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 +1,68 @@
+---
+base_model: Qwen/Qwen2.5-1.5B-Instruct
+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 [Qwen/Qwen2.5-1.5B-Instruct](https://huggingface.co/Qwen/Qwen2.5-1.5B-Instruct).
+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="lzy337/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/lzy37ld/huggingface/runs/7pjs2cfa) 
+
+
+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: 0.15.0.dev0
+- Transformers: 4.49.0.dev0
+- Pytorch: 2.5.1+cu124
+- Datasets: 3.2.0
+- 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": 2.2585060952830793,
+    "train_runtime": 36118.4141,
+    "train_samples": 72441,
+    "train_samples_per_second": 2.006,
+    "train_steps_per_second": 0.01
+}

config.json

@@ -0,0 +1,29 @@
+{
+  "_name_or_path": "Qwen/Qwen2.5-1.5B-Instruct",
+  "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": null,
+  "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 sha256:7019e90d563027891432f45b67fee830aff23e41dad2378c0f8745408cc513b1
+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": "<|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 sha256:5eee858c5123a4279c3e1f7b81247343f356ac767940b2692a928ad929543214
+size 11422063

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": "<|endoftext|>",
+  "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": 2.2585060952830793,
+    "train_runtime": 36118.4141,
+    "train_samples": 72441,
+    "train_samples_per_second": 2.006,
+    "train_steps_per_second": 0.01
+}

trainer_state.json

@@ -0,0 +1,1065 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9991717740599636,
+  "eval_steps": 100,
+  "global_step": 377,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "completion_length": 401.9587879180908,
+      "epoch": 0.013251615040583071,
+      "grad_norm": 0.8373466730117798,
+      "kl": 0.000208282470703125,
+      "learning_rate": 2.631578947368421e-06,
+      "loss": 0.0,
+      "reward": 0.6059523954987526,
+      "reward_std": 0.44210540093481543,
+      "rewards/accuracy_reward": 0.1523809561971575,
+      "rewards/format_reward": 0.45357143580913545,
+      "step": 5
+    },
+    {
+      "completion_length": 250.81726770401002,
+      "epoch": 0.026503230081166142,
+      "grad_norm": 3.4991674423217773,
+      "kl": 0.2951213836669922,
+      "learning_rate": 5.263157894736842e-06,
+      "loss": 0.0118,
+      "reward": 0.8930059671401978,
+      "reward_std": 0.3061427103355527,
+      "rewards/accuracy_reward": 0.0828869066084735,
+      "rewards/format_reward": 0.8101190589368343,
+      "step": 10
+    },
+    {
+      "completion_length": 168.98169927597047,
+      "epoch": 0.03975484512174921,
+      "grad_norm": 0.36309367418289185,
+      "kl": 0.07768707275390625,
+      "learning_rate": 7.894736842105265e-06,
+      "loss": 0.0031,
+      "reward": 1.0202381230890751,
+      "reward_std": 0.22592478841543198,
+      "rewards/accuracy_reward": 0.07678571618162096,
+      "rewards/format_reward": 0.9434523954987526,
+      "step": 15
+    },
+    {
+      "completion_length": 106.28705511093139,
+      "epoch": 0.053006460162332285,
+      "grad_norm": 0.6117094159126282,
+      "kl": 0.803338623046875,
+      "learning_rate": 1.0526315789473684e-05,
+      "loss": 0.0322,
+      "reward": 1.0802083522081376,
+      "reward_std": 0.2028849871829152,
+      "rewards/accuracy_reward": 0.11369047905318438,
+      "rewards/format_reward": 0.9665178783237934,
+      "step": 20
+    },
+    {
+      "completion_length": 97.13050751686096,
+      "epoch": 0.06625807520291535,
+      "grad_norm": 464.3936767578125,
+      "kl": 2.28043212890625,
+      "learning_rate": 1.3157894736842108e-05,
+      "loss": 0.0917,
+      "reward": 1.123511929810047,
+      "reward_std": 0.21935302754864097,
+      "rewards/accuracy_reward": 0.14895833635237068,
+      "rewards/format_reward": 0.9745535954833031,
+      "step": 25
+    },
+    {
+      "completion_length": 107.32217483520508,
+      "epoch": 0.07950969024349842,
+      "grad_norm": 0.5923934578895569,
+      "kl": 0.309405517578125,
+      "learning_rate": 1.578947368421053e-05,
+      "loss": 0.0124,
+      "reward": 1.1953869313001633,
+      "reward_std": 0.28015064969658854,
+      "rewards/accuracy_reward": 0.22440476599149406,
+      "rewards/format_reward": 0.9709821663796901,
+      "step": 30
+    },
+    {
+      "completion_length": 121.52857341766358,
+      "epoch": 0.0927613052840815,
+      "grad_norm": 1.9968442916870117,
+      "kl": 0.291650390625,
+      "learning_rate": 1.8421052631578947e-05,
+      "loss": 0.0117,
+      "reward": 1.1989583566784858,
+      "reward_std": 0.3172613775357604,
+      "rewards/accuracy_reward": 0.25104167093522844,
+      "rewards/format_reward": 0.9479166828095913,
+      "step": 35
+    },
+    {
+      "completion_length": 142.87068710327148,
+      "epoch": 0.10601292032466457,
+      "grad_norm": 0.3424147367477417,
+      "kl": 0.188043212890625,
+      "learning_rate": 1.9998282416292057e-05,
+      "loss": 0.0075,
+      "reward": 1.219642886519432,
+      "reward_std": 0.29310949984937906,
+      "rewards/accuracy_reward": 0.25595238553360106,
+      "rewards/format_reward": 0.963690496981144,
+      "step": 40
+    },
+    {
+      "completion_length": 172.70342588424683,
+      "epoch": 0.11926453536524764,
+      "grad_norm": 51.79156494140625,
+      "kl": 0.2910400390625,
+      "learning_rate": 1.9978966374934255e-05,
+      "loss": 0.0116,
+      "reward": 1.2520833611488342,
+      "reward_std": 0.29622648563236,
+      "rewards/accuracy_reward": 0.28556547965854406,
+      "rewards/format_reward": 0.9665178760886193,
+      "step": 45
+    },
+    {
+      "completion_length": 199.47649230957032,
+      "epoch": 0.1325161504058307,
+      "grad_norm": 1.2901036739349365,
+      "kl": 0.60821533203125,
+      "learning_rate": 1.993822891578708e-05,
+      "loss": 0.0243,
+      "reward": 1.2258928805589675,
+      "reward_std": 0.3094880359247327,
+      "rewards/accuracy_reward": 0.2681547679938376,
+      "rewards/format_reward": 0.9577381119132042,
+      "step": 50
+    },
+    {
+      "completion_length": 184.5258967399597,
+      "epoch": 0.1457677654464138,
+      "grad_norm": 0.8856645822525024,
+      "kl": 0.5405517578125,
+      "learning_rate": 1.9876157488201426e-05,
+      "loss": 0.0216,
+      "reward": 1.1828869305551053,
+      "reward_std": 0.40790193118155005,
+      "rewards/accuracy_reward": 0.3052083390764892,
+      "rewards/format_reward": 0.8776785850524902,
+      "step": 55
+    },
+    {
+      "completion_length": 157.01279964447022,
+      "epoch": 0.15901938048699685,
+      "grad_norm": 2.5129318237304688,
+      "kl": 3.7045166015625,
+      "learning_rate": 1.9792885338240375e-05,
+      "loss": 0.1482,
+      "reward": 1.2296131163835526,
+      "reward_std": 0.35207056868821385,
+      "rewards/accuracy_reward": 0.3175595284905285,
+      "rewards/format_reward": 0.9120535857975482,
+      "step": 60
+    },
+    {
+      "completion_length": 117.98095417022705,
+      "epoch": 0.17227099552757993,
+      "grad_norm": 3.385244369506836,
+      "kl": 4187.849670410156,
+      "learning_rate": 1.9688591222645607e-05,
+      "loss": 167.6914,
+      "reward": 1.2738095462322234,
+      "reward_std": 0.27289785919710996,
+      "rewards/accuracy_reward": 0.30744048254564404,
+      "rewards/format_reward": 0.9663690678775311,
+      "step": 65
+    },
+    {
+      "completion_length": 102.06607322692871,
+      "epoch": 0.185522610568163,
+      "grad_norm": 0.4773934483528137,
+      "kl": 0.56761474609375,
+      "learning_rate": 1.9563499025108e-05,
+      "loss": 0.0227,
+      "reward": 1.2117559798061848,
+      "reward_std": 0.2911853768862784,
+      "rewards/accuracy_reward": 0.2623511940706521,
+      "rewards/format_reward": 0.9494047820568084,
+      "step": 70
+    },
+    {
+      "completion_length": 132.6630976676941,
+      "epoch": 0.19877422560874605,
+      "grad_norm": 126.72654724121094,
+      "kl": 0.25335693359375,
+      "learning_rate": 1.941787727566613e-05,
+      "loss": 0.0101,
+      "reward": 1.2194940701127053,
+      "reward_std": 0.29773683892562985,
+      "rewards/accuracy_reward": 0.2714285757392645,
+      "rewards/format_reward": 0.9480654872953892,
+      "step": 75
+    },
+    {
+      "completion_length": 156.06845407485963,
+      "epoch": 0.21202584064932914,
+      "grad_norm": 0.282274454832077,
+      "kl": 0.1987060546875,
+      "learning_rate": 1.9252038574264403e-05,
+      "loss": 0.0079,
+      "reward": 1.226041679829359,
+      "reward_std": 0.321970590762794,
+      "rewards/accuracy_reward": 0.2833333383779973,
+      "rewards/format_reward": 0.9427083477377891,
+      "step": 80
+    },
+    {
+      "completion_length": 152.93095617294313,
+      "epoch": 0.2252774556899122,
+      "grad_norm": 0.2814752459526062,
+      "kl": 0.199560546875,
+      "learning_rate": 1.90663389197082e-05,
+      "loss": 0.008,
+      "reward": 1.2322916850447654,
+      "reward_std": 0.31209713909775016,
+      "rewards/accuracy_reward": 0.2824404813349247,
+      "rewards/format_reward": 0.9498512066900731,
+      "step": 85
+    },
+    {
+      "completion_length": 143.89285945892334,
+      "epoch": 0.23852907073049529,
+      "grad_norm": 8.334383010864258,
+      "kl": 0.5559326171875,
+      "learning_rate": 1.8861176945456542e-05,
+      "loss": 0.0222,
+      "reward": 1.250595250725746,
+      "reward_std": 0.26953114988282323,
+      "rewards/accuracy_reward": 0.27321428968571126,
+      "rewards/format_reward": 0.9773809745907783,
+      "step": 90
+    },
+    {
+      "completion_length": 159.24092626571655,
+      "epoch": 0.25178068577107837,
+      "grad_norm": 1.4891200065612793,
+      "kl": 1.0826171875,
+      "learning_rate": 1.8636993063892822e-05,
+      "loss": 0.0433,
+      "reward": 1.2684524029493331,
+      "reward_std": 0.28893801737576724,
+      "rewards/accuracy_reward": 0.3023809571750462,
+      "rewards/format_reward": 0.9660714507102967,
+      "step": 95
+    },
+    {
+      "completion_length": 160.03378248214722,
+      "epoch": 0.2650323008116614,
+      "grad_norm": 20.321685791015625,
+      "kl": 1.442724609375,
+      "learning_rate": 1.8394268520910467e-05,
+      "loss": 0.0577,
+      "reward": 1.2886904954910279,
+      "reward_std": 0.29596499148756267,
+      "rewards/accuracy_reward": 0.3220238142646849,
+      "rewards/format_reward": 0.9666666857898235,
+      "step": 100
+    },
+    {
+      "epoch": 0.2650323008116614,
+      "eval_completion_length": 136.75324919729522,
+      "eval_kl": 0.21012369791666666,
+      "eval_loss": 0.008404189720749855,
+      "eval_reward": 1.3304473486813633,
+      "eval_reward_std": 0.2531851855191318,
+      "eval_rewards/accuracy_reward": 0.3362193459814245,
+      "eval_rewards/format_reward": 0.9942279945720326,
+      "eval_runtime": 50.5376,
+      "eval_samples_per_second": 1.959,
+      "eval_steps_per_second": 0.653,
+      "step": 100
+    },
+    {
+      "completion_length": 141.8834858894348,
+      "epoch": 0.2782839158522445,
+      "grad_norm": 1.1371376514434814,
+      "kl": 0.77235107421875,
+      "learning_rate": 1.8133524362843105e-05,
+      "loss": 0.031,
+      "reward": 1.2941964536905288,
+      "reward_std": 0.24222191106528043,
+      "rewards/accuracy_reward": 0.3052083404269069,
+      "rewards/format_reward": 0.9889881186187267,
+      "step": 105
+    },
+    {
+      "completion_length": 185.20193843841554,
+      "epoch": 0.2915355308928276,
+      "grad_norm": 1.1394041776657104,
+      "kl": 1.09859619140625,
+      "learning_rate": 1.7855320317956785e-05,
+      "loss": 0.0439,
+      "reward": 1.2558035954833031,
+      "reward_std": 0.31485513523221015,
+      "rewards/accuracy_reward": 0.30238095819950106,
+      "rewards/format_reward": 0.9534226365387439,
+      "step": 110
+    },
+    {
+      "completion_length": 209.79643230438234,
+      "epoch": 0.3047871459334106,
+      "grad_norm": 0.7815774083137512,
+      "kl": 0.42103271484375,
+      "learning_rate": 1.7560253594905425e-05,
+      "loss": 0.0168,
+      "reward": 1.265327402949333,
+      "reward_std": 0.32841875310987234,
+      "rewards/accuracy_reward": 0.32113095666281877,
+      "rewards/format_reward": 0.9441964447498321,
+      "step": 115
+    },
+    {
+      "completion_length": 210.51429042816162,
+      "epoch": 0.3180387609739937,
+      "grad_norm": 0.23798367381095886,
+      "kl": 0.211962890625,
+      "learning_rate": 1.7248957600728664e-05,
+      "loss": 0.0085,
+      "reward": 1.247619064897299,
+      "reward_std": 0.34062892515212295,
+      "rewards/accuracy_reward": 0.33303571986034514,
+      "rewards/format_reward": 0.9145833447575569,
+      "step": 120
+    },
+    {
+      "completion_length": 189.7093783378601,
+      "epoch": 0.3312903760145768,
+      "grad_norm": 0.28196191787719727,
+      "kl": 0.208447265625,
+      "learning_rate": 1.6922100581144228e-05,
+      "loss": 0.0083,
+      "reward": 1.2735119313001633,
+      "reward_std": 0.33266823515295985,
+      "rewards/accuracy_reward": 0.3342261975631118,
+      "rewards/format_reward": 0.9392857261002063,
+      "step": 125
+    },
+    {
+      "completion_length": 196.90774164199829,
+      "epoch": 0.34454199105515987,
+      "grad_norm": 0.2767742872238159,
+      "kl": 0.2078125,
+      "learning_rate": 1.658038418605361e-05,
+      "loss": 0.0083,
+      "reward": 1.2672619305551052,
+      "reward_std": 0.34803154738619924,
+      "rewards/accuracy_reward": 0.33422619649209084,
+      "rewards/format_reward": 0.9330357283353805,
+      "step": 130
+    },
+    {
+      "completion_length": 199.43452701568603,
+      "epoch": 0.3577936060957429,
+      "grad_norm": 0.3050677180290222,
+      "kl": 0.1993896484375,
+      "learning_rate": 1.6224541963340392e-05,
+      "loss": 0.008,
+      "reward": 1.2619047835469246,
+      "reward_std": 0.3566672120243311,
+      "rewards/accuracy_reward": 0.3340773852542043,
+      "rewards/format_reward": 0.9278273940086365,
+      "step": 135
+    },
+    {
+      "completion_length": 189.28437843322754,
+      "epoch": 0.371045221136326,
+      "grad_norm": 0.30024826526641846,
+      "kl": 0.21253662109375,
+      "learning_rate": 1.5855337784194576e-05,
+      "loss": 0.0085,
+      "reward": 1.2757440686225892,
+      "reward_std": 0.32623464316129686,
+      "rewards/accuracy_reward": 0.337351196911186,
+      "rewards/format_reward": 0.938392873108387,
+      "step": 140
+    },
+    {
+      "completion_length": 180.53869304656982,
+      "epoch": 0.3842968361769091,
+      "grad_norm": 3.9169106483459473,
+      "kl": 1.5111572265625,
+      "learning_rate": 1.5473564203343173e-05,
+      "loss": 0.0605,
+      "reward": 1.3023809731006621,
+      "reward_std": 0.3125137767754495,
+      "rewards/accuracy_reward": 0.3575892925262451,
+      "rewards/format_reward": 0.9447916820645332,
+      "step": 145
+    },
+    {
+      "completion_length": 168.65536031723022,
+      "epoch": 0.3975484512174921,
+      "grad_norm": 12.137922286987305,
+      "kl": 2.247265625,
+      "learning_rate": 1.5080040757707045e-05,
+      "loss": 0.0899,
+      "reward": 1.2834821715950966,
+      "reward_std": 0.317144919373095,
+      "rewards/accuracy_reward": 0.33675595847889783,
+      "rewards/format_reward": 0.9467262037098407,
+      "step": 150
+    },
+    {
+      "completion_length": 209.73408164978028,
+      "epoch": 0.4108000662580752,
+      "grad_norm": 1.1104848384857178,
+      "kl": 0.47215576171875,
+      "learning_rate": 1.4675612207136283e-05,
+      "loss": 0.0189,
+      "reward": 1.2372024029493331,
+      "reward_std": 0.39521638955920935,
+      "rewards/accuracy_reward": 0.33601190941408277,
+      "rewards/format_reward": 0.9011904887855053,
+      "step": 155
+    },
+    {
+      "completion_length": 192.39732513427734,
+      "epoch": 0.4240516812986583,
+      "grad_norm": 0.795322835445404,
+      "kl": 0.5870361328125,
+      "learning_rate": 1.4261146721000554e-05,
+      "loss": 0.0235,
+      "reward": 1.278125023841858,
+      "reward_std": 0.3505269827321172,
+      "rewards/accuracy_reward": 0.3505952453240752,
+      "rewards/format_reward": 0.9275297746062279,
+      "step": 160
+    },
+    {
+      "completion_length": 204.86577787399293,
+      "epoch": 0.43730329633924137,
+      "grad_norm": 1.8953642845153809,
+      "kl": 1.4674072265625,
+      "learning_rate": 1.3837534014527292e-05,
+      "loss": 0.0587,
+      "reward": 1.2894345439970494,
+      "reward_std": 0.36970352567732334,
+      "rewards/accuracy_reward": 0.37619048319756987,
+      "rewards/format_reward": 0.9132440634071827,
+      "step": 165
+    },
+    {
+      "completion_length": 195.65952739715576,
+      "epoch": 0.4505549113798244,
+      "grad_norm": 5.028355598449707,
+      "kl": 1.5417724609375,
+      "learning_rate": 1.3405683438888281e-05,
+      "loss": 0.0617,
+      "reward": 1.2653274051845074,
+      "reward_std": 0.35815710462629796,
+      "rewards/accuracy_reward": 0.34985119504854084,
+      "rewards/format_reward": 0.9154762044548989,
+      "step": 170
+    },
+    {
+      "completion_length": 208.46548023223878,
+      "epoch": 0.4638065264204075,
+      "grad_norm": 1.6401289701461792,
+      "kl": 1.90859375,
+      "learning_rate": 1.2966522029134622e-05,
+      "loss": 0.0763,
+      "reward": 1.2275297917425632,
+      "reward_std": 0.36908484864979985,
+      "rewards/accuracy_reward": 0.325446433480829,
+      "rewards/format_reward": 0.9020833462476731,
+      "step": 175
+    },
+    {
+      "completion_length": 178.60446815490724,
+      "epoch": 0.47705814146099057,
+      "grad_norm": 37.09467315673828,
+      "kl": 1.423974609375,
+      "learning_rate": 1.252099251417048e-05,
+      "loss": 0.0569,
+      "reward": 1.2782738268375398,
+      "reward_std": 0.33831334542483094,
+      "rewards/accuracy_reward": 0.3473214345984161,
+      "rewards/format_reward": 0.9309523932635784,
+      "step": 180
+    },
+    {
+      "completion_length": 184.88735513687135,
+      "epoch": 0.49030975650157366,
+      "grad_norm": 1.528001070022583,
+      "kl": 0.662109375,
+      "learning_rate": 1.2070051293037493e-05,
+      "loss": 0.0265,
+      "reward": 1.2159226439893245,
+      "reward_std": 0.3878694591112435,
+      "rewards/accuracy_reward": 0.3151785763911903,
+      "rewards/format_reward": 0.9007440641522407,
+      "step": 185
+    },
+    {
+      "completion_length": 190.33065776824952,
+      "epoch": 0.5035613715421567,
+      "grad_norm": 6.517220973968506,
+      "kl": 1.2845458984375,
+      "learning_rate": 1.1614666381854107e-05,
+      "loss": 0.0514,
+      "reward": 1.0744047850370406,
+      "reward_std": 0.5224868061020971,
+      "rewards/accuracy_reward": 0.28303572041913866,
+      "rewards/format_reward": 0.7913690611720086,
+      "step": 190
+    },
+    {
+      "completion_length": 132.4203896522522,
+      "epoch": 0.5168129865827398,
+      "grad_norm": 0.33952298760414124,
+      "kl": 0.3914306640625,
+      "learning_rate": 1.115581533581701e-05,
+      "loss": 0.0157,
+      "reward": 1.3244047820568086,
+      "reward_std": 0.2432941493578255,
+      "rewards/accuracy_reward": 0.34761905362829565,
+      "rewards/format_reward": 0.976785735040903,
+      "step": 195
+    },
+    {
+      "completion_length": 127.37782945632935,
+      "epoch": 0.5300646016233228,
+      "grad_norm": 1.2143195867538452,
+      "kl": 0.28424072265625,
+      "learning_rate": 1.0694483150725458e-05,
+      "loss": 0.0114,
+      "reward": 1.3541666924953462,
+      "reward_std": 0.23322852002456784,
+      "rewards/accuracy_reward": 0.36949405539780855,
+      "rewards/format_reward": 0.9846726387739182,
+      "step": 200
+    },
+    {
+      "epoch": 0.5300646016233228,
+      "eval_completion_length": 145.582976138953,
+      "eval_kl": 0.26819957386363635,
+      "eval_loss": 0.0107291080057621,
+      "eval_reward": 1.3203463391824202,
+      "eval_reward_std": 0.2810380467862794,
+      "eval_rewards/accuracy_reward": 0.3492063595490022,
+      "eval_rewards/format_reward": 0.971139972860163,
+      "eval_runtime": 75.3601,
+      "eval_samples_per_second": 1.314,
+      "eval_steps_per_second": 0.438,
+      "step": 200
+    },
+    {
+      "completion_length": 161.00729446411134,
+      "epoch": 0.543316216663906,
+      "grad_norm": 4.276285648345947,
+      "kl": 0.426806640625,
+      "learning_rate": 1.0231660148533183e-05,
+      "loss": 0.0171,
+      "reward": 1.266369068622589,
+      "reward_std": 0.31076475251466035,
+      "rewards/accuracy_reward": 0.31622024429962037,
+      "rewards/format_reward": 0.9501488223671913,
+      "step": 205
+    },
+    {
+      "completion_length": 216.8572956085205,
+      "epoch": 0.556567831704489,
+      "grad_norm": 2.877614974975586,
+      "kl": 2.30439453125,
+      "learning_rate": 9.768339851466818e-06,
+      "loss": 0.0922,
+      "reward": 1.1943452663719654,
+      "reward_std": 0.4090527132153511,
+      "rewards/accuracy_reward": 0.29375000577419996,
+      "rewards/format_reward": 0.9005952522158622,
+      "step": 210
+    },
+    {
+      "completion_length": 201.9918186187744,
+      "epoch": 0.569819446745072,
+      "grad_norm": 0.5717713236808777,
+      "kl": 1.24404296875,
+      "learning_rate": 9.305516849274542e-06,
+      "loss": 0.0498,
+      "reward": 1.2285714492201805,
+      "reward_std": 0.34389270916581155,
+      "rewards/accuracy_reward": 0.2970238159410655,
+      "rewards/format_reward": 0.9315476305782795,
+      "step": 215
+    },
+    {
+      "completion_length": 180.8418180465698,
+      "epoch": 0.5830710617856552,
+      "grad_norm": 0.2742576599121094,
+      "kl": 0.54093017578125,
+      "learning_rate": 8.844184664182993e-06,
+      "loss": 0.0216,
+      "reward": 1.2680059686303138,
+      "reward_std": 0.2862160621210933,
+      "rewards/accuracy_reward": 0.30416667107492684,
+      "rewards/format_reward": 0.963839304447174,
+      "step": 220
+    },
+    {
+      "completion_length": 188.11295013427736,
+      "epoch": 0.5963226768262382,
+      "grad_norm": 9.134293556213379,
+      "kl": 0.57703857421875,
+      "learning_rate": 8.385333618145896e-06,
+      "loss": 0.0231,
+      "reward": 1.2605654999613762,
+      "reward_std": 0.2837338815443218,
+      "rewards/accuracy_reward": 0.30133929173462093,
+      "rewards/format_reward": 0.9592262096703053,
+      "step": 225
+    },
+    {
+      "completion_length": 207.8897367477417,
+      "epoch": 0.6095742918668212,
+      "grad_norm": 8.929195404052734,
+      "kl": 0.525146484375,
+      "learning_rate": 7.929948706962508e-06,
+      "loss": 0.021,
+      "reward": 1.2611607387661934,
+      "reward_std": 0.3355352971702814,
+      "rewards/accuracy_reward": 0.3227678622119129,
+      "rewards/format_reward": 0.9383928693830967,
+      "step": 230
+    },
+    {
+      "completion_length": 222.4827423095703,
+      "epoch": 0.6228259069074044,
+      "grad_norm": 7.437636375427246,
+      "kl": 0.644091796875,
+      "learning_rate": 7.479007485829523e-06,
+      "loss": 0.0258,
+      "reward": 1.218452400714159,
+      "reward_std": 0.35496083125472067,
+      "rewards/accuracy_reward": 0.3040178632363677,
+      "rewards/format_reward": 0.9144345402717591,
+      "step": 235
+    },
+    {
+      "completion_length": 234.08661136627197,
+      "epoch": 0.6360775219479874,
+      "grad_norm": 1242.3927001953125,
+      "kl": 1.87177734375,
+      "learning_rate": 7.033477970865381e-06,
+      "loss": 0.0748,
+      "reward": 1.2133928798139095,
+      "reward_std": 0.41558921709656715,
+      "rewards/accuracy_reward": 0.3251488147303462,
+      "rewards/format_reward": 0.8882440619170666,
+      "step": 240
+    },
+    {
+      "completion_length": 196.01518211364746,
+      "epoch": 0.6493291369885705,
+      "grad_norm": 3.4196267127990723,
+      "kl": 0.9855224609375,
+      "learning_rate": 6.5943165611117244e-06,
+      "loss": 0.0394,
+      "reward": 1.2288690701127052,
+      "reward_std": 0.3623060153797269,
+      "rewards/accuracy_reward": 0.320238099899143,
+      "rewards/format_reward": 0.9086309649050236,
+      "step": 245
+    },
+    {
+      "completion_length": 154.4151819229126,
+      "epoch": 0.6625807520291536,
+      "grad_norm": 1.869827151298523,
+      "kl": 0.8129638671875,
+      "learning_rate": 6.16246598547271e-06,
+      "loss": 0.0325,
+      "reward": 1.2891369298100472,
+      "reward_std": 0.2997500140219927,
+      "rewards/accuracy_reward": 0.3443452435545623,
+      "rewards/format_reward": 0.9447916835546494,
+      "step": 250
+    },
+    {
+      "completion_length": 166.84628257751464,
+      "epoch": 0.6758323670697366,
+      "grad_norm": 3.790719509124756,
+      "kl": 1.06826171875,
+      "learning_rate": 5.7388532789994476e-06,
+      "loss": 0.0427,
+      "reward": 1.2696428745985031,
+      "reward_std": 0.3165645905770361,
+      "rewards/accuracy_reward": 0.32901786286383866,
+      "rewards/format_reward": 0.9406250134110451,
+      "step": 255
+    },
+    {
+      "completion_length": 175.49167013168335,
+      "epoch": 0.6890839821103197,
+      "grad_norm": 1.6303218603134155,
+      "kl": 0.989306640625,
+      "learning_rate": 5.324387792863719e-06,
+      "loss": 0.0396,
+      "reward": 1.2730654999613762,
+      "reward_std": 0.3291014650836587,
+      "rewards/accuracy_reward": 0.34583333805203437,
+      "rewards/format_reward": 0.9272321559488773,
+      "step": 260
+    },
+    {
+      "completion_length": 184.52321767807007,
+      "epoch": 0.7023355971509028,
+      "grad_norm": 0.9721196889877319,
+      "kl": 1.154931640625,
+      "learning_rate": 4.9199592422929545e-06,
+      "loss": 0.0462,
+      "reward": 1.2248512104153633,
+      "reward_std": 0.37469834852963685,
+      "rewards/accuracy_reward": 0.313839291408658,
+      "rewards/format_reward": 0.9110119171440602,
+      "step": 265
+    },
+    {
+      "completion_length": 168.1200927734375,
+      "epoch": 0.7155872121914858,
+      "grad_norm": 0.9256172776222229,
+      "kl": 0.98193359375,
+      "learning_rate": 4.52643579665683e-06,
+      "loss": 0.0393,
+      "reward": 1.2313988298177718,
+      "reward_std": 0.363844196125865,
+      "rewards/accuracy_reward": 0.31160714766010644,
+      "rewards/format_reward": 0.919791679829359,
+      "step": 270
+    },
+    {
+      "completion_length": 152.76533002853392,
+      "epoch": 0.7288388272320689,
+      "grad_norm": 73.29869079589844,
+      "kl": 1.819775390625,
+      "learning_rate": 4.144662215805426e-06,
+      "loss": 0.0729,
+      "reward": 1.2657738298177719,
+      "reward_std": 0.32892794273793696,
+      "rewards/accuracy_reward": 0.32366072190925477,
+      "rewards/format_reward": 0.9421131126582623,
+      "step": 275
+    },
+    {
+      "completion_length": 150.89062795639038,
+      "epoch": 0.742090442272652,
+      "grad_norm": 4.905460834503174,
+      "kl": 0.6436279296875,
+      "learning_rate": 3.7754580366596116e-06,
+      "loss": 0.0257,
+      "reward": 1.2663690716028213,
+      "reward_std": 0.30096426121890546,
+      "rewards/accuracy_reward": 0.3188988157082349,
+      "rewards/format_reward": 0.9474702529609204,
+      "step": 280
+    },
+    {
+      "completion_length": 157.10922889709474,
+      "epoch": 0.755342057313235,
+      "grad_norm": 13.410654067993164,
+      "kl": 0.77381591796875,
+      "learning_rate": 3.419615813946392e-06,
+      "loss": 0.031,
+      "reward": 1.2614583536982535,
+      "reward_std": 0.3189463253132999,
+      "rewards/accuracy_reward": 0.3148809571750462,
+      "rewards/format_reward": 0.9465773954987526,
+      "step": 285
+    },
+    {
+      "completion_length": 169.77351474761963,
+      "epoch": 0.7685936723538181,
+      "grad_norm": 6.9417853355407715,
+      "kl": 1.8572265625,
+      "learning_rate": 3.0778994188557722e-06,
+      "loss": 0.0743,
+      "reward": 1.2480655021965503,
+      "reward_std": 0.3294074974954128,
+      "rewards/accuracy_reward": 0.3102678618626669,
+      "rewards/format_reward": 0.9377976343035698,
+      "step": 290
+    },
+    {
+      "completion_length": 164.56265211105347,
+      "epoch": 0.7818452873944012,
+      "grad_norm": 6.690864562988281,
+      "kl": 0.8572021484375,
+      "learning_rate": 2.7510423992713374e-06,
+      "loss": 0.0343,
+      "reward": 1.2416666865348815,
+      "reward_std": 0.3249609712511301,
+      "rewards/accuracy_reward": 0.3025297665968537,
+      "rewards/format_reward": 0.9391369208693504,
+      "step": 295
+    },
+    {
+      "completion_length": 172.95506219863893,
+      "epoch": 0.7950969024349842,
+      "grad_norm": 5.398939609527588,
+      "kl": 0.9433837890625,
+      "learning_rate": 2.4397464050945753e-06,
+      "loss": 0.0377,
+      "reward": 1.2473214477300645,
+      "reward_std": 0.35795026188716295,
+      "rewards/accuracy_reward": 0.31547619523480536,
+      "rewards/format_reward": 0.9318452522158622,
+      "step": 300
+    },
+    {
+      "epoch": 0.7950969024349842,
+      "eval_completion_length": 177.73737774473247,
+      "eval_kl": 1.0707267992424243,
+      "eval_loss": 0.042819034308195114,
+      "eval_reward": 1.2929293115933735,
+      "eval_reward_std": 0.36596207681930426,
+      "eval_rewards/accuracy_reward": 0.36363637221582007,
+      "eval_rewards/format_reward": 0.9292929335073992,
+      "eval_runtime": 112.5811,
+      "eval_samples_per_second": 0.879,
+      "eval_steps_per_second": 0.293,
+      "step": 300
+    },
+    {
+      "completion_length": 171.8497058868408,
+      "epoch": 0.8083485174755674,
+      "grad_norm": 7.889118671417236,
+      "kl": 0.9488037109375,
+      "learning_rate": 2.144679682043217e-06,
+      "loss": 0.0379,
+      "reward": 1.2526785925030708,
+      "reward_std": 0.33547496125102044,
+      "rewards/accuracy_reward": 0.3202380993403494,
+      "rewards/format_reward": 0.9324404880404472,
+      "step": 305
+    },
+    {
+      "completion_length": 167.9309557914734,
+      "epoch": 0.8216001325161504,
+      "grad_norm": 1.831909418106079,
+      "kl": 0.905615234375,
+      "learning_rate": 1.8664756371568981e-06,
+      "loss": 0.0362,
+      "reward": 1.2857143044471742,
+      "reward_std": 0.3516055729240179,
+      "rewards/accuracy_reward": 0.35148810180835427,
+      "rewards/format_reward": 0.9342262051999569,
+      "step": 310
+    },
+    {
+      "completion_length": 165.18854503631593,
+      "epoch": 0.8348517475567335,
+      "grad_norm": 1.0715512037277222,
+      "kl": 0.8928466796875,
+      "learning_rate": 1.605731479089534e-06,
+      "loss": 0.0357,
+      "reward": 1.2788690701127052,
+      "reward_std": 0.346990068256855,
+      "rewards/accuracy_reward": 0.34241071944124996,
+      "rewards/format_reward": 0.9364583484828473,
+      "step": 315
+    },
+    {
+      "completion_length": 161.76652059555053,
+      "epoch": 0.8481033625973166,
+      "grad_norm": 1.1607029438018799,
+      "kl": 0.86220703125,
+      "learning_rate": 1.363006936107183e-06,
+      "loss": 0.0345,
+      "reward": 1.2660714492201806,
+      "reward_std": 0.33531030286103486,
+      "rewards/accuracy_reward": 0.3278273862786591,
+      "rewards/format_reward": 0.9382440611720085,
+      "step": 320
+    },
+    {
+      "completion_length": 166.52857475280763,
+      "epoch": 0.8613549776378996,
+      "grad_norm": 1.234681487083435,
+      "kl": 0.9093994140625,
+      "learning_rate": 1.1388230545434652e-06,
+      "loss": 0.0364,
+      "reward": 1.273363122344017,
+      "reward_std": 0.3373934781178832,
+      "rewards/accuracy_reward": 0.3424107195809484,
+      "rewards/format_reward": 0.9309523954987526,
+      "step": 325
+    },
+    {
+      "completion_length": 181.2095274925232,
+      "epoch": 0.8746065926784827,
+      "grad_norm": 1.4917526245117188,
+      "kl": 0.9361572265625,
+      "learning_rate": 9.336610802918044e-07,
+      "loss": 0.0374,
+      "reward": 1.2404762201011181,
+      "reward_std": 0.3512815716676414,
+      "rewards/accuracy_reward": 0.31741072023287414,
+      "rewards/format_reward": 0.9230654880404472,
+      "step": 330
+    },
+    {
+      "completion_length": 173.2355694770813,
+      "epoch": 0.8878582077190658,
+      "grad_norm": 1.493273138999939,
+      "kl": 0.842724609375,
+      "learning_rate": 7.479614257355972e-07,
+      "loss": 0.0337,
+      "reward": 1.2668154940009118,
+      "reward_std": 0.33156857155263425,
+      "rewards/accuracy_reward": 0.33214286509901286,
+      "rewards/format_reward": 0.9346726343035698,
+      "step": 335
+    },
+    {
+      "completion_length": 169.36815776824952,
+      "epoch": 0.9011098227596488,
+      "grad_norm": 0.7635859251022339,
+      "kl": 0.775341796875,
+      "learning_rate": 5.821227243338712e-07,
+      "loss": 0.031,
+      "reward": 1.2864583551883697,
+      "reward_std": 0.32649895939975976,
+      "rewards/accuracy_reward": 0.3480654824525118,
+      "rewards/format_reward": 0.9383928693830967,
+      "step": 340
+    },
+    {
+      "completion_length": 183.0156273841858,
+      "epoch": 0.9143614378002319,
+      "grad_norm": 0.6962347030639648,
+      "kl": 0.8108154296875,
+      "learning_rate": 4.3650097489200125e-07,
+      "loss": 0.0325,
+      "reward": 1.2919643111526966,
+      "reward_std": 0.3276719061657786,
+      "rewards/accuracy_reward": 0.35744048189371824,
+      "rewards/format_reward": 0.9345238246023655,
+      "step": 345
+    },
+    {
+      "completion_length": 178.9885450363159,
+      "epoch": 0.927613052840815,
+      "grad_norm": 0.9081234335899353,
+      "kl": 0.8029541015625,
+      "learning_rate": 3.114087773543939e-07,
+      "loss": 0.0321,
+      "reward": 1.3026785895228385,
+      "reward_std": 0.3466304335743189,
+      "rewards/accuracy_reward": 0.36904762461781504,
+      "rewards/format_reward": 0.9336309641599655,
+      "step": 350
+    },
+    {
+      "completion_length": 174.72396116256715,
+      "epoch": 0.940864667881398,
+      "grad_norm": 0.6305360198020935,
+      "kl": 0.83115234375,
+      "learning_rate": 2.071146617596276e-07,
+      "loss": 0.0332,
+      "reward": 1.2918155014514923,
+      "reward_std": 0.3342212336137891,
+      "rewards/accuracy_reward": 0.3565476250834763,
+      "rewards/format_reward": 0.935267872363329,
+      "step": 355
+    },
+    {
+      "completion_length": 173.80669994354247,
+      "epoch": 0.9541162829219811,
+      "grad_norm": 0.6665492653846741,
+      "kl": 0.8109375,
+      "learning_rate": 1.2384251179857642e-07,
+      "loss": 0.0324,
+      "reward": 1.2958333596587182,
+      "reward_std": 0.33413169481791555,
+      "rewards/accuracy_reward": 0.3613095307722688,
+      "rewards/format_reward": 0.9345238253474235,
+      "step": 360
+    },
+    {
+      "completion_length": 174.30803899765016,
+      "epoch": 0.9673678979625642,
+      "grad_norm": 0.6672953963279724,
+      "kl": 0.798046875,
+      "learning_rate": 6.177108421292266e-08,
+      "loss": 0.0319,
+      "reward": 1.2974702656269073,
+      "reward_std": 0.32824334893375634,
+      "rewards/accuracy_reward": 0.3608631026931107,
+      "rewards/format_reward": 0.9366071559488773,
+      "step": 365
+    },
+    {
+      "completion_length": 174.2739619255066,
+      "epoch": 0.9806195130031473,
+      "grad_norm": 0.7460318207740784,
+      "kl": 0.861328125,
+      "learning_rate": 2.1033625065747244e-08,
+      "loss": 0.0345,
+      "reward": 1.2961309775710106,
+      "reward_std": 0.32596146771684287,
+      "rewards/accuracy_reward": 0.3602678650058806,
+      "rewards/format_reward": 0.9358631066977978,
+      "step": 370
+    },
+    {
+      "completion_length": 172.85729579925538,
+      "epoch": 0.9938711280437303,
+      "grad_norm": 1.4766656160354614,
+      "kl": 0.814013671875,
+      "learning_rate": 1.7175837079452806e-09,
+      "loss": 0.0326,
+      "reward": 1.3000000178813935,
+      "reward_std": 0.34734019450843334,
+      "rewards/accuracy_reward": 0.36309524523094294,
+      "rewards/format_reward": 0.9369047738611698,
+      "step": 375
+    },
+    {
+      "completion_length": 158.41108989715576,
+      "epoch": 0.9991717740599636,
+      "kl": 0.7142333984375,
+      "reward": 1.345982164144516,
+      "reward_std": 0.3284614863805473,
+      "rewards/accuracy_reward": 0.3962053656578064,
+      "rewards/format_reward": 0.9497767984867096,
+      "step": 377,
+      "total_flos": 0.0,
+      "train_loss": 2.2585060952830793,
+      "train_runtime": 36118.4141,
+      "train_samples_per_second": 2.006,
+      "train_steps_per_second": 0.01
+    }
+  ],
+  "logging_steps": 5,
+  "max_steps": 377,
+  "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": 4,
+  "trial_name": null,
+  "trial_params": null
+}

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ddb8e3035ef96ccd8352d6828277cc441306cf43b8b0617a058217580026cf65
+size 7352