    {
+      "epoch": 0.7303609341825902,
+      "grad_norm": 0.5008230814445663,
+      "learning_rate": 5.671221150683772e-06,
+      "loss": 0.9552,
+      "step": 344
+    },
+    {
+      "epoch": 0.732484076433121,
+      "grad_norm": 0.5236977533272006,
+      "learning_rate": 5.588897719233341e-06,
+      "loss": 0.8969,
+      "step": 345
+    },
+    {
+      "epoch": 0.7346072186836518,
+      "grad_norm": 0.5453574252356792,
+      "learning_rate": 5.507039149096251e-06,
+      "loss": 0.9645,
+      "step": 346
+    },
+    {
+      "epoch": 0.7367303609341825,
+      "grad_norm": 0.5327242182882551,
+      "learning_rate": 5.425649483675827e-06,
+      "loss": 1.0094,
+      "step": 347
+    },
+    {
+      "epoch": 0.7388535031847133,
+      "grad_norm": 0.5555867838219068,
+      "learning_rate": 5.3447327432138755e-06,
+      "loss": 0.9228,
+      "step": 348
+    },
+    {
+      "epoch": 0.7409766454352441,
+      "grad_norm": 0.5412007429751795,
+      "learning_rate": 5.264292924592074e-06,
+      "loss": 0.938,
+      "step": 349
+    },
+    {
+      "epoch": 0.7430997876857749,
+      "grad_norm": 0.5532186476228068,
+      "learning_rate": 5.184334001134542e-06,
+      "loss": 0.8919,
+      "step": 350
+    },
+    {
+      "epoch": 0.7452229299363057,
+      "grad_norm": 0.5570955284278991,
+      "learning_rate": 5.104859922411587e-06,
+      "loss": 0.9117,
+      "step": 351
+    },
+    {
+      "epoch": 0.7473460721868365,
+      "grad_norm": 0.5382963451516724,
+      "learning_rate": 5.025874614044631e-06,
+      "loss": 0.9468,
+      "step": 352
+    },
+    {
+      "epoch": 0.7494692144373672,
+      "grad_norm": 0.5149195845272291,
+      "learning_rate": 4.9473819775122716e-06,
+      "loss": 0.9421,
+      "step": 353
+    },
+    {
+      "epoch": 0.7515923566878981,
+      "grad_norm": 0.49481641390186526,
+      "learning_rate": 4.86938588995759e-06,
+      "loss": 1.0069,
+      "step": 354
+    },
+    {
+      "epoch": 0.7537154989384289,
+      "grad_norm": 0.5531251592147758,
+      "learning_rate": 4.791890203996634e-06,
+      "loss": 0.8964,
+      "step": 355
+    },
+    {
+      "epoch": 0.7558386411889597,
+      "grad_norm": 0.5334093734684512,
+      "learning_rate": 4.714898747528129e-06,
+      "loss": 0.9546,
+      "step": 356
+    },
+    {
+      "epoch": 0.7579617834394905,
+      "grad_norm": 0.5251264229643601,
+      "learning_rate": 4.638415323544376e-06,
+      "loss": 0.9159,
+      "step": 357
+    },
+    {
+      "epoch": 0.7600849256900213,
+      "grad_norm": 0.5276709219009906,
+      "learning_rate": 4.562443709943424e-06,
+      "loss": 0.9533,
+      "step": 358
+    },
+    {
+      "epoch": 0.7622080679405521,
+      "grad_norm": 0.5538080314575125,
+      "learning_rate": 4.4869876593424484e-06,
+      "loss": 0.9681,
+      "step": 359
+    },
+    {
+      "epoch": 0.7643312101910829,
+      "grad_norm": 0.5273232151286709,
+      "learning_rate": 4.412050898892393e-06,
+      "loss": 0.9337,
+      "step": 360
+    },
+    {
+      "epoch": 0.7664543524416136,
+      "grad_norm": 0.5372645133897195,
+      "learning_rate": 4.337637130093879e-06,
+      "loss": 0.8776,
+      "step": 361
+    },
+    {
+      "epoch": 0.7685774946921444,
+      "grad_norm": 0.5373725094136548,
+      "learning_rate": 4.263750028614352e-06,
+      "loss": 0.9626,
+      "step": 362
+    },
+    {
+      "epoch": 0.7707006369426752,
+      "grad_norm": 0.561051915231661,
+      "learning_rate": 4.1903932441065315e-06,
+      "loss": 0.8999,
+      "step": 363
+    },
+    {
+      "epoch": 0.772823779193206,
+      "grad_norm": 0.5629541189967254,
+      "learning_rate": 4.117570400028129e-06,
+      "loss": 0.9654,
+      "step": 364
+    },
+    {
+      "epoch": 0.7749469214437368,
+      "grad_norm": 0.5304742998126838,
+      "learning_rate": 4.045285093462887e-06,
+      "loss": 0.9513,
+      "step": 365
+    },
+    {
+      "epoch": 0.7770700636942676,
+      "grad_norm": 0.5490072502884541,
+      "learning_rate": 3.973540894942876e-06,
+      "loss": 0.8785,
+      "step": 366
+    },
+    {
+      "epoch": 0.7791932059447984,
+      "grad_norm": 0.5405499109981703,
+      "learning_rate": 3.902341348272144e-06,
+      "loss": 0.9486,
+      "step": 367
+    },
+    {
+      "epoch": 0.7813163481953291,
+      "grad_norm": 0.5447610621007355,
+      "learning_rate": 3.831689970351659e-06,
+      "loss": 0.9657,
+      "step": 368
+    },
+    {
+      "epoch": 0.7834394904458599,
+      "grad_norm": 0.5648435971288772,
+      "learning_rate": 3.7615902510056175e-06,
+      "loss": 0.8971,
+      "step": 369
+    },
+    {
+      "epoch": 0.7855626326963907,
+      "grad_norm": 0.5405051921246805,
+      "learning_rate": 3.6920456528090147e-06,
+      "loss": 0.9082,
+      "step": 370
+    },
+    {
+      "epoch": 0.7876857749469215,
+      "grad_norm": 0.5598616702029527,
+      "learning_rate": 3.6230596109166738e-06,
+      "loss": 0.9141,
+      "step": 371
+    },
+    {
+      "epoch": 0.7898089171974523,
+      "grad_norm": 0.5488453974007677,
+      "learning_rate": 3.5546355328935113e-06,
+      "loss": 0.9212,
+      "step": 372
+    },
+    {
+      "epoch": 0.7919320594479831,
+      "grad_norm": 0.5273723147565813,
+      "learning_rate": 3.4867767985462513e-06,
+      "loss": 0.862,
+      "step": 373
+    },
+    {
+      "epoch": 0.7940552016985138,
+      "grad_norm": 0.5678469468879229,
+      "learning_rate": 3.4194867597564783e-06,
+      "loss": 0.901,
+      "step": 374
+    },
+    {
+      "epoch": 0.7961783439490446,
+      "grad_norm": 0.5168620463274789,
+      "learning_rate": 3.352768740315055e-06,
+      "loss": 0.9321,
+      "step": 375
+    },
+    {
+      "epoch": 0.7983014861995754,
+      "grad_norm": 0.5283148075602601,
+      "learning_rate": 3.28662603575796e-06,
+      "loss": 0.9785,
+      "step": 376
+    },
+    {
+      "epoch": 0.8004246284501062,
+      "grad_norm": 0.5926837734103355,
+      "learning_rate": 3.2210619132034893e-06,
+      "loss": 0.9131,
+      "step": 377
+    },
+    {
+      "epoch": 0.802547770700637,
+      "grad_norm": 0.5476291313153387,
+      "learning_rate": 3.1560796111909017e-06,
+      "loss": 0.9441,
+      "step": 378
+    },
+    {
+      "epoch": 0.8046709129511678,
+      "grad_norm": 0.5569899840466258,
+      "learning_rate": 3.0916823395204204e-06,
+      "loss": 0.9746,
+      "step": 379
+    },
+    {
+      "epoch": 0.8067940552016986,
+      "grad_norm": 0.5276441512668693,
+      "learning_rate": 3.027873279094709e-06,
+      "loss": 0.9141,
+      "step": 380
+    },
+    {
+      "epoch": 0.8089171974522293,
+      "grad_norm": 0.5128633309076918,
+      "learning_rate": 2.964655581761732e-06,
+      "loss": 1.0081,
+      "step": 381
+    },
+    {
+      "epoch": 0.8110403397027601,
+      "grad_norm": 0.5247424426249943,
+      "learning_rate": 2.9020323701590984e-06,
+      "loss": 0.9706,
+      "step": 382
+    },
+    {
+      "epoch": 0.8131634819532909,
+      "grad_norm": 0.5210955810775265,
+      "learning_rate": 2.8400067375597737e-06,
+      "loss": 0.9451,
+      "step": 383
+    },
+    {
+      "epoch": 0.8152866242038217,
+      "grad_norm": 0.5074415330649746,
+      "learning_rate": 2.7785817477193393e-06,
+      "loss": 0.9037,
+      "step": 384
+    },
+    {
+      "epoch": 0.8174097664543525,
+      "grad_norm": 0.5239683128122566,
+      "learning_rate": 2.7177604347246133e-06,
+      "loss": 0.8987,
+      "step": 385
+    },
+    {
+      "epoch": 0.8195329087048833,
+      "grad_norm": 0.5646191562969649,
+      "learning_rate": 2.657545802843815e-06,
+      "loss": 0.8748,
+      "step": 386
+    },
+    {
+      "epoch": 0.821656050955414,
+      "grad_norm": 0.5317319375012074,
+      "learning_rate": 2.597940826378141e-06,
+      "loss": 0.9628,
+      "step": 387
+    },
+    {
+      "epoch": 0.8237791932059448,
+      "grad_norm": 0.546387093554132,
+      "learning_rate": 2.5389484495148845e-06,
+      "loss": 0.9247,
+      "step": 388
+    },
+    {
+      "epoch": 0.8259023354564756,
+      "grad_norm": 0.5277851982282334,
+      "learning_rate": 2.480571586181969e-06,
+      "loss": 1.006,
+      "step": 389
+    },
+    {
+      "epoch": 0.8280254777070064,
+      "grad_norm": 0.5353277977152705,
+      "learning_rate": 2.422813119904035e-06,
+      "loss": 0.9211,
+      "step": 390
+    },
+    {
+      "epoch": 0.8301486199575372,
+      "grad_norm": 0.49924730578167487,
+      "learning_rate": 2.365675903660019e-06,
+      "loss": 0.9882,
+      "step": 391
+    },
+    {
+      "epoch": 0.832271762208068,
+      "grad_norm": 0.5456822409035776,
+      "learning_rate": 2.309162759742203e-06,
+      "loss": 0.9151,
+      "step": 392
+    },
+    {
+      "epoch": 0.8343949044585988,
+      "grad_norm": 0.5124444624168185,
+      "learning_rate": 2.2532764796168293e-06,
+      "loss": 0.9209,
+      "step": 393
+    },
+    {
+      "epoch": 0.8365180467091295,
+      "grad_norm": 0.5304159814259841,
+      "learning_rate": 2.1980198237862003e-06,
+      "loss": 0.9031,
+      "step": 394
+    },
+    {
+      "epoch": 0.8386411889596603,
+      "grad_norm": 0.5485951227041947,
+      "learning_rate": 2.143395521652352e-06,
+      "loss": 0.9073,
+      "step": 395
+    },
+    {
+      "epoch": 0.8407643312101911,
+      "grad_norm": 0.5327985525474135,
+      "learning_rate": 2.0894062713821875e-06,
+      "loss": 0.9278,
+      "step": 396
+    },
+    {
+      "epoch": 0.8428874734607219,
+      "grad_norm": 0.5634270489913519,
+      "learning_rate": 2.036054739774252e-06,
+      "loss": 0.9634,
+      "step": 397
+    },
+    {
+      "epoch": 0.8450106157112527,
+      "grad_norm": 0.5667711068580208,
+      "learning_rate": 1.983343562126969e-06,
+      "loss": 1.0032,
+      "step": 398
+    },
+    {
+      "epoch": 0.8471337579617835,
+      "grad_norm": 0.5139510204694832,
+      "learning_rate": 1.931275342108497e-06,
+      "loss": 0.9276,
+      "step": 399
+    },
+    {
+      "epoch": 0.8492569002123143,
+      "grad_norm": 0.5235132713958444,
+      "learning_rate": 1.8798526516280873e-06,
+      "loss": 0.982,
+      "step": 400
+    },
+    {
+      "epoch": 0.851380042462845,
+      "grad_norm": 0.5791809289796096,
+      "learning_rate": 1.8290780307090893e-06,
+      "loss": 0.9705,
+      "step": 401
+    },
+    {
+      "epoch": 0.8535031847133758,
+      "grad_norm": 0.5472732839017603,
+      "learning_rate": 1.7789539873634465e-06,
+      "loss": 0.8861,
+      "step": 402
+    },
+    {
+      "epoch": 0.8556263269639066,
+      "grad_norm": 0.5337802205731331,
+      "learning_rate": 1.7294829974678338e-06,
+      "loss": 0.9718,
+      "step": 403
+    },
+    {
+      "epoch": 0.8577494692144374,
+      "grad_norm": 0.4961616533163996,
+      "learning_rate": 1.6806675046413523e-06,
+      "loss": 0.9298,
+      "step": 404
+    },
+    {
+      "epoch": 0.8598726114649682,
+      "grad_norm": 0.5153845139562472,
+      "learning_rate": 1.6325099201248434e-06,
+      "loss": 0.9034,
+      "step": 405
+    },
+    {
+      "epoch": 0.861995753715499,
+      "grad_norm": 0.5218172282337198,
+      "learning_rate": 1.5850126226617611e-06,
+      "loss": 0.9617,
+      "step": 406
+    },
+    {
+      "epoch": 0.8641188959660298,
+      "grad_norm": 0.5285316346557032,
+      "learning_rate": 1.538177958380692e-06,
+      "loss": 0.9333,
+      "step": 407
+    },
+    {
+      "epoch": 0.8662420382165605,
+      "grad_norm": 0.5290650405446068,
+      "learning_rate": 1.4920082406794577e-06,
+      "loss": 0.9675,
+      "step": 408
+    },
+    {
+      "epoch": 0.8683651804670913,
+      "grad_norm": 0.5152626273461832,
+      "learning_rate": 1.4465057501108547e-06,
+      "loss": 0.9525,
+      "step": 409
+    },
+    {
+      "epoch": 0.8704883227176221,
+      "grad_norm": 0.5541821991065954,
+      "learning_rate": 1.4016727342700013e-06,
+      "loss": 0.9728,
+      "step": 410
+    },
+    {
+      "epoch": 0.8726114649681529,
+      "grad_norm": 0.5180251216693375,
+      "learning_rate": 1.35751140768331e-06,
+      "loss": 0.9356,
+      "step": 411
+    },
+    {
+      "epoch": 0.8747346072186837,
+      "grad_norm": 0.5315632109606441,
+      "learning_rate": 1.314023951699126e-06,
+      "loss": 0.9481,
+      "step": 412
+    },
+    {
+      "epoch": 0.8768577494692145,
+      "grad_norm": 0.5317476915375534,
+      "learning_rate": 1.2712125143799352e-06,
+      "loss": 0.944,
+      "step": 413
+    },
+    {
+      "epoch": 0.8789808917197452,
+      "grad_norm": 0.5101783632443054,
+      "learning_rate": 1.2290792103963134e-06,
+      "loss": 0.9386,
+      "step": 414
+    },
+    {
+      "epoch": 0.881104033970276,
+      "grad_norm": 0.5161699094910599,
+      "learning_rate": 1.1876261209224315e-06,
+      "loss": 0.9739,
+      "step": 415
+    },
+    {
+      "epoch": 0.8832271762208068,
+      "grad_norm": 0.5069008558918029,
+      "learning_rate": 1.146855293533276e-06,
+      "loss": 0.9697,
+      "step": 416
+    },
+    {
+      "epoch": 0.8853503184713376,
+      "grad_norm": 0.5169934743179082,
+      "learning_rate": 1.1067687421034988e-06,
+      "loss": 0.9522,
+      "step": 417
+    },
+    {
+      "epoch": 0.8874734607218684,
+      "grad_norm": 0.5526064666443606,
+      "learning_rate": 1.0673684467079593e-06,
+      "loss": 0.9658,
+      "step": 418
+    },
+    {
+      "epoch": 0.8895966029723992,
+      "grad_norm": 0.558402647849203,
+      "learning_rate": 1.028656353523888e-06,
+      "loss": 0.8869,
+      "step": 419
+    },
+    {
+      "epoch": 0.89171974522293,
+      "grad_norm": 0.5155561997802292,
+      "learning_rate": 9.906343747347884e-07,
+      "loss": 0.8905,
+      "step": 420
+    },
+    {
+      "epoch": 0.8938428874734607,
+      "grad_norm": 0.5037480840863279,
+      "learning_rate": 9.533043884359616e-07,
+      "loss": 0.9004,
+      "step": 421
+    },
+    {
+      "epoch": 0.8959660297239915,
+      "grad_norm": 0.5310521562339193,
+      "learning_rate": 9.166682385417518e-07,
+      "loss": 0.9336,
+      "step": 422
+    },
+    {
+      "epoch": 0.8980891719745223,
+      "grad_norm": 0.4870982236860482,
+      "learning_rate": 8.807277346944536e-07,
+      "loss": 0.9222,
+      "step": 423
+    },
+    {
+      "epoch": 0.9002123142250531,
+      "grad_norm": 0.560683937579732,
+      "learning_rate": 8.454846521749359e-07,
+      "loss": 0.9421,
+      "step": 424
+    },
+    {
+      "epoch": 0.9023354564755839,
+      "grad_norm": 0.5397196499293764,
+      "learning_rate": 8.109407318149458e-07,
+      "loss": 0.9597,
+      "step": 425
+    },
+    {
+      "epoch": 0.9044585987261147,
+      "grad_norm": 0.5386667740753814,
+      "learning_rate": 7.770976799111173e-07,
+      "loss": 0.9355,
+      "step": 426
+    },
+    {
+      "epoch": 0.9065817409766455,
+      "grad_norm": 0.5256283563805639,
+      "learning_rate": 7.439571681407054e-07,
+      "loss": 0.9276,
+      "step": 427
+    },
+    {
+      "epoch": 0.9087048832271762,
+      "grad_norm": 0.5150091072669828,
+      "learning_rate": 7.115208334789902e-07,
+      "loss": 0.929,
+      "step": 428
+    },
+    {
+      "epoch": 0.910828025477707,
+      "grad_norm": 0.47871606744915557,
+      "learning_rate": 6.797902781184323e-07,
+      "loss": 0.9375,
+      "step": 429
+    },
+    {
+      "epoch": 0.9129511677282378,
+      "grad_norm": 0.5108982128127894,
+      "learning_rate": 6.487670693895303e-07,
+      "loss": 0.9055,
+      "step": 430
+    },
+    {
+      "epoch": 0.9150743099787686,
+      "grad_norm": 0.5129525617567522,
+      "learning_rate": 6.184527396834083e-07,
+      "loss": 0.9465,
+      "step": 431
+    },
+    {
+      "epoch": 0.9171974522292994,
+      "grad_norm": 0.5275723160353611,
+      "learning_rate": 5.888487863761044e-07,
+      "loss": 0.9447,
+      "step": 432
+    },
+    {
+      "epoch": 0.9193205944798302,
+      "grad_norm": 0.4945517459629921,
+      "learning_rate": 5.599566717546351e-07,
+      "loss": 0.9416,
+      "step": 433
+    },
+    {
+      "epoch": 0.921443736730361,
+      "grad_norm": 0.5278346054176051,
+      "learning_rate": 5.317778229447412e-07,
+      "loss": 0.9072,
+      "step": 434
+    },
+    {
+      "epoch": 0.9235668789808917,
+      "grad_norm": 0.4860681928212429,
+      "learning_rate": 5.043136318404129e-07,
+      "loss": 0.8765,
+      "step": 435
+    },
+    {
+      "epoch": 0.9256900212314225,
+      "grad_norm": 0.5135706402057871,
+      "learning_rate": 4.775654550351194e-07,
+      "loss": 0.9753,
+      "step": 436
+    },
+    {
+      "epoch": 0.9278131634819533,
+      "grad_norm": 0.5260121053711417,
+      "learning_rate": 4.515346137548193e-07,
+      "loss": 0.9039,
+      "step": 437
+    },
+    {
+      "epoch": 0.9299363057324841,
+      "grad_norm": 0.5456543202246997,
+      "learning_rate": 4.2622239379268457e-07,
+      "loss": 0.8885,
+      "step": 438
+    },
+    {
+      "epoch": 0.9320594479830149,
+      "grad_norm": 0.5282444147947248,
+      "learning_rate": 4.016300454455946e-07,
+      "loss": 0.9193,
+      "step": 439
+    },
+    {
+      "epoch": 0.9341825902335457,
+      "grad_norm": 0.5418531883223643,
+      "learning_rate": 3.777587834523788e-07,
+      "loss": 0.9541,
+      "step": 440
+    },
+    {
+      "epoch": 0.9363057324840764,
+      "grad_norm": 0.5087399292427347,
+      "learning_rate": 3.546097869338083e-07,
+      "loss": 0.9162,
+      "step": 441
+    },
+    {
+      "epoch": 0.9384288747346072,
+      "grad_norm": 0.4906370994458251,
+      "learning_rate": 3.3218419933436217e-07,
+      "loss": 0.8623,
+      "step": 442
+    },
+    {
+      "epoch": 0.940552016985138,
+      "grad_norm": 0.5153741738005715,
+      "learning_rate": 3.104831283657378e-07,
+      "loss": 0.9199,
+      "step": 443
+    },
+    {
+      "epoch": 0.9426751592356688,
+      "grad_norm": 0.5175526473310568,
+      "learning_rate": 2.8950764595214284e-07,
+      "loss": 0.891,
+      "step": 444
+    },
+    {
+      "epoch": 0.9447983014861996,
+      "grad_norm": 0.5274728670609353,
+      "learning_rate": 2.6925878817734786e-07,
+      "loss": 0.929,
+      "step": 445
+    },
+    {
+      "epoch": 0.9469214437367304,
+      "grad_norm": 0.5510020089756117,
+      "learning_rate": 2.4973755523349684e-07,
+      "loss": 0.9207,
+      "step": 446
+    },
+    {
+      "epoch": 0.9490445859872612,
+      "grad_norm": 0.5452919257619138,
+      "learning_rate": 2.3094491137172046e-07,
+      "loss": 0.9215,
+      "step": 447
+    },
+    {
+      "epoch": 0.9511677282377919,
+      "grad_norm": 0.4989565927307046,
+      "learning_rate": 2.128817848544956e-07,
+      "loss": 0.9364,
+      "step": 448
+    },
+    {
+      "epoch": 0.9532908704883227,
+      "grad_norm": 0.49733089547176756,
+      "learning_rate": 1.955490679097921e-07,
+      "loss": 0.9647,
+      "step": 449
+    },
+    {
+      "epoch": 0.9554140127388535,
+      "grad_norm": 0.5314718507978706,
+      "learning_rate": 1.7894761668701297e-07,
+      "loss": 0.9407,
+      "step": 450
+    },
+    {
+      "epoch": 0.9575371549893843,
+      "grad_norm": 0.5233578815604037,
+      "learning_rate": 1.6307825121469165e-07,
+      "loss": 0.9486,
+      "step": 451
+    },
+    {
+      "epoch": 0.9596602972399151,
+      "grad_norm": 0.5394446108476952,
+      "learning_rate": 1.4794175535999266e-07,
+      "loss": 0.9446,
+      "step": 452
+    },
+    {
+      "epoch": 0.9617834394904459,
+      "grad_norm": 0.5403944209516118,
+      "learning_rate": 1.3353887678999588e-07,
+      "loss": 0.9515,
+      "step": 453
+    },
+    {
+      "epoch": 0.9639065817409767,
+      "grad_norm": 0.4814718907548529,
+      "learning_rate": 1.198703269347612e-07,
+      "loss": 0.9452,
+      "step": 454
+    },
+    {
+      "epoch": 0.9660297239915074,
+      "grad_norm": 0.5267438283654036,
+      "learning_rate": 1.0693678095218484e-07,
+      "loss": 0.9335,
+      "step": 455
+    },
+    {
+      "epoch": 0.9681528662420382,
+      "grad_norm": 0.5007166014022263,
+      "learning_rate": 9.473887769465617e-08,
+      "loss": 0.9017,
+      "step": 456
+    },
+    {
+      "epoch": 0.970276008492569,
+      "grad_norm": 0.4955960410973204,
+      "learning_rate": 8.327721967749779e-08,
+      "loss": 0.891,
+      "step": 457
+    },
+    {
+      "epoch": 0.9723991507430998,
+      "grad_norm": 0.5564203179114615,
+      "learning_rate": 7.255237304920448e-08,
+      "loss": 0.8917,
+      "step": 458
+    },
+    {
+      "epoch": 0.9745222929936306,
+      "grad_norm": 0.522629078249321,
+      "learning_rate": 6.256486756348212e-08,
+      "loss": 0.9278,
+      "step": 459
+    },
+    {
+      "epoch": 0.9766454352441614,
+      "grad_norm": 0.5165538019449568,
+      "learning_rate": 5.331519655307371e-08,
+      "loss": 0.9608,
+      "step": 460
+    },
+    {
+      "epoch": 0.9787685774946921,
+      "grad_norm": 0.5160598267580098,
+      "learning_rate": 4.4803816905397147e-08,
+      "loss": 0.9291,
+      "step": 461
+    },
+    {
+      "epoch": 0.9808917197452229,
+      "grad_norm": 0.5438197694625472,
+      "learning_rate": 3.7031149039971624e-08,
+      "loss": 0.9563,
+      "step": 462
+    },
+    {
+      "epoch": 0.9830148619957537,
+      "grad_norm": 0.5181344718101011,
+      "learning_rate": 2.9997576887660915e-08,
+      "loss": 0.8855,
+      "step": 463
+    },
+    {
+      "epoch": 0.9851380042462845,
+      "grad_norm": 0.5431722389214804,
+      "learning_rate": 2.3703447871698538e-08,
+      "loss": 0.9591,
+      "step": 464
+    },
+    {
+      "epoch": 0.9872611464968153,
+      "grad_norm": 0.5187571644495663,
+      "learning_rate": 1.814907289052814e-08,
+      "loss": 0.9953,
+      "step": 465
+    },
+    {
+      "epoch": 0.9893842887473461,
+      "grad_norm": 0.5927364512159669,
+      "learning_rate": 1.3334726302454136e-08,
+      "loss": 0.9342,
+      "step": 466
+    },
+    {
+      "epoch": 0.9915074309978769,
+      "grad_norm": 0.5096371256028509,
+      "learning_rate": 9.260645912084198e-09,
+      "loss": 0.9076,
+      "step": 467
+    },
+    {
+      "epoch": 0.9936305732484076,
+      "grad_norm": 0.4981275603649081,
+      "learning_rate": 5.927032958580325e-09,
+      "loss": 0.9197,
+      "step": 468
+    },
+    {
+      "epoch": 0.9957537154989384,
+      "grad_norm": 0.4782521123129465,
+      "learning_rate": 3.3340521057284577e-09,
+      "loss": 0.9014,
+      "step": 469
+    },
+    {
+      "epoch": 0.9978768577494692,
+      "grad_norm": 0.5314495327090503,
+      "learning_rate": 1.4818314337916584e-09,
+      "loss": 0.9241,
+      "step": 470
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.4933970721941701,
+      "learning_rate": 3.704624331951623e-10,
+      "loss": 0.9263,
+      "step": 471
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 471,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 157,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 70004199849984.0,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cd681504643138a679e55133a50cd258b4e3cf9ca5f57dfb9fcdc5045782b5d3
+size 8017