Model save

.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

README.md

@@ -0,0 +1,67 @@
+---
+library_name: transformers
+model_name: Qwen2.5-1.5B-Open-R1-GRPO
+tags:
+- generated_from_trainer
+- trl
+- grpo
+licence: license
+---
+
+# Model Card for Qwen2.5-1.5B-Open-R1-GRPO
+
+This model is a fine-tuned version of [None](https://huggingface.co/None).
+It has been trained using [TRL](https://github.com/huggingface/trl).
+
+## Quick start
+
+```python
+from transformers import pipeline
+
+question = "If you had a time machine, but could only go to the past or the future once and never return, which would you choose and why?"
+generator = pipeline("text-generation", model="KMasaki/Qwen2.5-1.5B-Open-R1-GRPO", device="cuda")
+output = generator([{"role": "user", "content": question}], max_new_tokens=128, return_full_text=False)[0]
+print(output["generated_text"])
+```
+
+## Training procedure
+
+[<img src="https://raw.githubusercontent.com/wandb/assets/main/wandb-github-badge-28.svg" alt="Visualize in Weights & Biases" width="150" height="24"/>](https://wandb.ai/kawamuramasaki/open-r1/runs/ppb2nomx) 
+
+
+This model was trained with GRPO, a method introduced in [DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models](https://huggingface.co/papers/2402.03300).
+
+### Framework versions
+
+- TRL: 0.16.0.dev0
+- Transformers: 4.49.0.dev0
+- Pytorch: 2.5.1
+- Datasets: 3.3.0
+- Tokenizers: 0.21.0
+
+## Citations
+
+Cite GRPO as:
+
+```bibtex
+@article{zhihong2024deepseekmath,
+    title        = {{DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models}},
+    author       = {Zhihong Shao and Peiyi Wang and Qihao Zhu and Runxin Xu and Junxiao Song and Mingchuan Zhang and Y. K. Li and Y. Wu and Daya Guo},
+    year         = 2024,
+    eprint       = {arXiv:2402.03300},
+}
+
+```
+
+Cite TRL as:
+    
+```bibtex
+@misc{vonwerra2022trl,
+	title        = {{TRL: Transformer Reinforcement Learning}},
+	author       = {Leandro von Werra and Younes Belkada and Lewis Tunstall and Edward Beeching and Tristan Thrush and Nathan Lambert and Shengyi Huang and Kashif Rasul and Quentin Gallouédec},
+	year         = 2020,
+	journal      = {GitHub repository},
+	publisher    = {GitHub},
+	howpublished = {\url{https://github.com/huggingface/trl}}
+}
+```

added_tokens.json

@@ -0,0 +1,24 @@
+{
+  "</tool_call>": 151658,
+  "<tool_call>": 151657,
+  "<|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
+}

all_results.json

@@ -0,0 +1,8 @@
+{
+    "total_flos": 0.0,
+    "train_loss": 0.08444789265832109,
+    "train_runtime": 62734.3355,
+    "train_samples": 72441,
+    "train_samples_per_second": 1.155,
+    "train_steps_per_second": 0.009
+}

config.json

@@ -0,0 +1,29 @@
+{
+  "_name_or_path": "data/Qwen2.5-1.5B-Open-R1-Distill/",
+  "architectures": [
+    "Qwen2ForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 151643,
+  "eos_token_id": 151645,
+  "hidden_act": "silu",
+  "hidden_size": 1536,
+  "initializer_range": 0.02,
+  "intermediate_size": 8960,
+  "max_position_embeddings": 32768,
+  "max_window_layers": 21,
+  "model_type": "qwen2",
+  "num_attention_heads": 12,
+  "num_hidden_layers": 28,
+  "num_key_value_heads": 2,
+  "rms_norm_eps": 1e-06,
+  "rope_scaling": null,
+  "rope_theta": 1000000.0,
+  "sliding_window": 32768,
+  "tie_word_embeddings": true,
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.49.0.dev0",
+  "use_cache": false,
+  "use_sliding_window": false,
+  "vocab_size": 151936
+}

generation_config.json

@@ -0,0 +1,14 @@
+{
+  "bos_token_id": 151643,
+  "do_sample": true,
+  "eos_token_id": [
+    151645,
+    151643
+  ],
+  "pad_token_id": 151643,
+  "repetition_penalty": 1.1,
+  "temperature": 0.7,
+  "top_k": 20,
+  "top_p": 0.8,
+  "transformers_version": "4.49.0.dev0"
+}

model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a8b776bab95c3facbe304aad773e6994088dc2a4c2aabb9c454bd9510e7e922c
+size 3554214752

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": "<|im_end|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
+}

tokenizer.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:63a2951d5edfa5cc0a2346ef872f8c77a2920274cfc3b503b04e3799104dee80
+size 11422060

tokenizer_config.json

@@ -0,0 +1,209 @@
+{
+  "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
+    }
+  },
+  "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    {%- else %}\n        {{- 'You are Qwen, created by Alibaba Cloud. You are a helpful assistant.' }}\n    {%- endif %}\n    {{- \"\\n\\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    {%- else %}\n        {{- '<|im_start|>system\\nYou are Qwen, created by Alibaba Cloud. You are a helpful assistant.<|im_end|>\\n' }}\n    {%- endif %}\n{%- endif %}\n{%- for message in messages %}\n    {%- if (message.role == \"user\") or (message.role == \"system\" and not loop.first) or (message.role == \"assistant\" and not message.tool_calls) %}\n        {{- '<|im_start|>' + message.role + '\\n' + message.content + '<|im_end|>' + '\\n' }}\n    {%- elif message.role == \"assistant\" %}\n        {{- '<|im_start|>' + message.role }}\n        {%- if message.content %}\n            {{- '\\n' + message.content }}\n        {%- endif %}\n        {%- for tool_call in message.tool_calls %}\n            {%- if tool_call.function is defined %}\n                {%- set tool_call = tool_call.function %}\n            {%- endif %}\n            {{- '\\n<tool_call>\\n{\"name\": \"' }}\n            {{- tool_call.name }}\n            {{- '\", \"arguments\": ' }}\n            {{- tool_call.arguments | tojson }}\n            {{- '}\\n</tool_call>' }}\n        {%- endfor %}\n        {{- '<|im_end|>\\n' }}\n    {%- elif message.role == \"tool\" %}\n        {%- if (loop.index0 == 0) 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{%- endif %}\n",
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "<|im_end|>",
+  "errors": "replace",
+  "extra_special_tokens": {},
+  "model_max_length": 131072,
+  "pad_token": "<|im_end|>",
+  "padding_side": "left",
+  "split_special_tokens": false,
+  "tokenizer_class": "Qwen2Tokenizer",
+  "unk_token": null
+}

train_results.json

@@ -0,0 +1,8 @@
+{
+    "total_flos": 0.0,
+    "train_loss": 0.08444789265832109,
+    "train_runtime": 62734.3355,
+    "train_samples": 72441,
+    "train_samples_per_second": 1.155,
+    "train_steps_per_second": 0.009
+}

trainer_state.json

@@ -0,0 +1,1587 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 100,
+  "global_step": 566,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "completion_length": 1012.7799636840821,
+      "epoch": 0.0088339222614841,
+      "grad_norm": 0.060452450066804886,
+      "kl": 0.0001143336296081543,
+      "learning_rate": 1.7543859649122807e-06,
+      "loss": 0.0,
+      "reward": 0.04791666807141155,
+      "reward_std": 0.06029603164643049,
+      "rewards/accuracy_reward": 0.04791666807141155,
+      "rewards/format_reward": 0.0,
+      "step": 5
+    },
+    {
+      "completion_length": 1008.932827758789,
+      "epoch": 0.0176678445229682,
+      "grad_norm": 0.08252555131912231,
+      "kl": 0.00027997493743896483,
+      "learning_rate": 3.5087719298245615e-06,
+      "loss": 0.0,
+      "reward": 0.06822916846722364,
+      "reward_std": 0.08430604650638998,
+      "rewards/accuracy_reward": 0.06822916846722364,
+      "rewards/format_reward": 0.0,
+      "step": 10
+    },
+    {
+      "completion_length": 972.009130859375,
+      "epoch": 0.026501766784452298,
+      "grad_norm": 0.20415905117988586,
+      "kl": 0.002631378173828125,
+      "learning_rate": 5.263157894736842e-06,
+      "loss": 0.0001,
+      "reward": 0.14088542051613331,
+      "reward_std": 0.1278162115253508,
+      "rewards/accuracy_reward": 0.14088542051613331,
+      "rewards/format_reward": 0.0,
+      "step": 15
+    },
+    {
+      "completion_length": 895.1536674499512,
+      "epoch": 0.0353356890459364,
+      "grad_norm": 0.09219123423099518,
+      "kl": 0.010625839233398438,
+      "learning_rate": 7.017543859649123e-06,
+      "loss": 0.0004,
+      "reward": 0.22161459152121096,
+      "reward_std": 0.14556703479029237,
+      "rewards/accuracy_reward": 0.22161459152121096,
+      "rewards/format_reward": 0.0,
+      "step": 20
+    },
+    {
+      "completion_length": 917.4883033752442,
+      "epoch": 0.044169611307420496,
+      "grad_norm": 0.11080280691385269,
+      "kl": 0.017144012451171874,
+      "learning_rate": 8.771929824561405e-06,
+      "loss": 0.0007,
+      "reward": 0.23333334033377467,
+      "reward_std": 0.16294219507835805,
+      "rewards/accuracy_reward": 0.23333334033377467,
+      "rewards/format_reward": 0.0,
+      "step": 25
+    },
+    {
+      "completion_length": 972.0820533752442,
+      "epoch": 0.053003533568904596,
+      "grad_norm": 0.09853794425725937,
+      "kl": 0.0230743408203125,
+      "learning_rate": 1.0526315789473684e-05,
+      "loss": 0.0009,
+      "reward": 0.22760417270474137,
+      "reward_std": 0.15662940898910166,
+      "rewards/accuracy_reward": 0.22760417270474137,
+      "rewards/format_reward": 0.0,
+      "step": 30
+    },
+    {
+      "completion_length": 962.4583534240722,
+      "epoch": 0.061837455830388695,
+      "grad_norm": 0.1029583290219307,
+      "kl": 0.03768310546875,
+      "learning_rate": 1.2280701754385966e-05,
+      "loss": 0.0015,
+      "reward": 0.20807292265817523,
+      "reward_std": 0.14881791989319026,
+      "rewards/accuracy_reward": 0.20807292265817523,
+      "rewards/format_reward": 0.0,
+      "step": 35
+    },
+    {
+      "completion_length": 1004.2820465087891,
+      "epoch": 0.0706713780918728,
+      "grad_norm": 0.1385573297739029,
+      "kl": 0.0864959716796875,
+      "learning_rate": 1.4035087719298246e-05,
+      "loss": 0.0035,
+      "reward": 0.2763020928483456,
+      "reward_std": 0.18439405504614115,
+      "rewards/accuracy_reward": 0.2763020928483456,
+      "rewards/format_reward": 0.0,
+      "step": 40
+    },
+    {
+      "completion_length": 800.9942935943603,
+      "epoch": 0.07950530035335689,
+      "grad_norm": 0.20295394957065582,
+      "kl": 0.1165924072265625,
+      "learning_rate": 1.578947368421053e-05,
+      "loss": 0.0047,
+      "reward": 0.31484376010484993,
+      "reward_std": 0.23337243953719736,
+      "rewards/accuracy_reward": 0.31484376010484993,
+      "rewards/format_reward": 0.0,
+      "step": 45
+    },
+    {
+      "completion_length": 624.4739780426025,
+      "epoch": 0.08833922261484099,
+      "grad_norm": 0.1329674869775772,
+      "kl": 0.157794189453125,
+      "learning_rate": 1.754385964912281e-05,
+      "loss": 0.0063,
+      "reward": 0.3552083443850279,
+      "reward_std": 0.20170472487807273,
+      "rewards/accuracy_reward": 0.3552083443850279,
+      "rewards/format_reward": 0.0,
+      "step": 50
+    },
+    {
+      "completion_length": 594.1015830993653,
+      "epoch": 0.09717314487632508,
+      "grad_norm": 0.1421622484922409,
+      "kl": 0.216668701171875,
+      "learning_rate": 1.929824561403509e-05,
+      "loss": 0.0087,
+      "reward": 0.36406251094304026,
+      "reward_std": 0.21222675214521586,
+      "rewards/accuracy_reward": 0.36406251094304026,
+      "rewards/format_reward": 0.0,
+      "step": 55
+    },
+    {
+      "completion_length": 587.8580898284912,
+      "epoch": 0.10600706713780919,
+      "grad_norm": 1.4284125566482544,
+      "kl": 0.44776611328125,
+      "learning_rate": 1.9998285788966027e-05,
+      "loss": 0.0179,
+      "reward": 0.38958334531635047,
+      "reward_std": 0.23174314848147332,
+      "rewards/accuracy_reward": 0.38958334531635047,
+      "rewards/format_reward": 0.0,
+      "step": 60
+    },
+    {
+      "completion_length": 645.9356914520264,
+      "epoch": 0.11484098939929328,
+      "grad_norm": 0.29591915011405945,
+      "kl": 134.6033447265625,
+      "learning_rate": 1.998781218310425e-05,
+      "loss": 5.3828,
+      "reward": 0.3755208417773247,
+      "reward_std": 0.2100224627647549,
+      "rewards/accuracy_reward": 0.3755208417773247,
+      "rewards/format_reward": 0.0,
+      "step": 65
+    },
+    {
+      "completion_length": 677.3062721252442,
+      "epoch": 0.12367491166077739,
+      "grad_norm": 0.6859232187271118,
+      "kl": 6.9478271484375,
+      "learning_rate": 1.9967827272672407e-05,
+      "loss": 0.2775,
+      "reward": 0.3325520935235545,
+      "reward_std": 0.22865285109728575,
+      "rewards/accuracy_reward": 0.3325520935235545,
+      "rewards/format_reward": 0.0,
+      "step": 70
+    },
+    {
+      "completion_length": 613.6190311431885,
+      "epoch": 0.13250883392226148,
+      "grad_norm": 46.3941650390625,
+      "kl": 2.58236083984375,
+      "learning_rate": 1.993835008912268e-05,
+      "loss": 0.1033,
+      "reward": 0.39244793001562356,
+      "reward_std": 0.21552497819066047,
+      "rewards/accuracy_reward": 0.39244793001562356,
+      "rewards/format_reward": 0.0,
+      "step": 75
+    },
+    {
+      "completion_length": 663.7047077178955,
+      "epoch": 0.1413427561837456,
+      "grad_norm": 0.6843287348747253,
+      "kl": 0.7177490234375,
+      "learning_rate": 1.9899408703314383e-05,
+      "loss": 0.0287,
+      "reward": 0.37213542554527523,
+      "reward_std": 0.22119648223742844,
+      "rewards/accuracy_reward": 0.37213542554527523,
+      "rewards/format_reward": 0.0,
+      "step": 80
+    },
+    {
+      "completion_length": 690.5789264678955,
+      "epoch": 0.1501766784452297,
+      "grad_norm": 0.629157304763794,
+      "kl": 0.58406982421875,
+      "learning_rate": 1.985104019878233e-05,
+      "loss": 0.0234,
+      "reward": 0.35416667682584374,
+      "reward_std": 0.20734282806515694,
+      "rewards/accuracy_reward": 0.35416667682584374,
+      "rewards/format_reward": 0.0,
+      "step": 85
+    },
+    {
+      "completion_length": 709.6250190734863,
+      "epoch": 0.15901060070671377,
+      "grad_norm": 1.965692162513733,
+      "kl": 0.44111328125,
+      "learning_rate": 1.9793290636422503e-05,
+      "loss": 0.0177,
+      "reward": 0.30989584284834565,
+      "reward_std": 0.20093054957687856,
+      "rewards/accuracy_reward": 0.30989584284834565,
+      "rewards/format_reward": 0.0,
+      "step": 90
+    },
+    {
+      "completion_length": 705.39273147583,
+      "epoch": 0.16784452296819788,
+      "grad_norm": 0.5004879832267761,
+      "kl": 34.3031982421875,
+      "learning_rate": 1.9726215010628717e-05,
+      "loss": 1.3714,
+      "reward": 0.334375007590279,
+      "reward_std": 0.2137310016900301,
+      "rewards/accuracy_reward": 0.334375007590279,
+      "rewards/format_reward": 0.0,
+      "step": 95
+    },
+    {
+      "completion_length": 705.0820510864257,
+      "epoch": 0.17667844522968199,
+      "grad_norm": 0.20255358517169952,
+      "kl": 0.5868896484375,
+      "learning_rate": 1.96498771969219e-05,
+      "loss": 0.0235,
+      "reward": 0.35494792852550744,
+      "reward_std": 0.2118946822360158,
+      "rewards/accuracy_reward": 0.35494792852550744,
+      "rewards/format_reward": 0.0,
+      "step": 100
+    },
+    {
+      "epoch": 0.17667844522968199,
+      "eval_completion_length": 683.6413835797991,
+      "eval_kl": 0.7349330357142857,
+      "eval_loss": 0.029816392809152603,
+      "eval_reward": 0.37202381661960054,
+      "eval_reward_std": 0.18448604004723684,
+      "eval_rewards/accuracy_reward": 0.37202381661960054,
+      "eval_rewards/format_reward": 0.0,
+      "eval_runtime": 67.6427,
+      "eval_samples_per_second": 1.464,
+      "eval_steps_per_second": 0.03,
+      "step": 100
+    },
+    {
+      "completion_length": 708.5942901611328,
+      "epoch": 0.1855123674911661,
+      "grad_norm": 0.5656086206436157,
+      "kl": 0.77958984375,
+      "learning_rate": 1.9564349891122017e-05,
+      "loss": 0.0312,
+      "reward": 0.33385417610406876,
+      "reward_std": 0.20352514665573834,
+      "rewards/accuracy_reward": 0.33385417610406876,
+      "rewards/format_reward": 0.0,
+      "step": 105
+    },
+    {
+      "completion_length": 667.9695518493652,
+      "epoch": 0.19434628975265017,
+      "grad_norm": 0.5076290965080261,
+      "kl": 0.9525390625,
+      "learning_rate": 1.946971454012051e-05,
+      "loss": 0.0381,
+      "reward": 0.3570312585681677,
+      "reward_std": 0.22535927146673201,
+      "rewards/accuracy_reward": 0.3570312585681677,
+      "rewards/format_reward": 0.0,
+      "step": 110
+    },
+    {
+      "completion_length": 647.0583518981933,
+      "epoch": 0.20318021201413428,
+      "grad_norm": 2.9529125690460205,
+      "kl": 1.4181396484375,
+      "learning_rate": 1.9366061264319112e-05,
+      "loss": 0.0567,
+      "reward": 0.33463542591780426,
+      "reward_std": 0.21494693392887712,
+      "rewards/accuracy_reward": 0.33463542591780426,
+      "rewards/format_reward": 0.0,
+      "step": 115
+    },
+    {
+      "completion_length": 637.5487171173096,
+      "epoch": 0.21201413427561838,
+      "grad_norm": 1.0454950332641602,
+      "kl": 1.4077880859375,
+      "learning_rate": 1.9253488771809024e-05,
+      "loss": 0.0563,
+      "reward": 0.3505208448972553,
+      "reward_std": 0.23165740915574135,
+      "rewards/accuracy_reward": 0.3505208448972553,
+      "rewards/format_reward": 0.0,
+      "step": 120
+    },
+    {
+      "completion_length": 628.239083480835,
+      "epoch": 0.22084805653710246,
+      "grad_norm": 1.0473271608352661,
+      "kl": 0.9690185546875,
+      "learning_rate": 1.9132104264372065e-05,
+      "loss": 0.0388,
+      "reward": 0.3643229269888252,
+      "reward_std": 0.21877426127903163,
+      "rewards/accuracy_reward": 0.3643229269888252,
+      "rewards/format_reward": 0.0,
+      "step": 125
+    },
+    {
+      "completion_length": 702.5557498931885,
+      "epoch": 0.22968197879858657,
+      "grad_norm": 1.5544979572296143,
+      "kl": 1.5296875,
+      "learning_rate": 1.9002023335393366e-05,
+      "loss": 0.0612,
+      "reward": 0.33255209350027143,
+      "reward_std": 0.22226272178813816,
+      "rewards/accuracy_reward": 0.33255209350027143,
+      "rewards/format_reward": 0.0,
+      "step": 130
+    },
+    {
+      "completion_length": 647.2054889678955,
+      "epoch": 0.23851590106007067,
+      "grad_norm": 0.40587472915649414,
+      "kl": 0.4884521484375,
+      "learning_rate": 1.8863369859782824e-05,
+      "loss": 0.0195,
+      "reward": 0.35833334461785854,
+      "reward_std": 0.22049089586362242,
+      "rewards/accuracy_reward": 0.35833334461785854,
+      "rewards/format_reward": 0.0,
+      "step": 135
+    },
+    {
+      "completion_length": 669.537519454956,
+      "epoch": 0.24734982332155478,
+      "grad_norm": 0.6191494464874268,
+      "kl": 0.4470458984375,
+      "learning_rate": 1.8716275876010135e-05,
+      "loss": 0.0179,
+      "reward": 0.3484375092200935,
+      "reward_std": 0.2258549713063985,
+      "rewards/accuracy_reward": 0.3484375092200935,
+      "rewards/format_reward": 0.0,
+      "step": 140
+    },
+    {
+      "completion_length": 619.0041847229004,
+      "epoch": 0.25618374558303886,
+      "grad_norm": 0.1737312227487564,
+      "kl": 0.46015625,
+      "learning_rate": 1.8560881460365726e-05,
+      "loss": 0.0184,
+      "reward": 0.37473959419876335,
+      "reward_std": 0.21089042853564025,
+      "rewards/accuracy_reward": 0.37473959419876335,
+      "rewards/format_reward": 0.0,
+      "step": 145
+    },
+    {
+      "completion_length": 607.3304874420166,
+      "epoch": 0.26501766784452296,
+      "grad_norm": 0.28501132130622864,
+      "kl": 0.23609619140625,
+      "learning_rate": 1.8397334593567347e-05,
+      "loss": 0.0094,
+      "reward": 0.40338542629033325,
+      "reward_std": 0.21479419860988855,
+      "rewards/accuracy_reward": 0.40338542629033325,
+      "rewards/format_reward": 0.0,
+      "step": 150
+    },
+    {
+      "completion_length": 688.8893409729004,
+      "epoch": 0.27385159010600707,
+      "grad_norm": 1.4284056425094604,
+      "kl": 0.72652587890625,
+      "learning_rate": 1.8225791019839375e-05,
+      "loss": 0.0291,
+      "reward": 0.34583334196358917,
+      "reward_std": 0.21674091052263975,
+      "rewards/accuracy_reward": 0.34583334196358917,
+      "rewards/format_reward": 0.0,
+      "step": 155
+    },
+    {
+      "completion_length": 728.9481967926025,
+      "epoch": 0.2826855123674912,
+      "grad_norm": 0.7071075439453125,
+      "kl": 1.098583984375,
+      "learning_rate": 1.8046414098598947e-05,
+      "loss": 0.044,
+      "reward": 0.28619792573153974,
+      "reward_std": 0.22066022832877935,
+      "rewards/accuracy_reward": 0.28619792573153974,
+      "rewards/format_reward": 0.0,
+      "step": 160
+    },
+    {
+      "completion_length": 648.483610534668,
+      "epoch": 0.2915194346289753,
+      "grad_norm": 1.560168743133545,
+      "kl": 0.747900390625,
+      "learning_rate": 1.785937464889027e-05,
+      "loss": 0.0299,
+      "reward": 0.3583333441987634,
+      "reward_std": 0.216434388095513,
+      "rewards/accuracy_reward": 0.3583333441987634,
+      "rewards/format_reward": 0.0,
+      "step": 165
+    },
+    {
+      "completion_length": 648.3966369628906,
+      "epoch": 0.3003533568904594,
+      "grad_norm": 0.7639321684837341,
+      "kl": 1.24566650390625,
+      "learning_rate": 1.766485078671514e-05,
+      "loss": 0.0498,
+      "reward": 0.3364583439193666,
+      "reward_std": 0.21424420201219618,
+      "rewards/accuracy_reward": 0.3364583439193666,
+      "rewards/format_reward": 0.0,
+      "step": 170
+    },
+    {
+      "completion_length": 688.9159088134766,
+      "epoch": 0.30918727915194344,
+      "grad_norm": 0.4135216772556305,
+      "kl": 0.71966552734375,
+      "learning_rate": 1.746302775541467e-05,
+      "loss": 0.0288,
+      "reward": 0.3195312586380169,
+      "reward_std": 0.22360584419220686,
+      "rewards/accuracy_reward": 0.3195312586380169,
+      "rewards/format_reward": 0.0,
+      "step": 175
+    },
+    {
+      "completion_length": 662.4427268981933,
+      "epoch": 0.31802120141342755,
+      "grad_norm": 0.40282484889030457,
+      "kl": 0.71461181640625,
+      "learning_rate": 1.7254097749263735e-05,
+      "loss": 0.0286,
+      "reward": 0.32317709233611824,
+      "reward_std": 0.22129276697523892,
+      "rewards/accuracy_reward": 0.32317709233611824,
+      "rewards/format_reward": 0.0,
+      "step": 180
+    },
+    {
+      "completion_length": 600.8096572875977,
+      "epoch": 0.32685512367491165,
+      "grad_norm": 0.5917491316795349,
+      "kl": 0.54002685546875,
+      "learning_rate": 1.703825973044602e-05,
+      "loss": 0.0216,
+      "reward": 0.392187512665987,
+      "reward_std": 0.2121715539135039,
+      "rewards/accuracy_reward": 0.392187512665987,
+      "rewards/format_reward": 0.0,
+      "step": 185
+    },
+    {
+      "completion_length": 679.6117378234864,
+      "epoch": 0.33568904593639576,
+      "grad_norm": 0.7881314754486084,
+      "kl": 0.559759521484375,
+      "learning_rate": 1.681571923958416e-05,
+      "loss": 0.0224,
+      "reward": 0.32734376001171767,
+      "reward_std": 0.21214440735056997,
+      "rewards/accuracy_reward": 0.32734376001171767,
+      "rewards/format_reward": 0.0,
+      "step": 190
+    },
+    {
+      "completion_length": 661.2002803802491,
+      "epoch": 0.34452296819787986,
+      "grad_norm": 0.851506233215332,
+      "kl": 0.61300048828125,
+      "learning_rate": 1.6586688200005193e-05,
+      "loss": 0.0245,
+      "reward": 0.3382812590803951,
+      "reward_std": 0.2068442706950009,
+      "rewards/accuracy_reward": 0.3382812590803951,
+      "rewards/format_reward": 0.0,
+      "step": 195
+    },
+    {
+      "completion_length": 617.6755367279053,
+      "epoch": 0.35335689045936397,
+      "grad_norm": 0.3794810473918915,
+      "kl": 0.53272705078125,
+      "learning_rate": 1.6351384715927897e-05,
+      "loss": 0.0213,
+      "reward": 0.30729167368263005,
+      "reward_std": 0.19483569236472248,
+      "rewards/accuracy_reward": 0.30729167368263005,
+      "rewards/format_reward": 0.0,
+      "step": 200
+    },
+    {
+      "epoch": 0.35335689045936397,
+      "eval_completion_length": 678.7470354352679,
+      "eval_kl": 1.0011160714285714,
+      "eval_loss": 0.039382919669151306,
+      "eval_reward": 0.33630953303405214,
+      "eval_reward_std": 0.20554600123848235,
+      "eval_rewards/accuracy_reward": 0.33630953303405214,
+      "eval_rewards/format_reward": 0.0,
+      "eval_runtime": 59.339,
+      "eval_samples_per_second": 1.668,
+      "eval_steps_per_second": 0.034,
+      "step": 200
+    },
+    {
+      "completion_length": 638.4328308105469,
+      "epoch": 0.3621908127208481,
+      "grad_norm": 1.1715658903121948,
+      "kl": 0.6474609375,
+      "learning_rate": 1.611003286476406e-05,
+      "loss": 0.0259,
+      "reward": 0.31484376001171765,
+      "reward_std": 0.18996641575358808,
+      "rewards/accuracy_reward": 0.31484376001171765,
+      "rewards/format_reward": 0.0,
+      "step": 205
+    },
+    {
+      "completion_length": 638.4638221740722,
+      "epoch": 0.3710247349823322,
+      "grad_norm": 1.3915163278579712,
+      "kl": 1.1618896484375,
+      "learning_rate": 1.5862862483731574e-05,
+      "loss": 0.0465,
+      "reward": 0.36171875833533707,
+      "reward_std": 0.2458243785891682,
+      "rewards/accuracy_reward": 0.36171875833533707,
+      "rewards/format_reward": 0.0,
+      "step": 210
+    },
+    {
+      "completion_length": 613.5057498931885,
+      "epoch": 0.37985865724381623,
+      "grad_norm": 0.31533026695251465,
+      "kl": 0.41390380859375,
+      "learning_rate": 1.5610108950982494e-05,
+      "loss": 0.0166,
+      "reward": 0.3513020935002714,
+      "reward_std": 0.20828528399579227,
+      "rewards/accuracy_reward": 0.3513020935002714,
+      "rewards/format_reward": 0.0,
+      "step": 215
+    },
+    {
+      "completion_length": 601.9278839111328,
+      "epoch": 0.38869257950530034,
+      "grad_norm": 0.4190266728401184,
+      "kl": 0.6114501953125,
+      "learning_rate": 1.535201296145451e-05,
+      "loss": 0.0245,
+      "reward": 0.34817709494382143,
+      "reward_std": 0.20177546716295183,
+      "rewards/accuracy_reward": 0.34817709494382143,
+      "rewards/format_reward": 0.0,
+      "step": 220
+    },
+    {
+      "completion_length": 573.6575702667236,
+      "epoch": 0.39752650176678445,
+      "grad_norm": 0.5026732683181763,
+      "kl": 0.285626220703125,
+      "learning_rate": 1.5088820297659314e-05,
+      "loss": 0.0114,
+      "reward": 0.3773437601514161,
+      "reward_std": 0.21816352186724544,
+      "rewards/accuracy_reward": 0.3773437601514161,
+      "rewards/format_reward": 0.0,
+      "step": 225
+    },
+    {
+      "completion_length": 651.2117401123047,
+      "epoch": 0.40636042402826855,
+      "grad_norm": 0.4990842640399933,
+      "kl": 0.66497802734375,
+      "learning_rate": 1.4820781595626116e-05,
+      "loss": 0.0266,
+      "reward": 0.35208334154449406,
+      "reward_std": 0.22502720509655774,
+      "rewards/accuracy_reward": 0.35208334154449406,
+      "rewards/format_reward": 0.0,
+      "step": 230
+    },
+    {
+      "completion_length": 629.6927223205566,
+      "epoch": 0.41519434628975266,
+      "grad_norm": 0.5089983344078064,
+      "kl": 0.58125,
+      "learning_rate": 1.4548152106223157e-05,
+      "loss": 0.0233,
+      "reward": 0.34505209205672144,
+      "reward_std": 0.22014827858656644,
+      "rewards/accuracy_reward": 0.34505209205672144,
+      "rewards/format_reward": 0.0,
+      "step": 235
+    },
+    {
+      "completion_length": 572.4205932617188,
+      "epoch": 0.42402826855123676,
+      "grad_norm": 0.24783939123153687,
+      "kl": 0.29090576171875,
+      "learning_rate": 1.4271191452084598e-05,
+      "loss": 0.0116,
+      "reward": 0.37552084419876336,
+      "reward_std": 0.21084183501079679,
+      "rewards/accuracy_reward": 0.37552084419876336,
+      "rewards/format_reward": 0.0,
+      "step": 240
+    },
+    {
+      "completion_length": 578.217204284668,
+      "epoch": 0.43286219081272087,
+      "grad_norm": 0.4650849997997284,
+      "kl": 0.48919677734375,
+      "learning_rate": 1.3990163380374195e-05,
+      "loss": 0.0196,
+      "reward": 0.3367187597323209,
+      "reward_std": 0.20844926801510155,
+      "rewards/accuracy_reward": 0.3367187597323209,
+      "rewards/format_reward": 0.0,
+      "step": 245
+    },
+    {
+      "completion_length": 624.8208534240723,
+      "epoch": 0.4416961130742049,
+      "grad_norm": 0.3488200604915619,
+      "kl": 0.59932861328125,
+      "learning_rate": 1.3705335511621229e-05,
+      "loss": 0.024,
+      "reward": 0.3312500087078661,
+      "reward_std": 0.2190620328299701,
+      "rewards/accuracy_reward": 0.3312500087078661,
+      "rewards/format_reward": 0.0,
+      "step": 250
+    },
+    {
+      "completion_length": 684.0169486999512,
+      "epoch": 0.450530035335689,
+      "grad_norm": 0.5376025438308716,
+      "kl": 0.8154296875,
+      "learning_rate": 1.3416979084867851e-05,
+      "loss": 0.0326,
+      "reward": 0.3700520934537053,
+      "reward_std": 0.22584032705053686,
+      "rewards/accuracy_reward": 0.3700520934537053,
+      "rewards/format_reward": 0.0,
+      "step": 255
+    },
+    {
+      "completion_length": 661.9047054290771,
+      "epoch": 0.45936395759717313,
+      "grad_norm": 0.3623233139514923,
+      "kl": 0.54368896484375,
+      "learning_rate": 1.3125368699370567e-05,
+      "loss": 0.0218,
+      "reward": 0.3877604260109365,
+      "reward_std": 0.20249409526586531,
+      "rewards/accuracy_reward": 0.3877604260109365,
+      "rewards/format_reward": 0.0,
+      "step": 260
+    },
+    {
+      "completion_length": 700.0099140167237,
+      "epoch": 0.46819787985865724,
+      "grad_norm": 1.1863069534301758,
+      "kl": 0.6424072265625,
+      "learning_rate": 1.2830782053101807e-05,
+      "loss": 0.0257,
+      "reward": 0.36041667927056553,
+      "reward_std": 0.22222592746838926,
+      "rewards/accuracy_reward": 0.36041667927056553,
+      "rewards/format_reward": 0.0,
+      "step": 265
+    },
+    {
+      "completion_length": 724.0802284240723,
+      "epoch": 0.47703180212014135,
+      "grad_norm": 0.7839351892471313,
+      "kl": 0.72081298828125,
+      "learning_rate": 1.2533499678300618e-05,
+      "loss": 0.0288,
+      "reward": 0.349218759033829,
+      "reward_std": 0.21494603347964586,
+      "rewards/accuracy_reward": 0.349218759033829,
+      "rewards/format_reward": 0.0,
+      "step": 270
+    },
+    {
+      "completion_length": 672.5812713623047,
+      "epoch": 0.48586572438162545,
+      "grad_norm": 1.1607145071029663,
+      "kl": 0.60625,
+      "learning_rate": 1.223380467432432e-05,
+      "loss": 0.0243,
+      "reward": 0.3570312611758709,
+      "reward_std": 0.2103473860770464,
+      "rewards/accuracy_reward": 0.3570312611758709,
+      "rewards/format_reward": 0.0,
+      "step": 275
+    },
+    {
+      "completion_length": 637.1333545684814,
+      "epoch": 0.49469964664310956,
+      "grad_norm": 0.8619014620780945,
+      "kl": 0.69351806640625,
+      "learning_rate": 1.1931982438055506e-05,
+      "loss": 0.0278,
+      "reward": 0.3385416764765978,
+      "reward_std": 0.22153624026104807,
+      "rewards/accuracy_reward": 0.3385416764765978,
+      "rewards/format_reward": 0.0,
+      "step": 280
+    },
+    {
+      "completion_length": 609.2237182617188,
+      "epoch": 0.5035335689045937,
+      "grad_norm": 0.5707701444625854,
+      "kl": 0.51883544921875,
+      "learning_rate": 1.1628320392121118e-05,
+      "loss": 0.0208,
+      "reward": 0.36796876024454833,
+      "reward_std": 0.21415131203830243,
+      "rewards/accuracy_reward": 0.36796876024454833,
+      "rewards/format_reward": 0.0,
+      "step": 285
+    },
+    {
+      "completion_length": 592.3945514678956,
+      "epoch": 0.5123674911660777,
+      "grad_norm": 1.3289060592651367,
+      "kl": 0.53450927734375,
+      "learning_rate": 1.1323107711182474e-05,
+      "loss": 0.0214,
+      "reward": 0.35989584419876336,
+      "reward_std": 0.21049386793747543,
+      "rewards/accuracy_reward": 0.35989584419876336,
+      "rewards/format_reward": 0.0,
+      "step": 290
+    },
+    {
+      "completion_length": 587.1635578155517,
+      "epoch": 0.5212014134275619,
+      "grad_norm": 0.402815043926239,
+      "kl": 0.42646484375,
+      "learning_rate": 1.1016635046556773e-05,
+      "loss": 0.0171,
+      "reward": 0.337760426918976,
+      "reward_std": 0.21049761725589633,
+      "rewards/accuracy_reward": 0.337760426918976,
+      "rewards/format_reward": 0.0,
+      "step": 295
+    },
+    {
+      "completion_length": 624.5687698364258,
+      "epoch": 0.5300353356890459,
+      "grad_norm": 0.8592470288276672,
+      "kl": 0.54482421875,
+      "learning_rate": 1.0709194249432471e-05,
+      "loss": 0.0218,
+      "reward": 0.33411459270864724,
+      "reward_std": 0.195986431511119,
+      "rewards/accuracy_reward": 0.33411459270864724,
+      "rewards/format_reward": 0.0,
+      "step": 300
+    },
+    {
+      "epoch": 0.5300353356890459,
+      "eval_completion_length": 638.3154994419643,
+      "eval_kl": 0.42020089285714285,
+      "eval_loss": 0.016094399616122246,
+      "eval_reward": 0.38988096373421804,
+      "eval_reward_std": 0.21601268649101257,
+      "eval_rewards/accuracy_reward": 0.38988096373421804,
+      "eval_rewards/format_reward": 0.0,
+      "eval_runtime": 58.9898,
+      "eval_samples_per_second": 1.678,
+      "eval_steps_per_second": 0.034,
+      "step": 300
+    },
+    {
+      "completion_length": 634.8643444061279,
+      "epoch": 0.5388692579505301,
+      "grad_norm": 0.3406358063220978,
+      "kl": 0.562890625,
+      "learning_rate": 1.0401078092941972e-05,
+      "loss": 0.0225,
+      "reward": 0.33932292871177194,
+      "reward_std": 0.21115265358239413,
+      "rewards/accuracy_reward": 0.33932292871177194,
+      "rewards/format_reward": 0.0,
+      "step": 305
+    },
+    {
+      "completion_length": 605.8942924499512,
+      "epoch": 0.5477031802120141,
+      "grad_norm": 0.28513938188552856,
+      "kl": 0.35257568359375,
+      "learning_rate": 1.0092579993356386e-05,
+      "loss": 0.0141,
+      "reward": 0.3729166786419228,
+      "reward_std": 0.2152359550818801,
+      "rewards/accuracy_reward": 0.3729166786419228,
+      "rewards/format_reward": 0.0,
+      "step": 310
+    },
+    {
+      "completion_length": 635.9922039031983,
+      "epoch": 0.5565371024734982,
+      "grad_norm": 0.5206744074821472,
+      "kl": 0.5402099609375,
+      "learning_rate": 9.783993730667833e-06,
+      "loss": 0.0216,
+      "reward": 0.37265626098960636,
+      "reward_std": 0.21411701366305352,
+      "rewards/accuracy_reward": 0.37265626098960636,
+      "rewards/format_reward": 0.0,
+      "step": 315
+    },
+    {
+      "completion_length": 618.0041858673096,
+      "epoch": 0.5653710247349824,
+      "grad_norm": 0.5200411081314087,
+      "kl": 0.32769775390625,
+      "learning_rate": 9.475613168825374e-06,
+      "loss": 0.0131,
+      "reward": 0.34531251043081285,
+      "reward_std": 0.20256608743220567,
+      "rewards/accuracy_reward": 0.34531251043081285,
+      "rewards/format_reward": 0.0,
+      "step": 320
+    },
+    {
+      "completion_length": 626.5060115814209,
+      "epoch": 0.5742049469964664,
+      "grad_norm": 0.40177756547927856,
+      "kl": 0.62669677734375,
+      "learning_rate": 9.167731975890977e-06,
+      "loss": 0.0251,
+      "reward": 0.37265626145526765,
+      "reward_std": 0.22389861508272588,
+      "rewards/accuracy_reward": 0.37265626145526765,
+      "rewards/format_reward": 0.0,
+      "step": 325
+    },
+    {
+      "completion_length": 652.8005428314209,
+      "epoch": 0.5830388692579506,
+      "grad_norm": 0.712002158164978,
+      "kl": 0.62230224609375,
+      "learning_rate": 8.860643344382057e-06,
+      "loss": 0.0249,
+      "reward": 0.2976562581025064,
+      "reward_std": 0.2143249427434057,
+      "rewards/accuracy_reward": 0.2976562581025064,
+      "rewards/format_reward": 0.0,
+      "step": 330
+    },
+    {
+      "completion_length": 645.1703327178955,
+      "epoch": 0.5918727915194346,
+      "grad_norm": 0.6790205240249634,
+      "kl": 0.642987060546875,
+      "learning_rate": 8.554639712066837e-06,
+      "loss": 0.0257,
+      "reward": 0.32760417654644697,
+      "reward_std": 0.21297967862337827,
+      "rewards/accuracy_reward": 0.32760417654644697,
+      "rewards/format_reward": 0.0,
+      "step": 335
+    },
+    {
+      "completion_length": 640.1013217926026,
+      "epoch": 0.6007067137809188,
+      "grad_norm": 0.4978055953979492,
+      "kl": 0.48905029296875,
+      "learning_rate": 8.250012483478478e-06,
+      "loss": 0.0196,
+      "reward": 0.32291667591780426,
+      "reward_std": 0.2141881065443158,
+      "rewards/accuracy_reward": 0.32291667591780426,
+      "rewards/format_reward": 0.0,
+      "step": 340
+    },
+    {
+      "completion_length": 635.7453281402588,
+      "epoch": 0.6095406360424028,
+      "grad_norm": 0.5640181303024292,
+      "kl": 0.632666015625,
+      "learning_rate": 7.947051752413131e-06,
+      "loss": 0.0253,
+      "reward": 0.3463541771750897,
+      "reward_std": 0.21547603448852898,
+      "rewards/accuracy_reward": 0.3463541771750897,
+      "rewards/format_reward": 0.0,
+      "step": 345
+    },
+    {
+      "completion_length": 629.6963756561279,
+      "epoch": 0.6183745583038869,
+      "grad_norm": 0.6668252348899841,
+      "kl": 0.39061279296875,
+      "learning_rate": 7.646046025676198e-06,
+      "loss": 0.0156,
+      "reward": 0.34427084408234804,
+      "reward_std": 0.2008657043799758,
+      "rewards/accuracy_reward": 0.34427084408234804,
+      "rewards/format_reward": 0.0,
+      "step": 350
+    },
+    {
+      "completion_length": 646.0781440734863,
+      "epoch": 0.627208480565371,
+      "grad_norm": 0.3942064344882965,
+      "kl": 0.6318359375,
+      "learning_rate": 7.34728194833988e-06,
+      "loss": 0.0253,
+      "reward": 0.310677091171965,
+      "reward_std": 0.2146850608289242,
+      "rewards/accuracy_reward": 0.310677091171965,
+      "rewards/format_reward": 0.0,
+      "step": 355
+    },
+    {
+      "completion_length": 627.9213706970215,
+      "epoch": 0.6360424028268551,
+      "grad_norm": 1.0582332611083984,
+      "kl": 0.5762451171875,
+      "learning_rate": 7.051044030773619e-06,
+      "loss": 0.023,
+      "reward": 0.3656250102445483,
+      "reward_std": 0.22338166548870503,
+      "rewards/accuracy_reward": 0.3656250102445483,
+      "rewards/format_reward": 0.0,
+      "step": 360
+    },
+    {
+      "completion_length": 627.3216335296631,
+      "epoch": 0.6448763250883393,
+      "grad_norm": 0.6634775400161743,
+      "kl": 0.5901611328125,
+      "learning_rate": 6.757614377707409e-06,
+      "loss": 0.0236,
+      "reward": 0.3403645941521972,
+      "reward_std": 0.21042312644422054,
+      "rewards/accuracy_reward": 0.3403645941521972,
+      "rewards/format_reward": 0.0,
+      "step": 365
+    },
+    {
+      "completion_length": 640.6731971740722,
+      "epoch": 0.6537102473498233,
+      "grad_norm": 1.3277320861816406,
+      "kl": 0.5605224609375,
+      "learning_rate": 6.467272419585984e-06,
+      "loss": 0.0224,
+      "reward": 0.3505208439193666,
+      "reward_std": 0.2101976940408349,
+      "rewards/accuracy_reward": 0.3505208439193666,
+      "rewards/format_reward": 0.0,
+      "step": 370
+    },
+    {
+      "completion_length": 633.1187690734863,
+      "epoch": 0.6625441696113075,
+      "grad_norm": 1.4555702209472656,
+      "kl": 0.60972900390625,
+      "learning_rate": 6.18029464646968e-06,
+      "loss": 0.0244,
+      "reward": 0.37031251126900316,
+      "reward_std": 0.20959910564124584,
+      "rewards/accuracy_reward": 0.37031251126900316,
+      "rewards/format_reward": 0.0,
+      "step": 375
+    },
+    {
+      "completion_length": 625.6945514678955,
+      "epoch": 0.6713780918727915,
+      "grad_norm": 0.3421310782432556,
+      "kl": 0.515472412109375,
+      "learning_rate": 5.896954344735426e-06,
+      "loss": 0.0206,
+      "reward": 0.38437501098960636,
+      "reward_std": 0.21239073267206549,
+      "rewards/accuracy_reward": 0.38437501098960636,
+      "rewards/format_reward": 0.0,
+      "step": 380
+    },
+    {
+      "completion_length": 627.0758010864258,
+      "epoch": 0.6802120141342756,
+      "grad_norm": 0.7399106621742249,
+      "kl": 0.58017578125,
+      "learning_rate": 5.617521336828556e-06,
+      "loss": 0.0232,
+      "reward": 0.3838541777804494,
+      "reward_std": 0.21406057351268828,
+      "rewards/accuracy_reward": 0.3838541777804494,
+      "rewards/format_reward": 0.0,
+      "step": 385
+    },
+    {
+      "completion_length": 623.6627811431885,
+      "epoch": 0.6890459363957597,
+      "grad_norm": 1.0149897336959839,
+      "kl": 0.44697265625,
+      "learning_rate": 5.342261724313292e-06,
+      "loss": 0.0179,
+      "reward": 0.383593762665987,
+      "reward_std": 0.21393338404595852,
+      "rewards/accuracy_reward": 0.383593762665987,
+      "rewards/format_reward": 0.0,
+      "step": 390
+    },
+    {
+      "completion_length": 640.2651237487793,
+      "epoch": 0.6978798586572438,
+      "grad_norm": 0.541637659072876,
+      "kl": 0.5770751953125,
+      "learning_rate": 5.0714376344666095e-06,
+      "loss": 0.0231,
+      "reward": 0.36197917819954456,
+      "reward_std": 0.21014375062659382,
+      "rewards/accuracy_reward": 0.36197917819954456,
+      "rewards/format_reward": 0.0,
+      "step": 395
+    },
+    {
+      "completion_length": 630.2484580993653,
+      "epoch": 0.7067137809187279,
+      "grad_norm": 0.6355808973312378,
+      "kl": 0.53424072265625,
+      "learning_rate": 4.8053069706567555e-06,
+      "loss": 0.0214,
+      "reward": 0.35494792759418486,
+      "reward_std": 0.22305834330618382,
+      "rewards/accuracy_reward": 0.35494792759418486,
+      "rewards/format_reward": 0.0,
+      "step": 400
+    },
+    {
+      "epoch": 0.7067137809187279,
+      "eval_completion_length": 659.1681780133929,
+      "eval_kl": 0.8136160714285714,
+      "eval_loss": 0.031622376292943954,
+      "eval_reward": 0.39285715988704134,
+      "eval_reward_std": 0.2214358768292836,
+      "eval_rewards/accuracy_reward": 0.39285715988704134,
+      "eval_rewards/format_reward": 0.0,
+      "eval_runtime": 59.3442,
+      "eval_samples_per_second": 1.668,
+      "eval_steps_per_second": 0.034,
+      "step": 400
+    },
+    {
+      "completion_length": 627.2481983184814,
+      "epoch": 0.715547703180212,
+      "grad_norm": 0.3775692880153656,
+      "kl": 0.71290283203125,
+      "learning_rate": 4.5441231667441724e-06,
+      "loss": 0.0285,
+      "reward": 0.3666666762903333,
+      "reward_std": 0.20955641102045774,
+      "rewards/accuracy_reward": 0.3666666762903333,
+      "rewards/format_reward": 0.0,
+      "step": 405
+    },
+    {
+      "completion_length": 613.2617389678956,
+      "epoch": 0.7243816254416962,
+      "grad_norm": 0.5985289812088013,
+      "kl": 0.4572998046875,
+      "learning_rate": 4.288134945738684e-06,
+      "loss": 0.0183,
+      "reward": 0.37031251285225153,
+      "reward_std": 0.22802586145699025,
+      "rewards/accuracy_reward": 0.37031251285225153,
+      "rewards/format_reward": 0.0,
+      "step": 410
+    },
+    {
+      "completion_length": 598.452883529663,
+      "epoch": 0.7332155477031802,
+      "grad_norm": 0.6456860303878784,
+      "kl": 0.45269775390625,
+      "learning_rate": 4.037586082942805e-06,
+      "loss": 0.0181,
+      "reward": 0.40260417833924295,
+      "reward_std": 0.20974129284732043,
+      "rewards/accuracy_reward": 0.40260417833924295,
+      "rewards/format_reward": 0.0,
+      "step": 415
+    },
+    {
+      "completion_length": 630.413037109375,
+      "epoch": 0.7420494699646644,
+      "grad_norm": 0.651096522808075,
+      "kl": 0.5280029296875,
+      "learning_rate": 3.7927151738066693e-06,
+      "loss": 0.0211,
+      "reward": 0.36588542740792035,
+      "reward_std": 0.21933015692047775,
+      "rewards/accuracy_reward": 0.36588542740792035,
+      "rewards/format_reward": 0.0,
+      "step": 420
+    },
+    {
+      "completion_length": 613.5330924987793,
+      "epoch": 0.7508833922261484,
+      "grad_norm": 0.45714229345321655,
+      "kl": 0.4800537109375,
+      "learning_rate": 3.553755406715724e-06,
+      "loss": 0.0192,
+      "reward": 0.3856770938495174,
+      "reward_std": 0.21619091741740704,
+      "rewards/accuracy_reward": 0.3856770938495174,
+      "rewards/format_reward": 0.0,
+      "step": 425
+    },
+    {
+      "completion_length": 631.0403835296631,
+      "epoch": 0.7597173144876325,
+      "grad_norm": 1.2694158554077148,
+      "kl": 0.59678955078125,
+      "learning_rate": 3.320934340927513e-06,
+      "loss": 0.0239,
+      "reward": 0.37682292722165583,
+      "reward_std": 0.22145710308104755,
+      "rewards/accuracy_reward": 0.37682292722165583,
+      "rewards/format_reward": 0.0,
+      "step": 430
+    },
+    {
+      "completion_length": 653.9471565246582,
+      "epoch": 0.7685512367491166,
+      "grad_norm": 1.0589816570281982,
+      "kl": 0.6340087890625,
+      "learning_rate": 3.094473689869002e-06,
+      "loss": 0.0254,
+      "reward": 0.36328125931322575,
+      "reward_std": 0.22120612636208534,
+      "rewards/accuracy_reward": 0.36328125931322575,
+      "rewards/format_reward": 0.0,
+      "step": 435
+    },
+    {
+      "completion_length": 650.4448143005371,
+      "epoch": 0.7773851590106007,
+      "grad_norm": 0.4063872992992401,
+      "kl": 0.504638671875,
+      "learning_rate": 2.8745891100008683e-06,
+      "loss": 0.0202,
+      "reward": 0.3583333434537053,
+      "reward_std": 0.20930543607100843,
+      "rewards/accuracy_reward": 0.3583333434537053,
+      "rewards/format_reward": 0.0,
+      "step": 440
+    },
+    {
+      "completion_length": 653.0390823364257,
+      "epoch": 0.7862190812720848,
+      "grad_norm": 0.749354362487793,
+      "kl": 0.5887939453125,
+      "learning_rate": 2.6614899954497797e-06,
+      "loss": 0.0236,
+      "reward": 0.37760417694225906,
+      "reward_std": 0.2196731221396476,
+      "rewards/accuracy_reward": 0.37760417694225906,
+      "rewards/format_reward": 0.0,
+      "step": 445
+    },
+    {
+      "completion_length": 633.9838752746582,
+      "epoch": 0.7950530035335689,
+      "grad_norm": 0.5288252830505371,
+      "kl": 0.54503173828125,
+      "learning_rate": 2.455379278604226e-06,
+      "loss": 0.0218,
+      "reward": 0.37343751238659023,
+      "reward_std": 0.21816726946271955,
+      "rewards/accuracy_reward": 0.37343751238659023,
+      "rewards/format_reward": 0.0,
+      "step": 450
+    },
+    {
+      "completion_length": 625.8023654937745,
+      "epoch": 0.803886925795053,
+      "grad_norm": 0.5270794034004211,
+      "kl": 0.57318115234375,
+      "learning_rate": 2.256453236863815e-06,
+      "loss": 0.0229,
+      "reward": 0.38072917629033326,
+      "reward_std": 0.213863542675972,
+      "rewards/accuracy_reward": 0.38072917629033326,
+      "rewards/format_reward": 0.0,
+      "step": 455
+    },
+    {
+      "completion_length": 626.2771045684815,
+      "epoch": 0.8127208480565371,
+      "grad_norm": 0.6814591884613037,
+      "kl": 0.55157470703125,
+      "learning_rate": 2.064901305726055e-06,
+      "loss": 0.0221,
+      "reward": 0.370572925824672,
+      "reward_std": 0.22363763963803648,
+      "rewards/accuracy_reward": 0.370572925824672,
+      "rewards/format_reward": 0.0,
+      "step": 460
+    },
+    {
+      "completion_length": 623.0877819061279,
+      "epoch": 0.8215547703180212,
+      "grad_norm": 0.5188084244728088,
+      "kl": 0.58336181640625,
+      "learning_rate": 1.880905898388612e-06,
+      "loss": 0.0233,
+      "reward": 0.37890625949949025,
+      "reward_std": 0.2149899779818952,
+      "rewards/accuracy_reward": 0.37890625949949025,
+      "rewards/format_reward": 0.0,
+      "step": 465
+    },
+    {
+      "completion_length": 627.9685092926026,
+      "epoch": 0.8303886925795053,
+      "grad_norm": 0.41752591729164124,
+      "kl": 0.517425537109375,
+      "learning_rate": 1.7046422320388556e-06,
+      "loss": 0.0207,
+      "reward": 0.38437501108273864,
+      "reward_std": 0.23573571695014833,
+      "rewards/accuracy_reward": 0.38437501108273864,
+      "rewards/format_reward": 0.0,
+      "step": 470
+    },
+    {
+      "completion_length": 630.6453319549561,
+      "epoch": 0.8392226148409894,
+      "grad_norm": 0.5433622598648071,
+      "kl": 0.56822509765625,
+      "learning_rate": 1.5362781609960853e-06,
+      "loss": 0.0227,
+      "reward": 0.3841145928716287,
+      "reward_std": 0.22084475858137012,
+      "rewards/accuracy_reward": 0.3841145928716287,
+      "rewards/format_reward": 0.0,
+      "step": 475
+    },
+    {
+      "completion_length": 629.8807479858399,
+      "epoch": 0.8480565371024735,
+      "grad_norm": 0.5418740510940552,
+      "kl": 0.5584716796875,
+      "learning_rate": 1.375974016865359e-06,
+      "loss": 0.0223,
+      "reward": 0.3921875139232725,
+      "reward_std": 0.2101119550410658,
+      "rewards/accuracy_reward": 0.3921875139232725,
+      "rewards/format_reward": 0.0,
+      "step": 480
+    },
+    {
+      "completion_length": 623.6669467926025,
+      "epoch": 0.8568904593639576,
+      "grad_norm": 0.4898351728916168,
+      "kl": 0.55391845703125,
+      "learning_rate": 1.2238824558551365e-06,
+      "loss": 0.0222,
+      "reward": 0.40130209433846176,
+      "reward_std": 0.2192101170774549,
+      "rewards/accuracy_reward": 0.40130209433846176,
+      "rewards/format_reward": 0.0,
+      "step": 485
+    },
+    {
+      "completion_length": 633.0653854370117,
+      "epoch": 0.8657243816254417,
+      "grad_norm": 0.5053507685661316,
+      "kl": 0.5678955078125,
+      "learning_rate": 1.080148313404127e-06,
+      "loss": 0.0227,
+      "reward": 0.3671875108964741,
+      "reward_std": 0.21440658504143356,
+      "rewards/accuracy_reward": 0.3671875108964741,
+      "rewards/format_reward": 0.0,
+      "step": 490
+    },
+    {
+      "completion_length": 640.2271011352539,
+      "epoch": 0.8745583038869258,
+      "grad_norm": 0.6309220790863037,
+      "kl": 0.57041015625,
+      "learning_rate": 9.449084662557984e-07,
+      "loss": 0.0228,
+      "reward": 0.35885417805984615,
+      "reward_std": 0.2127078040037304,
+      "rewards/accuracy_reward": 0.35885417805984615,
+      "rewards/format_reward": 0.0,
+      "step": 495
+    },
+    {
+      "completion_length": 629.1083526611328,
+      "epoch": 0.8833922261484098,
+      "grad_norm": 0.4867832660675049,
+      "kl": 0.56951904296875,
+      "learning_rate": 8.182917021118664e-07,
+      "loss": 0.0228,
+      "reward": 0.3739583442918956,
+      "reward_std": 0.2111146087292582,
+      "rewards/accuracy_reward": 0.3739583442918956,
+      "rewards/format_reward": 0.0,
+      "step": 500
+    },
+    {
+      "epoch": 0.8833922261484098,
+      "eval_completion_length": 641.7425798688616,
+      "eval_kl": 0.505859375,
+      "eval_loss": 0.019658148288726807,
+      "eval_reward": 0.4166666737624577,
+      "eval_reward_std": 0.21701403175081527,
+      "eval_rewards/accuracy_reward": 0.4166666737624577,
+      "eval_rewards/format_reward": 0.0,
+      "eval_runtime": 60.7992,
+      "eval_samples_per_second": 1.628,
+      "eval_steps_per_second": 0.033,
+      "step": 500
+    },
+    {
+      "completion_length": 630.6286685943603,
+      "epoch": 0.892226148409894,
+      "grad_norm": 0.7990770936012268,
+      "kl": 0.5071533203125,
+      "learning_rate": 7.004185969889188e-07,
+      "loss": 0.0203,
+      "reward": 0.37343750884756444,
+      "reward_std": 0.20867379917763174,
+      "rewards/accuracy_reward": 0.37343750884756444,
+      "rewards/format_reward": 0.0,
+      "step": 505
+    },
+    {
+      "completion_length": 608.6307468414307,
+      "epoch": 0.901060070671378,
+      "grad_norm": 0.6939222812652588,
+      "kl": 0.487548828125,
+      "learning_rate": 5.914014003949408e-07,
+      "loss": 0.0195,
+      "reward": 0.39921876005828383,
+      "reward_std": 0.18637881381437182,
+      "rewards/accuracy_reward": 0.39921876005828383,
+      "rewards/format_reward": 0.0,
+      "step": 510
+    },
+    {
+      "completion_length": 609.941683959961,
+      "epoch": 0.9098939929328622,
+      "grad_norm": 0.7502965927124023,
+      "kl": 0.49111328125,
+      "learning_rate": 4.913439284351207e-07,
+      "loss": 0.0197,
+      "reward": 0.39531251043081284,
+      "reward_std": 0.22331592459231614,
+      "rewards/accuracy_reward": 0.39531251043081284,
+      "rewards/format_reward": 0.0,
+      "step": 515
+    },
+    {
+      "completion_length": 600.0021045684814,
+      "epoch": 0.9187279151943463,
+      "grad_norm": 0.714733898639679,
+      "kl": 0.51158447265625,
+      "learning_rate": 4.003414649486892e-07,
+      "loss": 0.0205,
+      "reward": 0.4018229302018881,
+      "reward_std": 0.21587419435381888,
+      "rewards/accuracy_reward": 0.4018229302018881,
+      "rewards/format_reward": 0.0,
+      "step": 520
+    },
+    {
+      "completion_length": 619.1349174499512,
+      "epoch": 0.9275618374558304,
+      "grad_norm": 0.5860229134559631,
+      "kl": 0.521954345703125,
+      "learning_rate": 3.184806707709698e-07,
+      "loss": 0.0209,
+      "reward": 0.38125000945292414,
+      "reward_std": 0.20484522059559823,
+      "rewards/accuracy_reward": 0.38125000945292414,
+      "rewards/format_reward": 0.0,
+      "step": 525
+    },
+    {
+      "completion_length": 618.1958557128906,
+      "epoch": 0.9363957597173145,
+      "grad_norm": 0.5392491221427917,
+      "kl": 0.5330322265625,
+      "learning_rate": 2.458395012070369e-07,
+      "loss": 0.0213,
+      "reward": 0.3861979279667139,
+      "reward_std": 0.23485649805516004,
+      "rewards/accuracy_reward": 0.3861979279667139,
+      "rewards/format_reward": 0.0,
+      "step": 530
+    },
+    {
+      "completion_length": 631.1799655914307,
+      "epoch": 0.9452296819787986,
+      "grad_norm": 0.5831504464149475,
+      "kl": 0.5838623046875,
+      "learning_rate": 1.8248713179557788e-07,
+      "loss": 0.0234,
+      "reward": 0.39375001257285475,
+      "reward_std": 0.23128639576025306,
+      "rewards/accuracy_reward": 0.39375001257285475,
+      "rewards/format_reward": 0.0,
+      "step": 535
+    },
+    {
+      "completion_length": 624.198452758789,
+      "epoch": 0.9540636042402827,
+      "grad_norm": 0.48825210332870483,
+      "kl": 0.538427734375,
+      "learning_rate": 1.2848389243363514e-07,
+      "loss": 0.0215,
+      "reward": 0.38046876033768057,
+      "reward_std": 0.21035578032024205,
+      "rewards/accuracy_reward": 0.38046876033768057,
+      "rewards/format_reward": 0.0,
+      "step": 540
+    },
+    {
+      "completion_length": 629.3713752746582,
+      "epoch": 0.9628975265017667,
+      "grad_norm": 0.5276216864585876,
+      "kl": 0.57401123046875,
+      "learning_rate": 8.388120992499083e-08,
+      "loss": 0.023,
+      "reward": 0.3695312602445483,
+      "reward_std": 0.21676895981654526,
+      "rewards/accuracy_reward": 0.3695312602445483,
+      "rewards/format_reward": 0.0,
+      "step": 545
+    },
+    {
+      "completion_length": 618.4643447875976,
+      "epoch": 0.9717314487632509,
+      "grad_norm": 0.5697087645530701,
+      "kl": 0.556787109375,
+      "learning_rate": 4.8721559006873473e-08,
+      "loss": 0.0223,
+      "reward": 0.39375000931322574,
+      "reward_std": 0.19753997740335763,
+      "rewards/accuracy_reward": 0.39375000931322574,
+      "rewards/format_reward": 0.0,
+      "step": 550
+    },
+    {
+      "completion_length": 629.5317882537842,
+      "epoch": 0.980565371024735,
+      "grad_norm": 0.5330089926719666,
+      "kl": 0.61251220703125,
+      "learning_rate": 2.3038421901651064e-08,
+      "loss": 0.0245,
+      "reward": 0.3750000110827386,
+      "reward_std": 0.21318545560352503,
+      "rewards/accuracy_reward": 0.3750000110827386,
+      "rewards/format_reward": 0.0,
+      "step": 555
+    },
+    {
+      "completion_length": 625.5620018005371,
+      "epoch": 0.9893992932862191,
+      "grad_norm": 0.6242526173591614,
+      "kl": 0.58502197265625,
+      "learning_rate": 6.856256432000719e-09,
+      "loss": 0.0234,
+      "reward": 0.4000000122003257,
+      "reward_std": 0.24645261880941688,
+      "rewards/accuracy_reward": 0.4000000122003257,
+      "rewards/format_reward": 0.0,
+      "step": 560
+    },
+    {
+      "completion_length": 617.1008010864258,
+      "epoch": 0.9982332155477032,
+      "grad_norm": 0.6724653840065002,
+      "kl": 0.59691162109375,
+      "learning_rate": 1.904727299473219e-10,
+      "loss": 0.0239,
+      "reward": 0.3833333443384618,
+      "reward_std": 0.21862742602825164,
+      "rewards/accuracy_reward": 0.3833333443384618,
+      "rewards/format_reward": 0.0,
+      "step": 565
+    },
+    {
+      "completion_length": 550.541669845581,
+      "epoch": 1.0,
+      "kl": 0.6014404296875,
+      "reward": 0.385416679084301,
+      "reward_std": 0.21174714877270162,
+      "rewards/accuracy_reward": 0.385416679084301,
+      "rewards/format_reward": 0.0,
+      "step": 566,
+      "total_flos": 0.0,
+      "train_loss": 0.08444789265832109,
+      "train_runtime": 62734.3355,
+      "train_samples_per_second": 1.155,
+      "train_steps_per_second": 0.009
+    }
+  ],
+  "logging_steps": 5,
+  "max_steps": 566,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": false,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 0.0,
+  "train_batch_size": 16,
+  "trial_name": null,
+  "trial_params": null
+}

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e5290bf3fb18b69bacac21c486d37213a0964b6e84c004efcfca891121457aad
+size 7608