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-Math-1.5B
+library_name: transformers
+model_name: Qwen-2.5-7B-Simple-RL
+tags:
+- generated_from_trainer
+- trl
+- grpo
+licence: license
+---
+
+# Model Card for Qwen-2.5-7B-Simple-RL
+
+This model is a fine-tuned version of [Qwen/Qwen2.5-Math-1.5B](https://huggingface.co/Qwen/Qwen2.5-Math-1.5B).
+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="Bradley/Qwen-2.5-7B-Simple-RL", 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/bradley/huggingface/runs/ju8pbdvp) 
+
+
+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": 0.0,
+    "train_runtime": 1.0266,
+    "train_samples": 7500,
+    "train_samples_per_second": 7305.492,
+    "train_steps_per_second": 56.496
+}

config.json

@@ -0,0 +1,30 @@
+{
+  "_name_or_path": "Qwen/Qwen2.5-Math-1.5B",
+  "architectures": [
+    "Qwen2ForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 151643,
+  "eos_token_id": 151643,
+  "hidden_act": "silu",
+  "hidden_size": 1536,
+  "initializer_range": 0.02,
+  "intermediate_size": 8960,
+  "max_position_embeddings": 4096,
+  "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": 10000,
+  "sliding_window": 4096,
+  "tie_word_embeddings": true,
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.49.0.dev0",
+  "use_cache": false,
+  "use_mrope": false,
+  "use_sliding_window": false,
+  "vocab_size": 151936
+}

generation_config.json

@@ -0,0 +1,6 @@
+{
+  "bos_token_id": 151643,
+  "eos_token_id": 151643,
+  "max_new_tokens": 2048,
+  "transformers_version": "4.49.0.dev0"
+}

model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ed5a684ef0ef5612c79d240e03a157adece0803f8ad596d73aa6d434a4a70990
+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": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": {
+    "content": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
+}

tokenizer.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9c5ae00e602b8860cbd784ba82a8aa14e8feecec692e7076590d014d7b7fdafa
+size 11421896

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        {{- 'Please reason step by step, and put your final answer within \\\\boxed{}.' }}\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\\nPlease reason step by step, and put your final answer within \\\\boxed{}.<|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": "<|endoftext|>",
+  "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": 0.0,
+    "train_runtime": 1.0266,
+    "train_samples": 7500,
+    "train_samples_per_second": 7305.492,
+    "train_steps_per_second": 56.496
+}

trainer_state.json

@@ -0,0 +1,796 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9893390191897654,
+  "eval_steps": 100,
+  "global_step": 58,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "completion_length": 612.3248062133789,
+      "epoch": 0.017057569296375266,
+      "grad_norm": 0.8495666980743408,
+      "kl": 0.0,
+      "learning_rate": 5e-07,
+      "loss": 0.0,
+      "reward": 0.3694196529686451,
+      "reward_std": 0.3680446147918701,
+      "rewards/accuracy_reward": 0.368303582072258,
+      "rewards/format_reward": 0.0011160714784637094,
+      "step": 1
+    },
+    {
+      "completion_length": 599.4174499511719,
+      "epoch": 0.03411513859275053,
+      "grad_norm": 0.458848774433136,
+      "kl": 0.0,
+      "learning_rate": 1e-06,
+      "loss": 0.0,
+      "reward": 0.349330373108387,
+      "reward_std": 0.3545062802731991,
+      "rewards/accuracy_reward": 0.3470982313156128,
+      "rewards/format_reward": 0.0022321429569274187,
+      "step": 2
+    },
+    {
+      "completion_length": 623.0669860839844,
+      "epoch": 0.0511727078891258,
+      "grad_norm": 0.4115377366542816,
+      "kl": 5.65648078918457e-05,
+      "learning_rate": 1.5e-06,
+      "loss": 0.0,
+      "reward": 0.3549107313156128,
+      "reward_std": 0.38977522775530815,
+      "rewards/accuracy_reward": 0.3515625186264515,
+      "rewards/format_reward": 0.003348214435391128,
+      "step": 3
+    },
+    {
+      "completion_length": 611.1585159301758,
+      "epoch": 0.06823027718550106,
+      "grad_norm": 0.47733554244041443,
+      "kl": 6.213784217834473e-05,
+      "learning_rate": 2e-06,
+      "loss": 0.0,
+      "reward": 0.3303571566939354,
+      "reward_std": 0.3627745509147644,
+      "rewards/accuracy_reward": 0.3292410857975483,
+      "rewards/format_reward": 0.0011160714784637094,
+      "step": 4
+    },
+    {
+      "completion_length": 580.760066986084,
+      "epoch": 0.08528784648187633,
+      "grad_norm": 0.4139915406703949,
+      "kl": 7.668137550354004e-05,
+      "learning_rate": 2.5e-06,
+      "loss": 0.0,
+      "reward": 0.36941965855658054,
+      "reward_std": 0.37279706075787544,
+      "rewards/accuracy_reward": 0.36607144214212894,
+      "rewards/format_reward": 0.003348214435391128,
+      "step": 5
+    },
+    {
+      "completion_length": 589.4765853881836,
+      "epoch": 0.1023454157782516,
+      "grad_norm": 0.9889692664146423,
+      "kl": 0.00010079145431518555,
+      "learning_rate": 3e-06,
+      "loss": 0.0,
+      "reward": 0.333705373108387,
+      "reward_std": 0.35657742619514465,
+      "rewards/accuracy_reward": 0.333705373108387,
+      "rewards/format_reward": 0.0,
+      "step": 6
+    },
+    {
+      "completion_length": 586.4297180175781,
+      "epoch": 0.11940298507462686,
+      "grad_norm": 1.1514272689819336,
+      "kl": 0.00018024444580078125,
+      "learning_rate": 2.9972633313349763e-06,
+      "loss": 0.0,
+      "reward": 0.3381696604192257,
+      "reward_std": 0.3602691851556301,
+      "rewards/accuracy_reward": 0.3370535857975483,
+      "rewards/format_reward": 0.0011160714784637094,
+      "step": 7
+    },
+    {
+      "completion_length": 607.0881881713867,
+      "epoch": 0.13646055437100213,
+      "grad_norm": 0.3025471866130829,
+      "kl": 0.0006165504455566406,
+      "learning_rate": 2.989063311147081e-06,
+      "loss": 0.0,
+      "reward": 0.3727678768336773,
+      "reward_std": 0.365192923694849,
+      "rewards/accuracy_reward": 0.3716518059372902,
+      "rewards/format_reward": 0.0011160714784637094,
+      "step": 8
+    },
+    {
+      "completion_length": 582.1261444091797,
+      "epoch": 0.1535181236673774,
+      "grad_norm": 0.3065524995326996,
+      "kl": 0.0008487701416015625,
+      "learning_rate": 2.9754298604207156e-06,
+      "loss": 0.0,
+      "reward": 0.4140625186264515,
+      "reward_std": 0.4111504293978214,
+      "rewards/accuracy_reward": 0.4129464477300644,
+      "rewards/format_reward": 0.0011160714784637094,
+      "step": 9
+    },
+    {
+      "completion_length": 609.3214645385742,
+      "epoch": 0.17057569296375266,
+      "grad_norm": 0.22064943611621857,
+      "kl": 0.00093841552734375,
+      "learning_rate": 2.956412726139078e-06,
+      "loss": 0.0,
+      "reward": 0.408482164144516,
+      "reward_std": 0.36822360940277576,
+      "rewards/accuracy_reward": 0.4073660932481289,
+      "rewards/format_reward": 0.0011160714784637094,
+      "step": 10
+    },
+    {
+      "completion_length": 639.6484680175781,
+      "epoch": 0.18763326226012794,
+      "grad_norm": 0.9334643483161926,
+      "kl": 0.003772258758544922,
+      "learning_rate": 2.9320812997628183e-06,
+      "loss": 0.0002,
+      "reward": 0.3917410857975483,
+      "reward_std": 0.3854985646903515,
+      "rewards/accuracy_reward": 0.3895089440047741,
+      "rewards/format_reward": 0.0022321429569274187,
+      "step": 11
+    },
+    {
+      "completion_length": 571.6004638671875,
+      "epoch": 0.2046908315565032,
+      "grad_norm": 0.308927059173584,
+      "kl": 0.0036773681640625,
+      "learning_rate": 2.9025243640281224e-06,
+      "loss": 0.0001,
+      "reward": 0.4854910932481289,
+      "reward_std": 0.3801889941096306,
+      "rewards/accuracy_reward": 0.4843750186264515,
+      "rewards/format_reward": 0.0011160714784637094,
+      "step": 12
+    },
+    {
+      "completion_length": 596.0346221923828,
+      "epoch": 0.22174840085287847,
+      "grad_norm": 0.3541490435600281,
+      "kl": 0.0049343109130859375,
+      "learning_rate": 2.8678497689881355e-06,
+      "loss": 0.0002,
+      "reward": 0.4654018096625805,
+      "reward_std": 0.3858399875462055,
+      "rewards/accuracy_reward": 0.4620535895228386,
+      "rewards/format_reward": 0.003348214435391128,
+      "step": 13
+    },
+    {
+      "completion_length": 630.7388610839844,
+      "epoch": 0.23880597014925373,
+      "grad_norm": 0.17941774427890778,
+      "kl": 0.004192352294921875,
+      "learning_rate": 2.8281840384798147e-06,
+      "loss": 0.0002,
+      "reward": 0.4642857350409031,
+      "reward_std": 0.3576914146542549,
+      "rewards/accuracy_reward": 0.4631696604192257,
+      "rewards/format_reward": 0.0011160714784637094,
+      "step": 14
+    },
+    {
+      "completion_length": 619.506721496582,
+      "epoch": 0.255863539445629,
+      "grad_norm": 0.2217731475830078,
+      "kl": 0.0070285797119140625,
+      "learning_rate": 2.7836719084521715e-06,
+      "loss": 0.0003,
+      "reward": 0.4229910895228386,
+      "reward_std": 0.3531936705112457,
+      "rewards/accuracy_reward": 0.4229910895228386,
+      "rewards/format_reward": 0.0,
+      "step": 15
+    },
+    {
+      "completion_length": 591.1908798217773,
+      "epoch": 0.27292110874200426,
+      "grad_norm": 0.23496699333190918,
+      "kl": 0.008819580078125,
+      "learning_rate": 2.7344757988404844e-06,
+      "loss": 0.0004,
+      "reward": 0.5089285857975483,
+      "reward_std": 0.3891791105270386,
+      "rewards/accuracy_reward": 0.5078125149011612,
+      "rewards/format_reward": 0.0011160714784637094,
+      "step": 16
+    },
+    {
+      "completion_length": 621.5346145629883,
+      "epoch": 0.2899786780383795,
+      "grad_norm": 1.1953942775726318,
+      "kl": 0.0303192138671875,
+      "learning_rate": 2.680775220913575e-06,
+      "loss": 0.0012,
+      "reward": 0.4888393096625805,
+      "reward_std": 0.3780059143900871,
+      "rewards/accuracy_reward": 0.4854910932481289,
+      "rewards/format_reward": 0.003348214435391128,
+      "step": 17
+    },
+    {
+      "completion_length": 649.108283996582,
+      "epoch": 0.3070362473347548,
+      "grad_norm": 0.8823431730270386,
+      "kl": 0.03516387939453125,
+      "learning_rate": 2.6227661222566517e-06,
+      "loss": 0.0014,
+      "reward": 0.4944196678698063,
+      "reward_std": 0.3778855614364147,
+      "rewards/accuracy_reward": 0.4944196678698063,
+      "rewards/format_reward": 0.0,
+      "step": 18
+    },
+    {
+      "completion_length": 627.3359603881836,
+      "epoch": 0.32409381663113007,
+      "grad_norm": 0.2409147024154663,
+      "kl": 0.0187530517578125,
+      "learning_rate": 2.5606601717798212e-06,
+      "loss": 0.0008,
+      "reward": 0.5446428842842579,
+      "reward_std": 0.3593517243862152,
+      "rewards/accuracy_reward": 0.5435268133878708,
+      "rewards/format_reward": 0.0011160714784637094,
+      "step": 19
+    },
+    {
+      "completion_length": 587.4598388671875,
+      "epoch": 0.3411513859275053,
+      "grad_norm": 0.36490699648857117,
+      "kl": 0.020496368408203125,
+      "learning_rate": 2.4946839873611927e-06,
+      "loss": 0.0008,
+      "reward": 0.5669643171131611,
+      "reward_std": 0.33723995834589005,
+      "rewards/accuracy_reward": 0.5658482499420643,
+      "rewards/format_reward": 0.0011160714784637094,
+      "step": 20
+    },
+    {
+      "completion_length": 639.7868499755859,
+      "epoch": 0.3582089552238806,
+      "grad_norm": 2.728965997695923,
+      "kl": 0.05742645263671875,
+      "learning_rate": 2.425078308942815e-06,
+      "loss": 0.0023,
+      "reward": 0.5747768059372902,
+      "reward_std": 0.31150117330253124,
+      "rewards/accuracy_reward": 0.5747768059372902,
+      "rewards/format_reward": 0.0,
+      "step": 21
+    },
+    {
+      "completion_length": 619.8281478881836,
+      "epoch": 0.3752665245202559,
+      "grad_norm": 0.12718994915485382,
+      "kl": 0.01346588134765625,
+      "learning_rate": 2.3520971200967337e-06,
+      "loss": 0.0005,
+      "reward": 0.5691964514553547,
+      "reward_std": 0.32910910062491894,
+      "rewards/accuracy_reward": 0.5691964514553547,
+      "rewards/format_reward": 0.0,
+      "step": 22
+    },
+    {
+      "completion_length": 639.185302734375,
+      "epoch": 0.39232409381663114,
+      "grad_norm": 0.15699955821037292,
+      "kl": 0.009624481201171875,
+      "learning_rate": 2.276006721266485e-06,
+      "loss": 0.0004,
+      "reward": 0.5256696678698063,
+      "reward_std": 0.3530396558344364,
+      "rewards/accuracy_reward": 0.5256696678698063,
+      "rewards/format_reward": 0.0,
+      "step": 23
+    },
+    {
+      "completion_length": 630.2265853881836,
+      "epoch": 0.4093816631130064,
+      "grad_norm": 0.09854375571012497,
+      "kl": 0.009571075439453125,
+      "learning_rate": 2.1970847580656528e-06,
+      "loss": 0.0004,
+      "reward": 0.5747768133878708,
+      "reward_std": 0.315556762740016,
+      "rewards/accuracy_reward": 0.5747768133878708,
+      "rewards/format_reward": 0.0,
+      "step": 24
+    },
+    {
+      "completion_length": 636.4408798217773,
+      "epoch": 0.42643923240938164,
+      "grad_norm": 0.12171110510826111,
+      "kl": 0.0113525390625,
+      "learning_rate": 2.1156192081791355e-06,
+      "loss": 0.0005,
+      "reward": 0.533482164144516,
+      "reward_std": 0.3439404182136059,
+      "rewards/accuracy_reward": 0.533482164144516,
+      "rewards/format_reward": 0.0,
+      "step": 25
+    },
+    {
+      "completion_length": 613.5491256713867,
+      "epoch": 0.44349680170575695,
+      "grad_norm": 0.10672565549612045,
+      "kl": 0.007846832275390625,
+      "learning_rate": 2.0319073305638034e-06,
+      "loss": 0.0003,
+      "reward": 0.5167411044239998,
+      "reward_std": 0.35071658343076706,
+      "rewards/accuracy_reward": 0.5156250298023224,
+      "rewards/format_reward": 0.0011160714784637094,
+      "step": 26
+    },
+    {
+      "completion_length": 624.9497985839844,
+      "epoch": 0.4605543710021322,
+      "grad_norm": 0.13481971621513367,
+      "kl": 0.0122833251953125,
+      "learning_rate": 1.9462545807828044e-06,
+      "loss": 0.0005,
+      "reward": 0.5837053842842579,
+      "reward_std": 0.3019485678523779,
+      "rewards/accuracy_reward": 0.5837053842842579,
+      "rewards/format_reward": 0.0,
+      "step": 27
+    },
+    {
+      "completion_length": 616.9285888671875,
+      "epoch": 0.47761194029850745,
+      "grad_norm": 0.1329677402973175,
+      "kl": 0.009983062744140625,
+      "learning_rate": 1.8589734964313368e-06,
+      "loss": 0.0004,
+      "reward": 0.5669643059372902,
+      "reward_std": 0.34032437205314636,
+      "rewards/accuracy_reward": 0.5669643059372902,
+      "rewards/format_reward": 0.0,
+      "step": 28
+    },
+    {
+      "completion_length": 634.3593902587891,
+      "epoch": 0.4946695095948827,
+      "grad_norm": 0.08903162181377411,
+      "kl": 0.00791168212890625,
+      "learning_rate": 1.7703825567208588e-06,
+      "loss": 0.0003,
+      "reward": 0.5680803768336773,
+      "reward_std": 0.2970337048172951,
+      "rewards/accuracy_reward": 0.5669643096625805,
+      "rewards/format_reward": 0.0011160714784637094,
+      "step": 29
+    },
+    {
+      "completion_length": 650.2902069091797,
+      "epoch": 0.511727078891258,
+      "grad_norm": 0.09270613640546799,
+      "kl": 0.010021209716796875,
+      "learning_rate": 1.6808050203829845e-06,
+      "loss": 0.0004,
+      "reward": 0.5368303805589676,
+      "reward_std": 0.316427243873477,
+      "rewards/accuracy_reward": 0.5368303805589676,
+      "rewards/format_reward": 0.0,
+      "step": 30
+    },
+    {
+      "completion_length": 635.8582916259766,
+      "epoch": 0.5287846481876333,
+      "grad_norm": 0.0931444764137268,
+      "kl": 0.007686614990234375,
+      "learning_rate": 1.5905677461334292e-06,
+      "loss": 0.0003,
+      "reward": 0.5435268245637417,
+      "reward_std": 0.31041241250932217,
+      "rewards/accuracy_reward": 0.5435268245637417,
+      "rewards/format_reward": 0.0,
+      "step": 31
+    },
+    {
+      "completion_length": 638.8214569091797,
+      "epoch": 0.5458422174840085,
+      "grad_norm": 0.1111321821808815,
+      "kl": 0.007816314697265625,
+      "learning_rate": 1.5e-06,
+      "loss": 0.0003,
+      "reward": 0.5691964626312256,
+      "reward_std": 0.3233349844813347,
+      "rewards/accuracy_reward": 0.5691964626312256,
+      "rewards/format_reward": 0.0,
+      "step": 32
+    },
+    {
+      "completion_length": 627.1317291259766,
+      "epoch": 0.5628997867803838,
+      "grad_norm": 0.1286892145872116,
+      "kl": 0.010196685791015625,
+      "learning_rate": 1.4094322538665708e-06,
+      "loss": 0.0004,
+      "reward": 0.5513393059372902,
+      "reward_std": 0.30831882171332836,
+      "rewards/accuracy_reward": 0.5513393059372902,
+      "rewards/format_reward": 0.0,
+      "step": 33
+    },
+    {
+      "completion_length": 623.7600708007812,
+      "epoch": 0.579957356076759,
+      "grad_norm": 0.10542038083076477,
+      "kl": 0.0088958740234375,
+      "learning_rate": 1.3191949796170155e-06,
+      "loss": 0.0004,
+      "reward": 0.510044664144516,
+      "reward_std": 0.3244265168905258,
+      "rewards/accuracy_reward": 0.5089285895228386,
+      "rewards/format_reward": 0.0011160714784637094,
+      "step": 34
+    },
+    {
+      "completion_length": 662.3705673217773,
+      "epoch": 0.5970149253731343,
+      "grad_norm": 0.16724978387355804,
+      "kl": 0.00736236572265625,
+      "learning_rate": 1.2296174432791415e-06,
+      "loss": 0.0003,
+      "reward": 0.558035746216774,
+      "reward_std": 0.3229935597628355,
+      "rewards/accuracy_reward": 0.558035746216774,
+      "rewards/format_reward": 0.0,
+      "step": 35
+    },
+    {
+      "completion_length": 609.7712326049805,
+      "epoch": 0.6140724946695096,
+      "grad_norm": 0.09450776875019073,
+      "kl": 0.009227752685546875,
+      "learning_rate": 1.141026503568664e-06,
+      "loss": 0.0004,
+      "reward": 0.5345982387661934,
+      "reward_std": 0.3159206211566925,
+      "rewards/accuracy_reward": 0.533482164144516,
+      "rewards/format_reward": 0.0011160714784637094,
+      "step": 36
+    },
+    {
+      "completion_length": 624.8783798217773,
+      "epoch": 0.6311300639658849,
+      "grad_norm": 0.1119498461484909,
+      "kl": 0.0086822509765625,
+      "learning_rate": 1.0537454192171958e-06,
+      "loss": 0.0003,
+      "reward": 0.6026785932481289,
+      "reward_std": 0.2997346203774214,
+      "rewards/accuracy_reward": 0.6026785932481289,
+      "rewards/format_reward": 0.0,
+      "step": 37
+    },
+    {
+      "completion_length": 613.0368576049805,
+      "epoch": 0.6481876332622601,
+      "grad_norm": 0.09415058046579361,
+      "kl": 0.00948333740234375,
+      "learning_rate": 9.680926694361964e-07,
+      "loss": 0.0004,
+      "reward": 0.6015625298023224,
+      "reward_std": 0.2972884103655815,
+      "rewards/accuracy_reward": 0.6015625298023224,
+      "rewards/format_reward": 0.0,
+      "step": 38
+    },
+    {
+      "completion_length": 605.041316986084,
+      "epoch": 0.6652452025586354,
+      "grad_norm": 0.11349077522754669,
+      "kl": 0.01213836669921875,
+      "learning_rate": 8.843807918208651e-07,
+      "loss": 0.0005,
+      "reward": 0.5982143208384514,
+      "reward_std": 0.30514490231871605,
+      "rewards/accuracy_reward": 0.597098246216774,
+      "rewards/format_reward": 0.0011160714784637094,
+      "step": 39
+    },
+    {
+      "completion_length": 664.9520416259766,
+      "epoch": 0.6823027718550106,
+      "grad_norm": 0.07810351997613907,
+      "kl": 0.008647918701171875,
+      "learning_rate": 8.029152419343472e-07,
+      "loss": 0.0003,
+      "reward": 0.5022321715950966,
+      "reward_std": 0.2990181464701891,
+      "rewards/accuracy_reward": 0.5022321715950966,
+      "rewards/format_reward": 0.0,
+      "step": 40
+    },
+    {
+      "completion_length": 637.6194458007812,
+      "epoch": 0.6993603411513859,
+      "grad_norm": 0.11813782900571823,
+      "kl": 0.0095977783203125,
+      "learning_rate": 7.239932787335147e-07,
+      "loss": 0.0004,
+      "reward": 0.5513393096625805,
+      "reward_std": 0.30872474052011967,
+      "rewards/accuracy_reward": 0.5513393096625805,
+      "rewards/format_reward": 0.0,
+      "step": 41
+    },
+    {
+      "completion_length": 629.3381881713867,
+      "epoch": 0.7164179104477612,
+      "grad_norm": 0.08325987309217453,
+      "kl": 0.010807037353515625,
+      "learning_rate": 6.479028799032664e-07,
+      "loss": 0.0004,
+      "reward": 0.597098246216774,
+      "reward_std": 0.31203021854162216,
+      "rewards/accuracy_reward": 0.597098246216774,
+      "rewards/format_reward": 0.0,
+      "step": 42
+    },
+    {
+      "completion_length": 683.7757034301758,
+      "epoch": 0.7334754797441365,
+      "grad_norm": 0.07234999537467957,
+      "kl": 0.006565093994140625,
+      "learning_rate": 5.749216910571854e-07,
+      "loss": 0.0003,
+      "reward": 0.5457589589059353,
+      "reward_std": 0.29359665140509605,
+      "rewards/accuracy_reward": 0.5457589589059353,
+      "rewards/format_reward": 0.0,
+      "step": 43
+    },
+    {
+      "completion_length": 626.8995895385742,
+      "epoch": 0.7505330490405118,
+      "grad_norm": 0.14146152138710022,
+      "kl": 0.01007080078125,
+      "learning_rate": 5.053160126388076e-07,
+      "loss": 0.0004,
+      "reward": 0.5781250335276127,
+      "reward_std": 0.3283701930195093,
+      "rewards/accuracy_reward": 0.5781250335276127,
+      "rewards/format_reward": 0.0,
+      "step": 44
+    },
+    {
+      "completion_length": 607.5156555175781,
+      "epoch": 0.767590618336887,
+      "grad_norm": 0.10176991671323776,
+      "kl": 0.01148223876953125,
+      "learning_rate": 4.3933982822017883e-07,
+      "loss": 0.0005,
+      "reward": 0.584821455180645,
+      "reward_std": 0.2961436016485095,
+      "rewards/accuracy_reward": 0.584821455180645,
+      "rewards/format_reward": 0.0,
+      "step": 45
+    },
+    {
+      "completion_length": 642.0223541259766,
+      "epoch": 0.7846481876332623,
+      "grad_norm": 0.09057696908712387,
+      "kl": 0.008152008056640625,
+      "learning_rate": 3.772338777433482e-07,
+      "loss": 0.0003,
+      "reward": 0.544642873108387,
+      "reward_std": 0.30674307234585285,
+      "rewards/accuracy_reward": 0.544642873108387,
+      "rewards/format_reward": 0.0,
+      "step": 46
+    },
+    {
+      "completion_length": 631.8716888427734,
+      "epoch": 0.8017057569296375,
+      "grad_norm": 0.09820377081632614,
+      "kl": 0.010608673095703125,
+      "learning_rate": 3.192247790864249e-07,
+      "loss": 0.0004,
+      "reward": 0.561383955180645,
+      "reward_std": 0.32947295345366,
+      "rewards/accuracy_reward": 0.5602678805589676,
+      "rewards/format_reward": 0.0011160714784637094,
+      "step": 47
+    },
+    {
+      "completion_length": 628.2678909301758,
+      "epoch": 0.8187633262260128,
+      "grad_norm": 0.10683470219373703,
+      "kl": 0.0122833251953125,
+      "learning_rate": 2.6552420115951547e-07,
+      "loss": 0.0005,
+      "reward": 0.5602678805589676,
+      "reward_std": 0.3167324475944042,
+      "rewards/accuracy_reward": 0.5602678805589676,
+      "rewards/format_reward": 0.0,
+      "step": 48
+    },
+    {
+      "completion_length": 649.482177734375,
+      "epoch": 0.835820895522388,
+      "grad_norm": 0.07014506310224533,
+      "kl": 0.00775146484375,
+      "learning_rate": 2.163280915478289e-07,
+      "loss": 0.0003,
+      "reward": 0.5546875298023224,
+      "reward_std": 0.2984890937805176,
+      "rewards/accuracy_reward": 0.5546875298023224,
+      "rewards/format_reward": 0.0,
+      "step": 49
+    },
+    {
+      "completion_length": 646.8906478881836,
+      "epoch": 0.8528784648187633,
+      "grad_norm": 0.09115675836801529,
+      "kl": 0.008701324462890625,
+      "learning_rate": 1.718159615201853e-07,
+      "loss": 0.0003,
+      "reward": 0.5781250186264515,
+      "reward_std": 0.3092201482504606,
+      "rewards/accuracy_reward": 0.5781250186264515,
+      "rewards/format_reward": 0.0,
+      "step": 50
+    },
+    {
+      "completion_length": 605.0301666259766,
+      "epoch": 0.8699360341151386,
+      "grad_norm": 0.10600890219211578,
+      "kl": 0.0100250244140625,
+      "learning_rate": 1.321502310118649e-07,
+      "loss": 0.0004,
+      "reward": 0.6149553805589676,
+      "reward_std": 0.3139810301363468,
+      "rewards/accuracy_reward": 0.6149553805589676,
+      "rewards/format_reward": 0.0,
+      "step": 51
+    },
+    {
+      "completion_length": 659.0055999755859,
+      "epoch": 0.8869936034115139,
+      "grad_norm": 0.07781906425952911,
+      "kl": 0.00876617431640625,
+      "learning_rate": 9.747563597187792e-08,
+      "loss": 0.0004,
+      "reward": 0.577008955180645,
+      "reward_std": 0.3069279305636883,
+      "rewards/accuracy_reward": 0.577008955180645,
+      "rewards/format_reward": 0.0,
+      "step": 52
+    },
+    {
+      "completion_length": 645.2064971923828,
+      "epoch": 0.9040511727078892,
+      "grad_norm": 0.10724423825740814,
+      "kl": 0.009578704833984375,
+      "learning_rate": 6.791870023718161e-08,
+      "loss": 0.0004,
+      "reward": 0.5636160895228386,
+      "reward_std": 0.2895326167345047,
+      "rewards/accuracy_reward": 0.5636160895228386,
+      "rewards/format_reward": 0.0,
+      "step": 53
+    },
+    {
+      "completion_length": 611.3013534545898,
+      "epoch": 0.9211087420042644,
+      "grad_norm": 0.09349353611469269,
+      "kl": 0.011241912841796875,
+      "learning_rate": 4.358727386092198e-08,
+      "loss": 0.0004,
+      "reward": 0.6250000335276127,
+      "reward_std": 0.31463043205440044,
+      "rewards/accuracy_reward": 0.6250000335276127,
+      "rewards/format_reward": 0.0,
+      "step": 54
+    },
+    {
+      "completion_length": 656.0178833007812,
+      "epoch": 0.9381663113006397,
+      "grad_norm": 0.09912308305501938,
+      "kl": 0.009449005126953125,
+      "learning_rate": 2.4570139579284723e-08,
+      "loss": 0.0004,
+      "reward": 0.5904018133878708,
+      "reward_std": 0.34508523903787136,
+      "rewards/accuracy_reward": 0.5904018133878708,
+      "rewards/format_reward": 0.0,
+      "step": 55
+    },
+    {
+      "completion_length": 626.8180999755859,
+      "epoch": 0.9552238805970149,
+      "grad_norm": 0.3471205234527588,
+      "kl": 0.017642974853515625,
+      "learning_rate": 1.093668885291904e-08,
+      "loss": 0.0007,
+      "reward": 0.5825893059372902,
+      "reward_std": 0.31324212066829205,
+      "rewards/accuracy_reward": 0.5814732387661934,
+      "rewards/format_reward": 0.0011160714784637094,
+      "step": 56
+    },
+    {
+      "completion_length": 627.2902069091797,
+      "epoch": 0.9722814498933902,
+      "grad_norm": 0.08442260324954987,
+      "kl": 0.008678436279296875,
+      "learning_rate": 2.736668665023756e-09,
+      "loss": 0.0003,
+      "reward": 0.597098246216774,
+      "reward_std": 0.2932440135627985,
+      "rewards/accuracy_reward": 0.597098246216774,
+      "rewards/format_reward": 0.0,
+      "step": 57
+    },
+    {
+      "completion_length": 619.4442138671875,
+      "epoch": 0.9893390191897654,
+      "grad_norm": 0.078741155564785,
+      "kl": 0.008129119873046875,
+      "learning_rate": 0.0,
+      "loss": 0.0003,
+      "reward": 0.5970982424914837,
+      "reward_std": 0.28597242943942547,
+      "rewards/accuracy_reward": 0.5970982424914837,
+      "rewards/format_reward": 0.0,
+      "step": 58
+    },
+    {
+      "epoch": 0.9893390191897654,
+      "step": 58,
+      "total_flos": 0.0,
+      "train_loss": 0.0,
+      "train_runtime": 1.0266,
+      "train_samples_per_second": 7305.492,
+      "train_steps_per_second": 56.496
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 58,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 20,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "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:e55781e19058d9186bf905b094cd5cb5eec883bbdf40d80849f0ade9ff24c939
+size 7608