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="andrewsiah/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/hong-exploration/openr1/runs/plxee6ek) 
+
+
+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
+- 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": 0.08713241612633617,
+    "train_runtime": 26094.7605,
+    "train_samples": 72441,
+    "train_samples_per_second": 2.776,
+    "train_steps_per_second": 0.011
+}

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 sha256:503226a773136a687afaf4201ba575a9693f2ab24153e332c1274a962569cfa4
+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": 0.08713241612633617,
+    "train_runtime": 26094.7605,
+    "train_samples": 72441,
+    "train_samples_per_second": 2.776,
+    "train_steps_per_second": 0.011
+}

trainer_state.json

@@ -0,0 +1,804 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 100,
+  "global_step": 283,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "completion_length": 397.9532531738281,
+      "epoch": 0.0176678445229682,
+      "grad_norm": 0.6871901154518127,
+      "kl": 0.0005688667297363281,
+      "learning_rate": 3.448275862068966e-06,
+      "loss": 0.0,
+      "reward": 0.6176339596509933,
+      "reward_std": 0.43154375404119494,
+      "rewards/accuracy_reward": 0.14810268627479672,
+      "rewards/format_reward": 0.4695312701165676,
+      "step": 5
+    },
+    {
+      "completion_length": 241.06094856262206,
+      "epoch": 0.0353356890459364,
+      "grad_norm": 9.674223899841309,
+      "kl": 0.35438079833984376,
+      "learning_rate": 6.896551724137932e-06,
+      "loss": 0.0142,
+      "reward": 0.9390625447034836,
+      "reward_std": 0.28645481169223785,
+      "rewards/accuracy_reward": 0.07845982513390481,
+      "rewards/format_reward": 0.8606027245521546,
+      "step": 10
+    },
+    {
+      "completion_length": 159.62165946960448,
+      "epoch": 0.053003533568904596,
+      "grad_norm": 0.3432278037071228,
+      "kl": 0.055914306640625,
+      "learning_rate": 1.0344827586206898e-05,
+      "loss": 0.0022,
+      "reward": 1.0266741544008255,
+      "reward_std": 0.2444406483322382,
+      "rewards/accuracy_reward": 0.09319196930155157,
+      "rewards/format_reward": 0.933482187986374,
+      "step": 15
+    },
+    {
+      "completion_length": 106.59353141784668,
+      "epoch": 0.0706713780918728,
+      "grad_norm": 0.3545452356338501,
+      "kl": 0.127545166015625,
+      "learning_rate": 1.3793103448275863e-05,
+      "loss": 0.0051,
+      "reward": 1.0803571969270707,
+      "reward_std": 0.22546937856823207,
+      "rewards/accuracy_reward": 0.12232143525034189,
+      "rewards/format_reward": 0.9580357596278191,
+      "step": 20
+    },
+    {
+      "completion_length": 170.26206130981444,
+      "epoch": 0.08833922261484099,
+      "grad_norm": 0.2860734164714813,
+      "kl": 0.10814208984375,
+      "learning_rate": 1.7241379310344828e-05,
+      "loss": 0.0043,
+      "reward": 1.1982143491506576,
+      "reward_std": 0.27088444530963895,
+      "rewards/accuracy_reward": 0.22511161882430314,
+      "rewards/format_reward": 0.9731027260422707,
+      "step": 25
+    },
+    {
+      "completion_length": 254.76976509094237,
+      "epoch": 0.10600706713780919,
+      "grad_norm": 0.29131728410720825,
+      "kl": 2.9212158203125,
+      "learning_rate": 1.999923511388017e-05,
+      "loss": 0.1172,
+      "reward": 1.187165230512619,
+      "reward_std": 0.35072063505649564,
+      "rewards/accuracy_reward": 0.2654017977416515,
+      "rewards/format_reward": 0.9217634350061417,
+      "step": 30
+    },
+    {
+      "completion_length": 180.78471794128419,
+      "epoch": 0.12367491166077739,
+      "grad_norm": 0.25441914796829224,
+      "kl": 0.14281005859375,
+      "learning_rate": 1.9972476383747748e-05,
+      "loss": 0.0057,
+      "reward": 1.2573661237955094,
+      "reward_std": 0.27556890323758126,
+      "rewards/accuracy_reward": 0.27410715371370314,
+      "rewards/format_reward": 0.98325896859169,
+      "step": 35
+    },
+    {
+      "completion_length": 265.0565986633301,
+      "epoch": 0.1413427561837456,
+      "grad_norm": 10500.8896484375,
+      "kl": 45.15205078125,
+      "learning_rate": 1.9907590277344582e-05,
+      "loss": 1.8084,
+      "reward": 1.2706473737955093,
+      "reward_std": 0.307077270001173,
+      "rewards/accuracy_reward": 0.3092634052038193,
+      "rewards/format_reward": 0.9613839730620384,
+      "step": 40
+    },
+    {
+      "completion_length": 177.18170375823973,
+      "epoch": 0.15901060070671377,
+      "grad_norm": 11.025420188903809,
+      "kl": 0.8410888671875,
+      "learning_rate": 1.9804824871166254e-05,
+      "loss": 0.0336,
+      "reward": 1.197098270058632,
+      "reward_std": 0.3107326962053776,
+      "rewards/accuracy_reward": 0.2512276891618967,
+      "rewards/format_reward": 0.9458705827593803,
+      "step": 45
+    },
+    {
+      "completion_length": 197.19275417327881,
+      "epoch": 0.17667844522968199,
+      "grad_norm": 2.2072925567626953,
+      "kl": 0.9974609375,
+      "learning_rate": 1.9664573064143604e-05,
+      "loss": 0.0399,
+      "reward": 1.083928619325161,
+      "reward_std": 0.41752763763070105,
+      "rewards/accuracy_reward": 0.2433035835623741,
+      "rewards/format_reward": 0.8406250357627869,
+      "step": 50
+    },
+    {
+      "completion_length": 125.53616695404052,
+      "epoch": 0.19434628975265017,
+      "grad_norm": 0.9618227481842041,
+      "kl": 0.76201171875,
+      "learning_rate": 1.948737107548771e-05,
+      "loss": 0.0305,
+      "reward": 1.118080411851406,
+      "reward_std": 0.3598920326679945,
+      "rewards/accuracy_reward": 0.2600446529686451,
+      "rewards/format_reward": 0.8580357536673546,
+      "step": 55
+    },
+    {
+      "completion_length": 327.64934196472166,
+      "epoch": 0.21201413427561838,
+      "grad_norm": 10.108561515808105,
+      "kl": 6.38251953125,
+      "learning_rate": 1.9273896394584103e-05,
+      "loss": 0.2551,
+      "reward": 0.9488839730620384,
+      "reward_std": 0.4913848154246807,
+      "rewards/accuracy_reward": 0.24017858244478701,
+      "rewards/format_reward": 0.7087053894996643,
+      "step": 60
+    },
+    {
+      "completion_length": 242.94030113220214,
+      "epoch": 0.22968197879858657,
+      "grad_norm": 2.333926200866699,
+      "kl": 1.8529296875,
+      "learning_rate": 1.9024965190774262e-05,
+      "loss": 0.0742,
+      "reward": 0.9114955708384513,
+      "reward_std": 0.5022768549621105,
+      "rewards/accuracy_reward": 0.23303572423756122,
+      "rewards/format_reward": 0.6784598529338837,
+      "step": 65
+    },
+    {
+      "completion_length": 415.55738792419436,
+      "epoch": 0.24734982332155478,
+      "grad_norm": 1.1008508205413818,
+      "kl": 11.3341796875,
+      "learning_rate": 1.8741529192927528e-05,
+      "loss": 0.4533,
+      "reward": 0.920647357404232,
+      "reward_std": 0.4701464258134365,
+      "rewards/accuracy_reward": 0.21796876080334188,
+      "rewards/format_reward": 0.702678607404232,
+      "step": 70
+    },
+    {
+      "completion_length": 168.85536460876466,
+      "epoch": 0.26501766784452296,
+      "grad_norm": 0.3387093245983124,
+      "kl": 0.340380859375,
+      "learning_rate": 1.8424672050733577e-05,
+      "loss": 0.0136,
+      "reward": 1.2478795111179353,
+      "reward_std": 0.2676353581249714,
+      "rewards/accuracy_reward": 0.26607143841683867,
+      "rewards/format_reward": 0.9818080767989159,
+      "step": 75
+    },
+    {
+      "completion_length": 228.08896026611328,
+      "epoch": 0.2826855123674912,
+      "grad_norm": 0.9070476293563843,
+      "kl": 0.5783935546875,
+      "learning_rate": 1.8075605191627242e-05,
+      "loss": 0.0231,
+      "reward": 1.2245536267757415,
+      "reward_std": 0.3449659403413534,
+      "rewards/accuracy_reward": 0.287500012665987,
+      "rewards/format_reward": 0.9370536103844642,
+      "step": 80
+    },
+    {
+      "completion_length": 349.13952407836916,
+      "epoch": 0.3003533568904594,
+      "grad_norm": 3.210876703262329,
+      "kl": 2.055322265625,
+      "learning_rate": 1.7695663189185703e-05,
+      "loss": 0.0823,
+      "reward": 1.1111607626080513,
+      "reward_std": 0.4750510863959789,
+      "rewards/accuracy_reward": 0.2958705469965935,
+      "rewards/format_reward": 0.8152902171015739,
+      "step": 85
+    },
+    {
+      "completion_length": 356.24353942871096,
+      "epoch": 0.31802120141342755,
+      "grad_norm": 0.17505526542663574,
+      "kl": 0.28251953125,
+      "learning_rate": 1.7286298660705877e-05,
+      "loss": 0.0113,
+      "reward": 1.1301339864730835,
+      "reward_std": 0.49002460166811945,
+      "rewards/accuracy_reward": 0.32444198094308374,
+      "rewards/format_reward": 0.8056919991970062,
+      "step": 90
+    },
+    {
+      "completion_length": 333.82724838256837,
+      "epoch": 0.33568904593639576,
+      "grad_norm": 0.16171219944953918,
+      "kl": 0.26669921875,
+      "learning_rate": 1.6849076713469914e-05,
+      "loss": 0.0107,
+      "reward": 1.2251116514205933,
+      "reward_std": 0.3983605161309242,
+      "rewards/accuracy_reward": 0.33158483579754827,
+      "rewards/format_reward": 0.8935268297791481,
+      "step": 95
+    },
+    {
+      "completion_length": 238.18862686157226,
+      "epoch": 0.35335689045936397,
+      "grad_norm": 1.341346263885498,
+      "kl": 0.214111328125,
+      "learning_rate": 1.6385668960932143e-05,
+      "loss": 0.0086,
+      "reward": 1.2977679133415223,
+      "reward_std": 0.3174902819097042,
+      "rewards/accuracy_reward": 0.3381696566939354,
+      "rewards/format_reward": 0.9595982566475868,
+      "step": 100
+    },
+    {
+      "epoch": 0.35335689045936397,
+      "eval_completion_length": 224.875,
+      "eval_kl": 0.2919921875,
+      "eval_loss": 0.012192243710160255,
+      "eval_reward": 1.233258992433548,
+      "eval_reward_std": 0.3431517854332924,
+      "eval_rewards/accuracy_reward": 0.3002232313156128,
+      "eval_rewards/format_reward": 0.9330357611179352,
+      "eval_runtime": 37.6317,
+      "eval_samples_per_second": 2.631,
+      "eval_steps_per_second": 0.027,
+      "step": 100
+    },
+    {
+      "completion_length": 261.6398551940918,
+      "epoch": 0.3710247349823322,
+      "grad_norm": 1.9713914394378662,
+      "kl": 0.8895751953125,
+      "learning_rate": 1.5897847131705194e-05,
+      "loss": 0.0356,
+      "reward": 1.1974330961704254,
+      "reward_std": 0.4034945048391819,
+      "rewards/accuracy_reward": 0.30926340594887736,
+      "rewards/format_reward": 0.8881696864962578,
+      "step": 105
+    },
+    {
+      "completion_length": 317.6125152587891,
+      "epoch": 0.38869257950530034,
+      "grad_norm": 0.4451054632663727,
+      "kl": 1.036181640625,
+      "learning_rate": 1.5387476295779737e-05,
+      "loss": 0.0415,
+      "reward": 1.175223258137703,
+      "reward_std": 0.4380528204143047,
+      "rewards/accuracy_reward": 0.3095982287079096,
+      "rewards/format_reward": 0.865625037252903,
+      "step": 110
+    },
+    {
+      "completion_length": 335.4704399108887,
+      "epoch": 0.40636042402826855,
+      "grad_norm": 0.2159155160188675,
+      "kl": 0.2209228515625,
+      "learning_rate": 1.4856507733875837e-05,
+      "loss": 0.0088,
+      "reward": 1.2347098767757416,
+      "reward_std": 0.3829378016293049,
+      "rewards/accuracy_reward": 0.3314732301980257,
+      "rewards/format_reward": 0.9032366424798965,
+      "step": 115
+    },
+    {
+      "completion_length": 326.20805130004885,
+      "epoch": 0.42402826855123676,
+      "grad_norm": 1.607401728630066,
+      "kl": 0.16533203125,
+      "learning_rate": 1.4306971477188223e-05,
+      "loss": 0.0066,
+      "reward": 1.344531312584877,
+      "reward_std": 0.3044683691114187,
+      "rewards/accuracy_reward": 0.38002234175801275,
+      "rewards/format_reward": 0.9645089700818061,
+      "step": 120
+    },
+    {
+      "completion_length": 346.5212211608887,
+      "epoch": 0.4416961130742049,
+      "grad_norm": 0.5397958755493164,
+      "kl": 0.2360595703125,
+      "learning_rate": 1.3740968546047935e-05,
+      "loss": 0.0094,
+      "reward": 1.3156250536441803,
+      "reward_std": 0.3454729124903679,
+      "rewards/accuracy_reward": 0.3767857324331999,
+      "rewards/format_reward": 0.9388393282890319,
+      "step": 125
+    },
+    {
+      "completion_length": 458.170223236084,
+      "epoch": 0.45936395759717313,
+      "grad_norm": 2.1645452976226807,
+      "kl": 3.1985595703125,
+      "learning_rate": 1.3160662917174045e-05,
+      "loss": 0.128,
+      "reward": 1.0997768297791481,
+      "reward_std": 0.5340591460466385,
+      "rewards/accuracy_reward": 0.3318080510944128,
+      "rewards/format_reward": 0.7679687827825546,
+      "step": 130
+    },
+    {
+      "completion_length": 381.4783645629883,
+      "epoch": 0.47703180212014135,
+      "grad_norm": 251.16600036621094,
+      "kl": 4.827490234375,
+      "learning_rate": 1.2568273250226681e-05,
+      "loss": 0.1933,
+      "reward": 1.2683036237955094,
+      "reward_std": 0.38932468742132187,
+      "rewards/accuracy_reward": 0.3660714436322451,
+      "rewards/format_reward": 0.9022321864962578,
+      "step": 135
+    },
+    {
+      "completion_length": 331.04789505004885,
+      "epoch": 0.49469964664310956,
+      "grad_norm": 0.9643793702125549,
+      "kl": 0.825537109375,
+      "learning_rate": 1.1966064405292887e-05,
+      "loss": 0.033,
+      "reward": 1.3402902394533158,
+      "reward_std": 0.32668678648769855,
+      "rewards/accuracy_reward": 0.3882812689989805,
+      "rewards/format_reward": 0.95200896859169,
+      "step": 140
+    },
+    {
+      "completion_length": 323.53405685424804,
+      "epoch": 0.5123674911660777,
+      "grad_norm": 5.4946722984313965,
+      "kl": 1.706689453125,
+      "learning_rate": 1.1356338783736256e-05,
+      "loss": 0.0683,
+      "reward": 1.3253348886966705,
+      "reward_std": 0.32204796373844147,
+      "rewards/accuracy_reward": 0.3709821619093418,
+      "rewards/format_reward": 0.9543527230620384,
+      "step": 145
+    },
+    {
+      "completion_length": 334.69420700073243,
+      "epoch": 0.5300353356890459,
+      "grad_norm": 28.387596130371094,
+      "kl": 2.99873046875,
+      "learning_rate": 1.0741427525516463e-05,
+      "loss": 0.12,
+      "reward": 1.272656312584877,
+      "reward_std": 0.3838721476495266,
+      "rewards/accuracy_reward": 0.3592634070664644,
+      "rewards/format_reward": 0.913392898440361,
+      "step": 150
+    },
+    {
+      "completion_length": 350.9021369934082,
+      "epoch": 0.5477031802120141,
+      "grad_norm": 6.340213298797607,
+      "kl": 1.597265625,
+      "learning_rate": 1.012368159663363e-05,
+      "loss": 0.0639,
+      "reward": 0.9706473708152771,
+      "reward_std": 0.5445995971560478,
+      "rewards/accuracy_reward": 0.2651785839349031,
+      "rewards/format_reward": 0.7054687768220902,
+      "step": 155
+    },
+    {
+      "completion_length": 307.3893013000488,
+      "epoch": 0.5653710247349824,
+      "grad_norm": 1.4934422969818115,
+      "kl": 1.1126953125,
+      "learning_rate": 9.505462800772612e-06,
+      "loss": 0.0445,
+      "reward": 1.2178571939468383,
+      "reward_std": 0.4379304856061935,
+      "rewards/accuracy_reward": 0.33515626527369025,
+      "rewards/format_reward": 0.8827009320259094,
+      "step": 160
+    },
+    {
+      "completion_length": 311.874341583252,
+      "epoch": 0.5830388692579506,
+      "grad_norm": 1.8048169612884521,
+      "kl": 0.587841796875,
+      "learning_rate": 8.889134749511956e-06,
+      "loss": 0.0235,
+      "reward": 1.2705357730388642,
+      "reward_std": 0.4048257850110531,
+      "rewards/accuracy_reward": 0.3630580514669418,
+      "rewards/format_reward": 0.9074777245521546,
+      "step": 165
+    },
+    {
+      "completion_length": 340.8625144958496,
+      "epoch": 0.6007067137809188,
+      "grad_norm": 11.069897651672363,
+      "kl": 0.925537109375,
+      "learning_rate": 8.277053825620836e-06,
+      "loss": 0.037,
+      "reward": 1.2059152334928513,
+      "reward_std": 0.41814019829034804,
+      "rewards/accuracy_reward": 0.3325892988592386,
+      "rewards/format_reward": 0.8733259364962578,
+      "step": 170
+    },
+    {
+      "completion_length": 396.6475631713867,
+      "epoch": 0.6183745583038869,
+      "grad_norm": 2.382586717605591,
+      "kl": 2.17744140625,
+      "learning_rate": 7.671560173993588e-06,
+      "loss": 0.0871,
+      "reward": 1.1248884528875351,
+      "reward_std": 0.5201080188155174,
+      "rewards/accuracy_reward": 0.32968751452863215,
+      "rewards/format_reward": 0.795200927555561,
+      "step": 175
+    },
+    {
+      "completion_length": 372.08874435424804,
+      "epoch": 0.6360424028268551,
+      "grad_norm": 0.8225041031837463,
+      "kl": 1.33203125,
+      "learning_rate": 7.07496875466589e-06,
+      "loss": 0.0533,
+      "reward": 1.198325940966606,
+      "reward_std": 0.46074943244457245,
+      "rewards/accuracy_reward": 0.3479910880327225,
+      "rewards/format_reward": 0.850334855914116,
+      "step": 180
+    },
+    {
+      "completion_length": 355.8721153259277,
+      "epoch": 0.6537102473498233,
+      "grad_norm": 0.28054162859916687,
+      "kl": 0.681640625,
+      "learning_rate": 6.489560492119225e-06,
+      "loss": 0.0273,
+      "reward": 1.2812500566244125,
+      "reward_std": 0.39972887486219405,
+      "rewards/accuracy_reward": 0.3756696589291096,
+      "rewards/format_reward": 0.9055803954601288,
+      "step": 185
+    },
+    {
+      "completion_length": 368.4487907409668,
+      "epoch": 0.6713780918727915,
+      "grad_norm": 1.2593623399734497,
+      "kl": 1.3046875,
+      "learning_rate": 5.9175735547120975e-06,
+      "loss": 0.0522,
+      "reward": 1.2546875447034835,
+      "reward_std": 0.42497922107577324,
+      "rewards/accuracy_reward": 0.3743303745985031,
+      "rewards/format_reward": 0.8803571835160255,
+      "step": 190
+    },
+    {
+      "completion_length": 373.5961112976074,
+      "epoch": 0.6890459363957597,
+      "grad_norm": 0.49754491448402405,
+      "kl": 1.1783203125,
+      "learning_rate": 5.361194797579108e-06,
+      "loss": 0.0471,
+      "reward": 1.207700952887535,
+      "reward_std": 0.4664508134126663,
+      "rewards/accuracy_reward": 0.35491073206067086,
+      "rewards/format_reward": 0.85279021859169,
+      "step": 195
+    },
+    {
+      "completion_length": 334.22434463500974,
+      "epoch": 0.7067137809187279,
+      "grad_norm": 0.35578688979148865,
+      "kl": 0.481298828125,
+      "learning_rate": 4.8225514017138205e-06,
+      "loss": 0.0193,
+      "reward": 1.2475446999073028,
+      "reward_std": 0.41954655647277833,
+      "rewards/accuracy_reward": 0.354799123108387,
+      "rewards/format_reward": 0.8927455767989159,
+      "step": 200
+    },
+    {
+      "epoch": 0.7067137809187279,
+      "eval_completion_length": 301.23802947998047,
+      "eval_kl": 0.3828125,
+      "eval_loss": 0.014114470221102238,
+      "eval_reward": 1.3214286267757416,
+      "eval_reward_std": 0.3996920883655548,
+      "eval_rewards/accuracy_reward": 0.4129464402794838,
+      "eval_rewards/format_reward": 0.9084821790456772,
+      "eval_runtime": 43.0662,
+      "eval_samples_per_second": 2.299,
+      "eval_steps_per_second": 0.023,
+      "step": 200
+    },
+    {
+      "completion_length": 310.799568939209,
+      "epoch": 0.7243816254416962,
+      "grad_norm": 0.23750554025173187,
+      "kl": 0.35625,
+      "learning_rate": 4.303702741201431e-06,
+      "loss": 0.0143,
+      "reward": 1.285491129755974,
+      "reward_std": 0.4243697762489319,
+      "rewards/accuracy_reward": 0.38526787608861923,
+      "rewards/format_reward": 0.900223258137703,
+      "step": 205
+    },
+    {
+      "completion_length": 308.95849609375,
+      "epoch": 0.7420494699646644,
+      "grad_norm": 5.210378646850586,
+      "kl": 0.3337646484375,
+      "learning_rate": 3.8066325096949153e-06,
+      "loss": 0.0133,
+      "reward": 1.243192020058632,
+      "reward_std": 0.41392386630177497,
+      "rewards/accuracy_reward": 0.3464285880327225,
+      "rewards/format_reward": 0.8967634350061416,
+      "step": 210
+    },
+    {
+      "completion_length": 302.08226890563964,
+      "epoch": 0.7597173144876325,
+      "grad_norm": 76.71176147460938,
+      "kl": 6.380517578125,
+      "learning_rate": 3.3332411362372063e-06,
+      "loss": 0.2552,
+      "reward": 1.2647322028875352,
+      "reward_std": 0.4097910635173321,
+      "rewards/accuracy_reward": 0.36004466041922567,
+      "rewards/format_reward": 0.9046875461935997,
+      "step": 215
+    },
+    {
+      "completion_length": 298.6371768951416,
+      "epoch": 0.7773851590106007,
+      "grad_norm": 6.7656354904174805,
+      "kl": 1.5587890625,
+      "learning_rate": 2.8853385194256677e-06,
+      "loss": 0.0624,
+      "reward": 1.278571480512619,
+      "reward_std": 0.4139442786574364,
+      "rewards/accuracy_reward": 0.37700894586741923,
+      "rewards/format_reward": 0.9015625432133675,
+      "step": 220
+    },
+    {
+      "completion_length": 297.99209213256836,
+      "epoch": 0.7950530035335689,
+      "grad_norm": 4.592045307159424,
+      "kl": 2.12470703125,
+      "learning_rate": 2.464637107698046e-06,
+      "loss": 0.085,
+      "reward": 1.267522367835045,
+      "reward_std": 0.40733090564608576,
+      "rewards/accuracy_reward": 0.36272323206067086,
+      "rewards/format_reward": 0.9047991514205933,
+      "step": 225
+    },
+    {
+      "completion_length": 299.3953262329102,
+      "epoch": 0.8127208480565371,
+      "grad_norm": 2.270449638366699,
+      "kl": 1.3138671875,
+      "learning_rate": 2.072745352195794e-06,
+      "loss": 0.0526,
+      "reward": 1.2822545230388642,
+      "reward_std": 0.4120013751089573,
+      "rewards/accuracy_reward": 0.37901787385344504,
+      "rewards/format_reward": 0.9032366499304771,
+      "step": 230
+    },
+    {
+      "completion_length": 298.9074893951416,
+      "epoch": 0.8303886925795053,
+      "grad_norm": 2.1383728981018066,
+      "kl": 1.87900390625,
+      "learning_rate": 1.7111615572361628e-06,
+      "loss": 0.0752,
+      "reward": 1.2869420260190965,
+      "reward_std": 0.4166914038360119,
+      "rewards/accuracy_reward": 0.3809151966124773,
+      "rewards/format_reward": 0.906026828289032,
+      "step": 235
+    },
+    {
+      "completion_length": 305.2895233154297,
+      "epoch": 0.8480565371024735,
+      "grad_norm": 1.2467355728149414,
+      "kl": 1.32548828125,
+      "learning_rate": 1.381268151904298e-06,
+      "loss": 0.053,
+      "reward": 1.2613839715719224,
+      "reward_std": 0.40479170605540277,
+      "rewards/accuracy_reward": 0.3544643003493547,
+      "rewards/format_reward": 0.9069196864962578,
+      "step": 240
+    },
+    {
+      "completion_length": 299.97478942871095,
+      "epoch": 0.8657243816254417,
+      "grad_norm": 1.4973329305648804,
+      "kl": 0.99462890625,
+      "learning_rate": 1.0843264046665558e-06,
+      "loss": 0.0398,
+      "reward": 1.2777902424335479,
+      "reward_std": 0.4032285138964653,
+      "rewards/accuracy_reward": 0.3697544813156128,
+      "rewards/format_reward": 0.9080357551574707,
+      "step": 245
+    },
+    {
+      "completion_length": 294.9779163360596,
+      "epoch": 0.8833922261484098,
+      "grad_norm": 0.6018723845481873,
+      "kl": 0.6970703125,
+      "learning_rate": 8.214716012124491e-07,
+      "loss": 0.0279,
+      "reward": 1.2985491663217545,
+      "reward_std": 0.4030233770608902,
+      "rewards/accuracy_reward": 0.3878348398953676,
+      "rewards/format_reward": 0.9107143312692643,
+      "step": 250
+    },
+    {
+      "completion_length": 326.83907623291014,
+      "epoch": 0.901060070671378,
+      "grad_norm": 0.3602770268917084,
+      "kl": 0.62119140625,
+      "learning_rate": 5.937087039615619e-07,
+      "loss": 0.0249,
+      "reward": 1.228459867835045,
+      "reward_std": 0.43514843806624415,
+      "rewards/accuracy_reward": 0.3444196585565805,
+      "rewards/format_reward": 0.8840402200818062,
+      "step": 255
+    },
+    {
+      "completion_length": 330.93282928466795,
+      "epoch": 0.9187279151943463,
+      "grad_norm": 0.2966284155845642,
+      "kl": 0.5912109375,
+      "learning_rate": 4.019085098303077e-07,
+      "loss": 0.0237,
+      "reward": 1.2388393431901932,
+      "reward_std": 0.45445024222135544,
+      "rewards/accuracy_reward": 0.36796876713633536,
+      "rewards/format_reward": 0.8708705797791481,
+      "step": 260
+    },
+    {
+      "completion_length": 342.0903045654297,
+      "epoch": 0.9363957597173145,
+      "grad_norm": 0.3853587508201599,
+      "kl": 0.630712890625,
+      "learning_rate": 2.4680432094837394e-07,
+      "loss": 0.0252,
+      "reward": 1.1949777334928513,
+      "reward_std": 0.4680465690791607,
+      "rewards/accuracy_reward": 0.3334821581840515,
+      "rewards/format_reward": 0.8614955797791481,
+      "step": 265
+    },
+    {
+      "completion_length": 333.04599533081057,
+      "epoch": 0.9540636042402827,
+      "grad_norm": 0.3707776665687561,
+      "kl": 0.55869140625,
+      "learning_rate": 1.289891410535593e-07,
+      "loss": 0.0223,
+      "reward": 1.2250000566244126,
+      "reward_std": 0.4637945219874382,
+      "rewards/accuracy_reward": 0.357142872735858,
+      "rewards/format_reward": 0.8678571864962578,
+      "step": 270
+    },
+    {
+      "completion_length": 335.9395240783691,
+      "epoch": 0.9717314487632509,
+      "grad_norm": 0.2659561038017273,
+      "kl": 0.572705078125,
+      "learning_rate": 4.8913408283934874e-08,
+      "loss": 0.0229,
+      "reward": 1.216294702887535,
+      "reward_std": 0.44666091352701187,
+      "rewards/accuracy_reward": 0.3496651936322451,
+      "rewards/format_reward": 0.8666295036673546,
+      "step": 275
+    },
+    {
+      "completion_length": 334.8103942871094,
+      "epoch": 0.9893992932862191,
+      "grad_norm": 0.3593786060810089,
+      "kl": 0.59736328125,
+      "learning_rate": 6.883273035447335e-09,
+      "loss": 0.0239,
+      "reward": 1.2222098767757417,
+      "reward_std": 0.4681737795472145,
+      "rewards/accuracy_reward": 0.3558035880327225,
+      "rewards/format_reward": 0.8664062902331352,
+      "step": 280
+    },
+    {
+      "completion_length": 347.41717020670575,
+      "epoch": 1.0,
+      "kl": 0.5322265625,
+      "reward": 1.2327009389797847,
+      "reward_std": 0.44660818700989086,
+      "rewards/accuracy_reward": 0.3696056691308816,
+      "rewards/format_reward": 0.8630952735741934,
+      "step": 283,
+      "total_flos": 0.0,
+      "train_loss": 0.08713241612633617,
+      "train_runtime": 26094.7605,
+      "train_samples_per_second": 2.776,
+      "train_steps_per_second": 0.011
+    }
+  ],
+  "logging_steps": 5,
+  "max_steps": 283,
+  "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": 32,
+  "trial_name": null,
+  "trial_params": null
+}

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:17f991b10f84285759eaffd9bc7353083daa3dce24c0d3c65bad841165cbe325
+size 7544