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
+}

config.json

@@ -0,0 +1,30 @@
+{
+  "architectures": [
+    "Qwen3ForCausalLM"
+  ],
+  "attention_bias": false,
+  "attention_dropout": 0.0,
+  "bos_token_id": 151643,
+  "eos_token_id": 151643,
+  "head_dim": 128,
+  "hidden_act": "silu",
+  "hidden_size": 1024,
+  "initializer_range": 0.02,
+  "intermediate_size": 3072,
+  "max_position_embeddings": 32768,
+  "max_window_layers": 28,
+  "model_type": "qwen3",
+  "num_attention_heads": 16,
+  "num_hidden_layers": 28,
+  "num_key_value_heads": 8,
+  "rms_norm_eps": 1e-06,
+  "rope_scaling": null,
+  "rope_theta": 1000000,
+  "sliding_window": null,
+  "tie_word_embeddings": true,
+  "torch_dtype": "float16",
+  "transformers_version": "4.51.3",
+  "use_cache": true,
+  "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": "4.51.3"
+}

model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:34daee22f3da79f04a110d67751c885964c9e42cea74eb16b11e1445d65f31e8
+size 1192134784

optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ea006d31679633f09f87669e36037e0958ed216eafbe6f86af1823fac9c2175b
+size 1211416522

rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d4c917636c7a58af68a29056522a757e9f9b99005b776641aa157c536967817d
+size 14244

scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a9b8f34b96b9d90cfe89bdd9aafba61ba227bc92faead103f96edaa3b9fd8c78
+size 1064

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": "<|endoftext|>",
+    "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,
+  "chat_template": "{%- if tools %}\n    {{- '<|im_start|>system\\n' }}\n    {%- if messages[0].role == 'system' %}\n        {{- messages[0].content + '\\n\\n' }}\n    {%- endif %}\n    {{- \"# 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>\" }}\n    {%- for tool in tools %}\n        {{- \"\\n\" }}\n        {{- tool | tojson }}\n    {%- endfor %}\n    {{- \"\\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\" }}\n{%- else %}\n    {%- if messages[0].role == 'system' %}\n        {{- '<|im_start|>system\\n' + messages[0].content + '<|im_end|>\\n' }}\n    {%- endif %}\n{%- endif %}\n{%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %}\n{%- for message in messages[::-1] %}\n    {%- set index = (messages|length - 1) - loop.index0 %}\n    {%- if ns.multi_step_tool and message.role == \"user\" and not(message.content.startswith('<tool_response>') and message.content.endswith('</tool_response>')) %}\n        {%- set ns.multi_step_tool = false %}\n        {%- set ns.last_query_index = index %}\n    {%- endif %}\n{%- endfor %}\n{%- for message in messages %}\n    {%- if (message.role == \"user\") or (message.role == \"system\" and not loop.first) %}\n        {{- '<|im_start|>' + message.role + '\\n' + message.content + '<|im_end|>' + '\\n' }}\n    {%- elif message.role == \"assistant\" %}\n        {%- set content = message.content %}\n        {%- set reasoning_content = '' %}\n        {%- if message.reasoning_content is defined and message.reasoning_content is not none %}\n            {%- set reasoning_content = message.reasoning_content %}\n        {%- else %}\n            {%- if '</think>' in message.content %}\n                {%- set content = message.content.split('</think>')[-1].lstrip('\\n') %}\n                {%- set reasoning_content = message.content.split('</think>')[0].rstrip('\\n').split('<think>')[-1].lstrip('\\n') %}\n            {%- endif %}\n        {%- endif %}\n        {%- if loop.index0 > ns.last_query_index %}\n            {%- if loop.last or (not loop.last and reasoning_content) %}\n                {{- '<|im_start|>' + message.role + '\\n<think>\\n' + reasoning_content.strip('\\n') + '\\n</think>\\n\\n' + content.lstrip('\\n') }}\n            {%- else %}\n                {{- '<|im_start|>' + message.role + '\\n' + content }}\n            {%- endif %}\n        {%- else %}\n            {{- '<|im_start|>' + message.role + '\\n' + content }}\n        {%- endif %}\n        {%- if message.tool_calls %}\n            {%- for tool_call in message.tool_calls %}\n                {%- if (loop.first and content) or (not loop.first) %}\n                    {{- '\\n' }}\n                {%- endif %}\n                {%- if tool_call.function %}\n                    {%- set tool_call = tool_call.function %}\n                {%- endif %}\n                {{- '<tool_call>\\n{\"name\": \"' }}\n                {{- tool_call.name }}\n                {{- '\", \"arguments\": ' }}\n                {%- if tool_call.arguments is string %}\n                    {{- tool_call.arguments }}\n                {%- else %}\n                    {{- tool_call.arguments | tojson }}\n                {%- endif %}\n                {{- '}\\n</tool_call>' }}\n            {%- endfor %}\n        {%- endif %}\n        {{- '<|im_end|>\\n' }}\n    {%- elif message.role == \"tool\" %}\n        {%- if loop.first or (messages[loop.index0 - 1].role != \"tool\") %}\n            {{- '<|im_start|>user' }}\n        {%- endif %}\n        {{- '\\n<tool_response>\\n' }}\n        {{- message.content }}\n        {{- '\\n</tool_response>' }}\n        {%- if loop.last or (messages[loop.index0 + 1].role != \"tool\") %}\n            {{- '<|im_end|>\\n' }}\n        {%- endif %}\n    {%- endif %}\n{%- endfor %}\n{%- if add_generation_prompt %}\n    {{- '<|im_start|>assistant\\n' }}\n    {%- if enable_thinking is defined and enable_thinking is false %}\n        {{- '<think>\\n\\n</think>\\n\\n' }}\n    {%- endif %}\n{%- endif %}",
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "<|endoftext|>",
+  "errors": "replace",
+  "extra_special_tokens": {},
+  "model_max_length": 131072,
+  "pad_token": "<|endoftext|>",
+  "split_special_tokens": false,
+  "tokenizer_class": "Qwen2Tokenizer",
+  "unk_token": null
+}

trainer_state.json

@@ -0,0 +1,2734 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.5010271055664112,
+  "eval_steps": 500,
+  "global_step": 15000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.001670090351888037,
+      "grad_norm": 7.16015625,
+      "learning_rate": 1.9600000000000002e-05,
+      "loss": 0.6304,
+      "mean_token_accuracy": 0.811416609287262,
+      "num_tokens": 150666.0,
+      "step": 50
+    },
+    {
+      "epoch": 0.003340180703776074,
+      "grad_norm": 5.58984375,
+      "learning_rate": 1.999986736222114e-05,
+      "loss": 0.4501,
+      "mean_token_accuracy": 0.8562266689538955,
+      "num_tokens": 307563.0,
+      "step": 100
+    },
+    {
+      "epoch": 0.0050102710556641115,
+      "grad_norm": 5.0859375,
+      "learning_rate": 1.9999458569756985e-05,
+      "loss": 0.5118,
+      "mean_token_accuracy": 0.838346039056778,
+      "num_tokens": 454754.0,
+      "step": 150
+    },
+    {
+      "epoch": 0.006680361407552148,
+      "grad_norm": 4.37890625,
+      "learning_rate": 1.9998773578656093e-05,
+      "loss": 0.5075,
+      "mean_token_accuracy": 0.8403138118982315,
+      "num_tokens": 604733.0,
+      "step": 200
+    },
+    {
+      "epoch": 0.008350451759440186,
+      "grad_norm": 4.40234375,
+      "learning_rate": 1.999781240783885e-05,
+      "loss": 0.5057,
+      "mean_token_accuracy": 0.8395608925819397,
+      "num_tokens": 751094.0,
+      "step": 250
+    },
+    {
+      "epoch": 0.010020542111328223,
+      "grad_norm": 4.26953125,
+      "learning_rate": 1.9996575083854096e-05,
+      "loss": 0.5063,
+      "mean_token_accuracy": 0.8390217781066894,
+      "num_tokens": 906188.0,
+      "step": 300
+    },
+    {
+      "epoch": 0.01169063246321626,
+      "grad_norm": 5.03125,
+      "learning_rate": 1.9995061640878407e-05,
+      "loss": 0.5076,
+      "mean_token_accuracy": 0.8388406819105149,
+      "num_tokens": 1058839.0,
+      "step": 350
+    },
+    {
+      "epoch": 0.013360722815104297,
+      "grad_norm": 4.41015625,
+      "learning_rate": 1.9993272120715135e-05,
+      "loss": 0.5248,
+      "mean_token_accuracy": 0.83592469394207,
+      "num_tokens": 1210117.0,
+      "step": 400
+    },
+    {
+      "epoch": 0.015030813166992334,
+      "grad_norm": 4.4921875,
+      "learning_rate": 1.9991206572793253e-05,
+      "loss": 0.5103,
+      "mean_token_accuracy": 0.8381919747591019,
+      "num_tokens": 1365295.0,
+      "step": 450
+    },
+    {
+      "epoch": 0.016700903518880372,
+      "grad_norm": 4.69921875,
+      "learning_rate": 1.998886505416601e-05,
+      "loss": 0.5177,
+      "mean_token_accuracy": 0.838459021449089,
+      "num_tokens": 1519295.0,
+      "step": 500
+    },
+    {
+      "epoch": 0.018370993870768407,
+      "grad_norm": 4.6484375,
+      "learning_rate": 1.9986247629509327e-05,
+      "loss": 0.5784,
+      "mean_token_accuracy": 0.8219448757171631,
+      "num_tokens": 1670026.0,
+      "step": 550
+    },
+    {
+      "epoch": 0.020041084222656446,
+      "grad_norm": 4.25,
+      "learning_rate": 1.9983354371120035e-05,
+      "loss": 0.7304,
+      "mean_token_accuracy": 0.7821861469745636,
+      "num_tokens": 1830249.0,
+      "step": 600
+    },
+    {
+      "epoch": 0.02171117457454448,
+      "grad_norm": 3.994140625,
+      "learning_rate": 1.998018535891386e-05,
+      "loss": 0.7442,
+      "mean_token_accuracy": 0.7787324601411819,
+      "num_tokens": 1981040.0,
+      "step": 650
+    },
+    {
+      "epoch": 0.02338126492643252,
+      "grad_norm": 3.548828125,
+      "learning_rate": 1.9976740680423227e-05,
+      "loss": 0.7469,
+      "mean_token_accuracy": 0.7790761750936508,
+      "num_tokens": 2133645.0,
+      "step": 700
+    },
+    {
+      "epoch": 0.02505135527832056,
+      "grad_norm": 3.595703125,
+      "learning_rate": 1.997302043079484e-05,
+      "loss": 0.7073,
+      "mean_token_accuracy": 0.7876522129774094,
+      "num_tokens": 2286147.0,
+      "step": 750
+    },
+    {
+      "epoch": 0.026721445630208594,
+      "grad_norm": 3.787109375,
+      "learning_rate": 1.9969024712787045e-05,
+      "loss": 0.7117,
+      "mean_token_accuracy": 0.7861441165208817,
+      "num_tokens": 2435302.0,
+      "step": 800
+    },
+    {
+      "epoch": 0.028391535982096632,
+      "grad_norm": 4.07421875,
+      "learning_rate": 1.996475363676701e-05,
+      "loss": 0.7301,
+      "mean_token_accuracy": 0.7839706349372864,
+      "num_tokens": 2586271.0,
+      "step": 850
+    },
+    {
+      "epoch": 0.030061626333984667,
+      "grad_norm": 4.15234375,
+      "learning_rate": 1.9960207320707655e-05,
+      "loss": 0.7237,
+      "mean_token_accuracy": 0.7836953294277191,
+      "num_tokens": 2736739.0,
+      "step": 900
+    },
+    {
+      "epoch": 0.031731716685872706,
+      "grad_norm": 4.83984375,
+      "learning_rate": 1.9955385890184415e-05,
+      "loss": 0.7354,
+      "mean_token_accuracy": 0.7828176522254944,
+      "num_tokens": 2882560.0,
+      "step": 950
+    },
+    {
+      "epoch": 0.033401807037760745,
+      "grad_norm": 3.923828125,
+      "learning_rate": 1.995028947837175e-05,
+      "loss": 0.7236,
+      "mean_token_accuracy": 0.7848378765583038,
+      "num_tokens": 3043075.0,
+      "step": 1000
+    },
+    {
+      "epoch": 0.03507189738964878,
+      "grad_norm": 3.59375,
+      "learning_rate": 1.994491822603948e-05,
+      "loss": 0.7182,
+      "mean_token_accuracy": 0.7871662402153015,
+      "num_tokens": 3198694.0,
+      "step": 1050
+    },
+    {
+      "epoch": 0.036741987741536815,
+      "grad_norm": 4.04296875,
+      "learning_rate": 1.9939272281548897e-05,
+      "loss": 0.7076,
+      "mean_token_accuracy": 0.7869892084598541,
+      "num_tokens": 3351166.0,
+      "step": 1100
+    },
+    {
+      "epoch": 0.03841207809342485,
+      "grad_norm": 4.29296875,
+      "learning_rate": 1.9933351800848663e-05,
+      "loss": 0.7079,
+      "mean_token_accuracy": 0.7874267196655274,
+      "num_tokens": 3497030.0,
+      "step": 1150
+    },
+    {
+      "epoch": 0.04008216844531289,
+      "grad_norm": 3.650390625,
+      "learning_rate": 1.99271569474705e-05,
+      "loss": 0.6844,
+      "mean_token_accuracy": 0.7945140707492828,
+      "num_tokens": 3650753.0,
+      "step": 1200
+    },
+    {
+      "epoch": 0.04175225879720093,
+      "grad_norm": 4.3515625,
+      "learning_rate": 1.9920687892524676e-05,
+      "loss": 0.714,
+      "mean_token_accuracy": 0.7859313750267028,
+      "num_tokens": 3803593.0,
+      "step": 1250
+    },
+    {
+      "epoch": 0.04342234914908896,
+      "grad_norm": 3.603515625,
+      "learning_rate": 1.9913944814695285e-05,
+      "loss": 0.6818,
+      "mean_token_accuracy": 0.7935056626796723,
+      "num_tokens": 3954344.0,
+      "step": 1300
+    },
+    {
+      "epoch": 0.045092439500977,
+      "grad_norm": 3.8828125,
+      "learning_rate": 1.99069279002353e-05,
+      "loss": 0.6986,
+      "mean_token_accuracy": 0.7902963298559189,
+      "num_tokens": 4109865.0,
+      "step": 1350
+    },
+    {
+      "epoch": 0.04676252985286504,
+      "grad_norm": 4.125,
+      "learning_rate": 1.9899637342961434e-05,
+      "loss": 0.6922,
+      "mean_token_accuracy": 0.7937425565719605,
+      "num_tokens": 4258123.0,
+      "step": 1400
+    },
+    {
+      "epoch": 0.04843262020475308,
+      "grad_norm": 4.23046875,
+      "learning_rate": 1.9892073344248794e-05,
+      "loss": 0.6954,
+      "mean_token_accuracy": 0.7912928277254104,
+      "num_tokens": 4404817.0,
+      "step": 1450
+    },
+    {
+      "epoch": 0.05010271055664112,
+      "grad_norm": 3.623046875,
+      "learning_rate": 1.98842361130253e-05,
+      "loss": 0.6794,
+      "mean_token_accuracy": 0.7929424160718918,
+      "num_tokens": 4552500.0,
+      "step": 1500
+    },
+    {
+      "epoch": 0.05177280090852915,
+      "grad_norm": 4.046875,
+      "learning_rate": 1.987612586576593e-05,
+      "loss": 0.7069,
+      "mean_token_accuracy": 0.7885507994890213,
+      "num_tokens": 4700529.0,
+      "step": 1550
+    },
+    {
+      "epoch": 0.05344289126041719,
+      "grad_norm": 3.5078125,
+      "learning_rate": 1.986774282648674e-05,
+      "loss": 0.6928,
+      "mean_token_accuracy": 0.7903753596544266,
+      "num_tokens": 4854159.0,
+      "step": 1600
+    },
+    {
+      "epoch": 0.055112981612305226,
+      "grad_norm": 3.23828125,
+      "learning_rate": 1.9859087226738667e-05,
+      "loss": 0.6862,
+      "mean_token_accuracy": 0.7946210086345673,
+      "num_tokens": 5004881.0,
+      "step": 1650
+    },
+    {
+      "epoch": 0.056783071964193264,
+      "grad_norm": 3.8828125,
+      "learning_rate": 1.9850159305601142e-05,
+      "loss": 0.681,
+      "mean_token_accuracy": 0.7951335823535919,
+      "num_tokens": 5155074.0,
+      "step": 1700
+    },
+    {
+      "epoch": 0.0584531623160813,
+      "grad_norm": 3.435546875,
+      "learning_rate": 1.984095930967548e-05,
+      "loss": 0.6775,
+      "mean_token_accuracy": 0.7948978447914123,
+      "num_tokens": 5305327.0,
+      "step": 1750
+    },
+    {
+      "epoch": 0.060123252667969335,
+      "grad_norm": 4.3125,
+      "learning_rate": 1.983148749307807e-05,
+      "loss": 0.6815,
+      "mean_token_accuracy": 0.7944793283939362,
+      "num_tokens": 5453165.0,
+      "step": 1800
+    },
+    {
+      "epoch": 0.06179334301985737,
+      "grad_norm": 3.498046875,
+      "learning_rate": 1.9821744117433367e-05,
+      "loss": 0.6737,
+      "mean_token_accuracy": 0.7980403685569764,
+      "num_tokens": 5605423.0,
+      "step": 1850
+    },
+    {
+      "epoch": 0.06346343337174541,
+      "grad_norm": 4.46484375,
+      "learning_rate": 1.9811729451866647e-05,
+      "loss": 0.6914,
+      "mean_token_accuracy": 0.7934044080972672,
+      "num_tokens": 5751905.0,
+      "step": 1900
+    },
+    {
+      "epoch": 0.06513352372363344,
+      "grad_norm": 3.724609375,
+      "learning_rate": 1.980144377299658e-05,
+      "loss": 0.6879,
+      "mean_token_accuracy": 0.7944984430074692,
+      "num_tokens": 5900554.0,
+      "step": 1950
+    },
+    {
+      "epoch": 0.06680361407552149,
+      "grad_norm": 3.11328125,
+      "learning_rate": 1.9790887364927603e-05,
+      "loss": 0.6887,
+      "mean_token_accuracy": 0.7929914563894271,
+      "num_tokens": 6057264.0,
+      "step": 2000
+    },
+    {
+      "epoch": 0.06847370442740952,
+      "grad_norm": 3.974609375,
+      "learning_rate": 1.9780060519242047e-05,
+      "loss": 0.6795,
+      "mean_token_accuracy": 0.7947278439998626,
+      "num_tokens": 6207563.0,
+      "step": 2050
+    },
+    {
+      "epoch": 0.07014379477929757,
+      "grad_norm": 3.337890625,
+      "learning_rate": 1.9768963534992114e-05,
+      "loss": 0.695,
+      "mean_token_accuracy": 0.7921637284755707,
+      "num_tokens": 6362024.0,
+      "step": 2100
+    },
+    {
+      "epoch": 0.0718138851311856,
+      "grad_norm": 4.35546875,
+      "learning_rate": 1.975759671869159e-05,
+      "loss": 0.6813,
+      "mean_token_accuracy": 0.7945687365531922,
+      "num_tokens": 6511612.0,
+      "step": 2150
+    },
+    {
+      "epoch": 0.07348397548307363,
+      "grad_norm": 3.443359375,
+      "learning_rate": 1.9745960384307384e-05,
+      "loss": 0.6664,
+      "mean_token_accuracy": 0.7972371423244476,
+      "num_tokens": 6664163.0,
+      "step": 2200
+    },
+    {
+      "epoch": 0.07515406583496168,
+      "grad_norm": 3.927734375,
+      "learning_rate": 1.9734054853250876e-05,
+      "loss": 0.6745,
+      "mean_token_accuracy": 0.7965255242586136,
+      "num_tokens": 6818983.0,
+      "step": 2250
+    },
+    {
+      "epoch": 0.0768241561868497,
+      "grad_norm": 4.40625,
+      "learning_rate": 1.9721880454369012e-05,
+      "loss": 0.6775,
+      "mean_token_accuracy": 0.7949534428119659,
+      "num_tokens": 6969178.0,
+      "step": 2300
+    },
+    {
+      "epoch": 0.07849424653873775,
+      "grad_norm": 3.94140625,
+      "learning_rate": 1.9709437523935234e-05,
+      "loss": 0.6933,
+      "mean_token_accuracy": 0.7906970298290252,
+      "num_tokens": 7120862.0,
+      "step": 2350
+    },
+    {
+      "epoch": 0.08016433689062578,
+      "grad_norm": 4.1484375,
+      "learning_rate": 1.9696726405640194e-05,
+      "loss": 0.6678,
+      "mean_token_accuracy": 0.7977210181951523,
+      "num_tokens": 7273755.0,
+      "step": 2400
+    },
+    {
+      "epoch": 0.08183442724251382,
+      "grad_norm": 3.671875,
+      "learning_rate": 1.9683747450582257e-05,
+      "loss": 0.7022,
+      "mean_token_accuracy": 0.7893240767717361,
+      "num_tokens": 7423742.0,
+      "step": 2450
+    },
+    {
+      "epoch": 0.08350451759440186,
+      "grad_norm": 3.5078125,
+      "learning_rate": 1.9670501017257805e-05,
+      "loss": 0.6847,
+      "mean_token_accuracy": 0.7929366368055344,
+      "num_tokens": 7574222.0,
+      "step": 2500
+    },
+    {
+      "epoch": 0.0851746079462899,
+      "grad_norm": 3.384765625,
+      "learning_rate": 1.965698747155133e-05,
+      "loss": 0.6571,
+      "mean_token_accuracy": 0.7997360718250275,
+      "num_tokens": 7730085.0,
+      "step": 2550
+    },
+    {
+      "epoch": 0.08684469829817792,
+      "grad_norm": 3.466796875,
+      "learning_rate": 1.9643207186725324e-05,
+      "loss": 0.6672,
+      "mean_token_accuracy": 0.7982296937704086,
+      "num_tokens": 7881768.0,
+      "step": 2600
+    },
+    {
+      "epoch": 0.08851478865006597,
+      "grad_norm": 3.521484375,
+      "learning_rate": 1.962916054340999e-05,
+      "loss": 0.685,
+      "mean_token_accuracy": 0.7952978044748307,
+      "num_tokens": 8027897.0,
+      "step": 2650
+    },
+    {
+      "epoch": 0.090184879001954,
+      "grad_norm": 3.79296875,
+      "learning_rate": 1.961484792959271e-05,
+      "loss": 0.6778,
+      "mean_token_accuracy": 0.7966848593950272,
+      "num_tokens": 8175954.0,
+      "step": 2700
+    },
+    {
+      "epoch": 0.09185496935384205,
+      "grad_norm": 3.40234375,
+      "learning_rate": 1.9600269740607328e-05,
+      "loss": 0.6488,
+      "mean_token_accuracy": 0.8027435272932053,
+      "num_tokens": 8332915.0,
+      "step": 2750
+    },
+    {
+      "epoch": 0.09352505970573008,
+      "grad_norm": 3.408203125,
+      "learning_rate": 1.958542637912324e-05,
+      "loss": 0.673,
+      "mean_token_accuracy": 0.7967682188749313,
+      "num_tokens": 8485109.0,
+      "step": 2800
+    },
+    {
+      "epoch": 0.09519515005761811,
+      "grad_norm": 3.244140625,
+      "learning_rate": 1.9570318255134263e-05,
+      "loss": 0.6544,
+      "mean_token_accuracy": 0.7994941908121109,
+      "num_tokens": 8638846.0,
+      "step": 2850
+    },
+    {
+      "epoch": 0.09686524040950616,
+      "grad_norm": 3.62109375,
+      "learning_rate": 1.9554945785947314e-05,
+      "loss": 0.67,
+      "mean_token_accuracy": 0.7985292530059814,
+      "num_tokens": 8788759.0,
+      "step": 2900
+    },
+    {
+      "epoch": 0.09853533076139419,
+      "grad_norm": 3.466796875,
+      "learning_rate": 1.953930939617089e-05,
+      "loss": 0.6722,
+      "mean_token_accuracy": 0.7943358135223388,
+      "num_tokens": 8941895.0,
+      "step": 2950
+    },
+    {
+      "epoch": 0.10020542111328223,
+      "grad_norm": 3.6796875,
+      "learning_rate": 1.952340951770333e-05,
+      "loss": 0.6488,
+      "mean_token_accuracy": 0.8029305410385131,
+      "num_tokens": 9093399.0,
+      "step": 3000
+    },
+    {
+      "epoch": 0.10187551146517027,
+      "grad_norm": 3.673828125,
+      "learning_rate": 1.9507246589720887e-05,
+      "loss": 0.6591,
+      "mean_token_accuracy": 0.7998628175258636,
+      "num_tokens": 9246318.0,
+      "step": 3050
+    },
+    {
+      "epoch": 0.1035456018170583,
+      "grad_norm": 3.796875,
+      "learning_rate": 1.9490821058665597e-05,
+      "loss": 0.6842,
+      "mean_token_accuracy": 0.7952588319778442,
+      "num_tokens": 9393063.0,
+      "step": 3100
+    },
+    {
+      "epoch": 0.10521569216894634,
+      "grad_norm": 3.64453125,
+      "learning_rate": 1.9474133378232955e-05,
+      "loss": 0.6627,
+      "mean_token_accuracy": 0.7986302423477173,
+      "num_tokens": 9546557.0,
+      "step": 3150
+    },
+    {
+      "epoch": 0.10688578252083437,
+      "grad_norm": 3.662109375,
+      "learning_rate": 1.945718400935938e-05,
+      "loss": 0.6564,
+      "mean_token_accuracy": 0.8012412291765213,
+      "num_tokens": 9698417.0,
+      "step": 3200
+    },
+    {
+      "epoch": 0.10855587287272242,
+      "grad_norm": 3.21484375,
+      "learning_rate": 1.9439973420209467e-05,
+      "loss": 0.67,
+      "mean_token_accuracy": 0.7971057319641113,
+      "num_tokens": 9846015.0,
+      "step": 3250
+    },
+    {
+      "epoch": 0.11022596322461045,
+      "grad_norm": 3.294921875,
+      "learning_rate": 1.9422502086163093e-05,
+      "loss": 0.6672,
+      "mean_token_accuracy": 0.797788228392601,
+      "num_tokens": 10001053.0,
+      "step": 3300
+    },
+    {
+      "epoch": 0.11189605357649848,
+      "grad_norm": 3.607421875,
+      "learning_rate": 1.9404770489802242e-05,
+      "loss": 0.6529,
+      "mean_token_accuracy": 0.7996513998508453,
+      "num_tokens": 10151845.0,
+      "step": 3350
+    },
+    {
+      "epoch": 0.11356614392838653,
+      "grad_norm": 3.240234375,
+      "learning_rate": 1.9386779120897712e-05,
+      "loss": 0.6268,
+      "mean_token_accuracy": 0.8078747218847275,
+      "num_tokens": 10303872.0,
+      "step": 3400
+    },
+    {
+      "epoch": 0.11523623428027456,
+      "grad_norm": 3.38671875,
+      "learning_rate": 1.936852847639556e-05,
+      "loss": 0.6608,
+      "mean_token_accuracy": 0.8000670689344406,
+      "num_tokens": 10456152.0,
+      "step": 3450
+    },
+    {
+      "epoch": 0.1169063246321626,
+      "grad_norm": 3.337890625,
+      "learning_rate": 1.9350019060403398e-05,
+      "loss": 0.6469,
+      "mean_token_accuracy": 0.8042119026184082,
+      "num_tokens": 10605108.0,
+      "step": 3500
+    },
+    {
+      "epoch": 0.11857641498405064,
+      "grad_norm": 3.521484375,
+      "learning_rate": 1.933125138417645e-05,
+      "loss": 0.6506,
+      "mean_token_accuracy": 0.8032512718439102,
+      "num_tokens": 10749490.0,
+      "step": 3550
+    },
+    {
+      "epoch": 0.12024650533593867,
+      "grad_norm": 3.251953125,
+      "learning_rate": 1.9312225966103445e-05,
+      "loss": 0.6753,
+      "mean_token_accuracy": 0.7979339599609375,
+      "num_tokens": 10899284.0,
+      "step": 3600
+    },
+    {
+      "epoch": 0.12191659568782671,
+      "grad_norm": 3.548828125,
+      "learning_rate": 1.9292943331692285e-05,
+      "loss": 0.6663,
+      "mean_token_accuracy": 0.7995783388614655,
+      "num_tokens": 11046487.0,
+      "step": 3650
+    },
+    {
+      "epoch": 0.12358668603971475,
+      "grad_norm": 3.27734375,
+      "learning_rate": 1.9273404013555544e-05,
+      "loss": 0.6481,
+      "mean_token_accuracy": 0.8006290888786316,
+      "num_tokens": 11197224.0,
+      "step": 3700
+    },
+    {
+      "epoch": 0.1252567763916028,
+      "grad_norm": 3.7109375,
+      "learning_rate": 1.9253608551395745e-05,
+      "loss": 0.6645,
+      "mean_token_accuracy": 0.7999362868070602,
+      "num_tokens": 11349738.0,
+      "step": 3750
+    },
+    {
+      "epoch": 0.12692686674349082,
+      "grad_norm": 4.05078125,
+      "learning_rate": 1.923355749199046e-05,
+      "loss": 0.644,
+      "mean_token_accuracy": 0.8039193218946457,
+      "num_tokens": 11501057.0,
+      "step": 3800
+    },
+    {
+      "epoch": 0.12859695709537886,
+      "grad_norm": 3.26171875,
+      "learning_rate": 1.92132513891772e-05,
+      "loss": 0.6398,
+      "mean_token_accuracy": 0.8046829622983932,
+      "num_tokens": 11656329.0,
+      "step": 3850
+    },
+    {
+      "epoch": 0.1302670474472669,
+      "grad_norm": 3.41796875,
+      "learning_rate": 1.919269080383813e-05,
+      "loss": 0.6595,
+      "mean_token_accuracy": 0.7993488019704819,
+      "num_tokens": 11800257.0,
+      "step": 3900
+    },
+    {
+      "epoch": 0.13193713779915495,
+      "grad_norm": 3.822265625,
+      "learning_rate": 1.9171876303884555e-05,
+      "loss": 0.6533,
+      "mean_token_accuracy": 0.8011213219165803,
+      "num_tokens": 11952176.0,
+      "step": 3950
+    },
+    {
+      "epoch": 0.13360722815104298,
+      "grad_norm": 3.0859375,
+      "learning_rate": 1.9150808464241257e-05,
+      "loss": 0.6443,
+      "mean_token_accuracy": 0.8050255984067917,
+      "num_tokens": 12103443.0,
+      "step": 4000
+    },
+    {
+      "epoch": 0.135277318502931,
+      "grad_norm": 3.5078125,
+      "learning_rate": 1.9129487866830602e-05,
+      "loss": 0.6533,
+      "mean_token_accuracy": 0.8013356167078018,
+      "num_tokens": 12254997.0,
+      "step": 4050
+    },
+    {
+      "epoch": 0.13694740885481904,
+      "grad_norm": 3.30078125,
+      "learning_rate": 1.9107915100556473e-05,
+      "loss": 0.6532,
+      "mean_token_accuracy": 0.8014611840248108,
+      "num_tokens": 12408985.0,
+      "step": 4100
+    },
+    {
+      "epoch": 0.13861749920670707,
+      "grad_norm": 3.14453125,
+      "learning_rate": 1.908609076128799e-05,
+      "loss": 0.6575,
+      "mean_token_accuracy": 0.8021772617101669,
+      "num_tokens": 12559124.0,
+      "step": 4150
+    },
+    {
+      "epoch": 0.14028758955859513,
+      "grad_norm": 3.515625,
+      "learning_rate": 1.906401545184307e-05,
+      "loss": 0.6475,
+      "mean_token_accuracy": 0.8028253322839737,
+      "num_tokens": 12707796.0,
+      "step": 4200
+    },
+    {
+      "epoch": 0.14195767991048316,
+      "grad_norm": 3.646484375,
+      "learning_rate": 1.9041689781971765e-05,
+      "loss": 0.6506,
+      "mean_token_accuracy": 0.8031806921958924,
+      "num_tokens": 12858676.0,
+      "step": 4250
+    },
+    {
+      "epoch": 0.1436277702623712,
+      "grad_norm": 4.37109375,
+      "learning_rate": 1.901911436833942e-05,
+      "loss": 0.6558,
+      "mean_token_accuracy": 0.8009441828727722,
+      "num_tokens": 13007021.0,
+      "step": 4300
+    },
+    {
+      "epoch": 0.14529786061425923,
+      "grad_norm": 3.494140625,
+      "learning_rate": 1.899628983450964e-05,
+      "loss": 0.629,
+      "mean_token_accuracy": 0.8062937146425248,
+      "num_tokens": 13155780.0,
+      "step": 4350
+    },
+    {
+      "epoch": 0.14696795096614726,
+      "grad_norm": 2.982421875,
+      "learning_rate": 1.8973216810927077e-05,
+      "loss": 0.6461,
+      "mean_token_accuracy": 0.8051383906602859,
+      "num_tokens": 13302653.0,
+      "step": 4400
+    },
+    {
+      "epoch": 0.14863804131803532,
+      "grad_norm": 3.529296875,
+      "learning_rate": 1.89498959349e-05,
+      "loss": 0.6367,
+      "mean_token_accuracy": 0.805459765791893,
+      "num_tokens": 13455088.0,
+      "step": 4450
+    },
+    {
+      "epoch": 0.15030813166992335,
+      "grad_norm": 3.228515625,
+      "learning_rate": 1.8926327850582706e-05,
+      "loss": 0.6288,
+      "mean_token_accuracy": 0.8068876641988755,
+      "num_tokens": 13609983.0,
+      "step": 4500
+    },
+    {
+      "epoch": 0.15197822202181138,
+      "grad_norm": 3.72265625,
+      "learning_rate": 1.8902513208957712e-05,
+      "loss": 0.6521,
+      "mean_token_accuracy": 0.8004961597919464,
+      "num_tokens": 13759536.0,
+      "step": 4550
+    },
+    {
+      "epoch": 0.1536483123736994,
+      "grad_norm": 3.83203125,
+      "learning_rate": 1.8878452667817785e-05,
+      "loss": 0.6314,
+      "mean_token_accuracy": 0.8060730373859406,
+      "num_tokens": 13914544.0,
+      "step": 4600
+    },
+    {
+      "epoch": 0.15531840272558745,
+      "grad_norm": 3.88671875,
+      "learning_rate": 1.8854146891747788e-05,
+      "loss": 0.6397,
+      "mean_token_accuracy": 0.805094587802887,
+      "num_tokens": 14064759.0,
+      "step": 4650
+    },
+    {
+      "epoch": 0.1569884930774755,
+      "grad_norm": 3.373046875,
+      "learning_rate": 1.8829596552106276e-05,
+      "loss": 0.6477,
+      "mean_token_accuracy": 0.8041272169351578,
+      "num_tokens": 14213937.0,
+      "step": 4700
+    },
+    {
+      "epoch": 0.15865858342936354,
+      "grad_norm": 3.1484375,
+      "learning_rate": 1.8804802327007005e-05,
+      "loss": 0.63,
+      "mean_token_accuracy": 0.8083387905359268,
+      "num_tokens": 14363168.0,
+      "step": 4750
+    },
+    {
+      "epoch": 0.16032867378125157,
+      "grad_norm": 3.6328125,
+      "learning_rate": 1.8779764901300174e-05,
+      "loss": 0.633,
+      "mean_token_accuracy": 0.8068809920549392,
+      "num_tokens": 14511952.0,
+      "step": 4800
+    },
+    {
+      "epoch": 0.1619987641331396,
+      "grad_norm": 3.693359375,
+      "learning_rate": 1.875448496655351e-05,
+      "loss": 0.6529,
+      "mean_token_accuracy": 0.8010576856136322,
+      "num_tokens": 14659521.0,
+      "step": 4850
+    },
+    {
+      "epoch": 0.16366885448502763,
+      "grad_norm": 3.970703125,
+      "learning_rate": 1.8728963221033168e-05,
+      "loss": 0.6299,
+      "mean_token_accuracy": 0.8058782732486724,
+      "num_tokens": 14812856.0,
+      "step": 4900
+    },
+    {
+      "epoch": 0.16533894483691566,
+      "grad_norm": 3.056640625,
+      "learning_rate": 1.870320036968445e-05,
+      "loss": 0.6102,
+      "mean_token_accuracy": 0.8117638313770295,
+      "num_tokens": 14968517.0,
+      "step": 4950
+    },
+    {
+      "epoch": 0.16700903518880372,
+      "grad_norm": 3.572265625,
+      "learning_rate": 1.867719712411233e-05,
+      "loss": 0.6454,
+      "mean_token_accuracy": 0.8036060810089112,
+      "num_tokens": 15120586.0,
+      "step": 5000
+    },
+    {
+      "epoch": 0.16867912554069175,
+      "grad_norm": 3.0390625,
+      "learning_rate": 1.8650954202561786e-05,
+      "loss": 0.6341,
+      "mean_token_accuracy": 0.8054838758707047,
+      "num_tokens": 15272994.0,
+      "step": 5050
+    },
+    {
+      "epoch": 0.1703492158925798,
+      "grad_norm": 2.947265625,
+      "learning_rate": 1.8624472329897985e-05,
+      "loss": 0.6328,
+      "mean_token_accuracy": 0.8070685297250748,
+      "num_tokens": 15423469.0,
+      "step": 5100
+    },
+    {
+      "epoch": 0.17201930624446782,
+      "grad_norm": 3.240234375,
+      "learning_rate": 1.8597752237586242e-05,
+      "loss": 0.6228,
+      "mean_token_accuracy": 0.809105772972107,
+      "num_tokens": 15568404.0,
+      "step": 5150
+    },
+    {
+      "epoch": 0.17368939659635585,
+      "grad_norm": 3.26171875,
+      "learning_rate": 1.8570794663671824e-05,
+      "loss": 0.6128,
+      "mean_token_accuracy": 0.811137170791626,
+      "num_tokens": 15721906.0,
+      "step": 5200
+    },
+    {
+      "epoch": 0.1753594869482439,
+      "grad_norm": 3.345703125,
+      "learning_rate": 1.8543600352759563e-05,
+      "loss": 0.6207,
+      "mean_token_accuracy": 0.8088931626081467,
+      "num_tokens": 15867611.0,
+      "step": 5250
+    },
+    {
+      "epoch": 0.17702957730013194,
+      "grad_norm": 3.76171875,
+      "learning_rate": 1.851617005599329e-05,
+      "loss": 0.6198,
+      "mean_token_accuracy": 0.810358560681343,
+      "num_tokens": 16017301.0,
+      "step": 5300
+    },
+    {
+      "epoch": 0.17869966765201997,
+      "grad_norm": 3.484375,
+      "learning_rate": 1.8488504531035084e-05,
+      "loss": 0.6354,
+      "mean_token_accuracy": 0.806412605047226,
+      "num_tokens": 16174153.0,
+      "step": 5350
+    },
+    {
+      "epoch": 0.180369758003908,
+      "grad_norm": 3.6015625,
+      "learning_rate": 1.8460604542044344e-05,
+      "loss": 0.6206,
+      "mean_token_accuracy": 0.809535750746727,
+      "num_tokens": 16326166.0,
+      "step": 5400
+    },
+    {
+      "epoch": 0.18203984835579604,
+      "grad_norm": 4.4765625,
+      "learning_rate": 1.8432470859656687e-05,
+      "loss": 0.6399,
+      "mean_token_accuracy": 0.8042701333761215,
+      "num_tokens": 16476025.0,
+      "step": 5450
+    },
+    {
+      "epoch": 0.1837099387076841,
+      "grad_norm": 3.203125,
+      "learning_rate": 1.8404104260962662e-05,
+      "loss": 0.6148,
+      "mean_token_accuracy": 0.811469207406044,
+      "num_tokens": 16622887.0,
+      "step": 5500
+    },
+    {
+      "epoch": 0.18538002905957213,
+      "grad_norm": 3.763671875,
+      "learning_rate": 1.8375505529486283e-05,
+      "loss": 0.6266,
+      "mean_token_accuracy": 0.8079740327596664,
+      "num_tokens": 16772669.0,
+      "step": 5550
+    },
+    {
+      "epoch": 0.18705011941146016,
+      "grad_norm": 3.28515625,
+      "learning_rate": 1.8346675455163376e-05,
+      "loss": 0.6151,
+      "mean_token_accuracy": 0.8107228165864945,
+      "num_tokens": 16920269.0,
+      "step": 5600
+    },
+    {
+      "epoch": 0.1887202097633482,
+      "grad_norm": 4.6015625,
+      "learning_rate": 1.8317614834319783e-05,
+      "loss": 0.6321,
+      "mean_token_accuracy": 0.8072718501091003,
+      "num_tokens": 17067401.0,
+      "step": 5650
+    },
+    {
+      "epoch": 0.19039030011523622,
+      "grad_norm": 4.953125,
+      "learning_rate": 1.828832446964935e-05,
+      "loss": 0.635,
+      "mean_token_accuracy": 0.8083576399087906,
+      "num_tokens": 17210399.0,
+      "step": 5700
+    },
+    {
+      "epoch": 0.19206039046712428,
+      "grad_norm": 3.9453125,
+      "learning_rate": 1.825880517019176e-05,
+      "loss": 0.6101,
+      "mean_token_accuracy": 0.8119984632730484,
+      "num_tokens": 17355215.0,
+      "step": 5750
+    },
+    {
+      "epoch": 0.1937304808190123,
+      "grad_norm": 3.5390625,
+      "learning_rate": 1.8229057751310188e-05,
+      "loss": 0.6179,
+      "mean_token_accuracy": 0.8104551881551743,
+      "num_tokens": 17501111.0,
+      "step": 5800
+    },
+    {
+      "epoch": 0.19540057117090034,
+      "grad_norm": 3.25,
+      "learning_rate": 1.8199083034668774e-05,
+      "loss": 0.6245,
+      "mean_token_accuracy": 0.8110457295179367,
+      "num_tokens": 17655096.0,
+      "step": 5850
+    },
+    {
+      "epoch": 0.19707066152278838,
+      "grad_norm": 3.59375,
+      "learning_rate": 1.816888184820993e-05,
+      "loss": 0.6302,
+      "mean_token_accuracy": 0.8070375627279281,
+      "num_tokens": 17809403.0,
+      "step": 5900
+    },
+    {
+      "epoch": 0.1987407518746764,
+      "grad_norm": 3.259765625,
+      "learning_rate": 1.8138455026131477e-05,
+      "loss": 0.6078,
+      "mean_token_accuracy": 0.8126195806264878,
+      "num_tokens": 17959815.0,
+      "step": 5950
+    },
+    {
+      "epoch": 0.20041084222656447,
+      "grad_norm": 3.291015625,
+      "learning_rate": 1.81078034088636e-05,
+      "loss": 0.6,
+      "mean_token_accuracy": 0.8153526914119721,
+      "num_tokens": 18112482.0,
+      "step": 6000
+    },
+    {
+      "epoch": 0.2020809325784525,
+      "grad_norm": 3.619140625,
+      "learning_rate": 1.807692784304562e-05,
+      "loss": 0.6088,
+      "mean_token_accuracy": 0.8110599368810654,
+      "num_tokens": 18269537.0,
+      "step": 6050
+    },
+    {
+      "epoch": 0.20375102293034053,
+      "grad_norm": 3.349609375,
+      "learning_rate": 1.8045829181502642e-05,
+      "loss": 0.62,
+      "mean_token_accuracy": 0.8105544656515121,
+      "num_tokens": 18418714.0,
+      "step": 6100
+    },
+    {
+      "epoch": 0.20542111328222856,
+      "grad_norm": 3.341796875,
+      "learning_rate": 1.8014508283221953e-05,
+      "loss": 0.6249,
+      "mean_token_accuracy": 0.807090380191803,
+      "num_tokens": 18577084.0,
+      "step": 6150
+    },
+    {
+      "epoch": 0.2070912036341166,
+      "grad_norm": 3.5546875,
+      "learning_rate": 1.798296601332935e-05,
+      "loss": 0.6166,
+      "mean_token_accuracy": 0.8106195080280304,
+      "num_tokens": 18723440.0,
+      "step": 6200
+    },
+    {
+      "epoch": 0.20876129398600465,
+      "grad_norm": 3.4921875,
+      "learning_rate": 1.7951203243065184e-05,
+      "loss": 0.6169,
+      "mean_token_accuracy": 0.8116037601232529,
+      "num_tokens": 18870030.0,
+      "step": 6250
+    },
+    {
+      "epoch": 0.21043138433789269,
+      "grad_norm": 3.80859375,
+      "learning_rate": 1.7919220849760347e-05,
+      "loss": 0.6093,
+      "mean_token_accuracy": 0.8129760199785232,
+      "num_tokens": 19020540.0,
+      "step": 6300
+    },
+    {
+      "epoch": 0.21210147468978072,
+      "grad_norm": 3.212890625,
+      "learning_rate": 1.7887019716812015e-05,
+      "loss": 0.6202,
+      "mean_token_accuracy": 0.8104865264892578,
+      "num_tokens": 19171916.0,
+      "step": 6350
+    },
+    {
+      "epoch": 0.21377156504166875,
+      "grad_norm": 3.998046875,
+      "learning_rate": 1.785460073365924e-05,
+      "loss": 0.6243,
+      "mean_token_accuracy": 0.810001392364502,
+      "num_tokens": 19317653.0,
+      "step": 6400
+    },
+    {
+      "epoch": 0.21544165539355678,
+      "grad_norm": 3.6015625,
+      "learning_rate": 1.782196479575841e-05,
+      "loss": 0.5972,
+      "mean_token_accuracy": 0.815451757311821,
+      "num_tokens": 19469488.0,
+      "step": 6450
+    },
+    {
+      "epoch": 0.21711174574544484,
+      "grad_norm": 3.171875,
+      "learning_rate": 1.778911280455848e-05,
+      "loss": 0.6094,
+      "mean_token_accuracy": 0.8127148807048797,
+      "num_tokens": 19617524.0,
+      "step": 6500
+    },
+    {
+      "epoch": 0.21878183609733287,
+      "grad_norm": 3.478515625,
+      "learning_rate": 1.7756045667476095e-05,
+      "loss": 0.6365,
+      "mean_token_accuracy": 0.8052962213754654,
+      "num_tokens": 19766937.0,
+      "step": 6550
+    },
+    {
+      "epoch": 0.2204519264492209,
+      "grad_norm": 3.943359375,
+      "learning_rate": 1.7722764297870534e-05,
+      "loss": 0.6008,
+      "mean_token_accuracy": 0.8136671632528305,
+      "num_tokens": 19918307.0,
+      "step": 6600
+    },
+    {
+      "epoch": 0.22212201680110893,
+      "grad_norm": 3.169921875,
+      "learning_rate": 1.768926961501846e-05,
+      "loss": 0.6066,
+      "mean_token_accuracy": 0.8141713869571686,
+      "num_tokens": 20066541.0,
+      "step": 6650
+    },
+    {
+      "epoch": 0.22379210715299697,
+      "grad_norm": 3.19921875,
+      "learning_rate": 1.7655562544088534e-05,
+      "loss": 0.6034,
+      "mean_token_accuracy": 0.8118478727340698,
+      "num_tokens": 20221668.0,
+      "step": 6700
+    },
+    {
+      "epoch": 0.22546219750488503,
+      "grad_norm": 3.275390625,
+      "learning_rate": 1.762164401611587e-05,
+      "loss": 0.6189,
+      "mean_token_accuracy": 0.8125155258178711,
+      "num_tokens": 20368732.0,
+      "step": 6750
+    },
+    {
+      "epoch": 0.22713228785677306,
+      "grad_norm": 3.462890625,
+      "learning_rate": 1.7587514967976315e-05,
+      "loss": 0.6095,
+      "mean_token_accuracy": 0.8130715191364288,
+      "num_tokens": 20519573.0,
+      "step": 6800
+    },
+    {
+      "epoch": 0.2288023782086611,
+      "grad_norm": 3.05859375,
+      "learning_rate": 1.7553176342360564e-05,
+      "loss": 0.6381,
+      "mean_token_accuracy": 0.8038551956415176,
+      "num_tokens": 20668027.0,
+      "step": 6850
+    },
+    {
+      "epoch": 0.23047246856054912,
+      "grad_norm": 3.392578125,
+      "learning_rate": 1.7518629087748125e-05,
+      "loss": 0.6147,
+      "mean_token_accuracy": 0.8109205877780914,
+      "num_tokens": 20815967.0,
+      "step": 6900
+    },
+    {
+      "epoch": 0.23214255891243715,
+      "grad_norm": 3.642578125,
+      "learning_rate": 1.7483874158381137e-05,
+      "loss": 0.6093,
+      "mean_token_accuracy": 0.812092969417572,
+      "num_tokens": 20966880.0,
+      "step": 6950
+    },
+    {
+      "epoch": 0.2338126492643252,
+      "grad_norm": 3.2578125,
+      "learning_rate": 1.7448912514237976e-05,
+      "loss": 0.5986,
+      "mean_token_accuracy": 0.8138788425922394,
+      "num_tokens": 21117408.0,
+      "step": 7000
+    },
+    {
+      "epoch": 0.23548273961621324,
+      "grad_norm": 3.28125,
+      "learning_rate": 1.7413745121006777e-05,
+      "loss": 0.6143,
+      "mean_token_accuracy": 0.8106629079580308,
+      "num_tokens": 21258290.0,
+      "step": 7050
+    },
+    {
+      "epoch": 0.23715282996810128,
+      "grad_norm": 3.671875,
+      "learning_rate": 1.737837295005874e-05,
+      "loss": 0.6026,
+      "mean_token_accuracy": 0.8134672123193741,
+      "num_tokens": 21413317.0,
+      "step": 7100
+    },
+    {
+      "epoch": 0.2388229203199893,
+      "grad_norm": 3.42578125,
+      "learning_rate": 1.73427969784213e-05,
+      "loss": 0.5989,
+      "mean_token_accuracy": 0.81451107442379,
+      "num_tokens": 21559517.0,
+      "step": 7150
+    },
+    {
+      "epoch": 0.24049301067187734,
+      "grad_norm": 3.2890625,
+      "learning_rate": 1.7307018188751153e-05,
+      "loss": 0.6134,
+      "mean_token_accuracy": 0.8111221677064896,
+      "num_tokens": 21711355.0,
+      "step": 7200
+    },
+    {
+      "epoch": 0.2421631010237654,
+      "grad_norm": 3.314453125,
+      "learning_rate": 1.72710375693071e-05,
+      "loss": 0.5866,
+      "mean_token_accuracy": 0.8179266756772995,
+      "num_tokens": 21863362.0,
+      "step": 7250
+    },
+    {
+      "epoch": 0.24383319137565343,
+      "grad_norm": 3.07421875,
+      "learning_rate": 1.723485611392275e-05,
+      "loss": 0.5775,
+      "mean_token_accuracy": 0.8185840421915054,
+      "num_tokens": 22015880.0,
+      "step": 7300
+    },
+    {
+      "epoch": 0.24550328172754146,
+      "grad_norm": 3.33203125,
+      "learning_rate": 1.7198474821979075e-05,
+      "loss": 0.6035,
+      "mean_token_accuracy": 0.814561493396759,
+      "num_tokens": 22164719.0,
+      "step": 7350
+    },
+    {
+      "epoch": 0.2471733720794295,
+      "grad_norm": 3.439453125,
+      "learning_rate": 1.7161894698376814e-05,
+      "loss": 0.5944,
+      "mean_token_accuracy": 0.8162802219390869,
+      "num_tokens": 22316748.0,
+      "step": 7400
+    },
+    {
+      "epoch": 0.24884346243131752,
+      "grad_norm": 3.787109375,
+      "learning_rate": 1.7125116753508696e-05,
+      "loss": 0.5999,
+      "mean_token_accuracy": 0.8156676411628723,
+      "num_tokens": 22466344.0,
+      "step": 7450
+    },
+    {
+      "epoch": 0.2505135527832056,
+      "grad_norm": 3.5625,
+      "learning_rate": 1.708814200323155e-05,
+      "loss": 0.5842,
+      "mean_token_accuracy": 0.8202319222688675,
+      "num_tokens": 22619286.0,
+      "step": 7500
+    },
+    {
+      "epoch": 0.2521836431350936,
+      "grad_norm": 3.37890625,
+      "learning_rate": 1.705097146883823e-05,
+      "loss": 0.6108,
+      "mean_token_accuracy": 0.8124721819162368,
+      "num_tokens": 22769773.0,
+      "step": 7550
+    },
+    {
+      "epoch": 0.25385373348698165,
+      "grad_norm": 3.431640625,
+      "learning_rate": 1.7013606177029417e-05,
+      "loss": 0.6097,
+      "mean_token_accuracy": 0.812991201877594,
+      "num_tokens": 22923130.0,
+      "step": 7600
+    },
+    {
+      "epoch": 0.2555238238388697,
+      "grad_norm": 3.142578125,
+      "learning_rate": 1.6976047159885258e-05,
+      "loss": 0.5725,
+      "mean_token_accuracy": 0.8219118654727936,
+      "num_tokens": 23072994.0,
+      "step": 7650
+    },
+    {
+      "epoch": 0.2571939141907577,
+      "grad_norm": 2.869140625,
+      "learning_rate": 1.6938295454836855e-05,
+      "loss": 0.5903,
+      "mean_token_accuracy": 0.8173499464988708,
+      "num_tokens": 23220605.0,
+      "step": 7700
+    },
+    {
+      "epoch": 0.25886400454264574,
+      "grad_norm": 4.078125,
+      "learning_rate": 1.6900352104637616e-05,
+      "loss": 0.6188,
+      "mean_token_accuracy": 0.8102790033817291,
+      "num_tokens": 23370496.0,
+      "step": 7750
+    },
+    {
+      "epoch": 0.2605340948945338,
+      "grad_norm": 3.7421875,
+      "learning_rate": 1.686221815733444e-05,
+      "loss": 0.6076,
+      "mean_token_accuracy": 0.8124845671653748,
+      "num_tokens": 23524575.0,
+      "step": 7800
+    },
+    {
+      "epoch": 0.2622041852464218,
+      "grad_norm": 4.11328125,
+      "learning_rate": 1.682389466623879e-05,
+      "loss": 0.5937,
+      "mean_token_accuracy": 0.8181423616409301,
+      "num_tokens": 23673370.0,
+      "step": 7850
+    },
+    {
+      "epoch": 0.2638742755983099,
+      "grad_norm": 3.591796875,
+      "learning_rate": 1.678538268989757e-05,
+      "loss": 0.5962,
+      "mean_token_accuracy": 0.8151667928695678,
+      "num_tokens": 23824567.0,
+      "step": 7900
+    },
+    {
+      "epoch": 0.2655443659501979,
+      "grad_norm": 3.638671875,
+      "learning_rate": 1.6746683292063913e-05,
+      "loss": 0.595,
+      "mean_token_accuracy": 0.816837123632431,
+      "num_tokens": 23979656.0,
+      "step": 7950
+    },
+    {
+      "epoch": 0.26721445630208596,
+      "grad_norm": 3.806640625,
+      "learning_rate": 1.670779754166779e-05,
+      "loss": 0.571,
+      "mean_token_accuracy": 0.8221717000007629,
+      "num_tokens": 24124997.0,
+      "step": 8000
+    },
+    {
+      "epoch": 0.268884546653974,
+      "grad_norm": 4.38671875,
+      "learning_rate": 1.666872651278648e-05,
+      "loss": 0.5823,
+      "mean_token_accuracy": 0.8184084331989289,
+      "num_tokens": 24274928.0,
+      "step": 8050
+    },
+    {
+      "epoch": 0.270554637005862,
+      "grad_norm": 3.314453125,
+      "learning_rate": 1.66294712846149e-05,
+      "loss": 0.5908,
+      "mean_token_accuracy": 0.8170400822162628,
+      "num_tokens": 24425253.0,
+      "step": 8100
+    },
+    {
+      "epoch": 0.27222472735775005,
+      "grad_norm": 3.328125,
+      "learning_rate": 1.6590032941435812e-05,
+      "loss": 0.587,
+      "mean_token_accuracy": 0.8180436098575592,
+      "num_tokens": 24570789.0,
+      "step": 8150
+    },
+    {
+      "epoch": 0.2738948177096381,
+      "grad_norm": 2.984375,
+      "learning_rate": 1.6550412572589857e-05,
+      "loss": 0.5961,
+      "mean_token_accuracy": 0.8165696883201599,
+      "num_tokens": 24728204.0,
+      "step": 8200
+    },
+    {
+      "epoch": 0.2755649080615261,
+      "grad_norm": 3.09375,
+      "learning_rate": 1.6510611272445472e-05,
+      "loss": 0.5883,
+      "mean_token_accuracy": 0.8188752412796021,
+      "num_tokens": 24884334.0,
+      "step": 8250
+    },
+    {
+      "epoch": 0.27723499841341415,
+      "grad_norm": 3.611328125,
+      "learning_rate": 1.6470630140368663e-05,
+      "loss": 0.591,
+      "mean_token_accuracy": 0.8173883992433548,
+      "num_tokens": 25029094.0,
+      "step": 8300
+    },
+    {
+      "epoch": 0.2789050887653022,
+      "grad_norm": 3.271484375,
+      "learning_rate": 1.6430470280692644e-05,
+      "loss": 0.5848,
+      "mean_token_accuracy": 0.820589507818222,
+      "num_tokens": 25177586.0,
+      "step": 8350
+    },
+    {
+      "epoch": 0.28057517911719027,
+      "grad_norm": 3.2109375,
+      "learning_rate": 1.639013280268732e-05,
+      "loss": 0.5581,
+      "mean_token_accuracy": 0.8250171238183975,
+      "num_tokens": 25332667.0,
+      "step": 8400
+    },
+    {
+      "epoch": 0.2822452694690783,
+      "grad_norm": 3.505859375,
+      "learning_rate": 1.6349618820528662e-05,
+      "loss": 0.5978,
+      "mean_token_accuracy": 0.8156885832548142,
+      "num_tokens": 25482520.0,
+      "step": 8450
+    },
+    {
+      "epoch": 0.28391535982096633,
+      "grad_norm": 3.703125,
+      "learning_rate": 1.6308929453267918e-05,
+      "loss": 0.5633,
+      "mean_token_accuracy": 0.8238223212957382,
+      "num_tokens": 25639672.0,
+      "step": 8500
+    },
+    {
+      "epoch": 0.28558545017285436,
+      "grad_norm": 3.390625,
+      "learning_rate": 1.6268065824800723e-05,
+      "loss": 0.5888,
+      "mean_token_accuracy": 0.8175674903392792,
+      "num_tokens": 25793282.0,
+      "step": 8550
+    },
+    {
+      "epoch": 0.2872555405247424,
+      "grad_norm": 3.59375,
+      "learning_rate": 1.622702906383603e-05,
+      "loss": 0.579,
+      "mean_token_accuracy": 0.8193080085515976,
+      "num_tokens": 25946540.0,
+      "step": 8600
+    },
+    {
+      "epoch": 0.2889256308766304,
+      "grad_norm": 3.0390625,
+      "learning_rate": 1.6185820303864957e-05,
+      "loss": 0.5738,
+      "mean_token_accuracy": 0.8222082054615021,
+      "num_tokens": 26105903.0,
+      "step": 8650
+    },
+    {
+      "epoch": 0.29059572122851846,
+      "grad_norm": 3.185546875,
+      "learning_rate": 1.614444068312946e-05,
+      "loss": 0.595,
+      "mean_token_accuracy": 0.8175247597694397,
+      "num_tokens": 26254429.0,
+      "step": 8700
+    },
+    {
+      "epoch": 0.2922658115804065,
+      "grad_norm": 2.779296875,
+      "learning_rate": 1.610289134459091e-05,
+      "loss": 0.5878,
+      "mean_token_accuracy": 0.818896831870079,
+      "num_tokens": 26410820.0,
+      "step": 8750
+    },
+    {
+      "epoch": 0.2939359019322945,
+      "grad_norm": 3.423828125,
+      "learning_rate": 1.606117343589851e-05,
+      "loss": 0.5953,
+      "mean_token_accuracy": 0.8172416472434998,
+      "num_tokens": 26561505.0,
+      "step": 8800
+    },
+    {
+      "epoch": 0.29560599228418255,
+      "grad_norm": 3.47265625,
+      "learning_rate": 1.60192881093576e-05,
+      "loss": 0.5691,
+      "mean_token_accuracy": 0.8218266826868057,
+      "num_tokens": 26714736.0,
+      "step": 8850
+    },
+    {
+      "epoch": 0.29727608263607064,
+      "grad_norm": 3.501953125,
+      "learning_rate": 1.5977236521897824e-05,
+      "loss": 0.5887,
+      "mean_token_accuracy": 0.8170481222867966,
+      "num_tokens": 26864415.0,
+      "step": 8900
+    },
+    {
+      "epoch": 0.29894617298795867,
+      "grad_norm": 3.005859375,
+      "learning_rate": 1.593501983504119e-05,
+      "loss": 0.556,
+      "mean_token_accuracy": 0.8250821077823639,
+      "num_tokens": 27012496.0,
+      "step": 8950
+    },
+    {
+      "epoch": 0.3006162633398467,
+      "grad_norm": 3.279296875,
+      "learning_rate": 1.589263921486997e-05,
+      "loss": 0.596,
+      "mean_token_accuracy": 0.8154706001281739,
+      "num_tokens": 27165174.0,
+      "step": 9000
+    },
+    {
+      "epoch": 0.30228635369173473,
+      "grad_norm": 4.0078125,
+      "learning_rate": 1.5850095831994486e-05,
+      "loss": 0.5945,
+      "mean_token_accuracy": 0.8172478175163269,
+      "num_tokens": 27322736.0,
+      "step": 9050
+    },
+    {
+      "epoch": 0.30395644404362276,
+      "grad_norm": 3.69921875,
+      "learning_rate": 1.5807390861520805e-05,
+      "loss": 0.5964,
+      "mean_token_accuracy": 0.8172479033470154,
+      "num_tokens": 27470431.0,
+      "step": 9100
+    },
+    {
+      "epoch": 0.3056265343955108,
+      "grad_norm": 3.1796875,
+      "learning_rate": 1.5764525483018254e-05,
+      "loss": 0.5978,
+      "mean_token_accuracy": 0.8172558772563935,
+      "num_tokens": 27616926.0,
+      "step": 9150
+    },
+    {
+      "epoch": 0.3072966247473988,
+      "grad_norm": 3.384765625,
+      "learning_rate": 1.5721500880486847e-05,
+      "loss": 0.5727,
+      "mean_token_accuracy": 0.8231468296051025,
+      "num_tokens": 27766315.0,
+      "step": 9200
+    },
+    {
+      "epoch": 0.30896671509928686,
+      "grad_norm": 3.265625,
+      "learning_rate": 1.5678318242324587e-05,
+      "loss": 0.6048,
+      "mean_token_accuracy": 0.8166863393783569,
+      "num_tokens": 27912309.0,
+      "step": 9250
+    },
+    {
+      "epoch": 0.3106368054511749,
+      "grad_norm": 3.587890625,
+      "learning_rate": 1.5634978761294627e-05,
+      "loss": 0.5783,
+      "mean_token_accuracy": 0.8223829501867295,
+      "num_tokens": 28068124.0,
+      "step": 9300
+    },
+    {
+      "epoch": 0.3123068958030629,
+      "grad_norm": 3.619140625,
+      "learning_rate": 1.559148363449235e-05,
+      "loss": 0.5815,
+      "mean_token_accuracy": 0.8205723863840103,
+      "num_tokens": 28212240.0,
+      "step": 9350
+    },
+    {
+      "epoch": 0.313976986154951,
+      "grad_norm": 3.5078125,
+      "learning_rate": 1.554783406331226e-05,
+      "loss": 0.5708,
+      "mean_token_accuracy": 0.8217070680856705,
+      "num_tokens": 28361807.0,
+      "step": 9400
+    },
+    {
+      "epoch": 0.31564707650683904,
+      "grad_norm": 3.46875,
+      "learning_rate": 1.5504031253414854e-05,
+      "loss": 0.5704,
+      "mean_token_accuracy": 0.8224059015512466,
+      "num_tokens": 28520675.0,
+      "step": 9450
+    },
+    {
+      "epoch": 0.3173171668587271,
+      "grad_norm": 3.56640625,
+      "learning_rate": 1.5460076414693276e-05,
+      "loss": 0.5947,
+      "mean_token_accuracy": 0.8166932672262192,
+      "num_tokens": 28671012.0,
+      "step": 9500
+    },
+    {
+      "epoch": 0.3189872572106151,
+      "grad_norm": 3.40625,
+      "learning_rate": 1.5415970761239908e-05,
+      "loss": 0.5703,
+      "mean_token_accuracy": 0.823592340350151,
+      "num_tokens": 28826413.0,
+      "step": 9550
+    },
+    {
+      "epoch": 0.32065734756250314,
+      "grad_norm": 3.021484375,
+      "learning_rate": 1.5371715511312846e-05,
+      "loss": 0.5524,
+      "mean_token_accuracy": 0.8269542592763901,
+      "num_tokens": 28973595.0,
+      "step": 9600
+    },
+    {
+      "epoch": 0.32232743791439117,
+      "grad_norm": 4.5625,
+      "learning_rate": 1.5327311887302247e-05,
+      "loss": 0.587,
+      "mean_token_accuracy": 0.8197296333312988,
+      "num_tokens": 29118143.0,
+      "step": 9650
+    },
+    {
+      "epoch": 0.3239975282662792,
+      "grad_norm": 4.28125,
+      "learning_rate": 1.528276111569656e-05,
+      "loss": 0.573,
+      "mean_token_accuracy": 0.8212918490171432,
+      "num_tokens": 29271964.0,
+      "step": 9700
+    },
+    {
+      "epoch": 0.32566761861816723,
+      "grad_norm": 3.27734375,
+      "learning_rate": 1.5238064427048644e-05,
+      "loss": 0.5636,
+      "mean_token_accuracy": 0.8254357093572616,
+      "num_tokens": 29423354.0,
+      "step": 9750
+    },
+    {
+      "epoch": 0.32733770897005526,
+      "grad_norm": 3.314453125,
+      "learning_rate": 1.5193223055941793e-05,
+      "loss": 0.5585,
+      "mean_token_accuracy": 0.8258392840623856,
+      "num_tokens": 29573920.0,
+      "step": 9800
+    },
+    {
+      "epoch": 0.3290077993219433,
+      "grad_norm": 3.05078125,
+      "learning_rate": 1.5148238240955628e-05,
+      "loss": 0.5751,
+      "mean_token_accuracy": 0.8223698770999909,
+      "num_tokens": 29723005.0,
+      "step": 9850
+    },
+    {
+      "epoch": 0.3306778896738313,
+      "grad_norm": 3.27734375,
+      "learning_rate": 1.5103111224631875e-05,
+      "loss": 0.5801,
+      "mean_token_accuracy": 0.8200888675451279,
+      "num_tokens": 29869002.0,
+      "step": 9900
+    },
+    {
+      "epoch": 0.3323479800257194,
+      "grad_norm": 3.923828125,
+      "learning_rate": 1.5057843253440063e-05,
+      "loss": 0.5861,
+      "mean_token_accuracy": 0.8186714911460876,
+      "num_tokens": 30015782.0,
+      "step": 9950
+    },
+    {
+      "epoch": 0.33401807037760745,
+      "grad_norm": 3.38671875,
+      "learning_rate": 1.5012435577743085e-05,
+      "loss": 0.5844,
+      "mean_token_accuracy": 0.821356486082077,
+      "num_tokens": 30161096.0,
+      "step": 10000
+    },
+    {
+      "epoch": 0.3356881607294955,
+      "grad_norm": 4.05078125,
+      "learning_rate": 1.4966889451762658e-05,
+      "loss": 0.5633,
+      "mean_token_accuracy": 0.8244505578279495,
+      "num_tokens": 30310339.0,
+      "step": 10050
+    },
+    {
+      "epoch": 0.3373582510813835,
+      "grad_norm": 3.69921875,
+      "learning_rate": 1.492120613354469e-05,
+      "loss": 0.5602,
+      "mean_token_accuracy": 0.8251171362400055,
+      "num_tokens": 30462338.0,
+      "step": 10100
+    },
+    {
+      "epoch": 0.33902834143327154,
+      "grad_norm": 3.177734375,
+      "learning_rate": 1.487538688492452e-05,
+      "loss": 0.5581,
+      "mean_token_accuracy": 0.825383592247963,
+      "num_tokens": 30614667.0,
+      "step": 10150
+    },
+    {
+      "epoch": 0.3406984317851596,
+      "grad_norm": 3.109375,
+      "learning_rate": 1.482943297149207e-05,
+      "loss": 0.5503,
+      "mean_token_accuracy": 0.8295097362995147,
+      "num_tokens": 30763133.0,
+      "step": 10200
+    },
+    {
+      "epoch": 0.3423685221370476,
+      "grad_norm": 3.802734375,
+      "learning_rate": 1.4783345662556894e-05,
+      "loss": 0.573,
+      "mean_token_accuracy": 0.823209963440895,
+      "num_tokens": 30917281.0,
+      "step": 10250
+    },
+    {
+      "epoch": 0.34403861248893564,
+      "grad_norm": 3.8828125,
+      "learning_rate": 1.47371262311131e-05,
+      "loss": 0.5714,
+      "mean_token_accuracy": 0.8230595916509629,
+      "num_tokens": 31063586.0,
+      "step": 10300
+    },
+    {
+      "epoch": 0.34570870284082367,
+      "grad_norm": 3.333984375,
+      "learning_rate": 1.4690775953804206e-05,
+      "loss": 0.5487,
+      "mean_token_accuracy": 0.8284380221366883,
+      "num_tokens": 31219950.0,
+      "step": 10350
+    },
+    {
+      "epoch": 0.3473787931927117,
+      "grad_norm": 3.666015625,
+      "learning_rate": 1.4644296110887869e-05,
+      "loss": 0.5672,
+      "mean_token_accuracy": 0.8228234505653381,
+      "num_tokens": 31370295.0,
+      "step": 10400
+    },
+    {
+      "epoch": 0.3490488835445998,
+      "grad_norm": 2.96875,
+      "learning_rate": 1.4597687986200529e-05,
+      "loss": 0.5744,
+      "mean_token_accuracy": 0.8230278396606445,
+      "num_tokens": 31517590.0,
+      "step": 10450
+    },
+    {
+      "epoch": 0.3507189738964878,
+      "grad_norm": 3.1875,
+      "learning_rate": 1.4550952867121933e-05,
+      "loss": 0.5654,
+      "mean_token_accuracy": 0.824184655547142,
+      "num_tokens": 31668597.0,
+      "step": 10500
+    },
+    {
+      "epoch": 0.35238906424837585,
+      "grad_norm": 3.6953125,
+      "learning_rate": 1.4504092044539597e-05,
+      "loss": 0.5592,
+      "mean_token_accuracy": 0.8253324276208878,
+      "num_tokens": 31817864.0,
+      "step": 10550
+    },
+    {
+      "epoch": 0.3540591546002639,
+      "grad_norm": 3.45703125,
+      "learning_rate": 1.4457106812813127e-05,
+      "loss": 0.5505,
+      "mean_token_accuracy": 0.827171710729599,
+      "num_tokens": 31973663.0,
+      "step": 10600
+    },
+    {
+      "epoch": 0.3557292449521519,
+      "grad_norm": 3.0234375,
+      "learning_rate": 1.4409998469738486e-05,
+      "loss": 0.5603,
+      "mean_token_accuracy": 0.8264797705411911,
+      "num_tokens": 32122867.0,
+      "step": 10650
+    },
+    {
+      "epoch": 0.35739933530403994,
+      "grad_norm": 3.646484375,
+      "learning_rate": 1.4362768316512143e-05,
+      "loss": 0.5526,
+      "mean_token_accuracy": 0.8274721032381058,
+      "num_tokens": 32271390.0,
+      "step": 10700
+    },
+    {
+      "epoch": 0.359069425655928,
+      "grad_norm": 3.0390625,
+      "learning_rate": 1.431541765769512e-05,
+      "loss": 0.5704,
+      "mean_token_accuracy": 0.8233271539211273,
+      "num_tokens": 32422937.0,
+      "step": 10750
+    },
+    {
+      "epoch": 0.360739516007816,
+      "grad_norm": 3.361328125,
+      "learning_rate": 1.4267947801176972e-05,
+      "loss": 0.5592,
+      "mean_token_accuracy": 0.8274497729539871,
+      "num_tokens": 32575242.0,
+      "step": 10800
+    },
+    {
+      "epoch": 0.36240960635970404,
+      "grad_norm": 3.708984375,
+      "learning_rate": 1.4220360058139658e-05,
+      "loss": 0.563,
+      "mean_token_accuracy": 0.8254307335615159,
+      "num_tokens": 32729992.0,
+      "step": 10850
+    },
+    {
+      "epoch": 0.36407969671159207,
+      "grad_norm": 4.1640625,
+      "learning_rate": 1.4172655743021317e-05,
+      "loss": 0.5382,
+      "mean_token_accuracy": 0.8311225980520248,
+      "num_tokens": 32875452.0,
+      "step": 10900
+    },
+    {
+      "epoch": 0.36574978706348016,
+      "grad_norm": 3.388671875,
+      "learning_rate": 1.4124836173479972e-05,
+      "loss": 0.5514,
+      "mean_token_accuracy": 0.8262929075956345,
+      "num_tokens": 33030316.0,
+      "step": 10950
+    },
+    {
+      "epoch": 0.3674198774153682,
+      "grad_norm": 3.482421875,
+      "learning_rate": 1.4076902670357132e-05,
+      "loss": 0.5602,
+      "mean_token_accuracy": 0.8263766020536423,
+      "num_tokens": 33183725.0,
+      "step": 11000
+    },
+    {
+      "epoch": 0.3690899677672562,
+      "grad_norm": 4.14453125,
+      "learning_rate": 1.4028856557641307e-05,
+      "loss": 0.56,
+      "mean_token_accuracy": 0.8278713476657867,
+      "num_tokens": 33338154.0,
+      "step": 11050
+    },
+    {
+      "epoch": 0.37076005811914425,
+      "grad_norm": 3.494140625,
+      "learning_rate": 1.3980699162431421e-05,
+      "loss": 0.5614,
+      "mean_token_accuracy": 0.8253583014011383,
+      "num_tokens": 33490131.0,
+      "step": 11100
+    },
+    {
+      "epoch": 0.3724301484710323,
+      "grad_norm": 3.3828125,
+      "learning_rate": 1.3932431814900195e-05,
+      "loss": 0.5431,
+      "mean_token_accuracy": 0.8307267421483994,
+      "num_tokens": 33644521.0,
+      "step": 11150
+    },
+    {
+      "epoch": 0.3741002388229203,
+      "grad_norm": 3.28125,
+      "learning_rate": 1.3884055848257369e-05,
+      "loss": 0.5521,
+      "mean_token_accuracy": 0.8285236042737961,
+      "num_tokens": 33800815.0,
+      "step": 11200
+    },
+    {
+      "epoch": 0.37577032917480835,
+      "grad_norm": 3.08203125,
+      "learning_rate": 1.383557259871289e-05,
+      "loss": 0.5586,
+      "mean_token_accuracy": 0.8272363543510437,
+      "num_tokens": 33946173.0,
+      "step": 11250
+    },
+    {
+      "epoch": 0.3774404195266964,
+      "grad_norm": 4.18359375,
+      "learning_rate": 1.3786983405440013e-05,
+      "loss": 0.552,
+      "mean_token_accuracy": 0.8265141195058823,
+      "num_tokens": 34095920.0,
+      "step": 11300
+    },
+    {
+      "epoch": 0.3791105098785844,
+      "grad_norm": 3.103515625,
+      "learning_rate": 1.373828961053829e-05,
+      "loss": 0.5837,
+      "mean_token_accuracy": 0.8216396170854569,
+      "num_tokens": 34248446.0,
+      "step": 11350
+    },
+    {
+      "epoch": 0.38078060023047244,
+      "grad_norm": 3.419921875,
+      "learning_rate": 1.3689492558996529e-05,
+      "loss": 0.5463,
+      "mean_token_accuracy": 0.8293001532554627,
+      "num_tokens": 34400906.0,
+      "step": 11400
+    },
+    {
+      "epoch": 0.38245069058236053,
+      "grad_norm": 3.173828125,
+      "learning_rate": 1.3640593598655611e-05,
+      "loss": 0.5463,
+      "mean_token_accuracy": 0.8295714586973191,
+      "num_tokens": 34553639.0,
+      "step": 11450
+    },
+    {
+      "epoch": 0.38412078093424856,
+      "grad_norm": 3.44140625,
+      "learning_rate": 1.3591594080171286e-05,
+      "loss": 0.5626,
+      "mean_token_accuracy": 0.8255741602182388,
+      "num_tokens": 34705512.0,
+      "step": 11500
+    },
+    {
+      "epoch": 0.3857908712861366,
+      "grad_norm": 2.962890625,
+      "learning_rate": 1.3542495356976851e-05,
+      "loss": 0.5529,
+      "mean_token_accuracy": 0.8291010856628418,
+      "num_tokens": 34851485.0,
+      "step": 11550
+    },
+    {
+      "epoch": 0.3874609616380246,
+      "grad_norm": 3.091796875,
+      "learning_rate": 1.3493298785245778e-05,
+      "loss": 0.5588,
+      "mean_token_accuracy": 0.826916481256485,
+      "num_tokens": 35008468.0,
+      "step": 11600
+    },
+    {
+      "epoch": 0.38913105198991266,
+      "grad_norm": 3.2109375,
+      "learning_rate": 1.344400572385424e-05,
+      "loss": 0.5458,
+      "mean_token_accuracy": 0.8301201510429382,
+      "num_tokens": 35157833.0,
+      "step": 11650
+    },
+    {
+      "epoch": 0.3908011423418007,
+      "grad_norm": 4.0859375,
+      "learning_rate": 1.3394617534343596e-05,
+      "loss": 0.5461,
+      "mean_token_accuracy": 0.8286441028118133,
+      "num_tokens": 35310802.0,
+      "step": 11700
+    },
+    {
+      "epoch": 0.3924712326936887,
+      "grad_norm": 3.578125,
+      "learning_rate": 1.3345135580882764e-05,
+      "loss": 0.553,
+      "mean_token_accuracy": 0.8291445529460907,
+      "num_tokens": 35456677.0,
+      "step": 11750
+    },
+    {
+      "epoch": 0.39414132304557675,
+      "grad_norm": 4.30078125,
+      "learning_rate": 1.3295561230230555e-05,
+      "loss": 0.557,
+      "mean_token_accuracy": 0.827009784579277,
+      "num_tokens": 35604720.0,
+      "step": 11800
+    },
+    {
+      "epoch": 0.3958114133974648,
+      "grad_norm": 3.46875,
+      "learning_rate": 1.3245895851697914e-05,
+      "loss": 0.5387,
+      "mean_token_accuracy": 0.832523928284645,
+      "num_tokens": 35758780.0,
+      "step": 11850
+    },
+    {
+      "epoch": 0.3974815037493528,
+      "grad_norm": 3.318359375,
+      "learning_rate": 1.3196140817110104e-05,
+      "loss": 0.5501,
+      "mean_token_accuracy": 0.8280089855194092,
+      "num_tokens": 35908459.0,
+      "step": 11900
+    },
+    {
+      "epoch": 0.3991515941012409,
+      "grad_norm": 3.591796875,
+      "learning_rate": 1.3146297500768799e-05,
+      "loss": 0.5518,
+      "mean_token_accuracy": 0.8280233424901963,
+      "num_tokens": 36054122.0,
+      "step": 11950
+    },
+    {
+      "epoch": 0.40082168445312893,
+      "grad_norm": 3.591796875,
+      "learning_rate": 1.3096367279414142e-05,
+      "loss": 0.557,
+      "mean_token_accuracy": 0.8281977719068527,
+      "num_tokens": 36206481.0,
+      "step": 12000
+    },
+    {
+      "epoch": 0.40249177480501697,
+      "grad_norm": 3.64453125,
+      "learning_rate": 1.3046351532186718e-05,
+      "loss": 0.5341,
+      "mean_token_accuracy": 0.8324338626861573,
+      "num_tokens": 36357066.0,
+      "step": 12050
+    },
+    {
+      "epoch": 0.404161865156905,
+      "grad_norm": 4.046875,
+      "learning_rate": 1.2996251640589443e-05,
+      "loss": 0.5597,
+      "mean_token_accuracy": 0.8256447410583496,
+      "num_tokens": 36509298.0,
+      "step": 12100
+    },
+    {
+      "epoch": 0.40583195550879303,
+      "grad_norm": 3.501953125,
+      "learning_rate": 1.2946068988449416e-05,
+      "loss": 0.5461,
+      "mean_token_accuracy": 0.8304660183191299,
+      "num_tokens": 36664284.0,
+      "step": 12150
+    },
+    {
+      "epoch": 0.40750204586068106,
+      "grad_norm": 3.775390625,
+      "learning_rate": 1.2895804961879697e-05,
+      "loss": 0.5452,
+      "mean_token_accuracy": 0.8312527567148209,
+      "num_tokens": 36815575.0,
+      "step": 12200
+    },
+    {
+      "epoch": 0.4091721362125691,
+      "grad_norm": 2.853515625,
+      "learning_rate": 1.2845460949241018e-05,
+      "loss": 0.5481,
+      "mean_token_accuracy": 0.8277529460191727,
+      "num_tokens": 36974534.0,
+      "step": 12250
+    },
+    {
+      "epoch": 0.4108422265644571,
+      "grad_norm": 3.171875,
+      "learning_rate": 1.2795038341103432e-05,
+      "loss": 0.5543,
+      "mean_token_accuracy": 0.8282213443517685,
+      "num_tokens": 37125168.0,
+      "step": 12300
+    },
+    {
+      "epoch": 0.41251231691634516,
+      "grad_norm": 3.2734375,
+      "learning_rate": 1.2744538530207916e-05,
+      "loss": 0.547,
+      "mean_token_accuracy": 0.8285178303718567,
+      "num_tokens": 37278364.0,
+      "step": 12350
+    },
+    {
+      "epoch": 0.4141824072682332,
+      "grad_norm": 2.806640625,
+      "learning_rate": 1.2693962911427877e-05,
+      "loss": 0.568,
+      "mean_token_accuracy": 0.8246202218532562,
+      "num_tokens": 37431928.0,
+      "step": 12400
+    },
+    {
+      "epoch": 0.4158524976201213,
+      "grad_norm": 2.970703125,
+      "learning_rate": 1.264331288173065e-05,
+      "loss": 0.5553,
+      "mean_token_accuracy": 0.8278548353910447,
+      "num_tokens": 37584284.0,
+      "step": 12450
+    },
+    {
+      "epoch": 0.4175225879720093,
+      "grad_norm": 3.25390625,
+      "learning_rate": 1.2592589840138897e-05,
+      "loss": 0.5417,
+      "mean_token_accuracy": 0.8309043395519257,
+      "num_tokens": 37729858.0,
+      "step": 12500
+    },
+    {
+      "epoch": 0.41919267832389734,
+      "grad_norm": 3.705078125,
+      "learning_rate": 1.2541795187691971e-05,
+      "loss": 0.5514,
+      "mean_token_accuracy": 0.8308093577623368,
+      "num_tokens": 37876100.0,
+      "step": 12550
+    },
+    {
+      "epoch": 0.42086276867578537,
+      "grad_norm": 3.5,
+      "learning_rate": 1.2490930327407209e-05,
+      "loss": 0.5575,
+      "mean_token_accuracy": 0.8272964191436768,
+      "num_tokens": 38025185.0,
+      "step": 12600
+    },
+    {
+      "epoch": 0.4225328590276734,
+      "grad_norm": 3.21484375,
+      "learning_rate": 1.243999666424119e-05,
+      "loss": 0.5485,
+      "mean_token_accuracy": 0.8294407540559768,
+      "num_tokens": 38181591.0,
+      "step": 12650
+    },
+    {
+      "epoch": 0.42420294937956143,
+      "grad_norm": 2.900390625,
+      "learning_rate": 1.2388995605050914e-05,
+      "loss": 0.5449,
+      "mean_token_accuracy": 0.8301975548267364,
+      "num_tokens": 38327561.0,
+      "step": 12700
+    },
+    {
+      "epoch": 0.42587303973144947,
+      "grad_norm": 3.689453125,
+      "learning_rate": 1.2337928558554956e-05,
+      "loss": 0.5332,
+      "mean_token_accuracy": 0.8314021265506745,
+      "num_tokens": 38473244.0,
+      "step": 12750
+    },
+    {
+      "epoch": 0.4275431300833375,
+      "grad_norm": 3.4921875,
+      "learning_rate": 1.2286796935294551e-05,
+      "loss": 0.5471,
+      "mean_token_accuracy": 0.8304495525360107,
+      "num_tokens": 38624223.0,
+      "step": 12800
+    },
+    {
+      "epoch": 0.42921322043522553,
+      "grad_norm": 3.779296875,
+      "learning_rate": 1.2235602147594631e-05,
+      "loss": 0.5479,
+      "mean_token_accuracy": 0.8294032102823258,
+      "num_tokens": 38773521.0,
+      "step": 12850
+    },
+    {
+      "epoch": 0.43088331078711356,
+      "grad_norm": 3.548828125,
+      "learning_rate": 1.2184345609524813e-05,
+      "loss": 0.5482,
+      "mean_token_accuracy": 0.8288749468326568,
+      "num_tokens": 38926115.0,
+      "step": 12900
+    },
+    {
+      "epoch": 0.43255340113900165,
+      "grad_norm": 2.939453125,
+      "learning_rate": 1.2133028736860348e-05,
+      "loss": 0.5353,
+      "mean_token_accuracy": 0.8340490102767945,
+      "num_tokens": 39077937.0,
+      "step": 12950
+    },
+    {
+      "epoch": 0.4342234914908897,
+      "grad_norm": 3.244140625,
+      "learning_rate": 1.2081652947043003e-05,
+      "loss": 0.5501,
+      "mean_token_accuracy": 0.8298387557268143,
+      "num_tokens": 39229394.0,
+      "step": 13000
+    },
+    {
+      "epoch": 0.4358935818427777,
+      "grad_norm": 2.775390625,
+      "learning_rate": 1.2030219659141928e-05,
+      "loss": 0.5353,
+      "mean_token_accuracy": 0.8325220793485641,
+      "num_tokens": 39378856.0,
+      "step": 13050
+    },
+    {
+      "epoch": 0.43756367219466574,
+      "grad_norm": 3.701171875,
+      "learning_rate": 1.1978730293814438e-05,
+      "loss": 0.5334,
+      "mean_token_accuracy": 0.8330723947286606,
+      "num_tokens": 39531138.0,
+      "step": 13100
+    },
+    {
+      "epoch": 0.4392337625465538,
+      "grad_norm": 3.646484375,
+      "learning_rate": 1.1927186273266784e-05,
+      "loss": 0.5345,
+      "mean_token_accuracy": 0.8333028835058213,
+      "num_tokens": 39677471.0,
+      "step": 13150
+    },
+    {
+      "epoch": 0.4409038528984418,
+      "grad_norm": 3.234375,
+      "learning_rate": 1.1875589021214877e-05,
+      "loss": 0.5436,
+      "mean_token_accuracy": 0.8311423152685166,
+      "num_tokens": 39825562.0,
+      "step": 13200
+    },
+    {
+      "epoch": 0.44257394325032984,
+      "grad_norm": 3.224609375,
+      "learning_rate": 1.1823939962844944e-05,
+      "loss": 0.5326,
+      "mean_token_accuracy": 0.8338733261823654,
+      "num_tokens": 39979460.0,
+      "step": 13250
+    },
+    {
+      "epoch": 0.44424403360221787,
+      "grad_norm": 3.4765625,
+      "learning_rate": 1.1772240524774179e-05,
+      "loss": 0.5395,
+      "mean_token_accuracy": 0.8336213880777359,
+      "num_tokens": 40121774.0,
+      "step": 13300
+    },
+    {
+      "epoch": 0.4459141239541059,
+      "grad_norm": 3.6875,
+      "learning_rate": 1.172049213501133e-05,
+      "loss": 0.5434,
+      "mean_token_accuracy": 0.8318886595964432,
+      "num_tokens": 40272793.0,
+      "step": 13350
+    },
+    {
+      "epoch": 0.44758421430599393,
+      "grad_norm": 3.65234375,
+      "learning_rate": 1.166869622291726e-05,
+      "loss": 0.5425,
+      "mean_token_accuracy": 0.8311206418275833,
+      "num_tokens": 40425954.0,
+      "step": 13400
+    },
+    {
+      "epoch": 0.44925430465788196,
+      "grad_norm": 3.109375,
+      "learning_rate": 1.1616854219165452e-05,
+      "loss": 0.5352,
+      "mean_token_accuracy": 0.8340102022886277,
+      "num_tokens": 40578555.0,
+      "step": 13450
+    },
+    {
+      "epoch": 0.45092439500977005,
+      "grad_norm": 3.138671875,
+      "learning_rate": 1.1564967555702526e-05,
+      "loss": 0.5593,
+      "mean_token_accuracy": 0.8278708016872406,
+      "num_tokens": 40725781.0,
+      "step": 13500
+    },
+    {
+      "epoch": 0.4525944853616581,
+      "grad_norm": 3.943359375,
+      "learning_rate": 1.1513037665708639e-05,
+      "loss": 0.5105,
+      "mean_token_accuracy": 0.8397551566362381,
+      "num_tokens": 40874427.0,
+      "step": 13550
+    },
+    {
+      "epoch": 0.4542645757135461,
+      "grad_norm": 3.845703125,
+      "learning_rate": 1.146106598355794e-05,
+      "loss": 0.5195,
+      "mean_token_accuracy": 0.8360963726043701,
+      "num_tokens": 41027698.0,
+      "step": 13600
+    },
+    {
+      "epoch": 0.45593466606543415,
+      "grad_norm": 4.0546875,
+      "learning_rate": 1.1409053944778925e-05,
+      "loss": 0.5367,
+      "mean_token_accuracy": 0.8321689450740815,
+      "num_tokens": 41177523.0,
+      "step": 13650
+    },
+    {
+      "epoch": 0.4576047564173222,
+      "grad_norm": 3.27734375,
+      "learning_rate": 1.1357002986014803e-05,
+      "loss": 0.5253,
+      "mean_token_accuracy": 0.8352927279472351,
+      "num_tokens": 41329674.0,
+      "step": 13700
+    },
+    {
+      "epoch": 0.4592748467692102,
+      "grad_norm": 3.017578125,
+      "learning_rate": 1.1304914544983794e-05,
+      "loss": 0.5232,
+      "mean_token_accuracy": 0.835522665977478,
+      "num_tokens": 41476485.0,
+      "step": 13750
+    },
+    {
+      "epoch": 0.46094493712109824,
+      "grad_norm": 3.44921875,
+      "learning_rate": 1.125279006043945e-05,
+      "loss": 0.5674,
+      "mean_token_accuracy": 0.824871341586113,
+      "num_tokens": 41628555.0,
+      "step": 13800
+    },
+    {
+      "epoch": 0.4626150274729863,
+      "grad_norm": 3.52734375,
+      "learning_rate": 1.1200630972130867e-05,
+      "loss": 0.5282,
+      "mean_token_accuracy": 0.8343129450082779,
+      "num_tokens": 41775801.0,
+      "step": 13850
+    },
+    {
+      "epoch": 0.4642851178248743,
+      "grad_norm": 2.576171875,
+      "learning_rate": 1.1148438720762961e-05,
+      "loss": 0.5106,
+      "mean_token_accuracy": 0.8398624455928803,
+      "num_tokens": 41925501.0,
+      "step": 13900
+    },
+    {
+      "epoch": 0.46595520817676234,
+      "grad_norm": 4.1015625,
+      "learning_rate": 1.109621474795666e-05,
+      "loss": 0.5338,
+      "mean_token_accuracy": 0.8335544627904892,
+      "num_tokens": 42079129.0,
+      "step": 13950
+    },
+    {
+      "epoch": 0.4676252985286504,
+      "grad_norm": 3.314453125,
+      "learning_rate": 1.104396049620907e-05,
+      "loss": 0.5235,
+      "mean_token_accuracy": 0.8374920153617859,
+      "num_tokens": 42225486.0,
+      "step": 14000
+    },
+    {
+      "epoch": 0.46929538888053846,
+      "grad_norm": 3.83203125,
+      "learning_rate": 1.0991677408853643e-05,
+      "loss": 0.5387,
+      "mean_token_accuracy": 0.8317029625177383,
+      "num_tokens": 42379482.0,
+      "step": 14050
+    },
+    {
+      "epoch": 0.4709654792324265,
+      "grad_norm": 4.4296875,
+      "learning_rate": 1.093936693002032e-05,
+      "loss": 0.5135,
+      "mean_token_accuracy": 0.8377016627788544,
+      "num_tokens": 42531797.0,
+      "step": 14100
+    },
+    {
+      "epoch": 0.4726355695843145,
+      "grad_norm": 2.8984375,
+      "learning_rate": 1.0887030504595624e-05,
+      "loss": 0.5392,
+      "mean_token_accuracy": 0.8313734871149063,
+      "num_tokens": 42686643.0,
+      "step": 14150
+    },
+    {
+      "epoch": 0.47430565993620255,
+      "grad_norm": 3.212890625,
+      "learning_rate": 1.0834669578182757e-05,
+      "loss": 0.5475,
+      "mean_token_accuracy": 0.8304911911487579,
+      "num_tokens": 42836786.0,
+      "step": 14200
+    },
+    {
+      "epoch": 0.4759757502880906,
+      "grad_norm": 3.361328125,
+      "learning_rate": 1.0782285597061679e-05,
+      "loss": 0.5259,
+      "mean_token_accuracy": 0.8347867679595947,
+      "num_tokens": 42982714.0,
+      "step": 14250
+    },
+    {
+      "epoch": 0.4776458406399786,
+      "grad_norm": 4.04296875,
+      "learning_rate": 1.0729880008149142e-05,
+      "loss": 0.5485,
+      "mean_token_accuracy": 0.829542515873909,
+      "num_tokens": 43142510.0,
+      "step": 14300
+    },
+    {
+      "epoch": 0.47931593099186665,
+      "grad_norm": 3.369140625,
+      "learning_rate": 1.0677454258958746e-05,
+      "loss": 0.5244,
+      "mean_token_accuracy": 0.8368889981508255,
+      "num_tokens": 43288399.0,
+      "step": 14350
+    },
+    {
+      "epoch": 0.4809860213437547,
+      "grad_norm": 3.76171875,
+      "learning_rate": 1.0625009797560936e-05,
+      "loss": 0.5142,
+      "mean_token_accuracy": 0.8388600462675094,
+      "num_tokens": 43444562.0,
+      "step": 14400
+    },
+    {
+      "epoch": 0.4826561116956427,
+      "grad_norm": 2.94140625,
+      "learning_rate": 1.0572548072543021e-05,
+      "loss": 0.5284,
+      "mean_token_accuracy": 0.8348941618204117,
+      "num_tokens": 43591116.0,
+      "step": 14450
+    },
+    {
+      "epoch": 0.4843262020475308,
+      "grad_norm": 3.28125,
+      "learning_rate": 1.0520070532969142e-05,
+      "loss": 0.5208,
+      "mean_token_accuracy": 0.8390517139434814,
+      "num_tokens": 43740326.0,
+      "step": 14500
+    },
+    {
+      "epoch": 0.4859962923994188,
+      "grad_norm": 3.853515625,
+      "learning_rate": 1.0467578628340278e-05,
+      "loss": 0.5075,
+      "mean_token_accuracy": 0.8397658413648605,
+      "num_tokens": 43891895.0,
+      "step": 14550
+    },
+    {
+      "epoch": 0.48766638275130686,
+      "grad_norm": 3.06640625,
+      "learning_rate": 1.0415073808554178e-05,
+      "loss": 0.5151,
+      "mean_token_accuracy": 0.8379220014810562,
+      "num_tokens": 44046860.0,
+      "step": 14600
+    },
+    {
+      "epoch": 0.4893364731031949,
+      "grad_norm": 3.13671875,
+      "learning_rate": 1.0362557523865323e-05,
+      "loss": 0.5281,
+      "mean_token_accuracy": 0.8356298345327378,
+      "num_tokens": 44203029.0,
+      "step": 14650
+    },
+    {
+      "epoch": 0.4910065634550829,
+      "grad_norm": 3.08984375,
+      "learning_rate": 1.0310031224844878e-05,
+      "loss": 0.5365,
+      "mean_token_accuracy": 0.8323665148019791,
+      "num_tokens": 44355830.0,
+      "step": 14700
+    },
+    {
+      "epoch": 0.49267665380697095,
+      "grad_norm": 3.234375,
+      "learning_rate": 1.0257496362340622e-05,
+      "loss": 0.5168,
+      "mean_token_accuracy": 0.8371515053510666,
+      "num_tokens": 44506850.0,
+      "step": 14750
+    },
+    {
+      "epoch": 0.494346744158859,
+      "grad_norm": 3.34765625,
+      "learning_rate": 1.0204954387436847e-05,
+      "loss": 0.5228,
+      "mean_token_accuracy": 0.8363370823860169,
+      "num_tokens": 44658966.0,
+      "step": 14800
+    },
+    {
+      "epoch": 0.496016834510747,
+      "grad_norm": 3.560546875,
+      "learning_rate": 1.0152406751414328e-05,
+      "loss": 0.5151,
+      "mean_token_accuracy": 0.8392669379711151,
+      "num_tokens": 44805368.0,
+      "step": 14850
+    },
+    {
+      "epoch": 0.49768692486263505,
+      "grad_norm": 3.609375,
+      "learning_rate": 1.0099854905710183e-05,
+      "loss": 0.5285,
+      "mean_token_accuracy": 0.8332407021522522,
+      "num_tokens": 44951969.0,
+      "step": 14900
+    },
+    {
+      "epoch": 0.4993570152145231,
+      "grad_norm": 3.0078125,
+      "learning_rate": 1.0047300301877822e-05,
+      "loss": 0.5289,
+      "mean_token_accuracy": 0.8356048625707626,
+      "num_tokens": 45108518.0,
+      "step": 14950
+    },
+    {
+      "epoch": 0.5010271055664112,
+      "grad_norm": 3.59765625,
+      "learning_rate": 9.994744391546838e-06,
+      "loss": 0.5206,
+      "mean_token_accuracy": 0.8376101624965667,
+      "num_tokens": 45263447.0,
+      "step": 15000
+    }
+  ],
+  "logging_steps": 50,
+  "max_steps": 29938,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 1000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.6088218656099533e+17,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:657cbef08b1c3ef63f4593deeb2ba73805f587a3794c3091761b43f1d736d144
+size 5624