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 sha256:420684b17d3c1f8b6da8fc7bcbc84087147902762afa58587cbbd565b60f3395
+size 4997184968

model-00002-of-00013.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:554e139002bd50fdd84d4ff49240a161a3ccda10ec58983b40ba7d2d568e97a4
+size 4997741608

model-00003-of-00013.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:85a31a685ed49a33e39ccc87e23935c672771940e63cabd1924eb9417974011f
+size 4997742208

model-00004-of-00013.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2700a37bb687af4e89d922d3179718b261552a6d6d113e1e5f8f036ac240b57d
+size 4997743184

model-00005-of-00013.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a217aae1b4e51fb22ce106f6618984a093b4104fc9982b3b5d53c57dc9cc3381
+size 4997743184

model-00006-of-00013.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e8782c34f887cd526147449a9fcacd6c2991507b95368c25480645bd2b7fcaef
+size 4997743184

model-00007-of-00013.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6d230136d810f7231927bd7dafa800332f40fe42c5f37ef207e937a71488c14d
+size 4997743184

model-00008-of-00013.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7aeb9effc2ae54743d9f2b6c1e2e4236fd65d0ca584052a14cfecd95ec84982a
+size 4997743184

model-00009-of-00013.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9e04241e269e37d007585d2a9b0242a1082487641818b4f98d511fdecd7bf20d
+size 4997743184

model-00010-of-00013.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6786d6e4b5f0e29f8ab44ba7a408adc4e9f8db7d59e770e1e519a2a5ffe740ae
+size 4997743184

model-00011-of-00013.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:47f844c36b12a552310cbcfe22e67ecefdc5cd0a4a1c1b04978dedb795730af2
+size 4997743184

model-00012-of-00013.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:76d2c4fea5823cc0f9202303ae7596e249343d0a405fd22f56cbae3c71b578e0
+size 4997743184

model-00013-of-00013.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1a7dd44f1c986e102af53e6797a489d5edbb8c210b4f962f16bc0918ff1811c7
+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": 6.036282375751906,
+      "learning_rate": 2.0833333333333333e-07,
+      "loss": 1.4092,
+      "step": 2
+    },
+    {
+      "epoch": 0.006369426751592357,
+      "grad_norm": 5.891130120999039,
+      "learning_rate": 4.1666666666666667e-07,
+      "loss": 1.4008,
+      "step": 3
+    },
+    {
+      "epoch": 0.008492569002123142,
+      "grad_norm": 6.074487333844155,
+      "learning_rate": 6.25e-07,
+      "loss": 1.4031,
+      "step": 4
+    },
+    {
+      "epoch": 0.010615711252653927,
+      "grad_norm": 5.743607303282022,
+      "learning_rate": 8.333333333333333e-07,
+      "loss": 1.3855,
+      "step": 5
+    },
+    {
+      "epoch": 0.012738853503184714,
+      "grad_norm": 5.978505048824056,
+      "learning_rate": 1.0416666666666667e-06,
+      "loss": 1.4133,
+      "step": 6
+    },
+    {
+      "epoch": 0.014861995753715499,
+      "grad_norm": 5.6113465264480835,
+      "learning_rate": 1.25e-06,
+      "loss": 1.3079,
+      "step": 7
+    },
+    {
+      "epoch": 0.016985138004246284,
+      "grad_norm": 5.290707913968832,
+      "learning_rate": 1.4583333333333335e-06,
+      "loss": 1.3761,
+      "step": 8
+    },
+    {
+      "epoch": 0.01910828025477707,
+      "grad_norm": 5.4254172240562735,
+      "learning_rate": 1.6666666666666667e-06,
+      "loss": 1.3741,
+      "step": 9
+    },
+    {
+      "epoch": 0.021231422505307854,
+      "grad_norm": 5.17021866948228,
+      "learning_rate": 1.8750000000000003e-06,
+      "loss": 1.3834,
+      "step": 10
+    },
+    {
+      "epoch": 0.02335456475583864,
+      "grad_norm": 4.250457587049274,
+      "learning_rate": 2.0833333333333334e-06,
+      "loss": 1.3276,
+      "step": 11
+    },
+    {
+      "epoch": 0.025477707006369428,
+      "grad_norm": 4.0761343260047616,
+      "learning_rate": 2.2916666666666666e-06,
+      "loss": 1.3091,
+      "step": 12
+    },
+    {
+      "epoch": 0.027600849256900213,
+      "grad_norm": 4.085721585422903,
+      "learning_rate": 2.5e-06,
+      "loss": 1.3736,
+      "step": 13
+    },
+    {
+      "epoch": 0.029723991507430998,
+      "grad_norm": 3.4660539777465926,
+      "learning_rate": 2.7083333333333334e-06,
+      "loss": 1.2959,
+      "step": 14
+    },
+    {
+      "epoch": 0.03184713375796178,
+      "grad_norm": 3.2306408364760006,
+      "learning_rate": 2.916666666666667e-06,
+      "loss": 1.3213,
+      "step": 15
+    },
+    {
+      "epoch": 0.03397027600849257,
+      "grad_norm": 2.9881593116573764,
+      "learning_rate": 3.125e-06,
+      "loss": 1.2841,
+      "step": 16
+    },
+    {
+      "epoch": 0.036093418259023353,
+      "grad_norm": 2.998705518524036,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 1.2581,
+      "step": 17
+    },
+    {
+      "epoch": 0.03821656050955414,
+      "grad_norm": 2.732961535739617,
+      "learning_rate": 3.5416666666666673e-06,
+      "loss": 1.2221,
+      "step": 18
+    },
+    {
+      "epoch": 0.040339702760084924,
+      "grad_norm": 1.8117955120416394,
+      "learning_rate": 3.7500000000000005e-06,
+      "loss": 1.1951,
+      "step": 19
+    },
+    {
+      "epoch": 0.04246284501061571,
+      "grad_norm": 1.9886830864925722,
+      "learning_rate": 3.958333333333333e-06,
+      "loss": 1.097,
+      "step": 20
+    },
+    {
+      "epoch": 0.044585987261146494,
+      "grad_norm": 2.162707915070587,
+      "learning_rate": 4.166666666666667e-06,
+      "loss": 1.2057,
+      "step": 21
+    },
+    {
+      "epoch": 0.04670912951167728,
+      "grad_norm": 2.696607450468263,
+      "learning_rate": 4.3750000000000005e-06,
+      "loss": 1.1726,
+      "step": 22
+    },
+    {
+      "epoch": 0.04883227176220807,
+      "grad_norm": 1.8593899724728866,
+      "learning_rate": 4.583333333333333e-06,
+      "loss": 1.0822,
+      "step": 23
+    },
+    {
+      "epoch": 0.050955414012738856,
+      "grad_norm": 1.4860669687181005,
+      "learning_rate": 4.791666666666668e-06,
+      "loss": 1.0889,
+      "step": 24
+    },
+    {
+      "epoch": 0.05307855626326964,
+      "grad_norm": 1.6240718023385812,
+      "learning_rate": 5e-06,
+      "loss": 1.1117,
+      "step": 25
+    },
+    {
+      "epoch": 0.055201698513800426,
+      "grad_norm": 2.09213691220876,
+      "learning_rate": 4.999938256261134e-06,
+      "loss": 1.0979,
+      "step": 26
+    },
+    {
+      "epoch": 0.05732484076433121,
+      "grad_norm": 1.7482122008435805,
+      "learning_rate": 4.999753028094368e-06,
+      "loss": 1.1355,
+      "step": 27
+    },
+    {
+      "epoch": 0.059447983014861996,
+      "grad_norm": 1.883493540815875,
+      "learning_rate": 4.999444324649045e-06,
+      "loss": 1.1341,
+      "step": 28
+    },
+    {
+      "epoch": 0.06157112526539278,
+      "grad_norm": 1.8270811899866102,
+      "learning_rate": 4.99901216117357e-06,
+      "loss": 1.1051,
+      "step": 29
+    },
+    {
+      "epoch": 0.06369426751592357,
+      "grad_norm": 1.4502089970711367,
+      "learning_rate": 4.998456559014653e-06,
+      "loss": 1.1046,
+      "step": 30
+    },
+    {
+      "epoch": 0.06581740976645435,
+      "grad_norm": 1.425355888694126,
+      "learning_rate": 4.997777545616258e-06,
+      "loss": 1.0905,
+      "step": 31
+    },
+    {
+      "epoch": 0.06794055201698514,
+      "grad_norm": 1.2505755417516413,
+      "learning_rate": 4.996975154518245e-06,
+      "loss": 1.0822,
+      "step": 32
+    },
+    {
+      "epoch": 0.07006369426751592,
+      "grad_norm": 1.1388510764741677,
+      "learning_rate": 4.996049425354717e-06,
+      "loss": 1.0415,
+      "step": 33
+    },
+    {
+      "epoch": 0.07218683651804671,
+      "grad_norm": 1.4794550850120727,
+      "learning_rate": 4.995000403852057e-06,
+      "loss": 1.035,
+      "step": 34
+    },
+    {
+      "epoch": 0.07430997876857749,
+      "grad_norm": 0.9754315052343883,
+      "learning_rate": 4.993828141826672e-06,
+      "loss": 1.0415,
+      "step": 35
+    },
+    {
+      "epoch": 0.07643312101910828,
+      "grad_norm": 0.9081548956782857,
+      "learning_rate": 4.992532697182434e-06,
+      "loss": 1.0696,
+      "step": 36
+    },
+    {
+      "epoch": 0.07855626326963906,
+      "grad_norm": 0.8968088704858169,
+      "learning_rate": 4.991114133907822e-06,
+      "loss": 1.0056,
+      "step": 37
+    },
+    {
+      "epoch": 0.08067940552016985,
+      "grad_norm": 0.9873042697593618,
+      "learning_rate": 4.989572522072753e-06,
+      "loss": 0.9949,
+      "step": 38
+    },
+    {
+      "epoch": 0.08280254777070063,
+      "grad_norm": 0.9927345903093555,
+      "learning_rate": 4.9879079378251325e-06,
+      "loss": 0.9767,
+      "step": 39
+    },
+    {
+      "epoch": 0.08492569002123142,
+      "grad_norm": 0.9227421380705588,
+      "learning_rate": 4.986120463387084e-06,
+      "loss": 1.0253,
+      "step": 40
+    },
+    {
+      "epoch": 0.0870488322717622,
+      "grad_norm": 1.1001162646683043,
+      "learning_rate": 4.984210187050891e-06,
+      "loss": 0.9883,
+      "step": 41
+    },
+    {
+      "epoch": 0.08917197452229299,
+      "grad_norm": 0.8622396840543709,
+      "learning_rate": 4.982177203174636e-06,
+      "loss": 0.9848,
+      "step": 42
+    },
+    {
+      "epoch": 0.09129511677282377,
+      "grad_norm": 0.7528337414302051,
+      "learning_rate": 4.9800216121775404e-06,
+      "loss": 0.9826,
+      "step": 43
+    },
+    {
+      "epoch": 0.09341825902335456,
+      "grad_norm": 0.7513148560404738,
+      "learning_rate": 4.977743520535001e-06,
+      "loss": 1.0494,
+      "step": 44
+    },
+    {
+      "epoch": 0.09554140127388536,
+      "grad_norm": 0.8500263228426365,
+      "learning_rate": 4.975343040773335e-06,
+      "loss": 1.051,
+      "step": 45
+    },
+    {
+      "epoch": 0.09766454352441614,
+      "grad_norm": 0.7905864487569398,
+      "learning_rate": 4.972820291464219e-06,
+      "loss": 1.0334,
+      "step": 46
+    },
+    {
+      "epoch": 0.09978768577494693,
+      "grad_norm": 0.7942918416765442,
+      "learning_rate": 4.970175397218832e-06,
+      "loss": 0.9982,
+      "step": 47
+    },
+    {
+      "epoch": 0.10191082802547771,
+      "grad_norm": 0.7726621895136483,
+      "learning_rate": 4.967408488681702e-06,
+      "loss": 0.9995,
+      "step": 48
+    },
+    {
+      "epoch": 0.1040339702760085,
+      "grad_norm": 0.8071759021352843,
+      "learning_rate": 4.964519702524251e-06,
+      "loss": 0.9859,
+      "step": 49
+    },
+    {
+      "epoch": 0.10615711252653928,
+      "grad_norm": 0.8568099936673271,
+      "learning_rate": 4.9615091814380465e-06,
+      "loss": 1.0227,
+      "step": 50
+    },
+    {
+      "epoch": 0.10828025477707007,
+      "grad_norm": 0.7646989151242823,
+      "learning_rate": 4.958377074127751e-06,
+      "loss": 1.0093,
+      "step": 51
+    },
+    {
+      "epoch": 0.11040339702760085,
+      "grad_norm": 0.6998250690202109,
+      "learning_rate": 4.955123535303775e-06,
+      "loss": 1.0157,
+      "step": 52
+    },
+    {
+      "epoch": 0.11252653927813164,
+      "grad_norm": 0.7041148727209998,
+      "learning_rate": 4.951748725674643e-06,
+      "loss": 0.944,
+      "step": 53
+    },
+    {
+      "epoch": 0.11464968152866242,
+      "grad_norm": 0.7178789444758534,
+      "learning_rate": 4.948252811939044e-06,
+      "loss": 1.0034,
+      "step": 54
+    },
+    {
+      "epoch": 0.11677282377919321,
+      "grad_norm": 0.7138939260260085,
+      "learning_rate": 4.944635966777607e-06,
+      "loss": 1.0411,
+      "step": 55
+    },
+    {
+      "epoch": 0.11889596602972399,
+      "grad_norm": 0.7303515503729836,
+      "learning_rate": 4.940898368844366e-06,
+      "loss": 1.0668,
+      "step": 56
+    },
+    {
+      "epoch": 0.12101910828025478,
+      "grad_norm": 0.7112682591004899,
+      "learning_rate": 4.937040202757937e-06,
+      "loss": 1.0061,
+      "step": 57
+    },
+    {
+      "epoch": 0.12314225053078556,
+      "grad_norm": 0.7839574196233094,
+      "learning_rate": 4.933061659092401e-06,
+      "loss": 1.0048,
+      "step": 58
+    },
+    {
+      "epoch": 0.12526539278131635,
+      "grad_norm": 0.7944407734544054,
+      "learning_rate": 4.928962934367887e-06,
+      "loss": 1.0004,
+      "step": 59
+    },
+    {
+      "epoch": 0.12738853503184713,
+      "grad_norm": 0.7280265198789528,
+      "learning_rate": 4.924744231040865e-06,
+      "loss": 0.9948,
+      "step": 60
+    },
+    {
+      "epoch": 0.12951167728237792,
+      "grad_norm": 0.7307519901984729,
+      "learning_rate": 4.920405757494147e-06,
+      "loss": 0.9212,
+      "step": 61
+    },
+    {
+      "epoch": 0.1316348195329087,
+      "grad_norm": 0.6979758699830271,
+      "learning_rate": 4.915947728026599e-06,
+      "loss": 0.9826,
+      "step": 62
+    },
+    {
+      "epoch": 0.1337579617834395,
+      "grad_norm": 0.6627924762335937,
+      "learning_rate": 4.911370362842544e-06,
+      "loss": 1.0036,
+      "step": 63
+    },
+    {
+      "epoch": 0.13588110403397027,
+      "grad_norm": 0.7322381275827045,
+      "learning_rate": 4.906673888040895e-06,
+      "loss": 1.0255,
+      "step": 64
+    },
+    {
+      "epoch": 0.13800424628450106,
+      "grad_norm": 0.7558866470186772,
+      "learning_rate": 4.901858535603982e-06,
+      "loss": 1.0337,
+      "step": 65
+    },
+    {
+      "epoch": 0.14012738853503184,
+      "grad_norm": 0.6564342903823904,
+      "learning_rate": 4.896924543386099e-06,
+      "loss": 0.9731,
+      "step": 66
+    },
+    {
+      "epoch": 0.14225053078556263,
+      "grad_norm": 0.6465572624496233,
+      "learning_rate": 4.891872155101746e-06,
+      "loss": 0.9757,
+      "step": 67
+    },
+    {
+      "epoch": 0.14437367303609341,
+      "grad_norm": 0.7659232216669788,
+      "learning_rate": 4.886701620313595e-06,
+      "loss": 0.9733,
+      "step": 68
+    },
+    {
+      "epoch": 0.1464968152866242,
+      "grad_norm": 0.6777766233641306,
+      "learning_rate": 4.88141319442017e-06,
+      "loss": 0.9516,
+      "step": 69
+    },
+    {
+      "epoch": 0.14861995753715498,
+      "grad_norm": 0.7233595060808201,
+      "learning_rate": 4.876007138643216e-06,
+      "loss": 1.0614,
+      "step": 70
+    },
+    {
+      "epoch": 0.15074309978768577,
+      "grad_norm": 1.0421440952867649,
+      "learning_rate": 4.870483720014814e-06,
+      "loss": 1.002,
+      "step": 71
+    },
+    {
+      "epoch": 0.15286624203821655,
+      "grad_norm": 0.7222126514560366,
+      "learning_rate": 4.8648432113641765e-06,
+      "loss": 1.0246,
+      "step": 72
+    },
+    {
+      "epoch": 0.15498938428874734,
+      "grad_norm": 0.756215690480781,
+      "learning_rate": 4.8590858913041775e-06,
+      "loss": 0.9696,
+      "step": 73
+    },
+    {
+      "epoch": 0.15711252653927812,
+      "grad_norm": 0.8486802960108908,
+      "learning_rate": 4.853212044217591e-06,
+      "loss": 0.9469,
+      "step": 74
+    },
+    {
+      "epoch": 0.1592356687898089,
+      "grad_norm": 0.6567453740552108,
+      "learning_rate": 4.847221960243042e-06,
+      "loss": 0.9919,
+      "step": 75
+    },
+    {
+      "epoch": 0.1613588110403397,
+      "grad_norm": 0.7119655661591375,
+      "learning_rate": 4.8411159352606735e-06,
+      "loss": 0.9105,
+      "step": 76
+    },
+    {
+      "epoch": 0.16348195329087048,
+      "grad_norm": 0.6694111209083354,
+      "learning_rate": 4.834894270877536e-06,
+      "loss": 0.9626,
+      "step": 77
+    },
+    {
+      "epoch": 0.16560509554140126,
+      "grad_norm": 0.7090596038764783,
+      "learning_rate": 4.828557274412686e-06,
+      "loss": 0.9786,
+      "step": 78
+    },
+    {
+      "epoch": 0.16772823779193205,
+      "grad_norm": 0.657778836038884,
+      "learning_rate": 4.822105258882007e-06,
+      "loss": 0.9775,
+      "step": 79
+    },
+    {
+      "epoch": 0.16985138004246284,
+      "grad_norm": 0.7229446128005313,
+      "learning_rate": 4.815538542982751e-06,
+      "loss": 1.0749,
+      "step": 80
+    },
+    {
+      "epoch": 0.17197452229299362,
+      "grad_norm": 0.737338248354114,
+      "learning_rate": 4.808857451077788e-06,
+      "loss": 1.0018,
+      "step": 81
+    },
+    {
+      "epoch": 0.1740976645435244,
+      "grad_norm": 0.7508049484616365,
+      "learning_rate": 4.802062313179595e-06,
+      "loss": 1.0558,
+      "step": 82
+    },
+    {
+      "epoch": 0.1762208067940552,
+      "grad_norm": 0.7007354351103503,
+      "learning_rate": 4.795153464933948e-06,
+      "loss": 0.9794,
+      "step": 83
+    },
+    {
+      "epoch": 0.17834394904458598,
+      "grad_norm": 0.662357169178099,
+      "learning_rate": 4.7881312476033444e-06,
+      "loss": 1.009,
+      "step": 84
+    },
+    {
+      "epoch": 0.18046709129511676,
+      "grad_norm": 0.6849477961866517,
+      "learning_rate": 4.7809960080501464e-06,
+      "loss": 0.9875,
+      "step": 85
+    },
+    {
+      "epoch": 0.18259023354564755,
+      "grad_norm": 0.6965592321624478,
+      "learning_rate": 4.773748098719448e-06,
+      "loss": 1.0182,
+      "step": 86
+    },
+    {
+      "epoch": 0.18471337579617833,
+      "grad_norm": 0.6979169529090863,
+      "learning_rate": 4.766387877621667e-06,
+      "loss": 1.0177,
+      "step": 87
+    },
+    {
+      "epoch": 0.18683651804670912,
+      "grad_norm": 0.7041639181716062,
+      "learning_rate": 4.758915708314858e-06,
+      "loss": 0.9516,
+      "step": 88
+    },
+    {
+      "epoch": 0.18895966029723993,
+      "grad_norm": 0.6953439875992204,
+      "learning_rate": 4.751331959886758e-06,
+      "loss": 0.9314,
+      "step": 89
+    },
+    {
+      "epoch": 0.1910828025477707,
+      "grad_norm": 0.6613631812740659,
+      "learning_rate": 4.743637006936552e-06,
+      "loss": 1.0019,
+      "step": 90
+    },
+    {
+      "epoch": 0.1932059447983015,
+      "grad_norm": 0.7037122761576546,
+      "learning_rate": 4.735831229556374e-06,
+      "loss": 0.9249,
+      "step": 91
+    },
+    {
+      "epoch": 0.19532908704883228,
+      "grad_norm": 0.8473990006647267,
+      "learning_rate": 4.727915013312527e-06,
+      "loss": 0.9633,
+      "step": 92
+    },
+    {
+      "epoch": 0.19745222929936307,
+      "grad_norm": 0.749684663578654,
+      "learning_rate": 4.719888749226442e-06,
+      "loss": 0.9788,
+      "step": 93
+    },
+    {
+      "epoch": 0.19957537154989385,
+      "grad_norm": 0.7293523592209221,
+      "learning_rate": 4.711752833755362e-06,
+      "loss": 0.9594,
+      "step": 94
+    },
+    {
+      "epoch": 0.20169851380042464,
+      "grad_norm": 2.969828282102022,
+      "learning_rate": 4.70350766877276e-06,
+      "loss": 1.0121,
+      "step": 95
+    },
+    {
+      "epoch": 0.20382165605095542,
+      "grad_norm": 0.7749000799093052,
+      "learning_rate": 4.695153661548486e-06,
+      "loss": 0.9606,
+      "step": 96
+    },
+    {
+      "epoch": 0.2059447983014862,
+      "grad_norm": 0.7712144141473064,
+      "learning_rate": 4.686691224728652e-06,
+      "loss": 0.958,
+      "step": 97
+    },
+    {
+      "epoch": 0.208067940552017,
+      "grad_norm": 0.7434684636124179,
+      "learning_rate": 4.678120776315251e-06,
+      "loss": 0.9432,
+      "step": 98
+    },
+    {
+      "epoch": 0.21019108280254778,
+      "grad_norm": 0.7363360177457489,
+      "learning_rate": 4.669442739645506e-06,
+      "loss": 1.0115,
+      "step": 99
+    },
+    {
+      "epoch": 0.21231422505307856,
+      "grad_norm": 0.6716882264915858,
+      "learning_rate": 4.660657543370958e-06,
+      "loss": 1.0124,
+      "step": 100
+    },
+    {
+      "epoch": 0.21443736730360935,
+      "grad_norm": 0.7654149258441244,
+      "learning_rate": 4.651765621436303e-06,
+      "loss": 1.0489,
+      "step": 101
+    },
+    {
+      "epoch": 0.21656050955414013,
+      "grad_norm": 0.8574368765734778,
+      "learning_rate": 4.642767413057942e-06,
+      "loss": 0.9658,
+      "step": 102
+    },
+    {
+      "epoch": 0.21868365180467092,
+      "grad_norm": 0.7424741102042721,
+      "learning_rate": 4.6336633627023e-06,
+      "loss": 0.9133,
+      "step": 103
+    },
+    {
+      "epoch": 0.2208067940552017,
+      "grad_norm": 0.6770848463550634,
+      "learning_rate": 4.624453920063863e-06,
+      "loss": 0.9765,
+      "step": 104
+    },
+    {
+      "epoch": 0.2229299363057325,
+      "grad_norm": 0.6895226993875193,
+      "learning_rate": 4.6151395400429665e-06,
+      "loss": 0.9389,
+      "step": 105
+    },
+    {
+      "epoch": 0.22505307855626328,
+      "grad_norm": 0.7364839772071928,
+      "learning_rate": 4.605720682723331e-06,
+      "loss": 0.9421,
+      "step": 106
+    },
+    {
+      "epoch": 0.22717622080679406,
+      "grad_norm": 0.6971178564903094,
+      "learning_rate": 4.596197813349328e-06,
+      "loss": 0.9838,
+      "step": 107
+    },
+    {
+      "epoch": 0.22929936305732485,
+      "grad_norm": 0.6796322414548438,
+      "learning_rate": 4.586571402303006e-06,
+      "loss": 0.9504,
+      "step": 108
+    },
+    {
+      "epoch": 0.23142250530785563,
+      "grad_norm": 0.7544007171896271,
+      "learning_rate": 4.576841925080853e-06,
+      "loss": 0.9994,
+      "step": 109
+    },
+    {
+      "epoch": 0.23354564755838642,
+      "grad_norm": 0.6857660018521021,
+      "learning_rate": 4.56700986227031e-06,
+      "loss": 1.0145,
+      "step": 110
+    },
+    {
+      "epoch": 0.2356687898089172,
+      "grad_norm": 0.6859556243272815,
+      "learning_rate": 4.557075699526032e-06,
+      "loss": 0.9596,
+      "step": 111
+    },
+    {
+      "epoch": 0.23779193205944799,
+      "grad_norm": 0.7081004553238264,
+      "learning_rate": 4.547039927545899e-06,
+      "loss": 0.9772,
+      "step": 112
+    },
+    {
+      "epoch": 0.23991507430997877,
+      "grad_norm": 0.7462372407123932,
+      "learning_rate": 4.536903042046778e-06,
+      "loss": 0.9507,
+      "step": 113
+    },
+    {
+      "epoch": 0.24203821656050956,
+      "grad_norm": 0.72601271137818,
+      "learning_rate": 4.526665543740038e-06,
+      "loss": 0.9563,
+      "step": 114
+    },
+    {
+      "epoch": 0.24416135881104034,
+      "grad_norm": 0.6381109078445588,
+      "learning_rate": 4.516327938306818e-06,
+      "loss": 1.0078,
+      "step": 115
+    },
+    {
+      "epoch": 0.24628450106157113,
+      "grad_norm": 0.7584749957729562,
+      "learning_rate": 4.505890736373045e-06,
+      "loss": 0.9765,
+      "step": 116
+    },
+    {
+      "epoch": 0.2484076433121019,
+      "grad_norm": 0.7102458794768719,
+      "learning_rate": 4.495354453484216e-06,
+      "loss": 0.9804,
+      "step": 117
+    },
+    {
+      "epoch": 0.2505307855626327,
+      "grad_norm": 0.6844867181397675,
+      "learning_rate": 4.4847196100799305e-06,
+      "loss": 0.9821,
+      "step": 118
+    },
+    {
+      "epoch": 0.2526539278131635,
+      "grad_norm": 0.7200996740889256,
+      "learning_rate": 4.473986731468183e-06,
+      "loss": 1.0106,
+      "step": 119
+    },
+    {
+      "epoch": 0.25477707006369427,
+      "grad_norm": 0.7030653364009113,
+      "learning_rate": 4.463156347799419e-06,
+      "loss": 1.0781,
+      "step": 120
+    },
+    {
+      "epoch": 0.25690021231422505,
+      "grad_norm": 0.7662768767817681,
+      "learning_rate": 4.452228994040341e-06,
+      "loss": 0.9618,
+      "step": 121
+    },
+    {
+      "epoch": 0.25902335456475584,
+      "grad_norm": 0.8105947072663244,
+      "learning_rate": 4.4412052099474916e-06,
+      "loss": 0.9622,
+      "step": 122
+    },
+    {
+      "epoch": 0.2611464968152866,
+      "grad_norm": 0.7211984778916917,
+      "learning_rate": 4.430085540040587e-06,
+      "loss": 1.0164,
+      "step": 123
+    },
+    {
+      "epoch": 0.2632696390658174,
+      "grad_norm": 0.6967668355088058,
+      "learning_rate": 4.418870533575626e-06,
+      "loss": 0.9824,
+      "step": 124
+    },
+    {
+      "epoch": 0.2653927813163482,
+      "grad_norm": 0.6779361410685855,
+      "learning_rate": 4.40756074451775e-06,
+      "loss": 0.9395,
+      "step": 125
+    },
+    {
+      "epoch": 0.267515923566879,
+      "grad_norm": 0.6670664720622704,
+      "learning_rate": 4.396156731513888e-06,
+      "loss": 1.0333,
+      "step": 126
+    },
+    {
+      "epoch": 0.26963906581740976,
+      "grad_norm": 0.7468957605757105,
+      "learning_rate": 4.384659057865165e-06,
+      "loss": 0.9469,
+      "step": 127
+    },
+    {
+      "epoch": 0.27176220806794055,
+      "grad_norm": 0.618195093191427,
+      "learning_rate": 4.373068291499065e-06,
+      "loss": 0.8902,
+      "step": 128
+    },
+    {
+      "epoch": 0.27388535031847133,
+      "grad_norm": 0.6899363714434318,
+      "learning_rate": 4.36138500494139e-06,
+      "loss": 0.973,
+      "step": 129
+    },
+    {
+      "epoch": 0.2760084925690021,
+      "grad_norm": 0.6648290687836514,
+      "learning_rate": 4.349609775287977e-06,
+      "loss": 0.9859,
+      "step": 130
+    },
+    {
+      "epoch": 0.2781316348195329,
+      "grad_norm": 0.6449761569340436,
+      "learning_rate": 4.337743184176188e-06,
+      "loss": 0.9589,
+      "step": 131
+    },
+    {
+      "epoch": 0.2802547770700637,
+      "grad_norm": 0.6376297046774647,
+      "learning_rate": 4.325785817756186e-06,
+      "loss": 0.9838,
+      "step": 132
+    },
+    {
+      "epoch": 0.2823779193205945,
+      "grad_norm": 0.6604823512178201,
+      "learning_rate": 4.313738266661979e-06,
+      "loss": 0.9331,
+      "step": 133
+    },
+    {
+      "epoch": 0.28450106157112526,
+      "grad_norm": 0.6580376975505629,
+      "learning_rate": 4.301601125982246e-06,
+      "loss": 1.0069,
+      "step": 134
+    },
+    {
+      "epoch": 0.28662420382165604,
+      "grad_norm": 0.7094205388987409,
+      "learning_rate": 4.289374995230942e-06,
+      "loss": 1.0132,
+      "step": 135
+    },
+    {
+      "epoch": 0.28874734607218683,
+      "grad_norm": 0.7347016248235378,
+      "learning_rate": 4.277060478317687e-06,
+      "loss": 1.0009,
+      "step": 136
+    },
+    {
+      "epoch": 0.2908704883227176,
+      "grad_norm": 0.6790220682871185,
+      "learning_rate": 4.264658183517935e-06,
+      "loss": 0.9377,
+      "step": 137
+    },
+    {
+      "epoch": 0.2929936305732484,
+      "grad_norm": 0.6615615208650163,
+      "learning_rate": 4.252168723442927e-06,
+      "loss": 0.9493,
+      "step": 138
+    },
+    {
+      "epoch": 0.2951167728237792,
+      "grad_norm": 0.690676805692251,
+      "learning_rate": 4.23959271500943e-06,
+      "loss": 0.9369,
+      "step": 139
+    },
+    {
+      "epoch": 0.29723991507430997,
+      "grad_norm": 0.7025773548182347,
+      "learning_rate": 4.226930779409271e-06,
+      "loss": 0.9983,
+      "step": 140
+    },
+    {
+      "epoch": 0.29936305732484075,
+      "grad_norm": 0.7730748791240522,
+      "learning_rate": 4.214183542078646e-06,
+      "loss": 0.9393,
+      "step": 141
+    },
+    {
+      "epoch": 0.30148619957537154,
+      "grad_norm": 0.6778294406818572,
+      "learning_rate": 4.201351632667227e-06,
+      "loss": 1.0114,
+      "step": 142
+    },
+    {
+      "epoch": 0.3036093418259023,
+      "grad_norm": 0.6852307950573668,
+      "learning_rate": 4.1884356850070695e-06,
+      "loss": 1.0313,
+      "step": 143
+    },
+    {
+      "epoch": 0.3057324840764331,
+      "grad_norm": 0.7191284149312358,
+      "learning_rate": 4.175436337081289e-06,
+      "loss": 0.9071,
+      "step": 144
+    },
+    {
+      "epoch": 0.3078556263269639,
+      "grad_norm": 0.7394267321034559,
+      "learning_rate": 4.162354230992562e-06,
+      "loss": 0.9757,
+      "step": 145
+    },
+    {
+      "epoch": 0.3099787685774947,
+      "grad_norm": 0.711234245522065,
+      "learning_rate": 4.149190012931402e-06,
+      "loss": 0.9233,
+      "step": 146
+    },
+    {
+      "epoch": 0.31210191082802546,
+      "grad_norm": 0.7677511242459594,
+      "learning_rate": 4.135944333144244e-06,
+      "loss": 0.9492,
+      "step": 147
+    },
+    {
+      "epoch": 0.31422505307855625,
+      "grad_norm": 0.6497634346049875,
+      "learning_rate": 4.122617845901322e-06,
+      "loss": 1.0376,
+      "step": 148
+    },
+    {
+      "epoch": 0.31634819532908703,
+      "grad_norm": 0.6757359209440288,
+      "learning_rate": 4.109211209464354e-06,
+      "loss": 1.0303,
+      "step": 149
+    },
+    {
+      "epoch": 0.3184713375796178,
+      "grad_norm": 0.6695087962535804,
+      "learning_rate": 4.095725086054029e-06,
+      "loss": 0.956,
+      "step": 150
+    },
+    {
+      "epoch": 0.3205944798301486,
+      "grad_norm": 0.6942775399870352,
+      "learning_rate": 4.0821601418172926e-06,
+      "loss": 0.9945,
+      "step": 151
+    },
+    {
+      "epoch": 0.3227176220806794,
+      "grad_norm": 0.6878597656985246,
+      "learning_rate": 4.068517046794443e-06,
+      "loss": 0.9921,
+      "step": 152
+    },
+    {
+      "epoch": 0.3248407643312102,
+      "grad_norm": 0.7171780998162959,
+      "learning_rate": 4.054796474886038e-06,
+      "loss": 1.0233,
+      "step": 153
+    },
+    {
+      "epoch": 0.32696390658174096,
+      "grad_norm": 0.6763242560453191,
+      "learning_rate": 4.040999103819606e-06,
+      "loss": 1.0145,
+      "step": 154
+    },
+    {
+      "epoch": 0.32908704883227174,
+      "grad_norm": 0.6295779748002456,
+      "learning_rate": 4.0271256151161666e-06,
+      "loss": 1.0073,
+      "step": 155
+    },
+    {
+      "epoch": 0.33121019108280253,
+      "grad_norm": 0.7389589402690067,
+      "learning_rate": 4.013176694056571e-06,
+      "loss": 0.9892,
+      "step": 156
+    },
+    {
+      "epoch": 0.3333333333333333,
+      "grad_norm": 0.6655878627704468,
+      "learning_rate": 3.999153029647651e-06,
+      "loss": 1.0411,
+      "step": 157
+    },
+    {
+      "epoch": 0.3354564755838641,
+      "grad_norm": 0.6375860922410523,
+      "learning_rate": 3.985055314588185e-06,
+      "loss": 0.9458,
+      "step": 158
+    },
+    {
+      "epoch": 0.3375796178343949,
+      "grad_norm": 0.7839584481038233,
+      "learning_rate": 3.970884245234684e-06,
+      "loss": 1.0261,
+      "step": 159
+    },
+    {
+      "epoch": 0.33970276008492567,
+      "grad_norm": 0.7385426113864771,
+      "learning_rate": 3.956640521566989e-06,
+      "loss": 0.9587,
+      "step": 160
+    },
+    {
+      "epoch": 0.34182590233545646,
+      "grad_norm": 0.663828302901297,
+      "learning_rate": 3.942324847153706e-06,
+      "loss": 0.9254,
+      "step": 161
+    },
+    {
+      "epoch": 0.34394904458598724,
+      "grad_norm": 0.6883166230000658,
+      "learning_rate": 3.927937929117447e-06,
+      "loss": 0.967,
+      "step": 162
+    },
+    {
+      "epoch": 0.346072186836518,
+      "grad_norm": 0.7082555205500026,
+      "learning_rate": 3.913480478099898e-06,
+      "loss": 0.9725,
+      "step": 163
+    },
+    {
+      "epoch": 0.3481953290870488,
+      "grad_norm": 0.7642622554393386,
+      "learning_rate": 3.898953208226723e-06,
+      "loss": 1.014,
+      "step": 164
+    },
+    {
+      "epoch": 0.3503184713375796,
+      "grad_norm": 0.6590730477383634,
+      "learning_rate": 3.884356837072289e-06,
+      "loss": 1.0197,
+      "step": 165
+    },
+    {
+      "epoch": 0.3524416135881104,
+      "grad_norm": 0.6482468015769196,
+      "learning_rate": 3.869692085624218e-06,
+      "loss": 0.9289,
+      "step": 166
+    },
+    {
+      "epoch": 0.35456475583864117,
+      "grad_norm": 0.7139224901487669,
+      "learning_rate": 3.8549596782477775e-06,
+      "loss": 0.9623,
+      "step": 167
+    },
+    {
+      "epoch": 0.35668789808917195,
+      "grad_norm": 0.6395897912297176,
+      "learning_rate": 3.8401603426501e-06,
+      "loss": 1.0118,
+      "step": 168
+    },
+    {
+      "epoch": 0.35881104033970274,
+      "grad_norm": 0.6867653353820191,
+      "learning_rate": 3.825294809844234e-06,
+      "loss": 0.9839,
+      "step": 169
+    },
+    {
+      "epoch": 0.3609341825902335,
+      "grad_norm": 0.7001331824072736,
+      "learning_rate": 3.8103638141130406e-06,
+      "loss": 0.9182,
+      "step": 170
+    },
+    {
+      "epoch": 0.3630573248407643,
+      "grad_norm": 0.6522485155963109,
+      "learning_rate": 3.7953680929729215e-06,
+      "loss": 0.9555,
+      "step": 171
+    },
+    {
+      "epoch": 0.3651804670912951,
+      "grad_norm": 0.6985843893126511,
+      "learning_rate": 3.7803083871373876e-06,
+      "loss": 0.944,
+      "step": 172
+    },
+    {
+      "epoch": 0.3673036093418259,
+      "grad_norm": 0.7007945690155835,
+      "learning_rate": 3.7651854404804757e-06,
+      "loss": 0.9399,
+      "step": 173
+    },
+    {
+      "epoch": 0.36942675159235666,
+      "grad_norm": 0.6780715813196918,
+      "learning_rate": 3.7500000000000005e-06,
+      "loss": 0.9814,
+      "step": 174
+    },
+    {
+      "epoch": 0.37154989384288745,
+      "grad_norm": 0.6641562357723244,
+      "learning_rate": 3.734752815780659e-06,
+      "loss": 0.9523,
+      "step": 175
+    },
+    {
+      "epoch": 0.37367303609341823,
+      "grad_norm": 0.6906594521754901,
+      "learning_rate": 3.7194446409569813e-06,
+      "loss": 0.9904,
+      "step": 176
+    },
+    {
+      "epoch": 0.37579617834394907,
+      "grad_norm": 0.7593873633214692,
+      "learning_rate": 3.7040762316761247e-06,
+      "loss": 0.9489,
+      "step": 177
+    },
+    {
+      "epoch": 0.37791932059447986,
+      "grad_norm": 0.6770564421900779,
+      "learning_rate": 3.6886483470605293e-06,
+      "loss": 0.9897,
+      "step": 178
+    },
+    {
+      "epoch": 0.38004246284501064,
+      "grad_norm": 0.7648650056676896,
+      "learning_rate": 3.6731617491704195e-06,
+      "loss": 0.9292,
+      "step": 179
+    },
+    {
+      "epoch": 0.3821656050955414,
+      "grad_norm": 0.7222877847544774,
+      "learning_rate": 3.6576172029661583e-06,
+      "loss": 1.0195,
+      "step": 180
+    },
+    {
+      "epoch": 0.3842887473460722,
+      "grad_norm": 0.6910668492827662,
+      "learning_rate": 3.6420154762704685e-06,
+      "loss": 0.9875,
+      "step": 181
+    },
+    {
+      "epoch": 0.386411889596603,
+      "grad_norm": 0.6741685399464014,
+      "learning_rate": 3.6263573397305003e-06,
+      "loss": 1.0038,
+      "step": 182
+    },
+    {
+      "epoch": 0.3885350318471338,
+      "grad_norm": 0.7033531290025623,
+      "learning_rate": 3.6106435667797684e-06,
+      "loss": 0.9411,
+      "step": 183
+    },
+    {
+      "epoch": 0.39065817409766457,
+      "grad_norm": 0.6890574251939826,
+      "learning_rate": 3.5948749335999493e-06,
+      "loss": 0.92,
+      "step": 184
+    },
+    {
+      "epoch": 0.39278131634819535,
+      "grad_norm": 0.7466250540457848,
+      "learning_rate": 3.5790522190825368e-06,
+      "loss": 0.9907,
+      "step": 185
+    },
+    {
+      "epoch": 0.39490445859872614,
+      "grad_norm": 0.7277438842625494,
+      "learning_rate": 3.5631762047903746e-06,
+      "loss": 1.024,
+      "step": 186
+    },
+    {
+      "epoch": 0.3970276008492569,
+      "grad_norm": 0.6333155312409119,
+      "learning_rate": 3.5472476749190465e-06,
+      "loss": 1.0138,
+      "step": 187
+    },
+    {
+      "epoch": 0.3991507430997877,
+      "grad_norm": 0.8297340505830674,
+      "learning_rate": 3.531267416258144e-06,
+      "loss": 1.0227,
+      "step": 188
+    },
+    {
+      "epoch": 0.4012738853503185,
+      "grad_norm": 0.7089430122521235,
+      "learning_rate": 3.5152362181524013e-06,
+      "loss": 0.9357,
+      "step": 189
+    },
+    {
+      "epoch": 0.4033970276008493,
+      "grad_norm": 0.6823030263348598,
+      "learning_rate": 3.4991548724627054e-06,
+      "loss": 0.8842,
+      "step": 190
+    },
+    {
+      "epoch": 0.40552016985138006,
+      "grad_norm": 0.6912404603041468,
+      "learning_rate": 3.4830241735269852e-06,
+      "loss": 0.945,
+      "step": 191
+    },
+    {
+      "epoch": 0.40764331210191085,
+      "grad_norm": 0.7115082118821694,
+      "learning_rate": 3.4668449181209695e-06,
+      "loss": 0.9493,
+      "step": 192
+    },
+    {
+      "epoch": 0.40976645435244163,
+      "grad_norm": 0.6252959058347625,
+      "learning_rate": 3.450617905418834e-06,
+      "loss": 0.9824,
+      "step": 193
+    },
+    {
+      "epoch": 0.4118895966029724,
+      "grad_norm": 0.6803790151109083,
+      "learning_rate": 3.434343936953729e-06,
+      "loss": 0.9668,
+      "step": 194
+    },
+    {
+      "epoch": 0.4140127388535032,
+      "grad_norm": 0.638640242080679,
+      "learning_rate": 3.4180238165781806e-06,
+      "loss": 0.9268,
+      "step": 195
+    },
+    {
+      "epoch": 0.416135881104034,
+      "grad_norm": 0.6837218466024749,
+      "learning_rate": 3.4016583504243892e-06,
+      "loss": 0.917,
+      "step": 196
+    },
+    {
+      "epoch": 0.4182590233545648,
+      "grad_norm": 0.6427086547693028,
+      "learning_rate": 3.38524834686441e-06,
+      "loss": 0.9782,
+      "step": 197
+    },
+    {
+      "epoch": 0.42038216560509556,
+      "grad_norm": 0.6683767385691209,
+      "learning_rate": 3.3687946164702225e-06,
+      "loss": 1.0093,
+      "step": 198
+    },
+    {
+      "epoch": 0.42250530785562634,
+      "grad_norm": 0.6553328409851518,
+      "learning_rate": 3.3522979719736923e-06,
+      "loss": 0.9418,
+      "step": 199
+    },
+    {
+      "epoch": 0.42462845010615713,
+      "grad_norm": 0.6778853484731171,
+      "learning_rate": 3.3357592282264295e-06,
+      "loss": 0.9607,
+      "step": 200
+    },
+    {
+      "epoch": 0.4267515923566879,
+      "grad_norm": 0.6516881094720604,
+      "learning_rate": 3.319179202159532e-06,
+      "loss": 0.9789,
+      "step": 201
+    },
+    {
+      "epoch": 0.4288747346072187,
+      "grad_norm": 0.6404550926711216,
+      "learning_rate": 3.3025587127432414e-06,
+      "loss": 0.9735,
+      "step": 202
+    },
+    {
+      "epoch": 0.4309978768577495,
+      "grad_norm": 0.6074907430474463,
+      "learning_rate": 3.2858985809464823e-06,
+      "loss": 0.929,
+      "step": 203
+    },
+    {
+      "epoch": 0.43312101910828027,
+      "grad_norm": 0.6754111542436511,
+      "learning_rate": 3.2691996296963185e-06,
+      "loss": 0.9613,
+      "step": 204
+    },
+    {
+      "epoch": 0.43524416135881105,
+      "grad_norm": 0.6093435246887753,
+      "learning_rate": 3.2524626838372976e-06,
+      "loss": 1.0417,
+      "step": 205
+    },
+    {
+      "epoch": 0.43736730360934184,
+      "grad_norm": 0.6463846637401074,
+      "learning_rate": 3.2356885700907096e-06,
+      "loss": 0.9503,
+      "step": 206
+    },
+    {
+      "epoch": 0.4394904458598726,
+      "grad_norm": 0.7058617139491341,
+      "learning_rate": 3.218878117013756e-06,
+      "loss": 0.9583,
+      "step": 207
+    },
+    {
+      "epoch": 0.4416135881104034,
+      "grad_norm": 0.8080013671095668,
+      "learning_rate": 3.202032154958615e-06,
+      "loss": 0.9207,
+      "step": 208
+    },
+    {
+      "epoch": 0.4437367303609342,
+      "grad_norm": 0.646008011251542,
+      "learning_rate": 3.1851515160314346e-06,
+      "loss": 0.9202,
+      "step": 209
+    },
+    {
+      "epoch": 0.445859872611465,
+      "grad_norm": 0.6526157991983537,
+      "learning_rate": 3.168237034051222e-06,
+      "loss": 0.983,
+      "step": 210
+    },
+    {
+      "epoch": 0.44798301486199577,
+      "grad_norm": 0.6820529024780707,
+      "learning_rate": 3.151289544508664e-06,
+      "loss": 0.9554,
+      "step": 211
+    },
+    {
+      "epoch": 0.45010615711252655,
+      "grad_norm": 0.6888098712935105,
+      "learning_rate": 3.1343098845248564e-06,
+      "loss": 0.9312,
+      "step": 212
+    },
+    {
+      "epoch": 0.45222929936305734,
+      "grad_norm": 0.7112371797389818,
+      "learning_rate": 3.117298892809953e-06,
+      "loss": 0.9319,
+      "step": 213
+    },
+    {
+      "epoch": 0.4543524416135881,
+      "grad_norm": 0.6892727118283283,
+      "learning_rate": 3.100257409621738e-06,
+      "loss": 0.9165,
+      "step": 214
+    },
+    {
+      "epoch": 0.4564755838641189,
+      "grad_norm": 0.6882279635622599,
+      "learning_rate": 3.0831862767241206e-06,
+      "loss": 0.9258,
+      "step": 215
+    },
+    {
+      "epoch": 0.4585987261146497,
+      "grad_norm": 0.7044667471867824,
+      "learning_rate": 3.0660863373455594e-06,
+      "loss": 0.9744,
+      "step": 216
+    },
+    {
+      "epoch": 0.4607218683651805,
+      "grad_norm": 0.6748729735766027,
+      "learning_rate": 3.0489584361374073e-06,
+      "loss": 1.0145,
+      "step": 217
+    },
+    {
+      "epoch": 0.46284501061571126,
+      "grad_norm": 0.6013878264966025,
+      "learning_rate": 3.031803419132192e-06,
+      "loss": 0.9491,
+      "step": 218
+    },
+    {
+      "epoch": 0.46496815286624205,
+      "grad_norm": 0.6154706328562357,
+      "learning_rate": 3.014622133701826e-06,
+      "loss": 0.9307,
+      "step": 219
+    },
+    {
+      "epoch": 0.46709129511677283,
+      "grad_norm": 0.6288716332754415,
+      "learning_rate": 2.99741542851575e-06,
+      "loss": 0.95,
+      "step": 220
+    },
+    {
+      "epoch": 0.4692144373673036,
+      "grad_norm": 0.6406745016032289,
+      "learning_rate": 2.9801841534990115e-06,
+      "loss": 1.0193,
+      "step": 221
+    },
+    {
+      "epoch": 0.4713375796178344,
+      "grad_norm": 0.627509712489892,
+      "learning_rate": 2.96292915979029e-06,
+      "loss": 0.8955,
+      "step": 222
+    },
+    {
+      "epoch": 0.4734607218683652,
+      "grad_norm": 0.6829084713106287,
+      "learning_rate": 2.945651299699843e-06,
+      "loss": 0.9282,
+      "step": 223
+    },
+    {
+      "epoch": 0.47558386411889597,
+      "grad_norm": 0.6619023830231336,
+      "learning_rate": 2.9283514266674166e-06,
+      "loss": 1.0037,
+      "step": 224
+    },
+    {
+      "epoch": 0.47770700636942676,
+      "grad_norm": 0.6649971682870974,
+      "learning_rate": 2.9110303952200863e-06,
+      "loss": 0.9338,
+      "step": 225
+    },
+    {
+      "epoch": 0.47983014861995754,
+      "grad_norm": 0.691407091915056,
+      "learning_rate": 2.893689060930045e-06,
+      "loss": 0.9325,
+      "step": 226
+    },
+    {
+      "epoch": 0.4819532908704883,
+      "grad_norm": 0.6765826616693394,
+      "learning_rate": 2.876328280372346e-06,
+      "loss": 0.9666,
+      "step": 227
+    },
+    {
+      "epoch": 0.4840764331210191,
+      "grad_norm": 0.659756365707493,
+      "learning_rate": 2.8589489110825895e-06,
+      "loss": 0.9654,
+      "step": 228
+    },
+    {
+      "epoch": 0.4861995753715499,
+      "grad_norm": 0.7460102157125611,
+      "learning_rate": 2.8415518115145673e-06,
+      "loss": 0.9861,
+      "step": 229
+    },
+    {
+      "epoch": 0.4883227176220807,
+      "grad_norm": 0.6433037973844256,
+      "learning_rate": 2.824137840997858e-06,
+      "loss": 0.9584,
+      "step": 230
+    },
+    {
+      "epoch": 0.49044585987261147,
+      "grad_norm": 0.6856232642287533,
+      "learning_rate": 2.8067078596953793e-06,
+      "loss": 0.9164,
+      "step": 231
+    },
+    {
+      "epoch": 0.49256900212314225,
+      "grad_norm": 0.7270361036995455,
+      "learning_rate": 2.7892627285609035e-06,
+      "loss": 1.0485,
+      "step": 232
+    },
+    {
+      "epoch": 0.49469214437367304,
+      "grad_norm": 0.6968320986232821,
+      "learning_rate": 2.7718033092965267e-06,
+      "loss": 0.9211,
+      "step": 233
+    },
+    {
+      "epoch": 0.4968152866242038,
+      "grad_norm": 0.7067362359142023,
+      "learning_rate": 2.754330464310108e-06,
+      "loss": 0.9047,
+      "step": 234
+    },
+    {
+      "epoch": 0.4989384288747346,
+      "grad_norm": 0.7611967162098211,
+      "learning_rate": 2.7368450566726714e-06,
+      "loss": 0.9708,
+      "step": 235
+    },
+    {
+      "epoch": 0.5010615711252654,
+      "grad_norm": 0.6940560339648915,
+      "learning_rate": 2.7193479500757684e-06,
+      "loss": 0.8726,
+      "step": 236
+    },
+    {
+      "epoch": 0.5031847133757962,
+      "grad_norm": 0.7870025525904728,
+      "learning_rate": 2.7018400087888265e-06,
+      "loss": 0.9617,
+      "step": 237
+    },
+    {
+      "epoch": 0.505307855626327,
+      "grad_norm": 0.6559672876844489,
+      "learning_rate": 2.684322097616448e-06,
+      "loss": 0.9901,
+      "step": 238
+    },
+    {
+      "epoch": 0.5074309978768577,
+      "grad_norm": 0.7529609642384081,
+      "learning_rate": 2.6667950818556992e-06,
+      "loss": 0.9333,
+      "step": 239
+    },
+    {
+      "epoch": 0.5095541401273885,
+      "grad_norm": 0.652243500501896,
+      "learning_rate": 2.649259827253368e-06,
+      "loss": 1.0095,
+      "step": 240
+    },
+    {
+      "epoch": 0.5116772823779193,
+      "grad_norm": 0.660107946462241,
+      "learning_rate": 2.6317171999631992e-06,
+      "loss": 0.9448,
+      "step": 241
+    },
+    {
+      "epoch": 0.5138004246284501,
+      "grad_norm": 0.7440712860315413,
+      "learning_rate": 2.6141680665031116e-06,
+      "loss": 0.9075,
+      "step": 242
+    },
+    {
+      "epoch": 0.5159235668789809,
+      "grad_norm": 0.6624899509120282,
+      "learning_rate": 2.5966132937123963e-06,
+      "loss": 0.9143,
+      "step": 243
+    },
+    {
+      "epoch": 0.5180467091295117,
+      "grad_norm": 0.6587688885910037,
+      "learning_rate": 2.5790537487088975e-06,
+      "loss": 0.9923,
+      "step": 244
+    },
+    {
+      "epoch": 0.5201698513800425,
+      "grad_norm": 0.6302763737309945,
+      "learning_rate": 2.561490298846186e-06,
+      "loss": 1.0052,
+      "step": 245
+    },
+    {
+      "epoch": 0.5222929936305732,
+      "grad_norm": 0.6810085520210082,
+      "learning_rate": 2.5439238116707103e-06,
+      "loss": 0.9866,
+      "step": 246
+    },
+    {
+      "epoch": 0.524416135881104,
+      "grad_norm": 0.7140669070578711,
+      "learning_rate": 2.5263551548789494e-06,
+      "loss": 0.9656,
+      "step": 247
+    },
+    {
+      "epoch": 0.5265392781316348,
+      "grad_norm": 0.6289851811042011,
+      "learning_rate": 2.5087851962745468e-06,
+      "loss": 0.9364,
+      "step": 248
+    },
+    {
+      "epoch": 0.5286624203821656,
+      "grad_norm": 0.667727375732809,
+      "learning_rate": 2.4912148037254536e-06,
+      "loss": 0.9748,
+      "step": 249
+    },
+    {
+      "epoch": 0.5307855626326964,
+      "grad_norm": 0.7059493512489892,
+      "learning_rate": 2.4736448451210514e-06,
+      "loss": 0.9876,
+      "step": 250
+    },
+    {
+      "epoch": 0.5329087048832272,
+      "grad_norm": 0.7422236381441959,
+      "learning_rate": 2.45607618832929e-06,
+      "loss": 0.9398,
+      "step": 251
+    },
+    {
+      "epoch": 0.535031847133758,
+      "grad_norm": 0.7567396838017741,
+      "learning_rate": 2.4385097011538144e-06,
+      "loss": 0.9479,
+      "step": 252
+    },
+    {
+      "epoch": 0.5371549893842887,
+      "grad_norm": 0.6422364949883526,
+      "learning_rate": 2.420946251291103e-06,
+      "loss": 0.9708,
+      "step": 253
+    },
+    {
+      "epoch": 0.5392781316348195,
+      "grad_norm": 0.6121337470685102,
+      "learning_rate": 2.4033867062876054e-06,
+      "loss": 1.0116,
+      "step": 254
+    },
+    {
+      "epoch": 0.5414012738853503,
+      "grad_norm": 0.6625230272994517,
+      "learning_rate": 2.3858319334968892e-06,
+      "loss": 0.9333,
+      "step": 255
+    },
+    {
+      "epoch": 0.5435244161358811,
+      "grad_norm": 0.6529959967773957,
+      "learning_rate": 2.368282800036801e-06,
+      "loss": 0.9262,
+      "step": 256
+    },
+    {
+      "epoch": 0.5456475583864119,
+      "grad_norm": 0.706423713198206,
+      "learning_rate": 2.350740172746633e-06,
+      "loss": 0.9357,
+      "step": 257
+    },
+    {
+      "epoch": 0.5477707006369427,
+      "grad_norm": 0.6987077360671913,
+      "learning_rate": 2.3332049181443016e-06,
+      "loss": 0.9756,
+      "step": 258
+    },
+    {
+      "epoch": 0.5498938428874734,
+      "grad_norm": 0.7052371639528945,
+      "learning_rate": 2.315677902383553e-06,
+      "loss": 0.9487,
+      "step": 259
+    },
+    {
+      "epoch": 0.5520169851380042,
+      "grad_norm": 0.6250495885700643,
+      "learning_rate": 2.298159991211174e-06,
+      "loss": 0.9192,
+      "step": 260
+    },
+    {
+      "epoch": 0.554140127388535,
+      "grad_norm": 0.6982497766833491,
+      "learning_rate": 2.280652049924232e-06,
+      "loss": 0.9609,
+      "step": 261
+    },
+    {
+      "epoch": 0.5562632696390658,
+      "grad_norm": 0.6295040315340362,
+      "learning_rate": 2.2631549433273294e-06,
+      "loss": 1.0085,
+      "step": 262
+    },
+    {
+      "epoch": 0.5583864118895966,
+      "grad_norm": 0.6644616810444942,
+      "learning_rate": 2.2456695356898915e-06,
+      "loss": 0.8849,
+      "step": 263
+    },
+    {
+      "epoch": 0.5605095541401274,
+      "grad_norm": 0.663969598277379,
+      "learning_rate": 2.2281966907034737e-06,
+      "loss": 0.9424,
+      "step": 264
+    },
+    {
+      "epoch": 0.5626326963906582,
+      "grad_norm": 0.6693955418858921,
+      "learning_rate": 2.2107372714390973e-06,
+      "loss": 0.9954,
+      "step": 265
+    },
+    {
+      "epoch": 0.564755838641189,
+      "grad_norm": 0.7021462467608284,
+      "learning_rate": 2.193292140304621e-06,
+      "loss": 0.9928,
+      "step": 266
+    },
+    {
+      "epoch": 0.5668789808917197,
+      "grad_norm": 0.6692328908805102,
+      "learning_rate": 2.175862159002143e-06,
+      "loss": 0.9524,
+      "step": 267
+    },
+    {
+      "epoch": 0.5690021231422505,
+      "grad_norm": 0.6391930894120555,
+      "learning_rate": 2.158448188485433e-06,
+      "loss": 1.0429,
+      "step": 268
+    },
+    {
+      "epoch": 0.5711252653927813,
+      "grad_norm": 0.634093794488882,
+      "learning_rate": 2.141051088917411e-06,
+      "loss": 0.9421,
+      "step": 269
+    },
+    {
+      "epoch": 0.5732484076433121,
+      "grad_norm": 0.61983995227008,
+      "learning_rate": 2.1236717196276558e-06,
+      "loss": 0.9541,
+      "step": 270
+    },
+    {
+      "epoch": 0.5753715498938429,
+      "grad_norm": 0.6086341907488318,
+      "learning_rate": 2.106310939069956e-06,
+      "loss": 0.9437,
+      "step": 271
+    },
+    {
+      "epoch": 0.5774946921443737,
+      "grad_norm": 0.6658683338126193,
+      "learning_rate": 2.088969604779914e-06,
+      "loss": 0.9992,
+      "step": 272
+    },
+    {
+      "epoch": 0.5796178343949044,
+      "grad_norm": 0.6650210773963607,
+      "learning_rate": 2.0716485733325834e-06,
+      "loss": 0.9187,
+      "step": 273
+    },
+    {
+      "epoch": 0.5817409766454352,
+      "grad_norm": 0.6582276099759867,
+      "learning_rate": 2.054348700300158e-06,
+      "loss": 0.9959,
+      "step": 274
+    },
+    {
+      "epoch": 0.583864118895966,
+      "grad_norm": 0.6730751436716349,
+      "learning_rate": 2.037070840209711e-06,
+      "loss": 1.0034,
+      "step": 275
+    },
+    {
+      "epoch": 0.5859872611464968,
+      "grad_norm": 0.7084245464531143,
+      "learning_rate": 2.019815846500988e-06,
+      "loss": 0.9775,
+      "step": 276
+    },
+    {
+      "epoch": 0.5881104033970276,
+      "grad_norm": 0.6518940026744457,
+      "learning_rate": 2.0025845714842514e-06,
+      "loss": 0.9474,
+      "step": 277
+    },
+    {
+      "epoch": 0.5902335456475584,
+      "grad_norm": 0.6808230009128965,
+      "learning_rate": 1.9853778662981747e-06,
+      "loss": 0.9451,
+      "step": 278
+    },
+    {
+      "epoch": 0.5923566878980892,
+      "grad_norm": 0.6902948572901826,
+      "learning_rate": 1.968196580867808e-06,
+      "loss": 0.9856,
+      "step": 279
+    },
+    {
+      "epoch": 0.5944798301486199,
+      "grad_norm": 0.611810608377548,
+      "learning_rate": 1.951041563862593e-06,
+      "loss": 1.0013,
+      "step": 280
+    },
+    {
+      "epoch": 0.5966029723991507,
+      "grad_norm": 0.6954721611774066,
+      "learning_rate": 1.933913662654441e-06,
+      "loss": 0.9897,
+      "step": 281
+    },
+    {
+      "epoch": 0.5987261146496815,
+      "grad_norm": 0.6406168524428356,
+      "learning_rate": 1.91681372327588e-06,
+      "loss": 0.919,
+      "step": 282
+    },
+    {
+      "epoch": 0.6008492569002123,
+      "grad_norm": 0.6704361726473957,
+      "learning_rate": 1.899742590378263e-06,
+      "loss": 0.9811,
+      "step": 283
+    },
+    {
+      "epoch": 0.6029723991507431,
+      "grad_norm": 0.6778804788413157,
+      "learning_rate": 1.8827011071900476e-06,
+      "loss": 0.9735,
+      "step": 284
+    },
+    {
+      "epoch": 0.6050955414012739,
+      "grad_norm": 0.7177755805617885,
+      "learning_rate": 1.865690115475144e-06,
+      "loss": 0.9925,
+      "step": 285
+    },
+    {
+      "epoch": 0.6072186836518046,
+      "grad_norm": 0.6894390591882673,
+      "learning_rate": 1.8487104554913362e-06,
+      "loss": 0.9305,
+      "step": 286
+    },
+    {
+      "epoch": 0.6093418259023354,
+      "grad_norm": 0.6116905753834903,
+      "learning_rate": 1.8317629659487796e-06,
+      "loss": 0.9274,
+      "step": 287
+    },
+    {
+      "epoch": 0.6114649681528662,
+      "grad_norm": 0.6670439904291463,
+      "learning_rate": 1.8148484839685663e-06,
+      "loss": 0.9729,
+      "step": 288
+    },
+    {
+      "epoch": 0.613588110403397,
+      "grad_norm": 0.694254062629121,
+      "learning_rate": 1.7979678450413845e-06,
+      "loss": 0.9044,
+      "step": 289
+    },
+    {
+      "epoch": 0.6157112526539278,
+      "grad_norm": 0.6159818959254523,
+      "learning_rate": 1.781121882986245e-06,
+      "loss": 0.9642,
+      "step": 290
+    },
+    {
+      "epoch": 0.6178343949044586,
+      "grad_norm": 0.660521989314923,
+      "learning_rate": 1.764311429909291e-06,
+      "loss": 0.9684,
+      "step": 291
+    },
+    {
+      "epoch": 0.6199575371549894,
+      "grad_norm": 0.6190727862079969,
+      "learning_rate": 1.7475373161627034e-06,
+      "loss": 0.98,
+      "step": 292
+    },
+    {
+      "epoch": 0.6220806794055201,
+      "grad_norm": 0.6684171243946048,
+      "learning_rate": 1.730800370303683e-06,
+      "loss": 0.9953,
+      "step": 293
+    },
+    {
+      "epoch": 0.6242038216560509,
+      "grad_norm": 0.6655271015347415,
+      "learning_rate": 1.7141014190535183e-06,
+      "loss": 0.8907,
+      "step": 294
+    },
+    {
+      "epoch": 0.6263269639065817,
+      "grad_norm": 0.6631683337776253,
+      "learning_rate": 1.6974412872567598e-06,
+      "loss": 0.9509,
+      "step": 295
+    },
+    {
+      "epoch": 0.6284501061571125,
+      "grad_norm": 0.7133500119714207,
+      "learning_rate": 1.6808207978404685e-06,
+      "loss": 0.9914,
+      "step": 296
+    },
+    {
+      "epoch": 0.6305732484076433,
+      "grad_norm": 0.6848781673615809,
+      "learning_rate": 1.6642407717735711e-06,
+      "loss": 0.9732,
+      "step": 297
+    },
+    {
+      "epoch": 0.6326963906581741,
+      "grad_norm": 0.6309522450595455,
+      "learning_rate": 1.647702028026308e-06,
+      "loss": 0.9368,
+      "step": 298
+    },
+    {
+      "epoch": 0.6348195329087049,
+      "grad_norm": 0.6178347487450343,
+      "learning_rate": 1.6312053835297783e-06,
+      "loss": 0.983,
+      "step": 299
+    },
+    {
+      "epoch": 0.6369426751592356,
+      "grad_norm": 0.6737246544973767,
+      "learning_rate": 1.6147516531355912e-06,
+      "loss": 0.9203,
+      "step": 300
+    },
+    {
+      "epoch": 0.6390658174097664,
+      "grad_norm": 0.675352006077776,
+      "learning_rate": 1.5983416495756116e-06,
+      "loss": 0.9971,
+      "step": 301
+    },
+    {
+      "epoch": 0.6411889596602972,
+      "grad_norm": 0.6402934188286067,
+      "learning_rate": 1.58197618342182e-06,
+      "loss": 0.9614,
+      "step": 302
+    },
+    {
+      "epoch": 0.643312101910828,
+      "grad_norm": 0.7184921383563952,
+      "learning_rate": 1.565656063046272e-06,
+      "loss": 0.9516,
+      "step": 303
+    },
+    {
+      "epoch": 0.6454352441613588,
+      "grad_norm": 0.6704718578724861,
+      "learning_rate": 1.549382094581166e-06,
+      "loss": 0.9153,
+      "step": 304
+    },
+    {
+      "epoch": 0.6475583864118896,
+      "grad_norm": 0.6636811286993395,
+      "learning_rate": 1.5331550818790313e-06,
+      "loss": 0.9361,
+      "step": 305
+    },
+    {
+      "epoch": 0.6496815286624203,
+      "grad_norm": 0.6355031775992976,
+      "learning_rate": 1.516975826473015e-06,
+      "loss": 0.9661,
+      "step": 306
+    },
+    {
+      "epoch": 0.6518046709129511,
+      "grad_norm": 0.6864068487323046,
+      "learning_rate": 1.5008451275372948e-06,
+      "loss": 0.9647,
+      "step": 307
+    },
+    {
+      "epoch": 0.6539278131634819,
+      "grad_norm": 0.6194120153211208,
+      "learning_rate": 1.4847637818475992e-06,
+      "loss": 0.9683,
+      "step": 308
+    },
+    {
+      "epoch": 0.6560509554140127,
+      "grad_norm": 0.6364599124064249,
+      "learning_rate": 1.4687325837418564e-06,
+      "loss": 1.0339,
+      "step": 309
+    },
+    {
+      "epoch": 0.6581740976645435,
+      "grad_norm": 0.6254187950093356,
+      "learning_rate": 1.4527523250809544e-06,
+      "loss": 0.991,
+      "step": 310
+    },
+    {
+      "epoch": 0.6602972399150743,
+      "grad_norm": 0.7142238718221583,
+      "learning_rate": 1.4368237952096258e-06,
+      "loss": 0.9417,
+      "step": 311
+    },
+    {
+      "epoch": 0.6624203821656051,
+      "grad_norm": 0.6888278924765849,
+      "learning_rate": 1.4209477809174634e-06,
+      "loss": 0.9123,
+      "step": 312
+    },
+    {
+      "epoch": 0.6645435244161358,
+      "grad_norm": 0.7263883196119595,
+      "learning_rate": 1.4051250664000515e-06,
+      "loss": 0.9574,
+      "step": 313
+    },
+    {
+      "epoch": 0.6666666666666666,
+      "grad_norm": 0.6287509757808388,
+      "learning_rate": 1.389356433220232e-06,
+      "loss": 0.9763,
+      "step": 314
+    },
+    {
+      "epoch": 0.6687898089171974,
+      "grad_norm": 0.6255170909323848,
+      "learning_rate": 1.3736426602694999e-06,
+      "loss": 0.9185,
+      "step": 315
+    },
+    {
+      "epoch": 0.6709129511677282,
+      "grad_norm": 0.6159840521359649,
+      "learning_rate": 1.357984523729533e-06,
+      "loss": 0.9936,
+      "step": 316
+    },
+    {
+      "epoch": 0.673036093418259,
+      "grad_norm": 0.6342661101597554,
+      "learning_rate": 1.3423827970338422e-06,
+      "loss": 0.9582,
+      "step": 317
+    },
+    {
+      "epoch": 0.6751592356687898,
+      "grad_norm": 0.6556216469291356,
+      "learning_rate": 1.3268382508295813e-06,
+      "loss": 0.9382,
+      "step": 318
+    },
+    {
+      "epoch": 0.6772823779193206,
+      "grad_norm": 0.6455027818834712,
+      "learning_rate": 1.3113516529394705e-06,
+      "loss": 0.9464,
+      "step": 319
+    },
+    {
+      "epoch": 0.6794055201698513,
+      "grad_norm": 0.646423128738099,
+      "learning_rate": 1.2959237683238768e-06,
+      "loss": 0.8862,
+      "step": 320
+    },
+    {
+      "epoch": 0.6815286624203821,
+      "grad_norm": 0.6149278332084452,
+      "learning_rate": 1.2805553590430198e-06,
+      "loss": 0.9907,
+      "step": 321
+    },
+    {
+      "epoch": 0.6836518046709129,
+      "grad_norm": 0.661598136600806,
+      "learning_rate": 1.2652471842193415e-06,
+      "loss": 0.8954,
+      "step": 322
+    },
+    {
+      "epoch": 0.6857749469214437,
+      "grad_norm": 0.7260709760195354,
+      "learning_rate": 1.2500000000000007e-06,
+      "loss": 0.9774,
+      "step": 323
+    },
+    {
+      "epoch": 0.6878980891719745,
+      "grad_norm": 0.6539931348515254,
+      "learning_rate": 1.2348145595195247e-06,
+      "loss": 0.993,
+      "step": 324
+    },
+    {
+      "epoch": 0.6900212314225053,
+      "grad_norm": 0.6415119742490282,
+      "learning_rate": 1.2196916128626126e-06,
+      "loss": 0.9817,
+      "step": 325
+    },
+    {
+      "epoch": 0.692144373673036,
+      "grad_norm": 0.6229170743011443,
+      "learning_rate": 1.2046319070270793e-06,
+      "loss": 0.9355,
+      "step": 326
+    },
+    {
+      "epoch": 0.6942675159235668,
+      "grad_norm": 0.6265657687743286,
+      "learning_rate": 1.1896361858869598e-06,
+      "loss": 0.97,
+      "step": 327
+    },
+    {
+      "epoch": 0.6963906581740976,
+      "grad_norm": 0.6472861949921104,
+      "learning_rate": 1.174705190155766e-06,
+      "loss": 0.9052,
+      "step": 328
+    },
+    {
+      "epoch": 0.6985138004246284,
+      "grad_norm": 0.6491445376690814,
+      "learning_rate": 1.1598396573499005e-06,
+      "loss": 0.928,
+      "step": 329
+    },
+    {
+      "epoch": 0.7006369426751592,
+      "grad_norm": 0.7123592443926866,
+      "learning_rate": 1.145040321752223e-06,
+      "loss": 0.9875,
+      "step": 330
+    },
+    {
+      "epoch": 0.70276008492569,
+      "grad_norm": 0.6420954357672332,
+      "learning_rate": 1.1303079143757831e-06,
+      "loss": 0.9396,
+      "step": 331
+    },
+    {
+      "epoch": 0.7048832271762208,
+      "grad_norm": 0.687954637263409,
+      "learning_rate": 1.1156431629277118e-06,
+      "loss": 0.9622,
+      "step": 332
+    },
+    {
+      "epoch": 0.7070063694267515,
+      "grad_norm": 0.6587466991092564,
+      "learning_rate": 1.1010467917732783e-06,
+      "loss": 0.9569,
+      "step": 333
+    },
+    {
+      "epoch": 0.7091295116772823,
+      "grad_norm": 0.6343940288936673,
+      "learning_rate": 1.0865195219001028e-06,
+      "loss": 0.9477,
+      "step": 334
+    },
+    {
+      "epoch": 0.7112526539278131,
+      "grad_norm": 0.6054208136804363,
+      "learning_rate": 1.0720620708825536e-06,
+      "loss": 0.9379,
+      "step": 335
+    },
+    {
+      "epoch": 0.7133757961783439,
+      "grad_norm": 0.6680904987189861,
+      "learning_rate": 1.0576751528462936e-06,
+      "loss": 0.973,
+      "step": 336
+    },
+    {
+      "epoch": 0.7154989384288747,
+      "grad_norm": 0.6359560896597604,
+      "learning_rate": 1.043359478433012e-06,
+      "loss": 0.9951,
+      "step": 337
+    },
+    {
+      "epoch": 0.7176220806794055,
+      "grad_norm": 0.674548446558131,
+      "learning_rate": 1.0291157547653171e-06,
+      "loss": 0.9128,
+      "step": 338
+    },
+    {
+      "epoch": 0.7197452229299363,
+      "grad_norm": 0.6171616777905478,
+      "learning_rate": 1.0149446854118154e-06,
+      "loss": 0.9619,
+      "step": 339
+    },
+    {
+      "epoch": 0.721868365180467,
+      "grad_norm": 0.7018076786345825,
+      "learning_rate": 1.0008469703523493e-06,
+      "loss": 0.9473,
+      "step": 340
+    },
+    {
+      "epoch": 0.7239915074309978,
+      "grad_norm": 0.6108858922632827,
+      "learning_rate": 9.868233059434286e-07,
+      "loss": 0.979,
+      "step": 341
+    },
+    {
+      "epoch": 0.7261146496815286,
+      "grad_norm": 0.6059128161574572,
+      "learning_rate": 9.728743848838334e-07,
+      "loss": 0.9513,
+      "step": 342
+    },
+    {
+      "epoch": 0.7282377919320594,
+      "grad_norm": 0.639834131452764,
+      "learning_rate": 9.590008961803942e-07,
+      "loss": 1.0088,
+      "step": 343
+    },
+    {
+      "epoch": 0.7303609341825902,
+      "grad_norm": 0.5974204476419624,
+      "learning_rate": 9.452035251139621e-07,
+      "loss": 0.9731,
+      "step": 344
+    },
+    {
+      "epoch": 0.732484076433121,
+      "grad_norm": 0.6373372180798436,
+      "learning_rate": 9.314829532055569e-07,
+      "loss": 0.9176,
+      "step": 345
+    },
+    {
+      "epoch": 0.7346072186836518,
+      "grad_norm": 0.6323537425779472,
+      "learning_rate": 9.178398581827086e-07,
+      "loss": 0.9845,
+      "step": 346
+    },
+    {
+      "epoch": 0.7367303609341825,
+      "grad_norm": 0.6482901339045514,
+      "learning_rate": 9.042749139459711e-07,
+      "loss": 1.0278,
+      "step": 347
+    },
+    {
+      "epoch": 0.7388535031847133,
+      "grad_norm": 0.6726254274936714,
+      "learning_rate": 8.907887905356461e-07,
+      "loss": 0.9436,
+      "step": 348
+    },
+    {
+      "epoch": 0.7409766454352441,
+      "grad_norm": 0.6629110485150135,
+      "learning_rate": 8.77382154098679e-07,
+      "loss": 0.9551,
+      "step": 349
+    },
+    {
+      "epoch": 0.7430997876857749,
+      "grad_norm": 0.6671801531262402,
+      "learning_rate": 8.64055666855757e-07,
+      "loss": 0.9105,
+      "step": 350
+    },
+    {
+      "epoch": 0.7452229299363057,
+      "grad_norm": 0.6703424513016992,
+      "learning_rate": 8.508099870685979e-07,
+      "loss": 0.9302,
+      "step": 351
+    },
+    {
+      "epoch": 0.7473460721868365,
+      "grad_norm": 0.6361868915566896,
+      "learning_rate": 8.376457690074386e-07,
+      "loss": 0.9655,
+      "step": 352
+    },
+    {
+      "epoch": 0.7494692144373672,
+      "grad_norm": 0.6209038321111798,
+      "learning_rate": 8.245636629187121e-07,
+      "loss": 0.9607,
+      "step": 353
+    },
+    {
+      "epoch": 0.7515923566878981,
+      "grad_norm": 0.607143406990376,
+      "learning_rate": 8.115643149929318e-07,
+      "loss": 1.028,
+      "step": 354
+    },
+    {
+      "epoch": 0.7537154989384289,
+      "grad_norm": 0.6565024756948546,
+      "learning_rate": 7.986483673327724e-07,
+      "loss": 0.9162,
+      "step": 355
+    },
+    {
+      "epoch": 0.7558386411889597,
+      "grad_norm": 0.6064614848733034,
+      "learning_rate": 7.858164579213548e-07,
+      "loss": 0.9722,
+      "step": 356
+    },
+    {
+      "epoch": 0.7579617834394905,
+      "grad_norm": 0.6470366271469512,
+      "learning_rate": 7.730692205907295e-07,
+      "loss": 0.9369,
+      "step": 357
+    },
+    {
+      "epoch": 0.7600849256900213,
+      "grad_norm": 0.619321171290989,
+      "learning_rate": 7.604072849905708e-07,
+      "loss": 0.9738,
+      "step": 358
+    },
+    {
+      "epoch": 0.7622080679405521,
+      "grad_norm": 0.6651018030884595,
+      "learning_rate": 7.478312765570747e-07,
+      "loss": 0.9867,
+      "step": 359
+    },
+    {
+      "epoch": 0.7643312101910829,
+      "grad_norm": 0.643295699040654,
+      "learning_rate": 7.353418164820656e-07,
+      "loss": 0.9507,
+      "step": 360
+    },
+    {
+      "epoch": 0.7664543524416136,
+      "grad_norm": 0.6349317217486001,
+      "learning_rate": 7.229395216823132e-07,
+      "loss": 0.8986,
+      "step": 361
+    },
+    {
+      "epoch": 0.7685774946921444,
+      "grad_norm": 0.6623180797372981,
+      "learning_rate": 7.106250047690588e-07,
+      "loss": 0.9791,
+      "step": 362
+    },
+    {
+      "epoch": 0.7707006369426752,
+      "grad_norm": 0.6094671854634959,
+      "learning_rate": 6.983988740177553e-07,
+      "loss": 0.9179,
+      "step": 363
+    },
+    {
+      "epoch": 0.772823779193206,
+      "grad_norm": 0.6718517906357772,
+      "learning_rate": 6.862617333380214e-07,
+      "loss": 0.9912,
+      "step": 364
+    },
+    {
+      "epoch": 0.7749469214437368,
+      "grad_norm": 0.5913671123397655,
+      "learning_rate": 6.742141822438144e-07,
+      "loss": 0.969,
+      "step": 365
+    },
+    {
+      "epoch": 0.7770700636942676,
+      "grad_norm": 0.6223699232576859,
+      "learning_rate": 6.622568158238126e-07,
+      "loss": 0.8975,
+      "step": 366
+    },
+    {
+      "epoch": 0.7791932059447984,
+      "grad_norm": 0.6272157524107641,
+      "learning_rate": 6.503902247120239e-07,
+      "loss": 0.9675,
+      "step": 367
+    },
+    {
+      "epoch": 0.7813163481953291,
+      "grad_norm": 0.6511133863417863,
+      "learning_rate": 6.386149950586099e-07,
+      "loss": 0.9877,
+      "step": 368
+    },
+    {
+      "epoch": 0.7834394904458599,
+      "grad_norm": 0.6541100860291786,
+      "learning_rate": 6.269317085009364e-07,
+      "loss": 0.9177,
+      "step": 369
+    },
+    {
+      "epoch": 0.7855626326963907,
+      "grad_norm": 0.6243600418119797,
+      "learning_rate": 6.153409421348358e-07,
+      "loss": 0.9283,
+      "step": 370
+    },
+    {
+      "epoch": 0.7876857749469215,
+      "grad_norm": 0.6728138627005387,
+      "learning_rate": 6.038432684861123e-07,
+      "loss": 0.9366,
+      "step": 371
+    },
+    {
+      "epoch": 0.7898089171974523,
+      "grad_norm": 0.6359307108447517,
+      "learning_rate": 5.92439255482252e-07,
+      "loss": 0.9407,
+      "step": 372
+    },
+    {
+      "epoch": 0.7919320594479831,
+      "grad_norm": 0.6515305344696402,
+      "learning_rate": 5.811294664243752e-07,
+      "loss": 0.8832,
+      "step": 373
+    },
+    {
+      "epoch": 0.7940552016985138,
+      "grad_norm": 0.6699132219023968,
+      "learning_rate": 5.699144599594131e-07,
+      "loss": 0.9203,
+      "step": 374
+    },
+    {
+      "epoch": 0.7961783439490446,
+      "grad_norm": 0.6531325580443874,
+      "learning_rate": 5.587947900525093e-07,
+      "loss": 0.9486,
+      "step": 375
+    },
+    {
+      "epoch": 0.7983014861995754,
+      "grad_norm": 0.6235259540633097,
+      "learning_rate": 5.4777100595966e-07,
+      "loss": 0.9976,
+      "step": 376
+    },
+    {
+      "epoch": 0.8004246284501062,
+      "grad_norm": 0.721147004969793,
+      "learning_rate": 5.368436522005816e-07,
+      "loss": 0.9333,
+      "step": 377
+    },
+    {
+      "epoch": 0.802547770700637,
+      "grad_norm": 0.654820567386982,
+      "learning_rate": 5.26013268531817e-07,
+      "loss": 0.9609,
+      "step": 378
+    },
+    {
+      "epoch": 0.8046709129511678,
+      "grad_norm": 0.6455700462685393,
+      "learning_rate": 5.1528038992007e-07,
+      "loss": 0.9963,
+      "step": 379
+    },
+    {
+      "epoch": 0.8067940552016986,
+      "grad_norm": 0.6521250389244575,
+      "learning_rate": 5.046455465157849e-07,
+      "loss": 0.9317,
+      "step": 380
+    },
+    {
+      "epoch": 0.8089171974522293,
+      "grad_norm": 0.6233082624968019,
+      "learning_rate": 4.941092636269554e-07,
+      "loss": 1.0314,
+      "step": 381
+    },
+    {
+      "epoch": 0.8110403397027601,
+      "grad_norm": 0.6284738498607256,
+      "learning_rate": 4.836720616931831e-07,
+      "loss": 0.991,
+      "step": 382
+    },
+    {
+      "epoch": 0.8131634819532909,
+      "grad_norm": 0.6251411989698912,
+      "learning_rate": 4.7333445625996233e-07,
+      "loss": 0.9631,
+      "step": 383
+    },
+    {
+      "epoch": 0.8152866242038217,
+      "grad_norm": 0.6093691597832145,
+      "learning_rate": 4.6309695795322323e-07,
+      "loss": 0.9258,
+      "step": 384
+    },
+    {
+      "epoch": 0.8174097664543525,
+      "grad_norm": 0.6425399405048542,
+      "learning_rate": 4.5296007245410225e-07,
+      "loss": 0.9179,
+      "step": 385
+    },
+    {
+      "epoch": 0.8195329087048833,
+      "grad_norm": 0.6669561342230419,
+      "learning_rate": 4.429243004739692e-07,
+      "loss": 0.8939,
+      "step": 386
+    },
+    {
+      "epoch": 0.821656050955414,
+      "grad_norm": 0.6260259468897137,
+      "learning_rate": 4.329901377296902e-07,
+      "loss": 0.984,
+      "step": 387
+    },
+    {
+      "epoch": 0.8237791932059448,
+      "grad_norm": 0.6393791566126937,
+      "learning_rate": 4.2315807491914747e-07,
+      "loss": 0.9454,
+      "step": 388
+    },
+    {
+      "epoch": 0.8259023354564756,
+      "grad_norm": 0.6513774907103931,
+      "learning_rate": 4.1342859769699483e-07,
+      "loss": 1.0317,
+      "step": 389
+    },
+    {
+      "epoch": 0.8280254777070064,
+      "grad_norm": 0.6269281204448669,
+      "learning_rate": 4.038021866506725e-07,
+      "loss": 0.9402,
+      "step": 390
+    },
+    {
+      "epoch": 0.8301486199575372,
+      "grad_norm": 0.6049355089918522,
+      "learning_rate": 3.9427931727666983e-07,
+      "loss": 1.0111,
+      "step": 391
+    },
+    {
+      "epoch": 0.832271762208068,
+      "grad_norm": 0.6896126089841778,
+      "learning_rate": 3.848604599570338e-07,
+      "loss": 0.9388,
+      "step": 392
+    },
+    {
+      "epoch": 0.8343949044585988,
+      "grad_norm": 0.6323897929388586,
+      "learning_rate": 3.7554607993613823e-07,
+      "loss": 0.943,
+      "step": 393
+    },
+    {
+      "epoch": 0.8365180467091295,
+      "grad_norm": 0.643316655036959,
+      "learning_rate": 3.6633663729770007e-07,
+      "loss": 0.9258,
+      "step": 394
+    },
+    {
+      "epoch": 0.8386411889596603,
+      "grad_norm": 0.6880537419205733,
+      "learning_rate": 3.572325869420587e-07,
+      "loss": 0.9288,
+      "step": 395
+    },
+    {
+      "epoch": 0.8407643312101911,
+      "grad_norm": 0.6503968205421878,
+      "learning_rate": 3.48234378563698e-07,
+      "loss": 0.9486,
+      "step": 396
+    },
+    {
+      "epoch": 0.8428874734607219,
+      "grad_norm": 0.6553158860120801,
+      "learning_rate": 3.3934245662904207e-07,
+      "loss": 0.9833,
+      "step": 397
+    },
+    {
+      "epoch": 0.8450106157112527,
+      "grad_norm": 0.6537781810406033,
+      "learning_rate": 3.3055726035449486e-07,
+      "loss": 1.0238,
+      "step": 398
+    },
+    {
+      "epoch": 0.8471337579617835,
+      "grad_norm": 0.6064261172835044,
+      "learning_rate": 3.2187922368474954e-07,
+      "loss": 0.9503,
+      "step": 399
+    },
+    {
+      "epoch": 0.8492569002123143,
+      "grad_norm": 0.6358301794610944,
+      "learning_rate": 3.133087752713479e-07,
+      "loss": 1.0033,
+      "step": 400
+    },
+    {
+      "epoch": 0.851380042462845,
+      "grad_norm": 0.6910319036256287,
+      "learning_rate": 3.048463384515149e-07,
+      "loss": 0.99,
+      "step": 401
+    },
+    {
+      "epoch": 0.8535031847133758,
+      "grad_norm": 0.6770622312793074,
+      "learning_rate": 2.9649233122724106e-07,
+      "loss": 0.9066,
+      "step": 402
+    },
+    {
+      "epoch": 0.8556263269639066,
+      "grad_norm": 0.6570081359837786,
+      "learning_rate": 2.88247166244639e-07,
+      "loss": 0.9943,
+      "step": 403
+    },
+    {
+      "epoch": 0.8577494692144374,
+      "grad_norm": 0.6111215068802225,
+      "learning_rate": 2.8011125077355873e-07,
+      "loss": 0.949,
+      "step": 404
+    },
+    {
+      "epoch": 0.8598726114649682,
+      "grad_norm": 0.6463532654670587,
+      "learning_rate": 2.7208498668747393e-07,
+      "loss": 0.9262,
+      "step": 405
+    },
+    {
+      "epoch": 0.861995753715499,
+      "grad_norm": 0.6544581992958507,
+      "learning_rate": 2.641687704436269e-07,
+      "loss": 0.983,
+      "step": 406
+    },
+    {
+      "epoch": 0.8641188959660298,
+      "grad_norm": 0.6368211507084761,
+      "learning_rate": 2.563629930634487e-07,
+      "loss": 0.9573,
+      "step": 407
+    },
+    {
+      "epoch": 0.8662420382165605,
+      "grad_norm": 0.6382461780742053,
+      "learning_rate": 2.4866804011324296e-07,
+      "loss": 0.9887,
+      "step": 408
+    },
+    {
+      "epoch": 0.8683651804670913,
+      "grad_norm": 0.6250872686300754,
+      "learning_rate": 2.410842916851425e-07,
+      "loss": 0.9696,
+      "step": 409
+    },
+    {
+      "epoch": 0.8704883227176221,
+      "grad_norm": 0.6767356131769894,
+      "learning_rate": 2.3361212237833357e-07,
+      "loss": 0.9911,
+      "step": 410
+    },
+    {
+      "epoch": 0.8726114649681529,
+      "grad_norm": 0.6224360952672792,
+      "learning_rate": 2.2625190128055168e-07,
+      "loss": 0.9579,
+      "step": 411
+    },
+    {
+      "epoch": 0.8747346072186837,
+      "grad_norm": 0.6529103135781014,
+      "learning_rate": 2.1900399194985434e-07,
+      "loss": 0.9722,
+      "step": 412
+    },
+    {
+      "epoch": 0.8768577494692145,
+      "grad_norm": 0.6367464766206924,
+      "learning_rate": 2.118687523966559e-07,
+      "loss": 0.9636,
+      "step": 413
+    },
+    {
+      "epoch": 0.8789808917197452,
+      "grad_norm": 0.6316689138643529,
+      "learning_rate": 2.0484653506605223e-07,
+      "loss": 0.959,
+      "step": 414
+    },
+    {
+      "epoch": 0.881104033970276,
+      "grad_norm": 0.6451224349585934,
+      "learning_rate": 1.9793768682040526e-07,
+      "loss": 0.9981,
+      "step": 415
+    },
+    {
+      "epoch": 0.8832271762208068,
+      "grad_norm": 0.6301838097010446,
+      "learning_rate": 1.911425489222127e-07,
+      "loss": 0.9904,
+      "step": 416
+    },
+    {
+      "epoch": 0.8853503184713376,
+      "grad_norm": 0.7327431200834074,
+      "learning_rate": 1.8446145701724982e-07,
+      "loss": 0.9716,
+      "step": 417
+    },
+    {
+      "epoch": 0.8874734607218684,
+      "grad_norm": 0.6641289243508941,
+      "learning_rate": 1.778947411179932e-07,
+      "loss": 0.987,
+      "step": 418
+    },
+    {
+      "epoch": 0.8895966029723992,
+      "grad_norm": 0.6760362291459804,
+      "learning_rate": 1.7144272558731466e-07,
+      "loss": 0.9067,
+      "step": 419
+    },
+    {
+      "epoch": 0.89171974522293,
+      "grad_norm": 0.6288353137171142,
+      "learning_rate": 1.6510572912246476e-07,
+      "loss": 0.9142,
+      "step": 420
+    },
+    {
+      "epoch": 0.8938428874734607,
+      "grad_norm": 0.6217799653332664,
+      "learning_rate": 1.5888406473932694e-07,
+      "loss": 0.9244,
+      "step": 421
+    },
+    {
+      "epoch": 0.8959660297239915,
+      "grad_norm": 0.6524011013463125,
+      "learning_rate": 1.5277803975695865e-07,
+      "loss": 0.9552,
+      "step": 422
+    },
+    {
+      "epoch": 0.8980891719745223,
+      "grad_norm": 0.5937240986119188,
+      "learning_rate": 1.4678795578240897e-07,
+      "loss": 0.9441,
+      "step": 423
+    },
+    {
+      "epoch": 0.9002123142250531,
+      "grad_norm": 0.6690722396611295,
+      "learning_rate": 1.4091410869582267e-07,
+      "loss": 0.9643,
+      "step": 424
+    },
+    {
+      "epoch": 0.9023354564755839,
+      "grad_norm": 0.6537649069457198,
+      "learning_rate": 1.3515678863582432e-07,
+      "loss": 0.9784,
+      "step": 425
+    },
+    {
+      "epoch": 0.9044585987261147,
+      "grad_norm": 0.6662079340519091,
+      "learning_rate": 1.2951627998518623e-07,
+      "loss": 0.9598,
+      "step": 426
+    },
+    {
+      "epoch": 0.9065817409766455,
+      "grad_norm": 0.674794856138497,
+      "learning_rate": 1.2399286135678424e-07,
+      "loss": 0.9506,
+      "step": 427
+    },
+    {
+      "epoch": 0.9087048832271762,
+      "grad_norm": 0.6304822222858429,
+      "learning_rate": 1.1858680557983171e-07,
+      "loss": 0.9491,
+      "step": 428
+    },
+    {
+      "epoch": 0.910828025477707,
+      "grad_norm": 0.6036661319596233,
+      "learning_rate": 1.1329837968640539e-07,
+      "loss": 0.9582,
+      "step": 429
+    },
+    {
+      "epoch": 0.9129511677282378,
+      "grad_norm": 0.6336696118246462,
+      "learning_rate": 1.0812784489825506e-07,
+      "loss": 0.928,
+      "step": 430
+    },
+    {
+      "epoch": 0.9150743099787686,
+      "grad_norm": 0.6225535325876186,
+      "learning_rate": 1.030754566139014e-07,
+      "loss": 0.9707,
+      "step": 431
+    },
+    {
+      "epoch": 0.9171974522292994,
+      "grad_norm": 0.6414866827963036,
+      "learning_rate": 9.814146439601741e-08,
+      "loss": 0.9673,
+      "step": 432
+    },
+    {
+      "epoch": 0.9193205944798302,
+      "grad_norm": 0.598006805118166,
+      "learning_rate": 9.332611195910585e-08,
+      "loss": 0.9626,
+      "step": 433
+    },
+    {
+      "epoch": 0.921443736730361,
+      "grad_norm": 0.6369023150011093,
+      "learning_rate": 8.862963715745687e-08,
+      "loss": 0.9292,
+      "step": 434
+    },
+    {
+      "epoch": 0.9235668789808917,
+      "grad_norm": 0.6067154709316188,
+      "learning_rate": 8.405227197340216e-08,
+      "loss": 0.8989,
+      "step": 435
+    },
+    {
+      "epoch": 0.9256900212314225,
+      "grad_norm": 0.6123427935922671,
+      "learning_rate": 7.959424250585323e-08,
+      "loss": 0.9973,
+      "step": 436
+    },
+    {
+      "epoch": 0.9278131634819533,
+      "grad_norm": 0.631261585951247,
+      "learning_rate": 7.525576895913655e-08,
+      "loss": 0.9253,
+      "step": 437
+    },
+    {
+      "epoch": 0.9299363057324841,
+      "grad_norm": 0.6684228376208218,
+      "learning_rate": 7.10370656321141e-08,
+      "loss": 0.9093,
+      "step": 438
+    },
+    {
+      "epoch": 0.9320594479830149,
+      "grad_norm": 0.6341580390596356,
+      "learning_rate": 6.69383409075991e-08,
+      "loss": 0.9385,
+      "step": 439
+    },
+    {
+      "epoch": 0.9341825902335457,
+      "grad_norm": 0.6641492655666181,
+      "learning_rate": 6.295979724206314e-08,
+      "loss": 0.9785,
+      "step": 440
+    },
+    {
+      "epoch": 0.9363057324840764,
+      "grad_norm": 0.6274625041818331,
+      "learning_rate": 5.910163115563472e-08,
+      "loss": 0.941,
+      "step": 441
+    },
+    {
+      "epoch": 0.9384288747346072,
+      "grad_norm": 0.5986199966520086,
+      "learning_rate": 5.53640332223937e-08,
+      "loss": 0.8832,
+      "step": 442
+    },
+    {
+      "epoch": 0.940552016985138,
+      "grad_norm": 0.6287455293561965,
+      "learning_rate": 5.1747188060956296e-08,
+      "loss": 0.9405,
+      "step": 443
+    },
+    {
+      "epoch": 0.9426751592356688,
+      "grad_norm": 0.6486765094469326,
+      "learning_rate": 4.825127432535714e-08,
+      "loss": 0.9153,
+      "step": 444
+    },
+    {
+      "epoch": 0.9447983014861996,
+      "grad_norm": 0.616792972142488,
+      "learning_rate": 4.487646469622464e-08,
+      "loss": 0.9511,
+      "step": 445
+    },
+    {
+      "epoch": 0.9469214437367304,
+      "grad_norm": 0.6899660172126519,
+      "learning_rate": 4.1622925872249476e-08,
+      "loss": 0.9427,
+      "step": 446
+    },
+    {
+      "epoch": 0.9490445859872612,
+      "grad_norm": 0.6644667873221963,
+      "learning_rate": 3.849081856195341e-08,
+      "loss": 0.9439,
+      "step": 447
+    },
+    {
+      "epoch": 0.9511677282377919,
+      "grad_norm": 0.6168810973654132,
+      "learning_rate": 3.548029747574927e-08,
+      "loss": 0.9606,
+      "step": 448
+    },
+    {
+      "epoch": 0.9532908704883227,
+      "grad_norm": 0.6065862488438216,
+      "learning_rate": 3.259151131829869e-08,
+      "loss": 0.9844,
+      "step": 449
+    },
+    {
+      "epoch": 0.9554140127388535,
+      "grad_norm": 0.6486015909726873,
+      "learning_rate": 2.982460278116883e-08,
+      "loss": 0.9633,
+      "step": 450
+    },
+    {
+      "epoch": 0.9575371549893843,
+      "grad_norm": 0.631115227640127,
+      "learning_rate": 2.7179708535781945e-08,
+      "loss": 0.9675,
+      "step": 451
+    },
+    {
+      "epoch": 0.9596602972399151,
+      "grad_norm": 0.6844872355582275,
+      "learning_rate": 2.4656959226665445e-08,
+      "loss": 0.9678,
+      "step": 452
+    },
+    {
+      "epoch": 0.9617834394904459,
+      "grad_norm": 0.676038419107933,
+      "learning_rate": 2.2256479464999315e-08,
+      "loss": 0.9741,
+      "step": 453
+    },
+    {
+      "epoch": 0.9639065817409767,
+      "grad_norm": 0.5792729689539223,
+      "learning_rate": 1.99783878224602e-08,
+      "loss": 0.9668,
+      "step": 454
+    },
+    {
+      "epoch": 0.9660297239915074,
+      "grad_norm": 0.6539907426996568,
+      "learning_rate": 1.7822796825364142e-08,
+      "loss": 0.9546,
+      "step": 455
+    },
+    {
+      "epoch": 0.9681528662420382,
+      "grad_norm": 0.6230572975623042,
+      "learning_rate": 1.578981294910936e-08,
+      "loss": 0.9274,
+      "step": 456
+    },
+    {
+      "epoch": 0.970276008492569,
+      "grad_norm": 0.5968504324622771,
+      "learning_rate": 1.38795366129163e-08,
+      "loss": 0.9125,
+      "step": 457
+    },
+    {
+      "epoch": 0.9723991507430998,
+      "grad_norm": 0.6390123520113857,
+      "learning_rate": 1.2092062174867414e-08,
+      "loss": 0.9138,
+      "step": 458
+    },
+    {
+      "epoch": 0.9745222929936306,
+      "grad_norm": 0.6236803320324935,
+      "learning_rate": 1.042747792724702e-08,
+      "loss": 0.9492,
+      "step": 459
+    },
+    {
+      "epoch": 0.9766454352441614,
+      "grad_norm": 0.6338903091057604,
+      "learning_rate": 8.885866092178952e-09,
+      "loss": 0.9838,
+      "step": 460
+    },
+    {
+      "epoch": 0.9787685774946921,
+      "grad_norm": 0.6353390431277406,
+      "learning_rate": 7.467302817566192e-09,
+      "loss": 0.9526,
+      "step": 461
+    },
+    {
+      "epoch": 0.9808917197452229,
+      "grad_norm": 0.654122905953797,
+      "learning_rate": 6.171858173328604e-09,
+      "loss": 0.9822,
+      "step": 462
+    },
+    {
+      "epoch": 0.9830148619957537,
+      "grad_norm": 0.6506585984715718,
+      "learning_rate": 4.999596147943486e-09,
+      "loss": 0.9091,
+      "step": 463
+    },
+    {
+      "epoch": 0.9851380042462845,
+      "grad_norm": 0.6729520792369528,
+      "learning_rate": 3.95057464528309e-09,
+      "loss": 0.9826,
+      "step": 464
+    },
+    {
+      "epoch": 0.9872611464968153,
+      "grad_norm": 0.6314573444952737,
+      "learning_rate": 3.02484548175469e-09,
+      "loss": 1.0175,
+      "step": 465
+    },
+    {
+      "epoch": 0.9893842887473461,
+      "grad_norm": 0.6986521882784366,
+      "learning_rate": 2.222454383742356e-09,
+      "loss": 0.9587,
+      "step": 466
+    },
+    {
+      "epoch": 0.9915074309978769,
+      "grad_norm": 0.6226442078982108,
+      "learning_rate": 1.5434409853473664e-09,
+      "loss": 0.9285,
+      "step": 467
+    },
+    {
+      "epoch": 0.9936305732484076,
+      "grad_norm": 0.6278882096671412,
+      "learning_rate": 9.878388264300543e-10,
+      "loss": 0.944,
+      "step": 468
+    },
+    {
+      "epoch": 0.9957537154989384,
+      "grad_norm": 0.5896705695288925,
+      "learning_rate": 5.55675350954743e-10,
+      "loss": 0.9217,
+      "step": 469
+    },
+    {
+      "epoch": 0.9978768577494692,
+      "grad_norm": 0.6570338350651012,
+      "learning_rate": 2.4697190563194307e-10,
+      "loss": 0.9466,
+      "step": 470
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6095877208075582,
+      "learning_rate": 6.174373886586038e-11,
+      "loss": 0.9491,
+      "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:fb8e23c64c634b1182f062703df8841ca23b1e888c2c028da1ee4aa6d18865e0
+size 8017