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="TTTXXX01/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/yuanyige/OpenR1/runs/5f2hz5f5) 
+
+
+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: 4.50.0.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": 0.40895458704104165,
+    "train_runtime": 40740.9033,
+    "train_samples": 72441,
+    "train_samples_per_second": 1.778,
+    "train_steps_per_second": 0.007
+}

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": 32768,
+  "tie_word_embeddings": true,
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.50.0.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": "4.50.0.dev0"
+}

model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:76cf012a277e2f7845ffd1607a4832d5cf30f436e32b33261ef79bdb28ddf15f
+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": 0.40895458704104165,
+    "train_runtime": 40740.9033,
+    "train_samples": 72441,
+    "train_samples_per_second": 1.778,
+    "train_steps_per_second": 0.007
+}

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": 382.93059597015383,
+      "epoch": 0.0176678445229682,
+      "grad_norm": 0.40728917717933655,
+      "kl": 0.0010067939758300782,
+      "learning_rate": 3.448275862068966e-06,
+      "loss": 0.1437,
+      "reward": 0.6503348495811224,
+      "reward_std": 0.4365073021501303,
+      "rewards/accuracy_reward": 0.1477678638882935,
+      "rewards/format_reward": 0.5025669865310192,
+      "step": 5
+    },
+    {
+      "completion_length": 255.40380630493163,
+      "epoch": 0.0353356890459364,
+      "grad_norm": 0.23681369423866272,
+      "kl": 0.21269454956054687,
+      "learning_rate": 6.896551724137932e-06,
+      "loss": 0.1146,
+      "reward": 0.9584821842610836,
+      "reward_std": 0.2699831190519035,
+      "rewards/accuracy_reward": 0.076004467648454,
+      "rewards/format_reward": 0.8824777193367481,
+      "step": 10
+    },
+    {
+      "completion_length": 210.21340198516845,
+      "epoch": 0.053003533568904596,
+      "grad_norm": 1.0446707010269165,
+      "kl": 0.0361907958984375,
+      "learning_rate": 1.0344827586206898e-05,
+      "loss": 0.0624,
+      "reward": 1.0110491536557675,
+      "reward_std": 0.26576729528605936,
+      "rewards/accuracy_reward": 0.08939732502913103,
+      "rewards/format_reward": 0.9216518297791481,
+      "step": 15
+    },
+    {
+      "completion_length": 173.89855737686156,
+      "epoch": 0.0706713780918728,
+      "grad_norm": 0.2942740023136139,
+      "kl": 0.074810791015625,
+      "learning_rate": 1.3793103448275863e-05,
+      "loss": 0.032,
+      "reward": 1.073437547683716,
+      "reward_std": 0.23347075609490275,
+      "rewards/accuracy_reward": 0.11428571951109916,
+      "rewards/format_reward": 0.9591518238186836,
+      "step": 20
+    },
+    {
+      "completion_length": 220.5953229904175,
+      "epoch": 0.08833922261484099,
+      "grad_norm": 3.951831340789795,
+      "kl": 0.531695556640625,
+      "learning_rate": 1.7241379310344828e-05,
+      "loss": 0.0981,
+      "reward": 1.0161830835044383,
+      "reward_std": 0.3684789047576487,
+      "rewards/accuracy_reward": 0.16328125749714673,
+      "rewards/format_reward": 0.8529018230736256,
+      "step": 25
+    },
+    {
+      "completion_length": 261.87579288482664,
+      "epoch": 0.10600706713780919,
+      "grad_norm": 11.097806930541992,
+      "kl": 2.1247802734375,
+      "learning_rate": 1.999923511388017e-05,
+      "loss": 0.3122,
+      "reward": 0.7962053943425417,
+      "reward_std": 0.5473601501435041,
+      "rewards/accuracy_reward": 0.17433036593720316,
+      "rewards/format_reward": 0.6218750271946192,
+      "step": 30
+    },
+    {
+      "completion_length": 151.15904779434203,
+      "epoch": 0.12367491166077739,
+      "grad_norm": 8.74118709564209,
+      "kl": 0.6091796875,
+      "learning_rate": 1.9972476383747748e-05,
+      "loss": 0.1028,
+      "reward": 0.8780134357511997,
+      "reward_std": 0.5024180840700865,
+      "rewards/accuracy_reward": 0.18404018736910074,
+      "rewards/format_reward": 0.6939732424914837,
+      "step": 35
+    },
+    {
+      "completion_length": 113.44442443847656,
+      "epoch": 0.1413427561837456,
+      "grad_norm": 1.2186102867126465,
+      "kl": 0.50096435546875,
+      "learning_rate": 1.9907590277344582e-05,
+      "loss": 0.0013,
+      "reward": 1.1537946984171867,
+      "reward_std": 0.36896395664662124,
+      "rewards/accuracy_reward": 0.23616072488948703,
+      "rewards/format_reward": 0.9176339685916901,
+      "step": 40
+    },
+    {
+      "completion_length": 156.94252967834473,
+      "epoch": 0.15901060070671377,
+      "grad_norm": 3.4705989360809326,
+      "kl": 1.0330810546875,
+      "learning_rate": 1.9804824871166254e-05,
+      "loss": 0.0519,
+      "reward": 1.2366071969270707,
+      "reward_std": 0.28554527759552,
+      "rewards/accuracy_reward": 0.28035715613514184,
+      "rewards/format_reward": 0.9562500424683094,
+      "step": 45
+    },
+    {
+      "completion_length": 140.84342164993285,
+      "epoch": 0.17667844522968199,
+      "grad_norm": 3.2262566089630127,
+      "kl": 0.81142578125,
+      "learning_rate": 1.9664573064143604e-05,
+      "loss": 0.0412,
+      "reward": 1.210714339464903,
+      "reward_std": 0.3142393365502357,
+      "rewards/accuracy_reward": 0.2649553695693612,
+      "rewards/format_reward": 0.9457589693367481,
+      "step": 50
+    },
+    {
+      "completion_length": 135.22857761383057,
+      "epoch": 0.19434628975265017,
+      "grad_norm": 0.37104904651641846,
+      "kl": 0.295947265625,
+      "learning_rate": 1.948737107548771e-05,
+      "loss": 0.0008,
+      "reward": 1.2118304073810577,
+      "reward_std": 0.31936368104070423,
+      "rewards/accuracy_reward": 0.2750000124797225,
+      "rewards/format_reward": 0.9368304006755352,
+      "step": 55
+    },
+    {
+      "completion_length": 194.0175313949585,
+      "epoch": 0.21201413427561838,
+      "grad_norm": 79.59091186523438,
+      "kl": 2.442626953125,
+      "learning_rate": 1.9273896394584103e-05,
+      "loss": 0.4489,
+      "reward": 1.2225446961820126,
+      "reward_std": 0.3489018376916647,
+      "rewards/accuracy_reward": 0.2940848357975483,
+      "rewards/format_reward": 0.9284598678350449,
+      "step": 60
+    },
+    {
+      "completion_length": 299.74309253692627,
+      "epoch": 0.22968197879858657,
+      "grad_norm": 9.220332145690918,
+      "kl": 2.752783203125,
+      "learning_rate": 1.9024965190774262e-05,
+      "loss": 0.7219,
+      "reward": 1.0818080872297287,
+      "reward_std": 0.4944979056715965,
+      "rewards/accuracy_reward": 0.294196442887187,
+      "rewards/format_reward": 0.7876116424798966,
+      "step": 65
+    },
+    {
+      "completion_length": 347.3866226196289,
+      "epoch": 0.24734982332155478,
+      "grad_norm": 8.275785446166992,
+      "kl": 5.707421875,
+      "learning_rate": 1.8741529192927528e-05,
+      "loss": 0.9457,
+      "reward": 1.0705357603728771,
+      "reward_std": 0.511051194742322,
+      "rewards/accuracy_reward": 0.30234376471489666,
+      "rewards/format_reward": 0.7681920006871223,
+      "step": 70
+    },
+    {
+      "completion_length": 258.911171913147,
+      "epoch": 0.26501766784452296,
+      "grad_norm": 6.921362400054932,
+      "kl": 1.3368896484375,
+      "learning_rate": 1.8424672050733577e-05,
+      "loss": 0.4684,
+      "reward": 1.2554688058793544,
+      "reward_std": 0.38661765307188034,
+      "rewards/accuracy_reward": 0.3574776967987418,
+      "rewards/format_reward": 0.8979911111295223,
+      "step": 75
+    },
+    {
+      "completion_length": 274.3730037689209,
+      "epoch": 0.2826855123674912,
+      "grad_norm": 2.2994608879089355,
+      "kl": 0.790625,
+      "learning_rate": 1.8075605191627242e-05,
+      "loss": 0.4752,
+      "reward": 1.2005580842494965,
+      "reward_std": 0.4162170600146055,
+      "rewards/accuracy_reward": 0.3270089436322451,
+      "rewards/format_reward": 0.873549147695303,
+      "step": 80
+    },
+    {
+      "completion_length": 333.57389736175537,
+      "epoch": 0.3003533568904594,
+      "grad_norm": 0.5888619422912598,
+      "kl": 0.4383544921875,
+      "learning_rate": 1.7695663189185703e-05,
+      "loss": 0.6874,
+      "reward": 1.0303571939468383,
+      "reward_std": 0.513068362697959,
+      "rewards/accuracy_reward": 0.2581473340280354,
+      "rewards/format_reward": 0.7722098581492901,
+      "step": 85
+    },
+    {
+      "completion_length": 218.5093852996826,
+      "epoch": 0.31802120141342755,
+      "grad_norm": 607.1968994140625,
+      "kl": 4.6947509765625,
+      "learning_rate": 1.7286298660705877e-05,
+      "loss": 0.9755,
+      "reward": 1.2218750618398189,
+      "reward_std": 0.365594344958663,
+      "rewards/accuracy_reward": 0.31350447833538053,
+      "rewards/format_reward": 0.9083705767989159,
+      "step": 90
+    },
+    {
+      "completion_length": 181.7823745727539,
+      "epoch": 0.33568904593639576,
+      "grad_norm": 7.599749565124512,
+      "kl": 0.5159912109375,
+      "learning_rate": 1.6849076713469914e-05,
+      "loss": 0.2169,
+      "reward": 1.2810268446803093,
+      "reward_std": 0.3074988966807723,
+      "rewards/accuracy_reward": 0.3281250137835741,
+      "rewards/format_reward": 0.952901828289032,
+      "step": 95
+    },
+    {
+      "completion_length": 175.87612476348878,
+      "epoch": 0.35335689045936397,
+      "grad_norm": 1.6138108968734741,
+      "kl": 0.40145263671875,
+      "learning_rate": 1.6385668960932143e-05,
+      "loss": 0.1372,
+      "reward": 1.2872768446803093,
+      "reward_std": 0.28329721204936503,
+      "rewards/accuracy_reward": 0.325446442887187,
+      "rewards/format_reward": 0.9618303947150707,
+      "step": 100
+    },
+    {
+      "epoch": 0.35335689045936397,
+      "eval_completion_length": 172.60817337036133,
+      "eval_kl": 2.84375,
+      "eval_loss": 0.4997411072254181,
+      "eval_reward": 1.262276828289032,
+      "eval_reward_std": 0.4010584279894829,
+      "eval_rewards/accuracy_reward": 0.3392857238650322,
+      "eval_rewards/format_reward": 0.9229911118745804,
+      "eval_runtime": 42.145,
+      "eval_samples_per_second": 2.349,
+      "eval_steps_per_second": 0.024,
+      "step": 100
+    },
+    {
+      "completion_length": 202.6909688949585,
+      "epoch": 0.3710247349823322,
+      "grad_norm": 3.8880608081817627,
+      "kl": 1.9408203125,
+      "learning_rate": 1.5897847131705194e-05,
+      "loss": 0.5156,
+      "reward": 1.182031300663948,
+      "reward_std": 0.4068946409970522,
+      "rewards/accuracy_reward": 0.2914062632247806,
+      "rewards/format_reward": 0.8906250417232513,
+      "step": 105
+    },
+    {
+      "completion_length": 234.45369472503663,
+      "epoch": 0.38869257950530034,
+      "grad_norm": 10.51020336151123,
+      "kl": 3.070166015625,
+      "learning_rate": 1.5387476295779737e-05,
+      "loss": 0.6781,
+      "reward": 1.1487723655998707,
+      "reward_std": 0.4561993557959795,
+      "rewards/accuracy_reward": 0.290513406926766,
+      "rewards/format_reward": 0.858258967846632,
+      "step": 110
+    },
+    {
+      "completion_length": 270.6211072921753,
+      "epoch": 0.40636042402826855,
+      "grad_norm": 17.52117919921875,
+      "kl": 3.108984375,
+      "learning_rate": 1.4856507733875837e-05,
+      "loss": 0.7116,
+      "reward": 1.1513393305242061,
+      "reward_std": 0.4505001582205296,
+      "rewards/accuracy_reward": 0.28861608654260634,
+      "rewards/format_reward": 0.8627232544124126,
+      "step": 115
+    },
+    {
+      "completion_length": 293.4304817199707,
+      "epoch": 0.42402826855123676,
+      "grad_norm": 3.0901668071746826,
+      "kl": 1.826953125,
+      "learning_rate": 1.4306971477188223e-05,
+      "loss": 0.6357,
+      "reward": 1.160714340209961,
+      "reward_std": 0.4445430090650916,
+      "rewards/accuracy_reward": 0.29687501220032575,
+      "rewards/format_reward": 0.863839327543974,
+      "step": 120
+    },
+    {
+      "completion_length": 293.4064878463745,
+      "epoch": 0.4416961130742049,
+      "grad_norm": 1.0616240501403809,
+      "kl": 0.759033203125,
+      "learning_rate": 1.3740968546047935e-05,
+      "loss": 0.5418,
+      "reward": 1.1334821924567222,
+      "reward_std": 0.4580180231481791,
+      "rewards/accuracy_reward": 0.2792410831898451,
+      "rewards/format_reward": 0.8542411103844643,
+      "step": 125
+    },
+    {
+      "completion_length": 187.28449478149415,
+      "epoch": 0.45936395759717313,
+      "grad_norm": 1.032782793045044,
+      "kl": 1.75234375,
+      "learning_rate": 1.3160662917174045e-05,
+      "loss": 0.2673,
+      "reward": 1.285825952887535,
+      "reward_std": 0.3237128121778369,
+      "rewards/accuracy_reward": 0.32957590855658053,
+      "rewards/format_reward": 0.9562500439584255,
+      "step": 130
+    },
+    {
+      "completion_length": 249.5670877456665,
+      "epoch": 0.47703180212014135,
+      "grad_norm": 5.672813892364502,
+      "kl": 0.9608154296875,
+      "learning_rate": 1.2568273250226681e-05,
+      "loss": 0.42,
+      "reward": 1.1869420193135738,
+      "reward_std": 0.41579487193375825,
+      "rewards/accuracy_reward": 0.30133930053561925,
+      "rewards/format_reward": 0.8856027238070965,
+      "step": 135
+    },
+    {
+      "completion_length": 251.65201930999757,
+      "epoch": 0.49469964664310956,
+      "grad_norm": 3.7146189212799072,
+      "kl": 0.66943359375,
+      "learning_rate": 1.1966064405292887e-05,
+      "loss": 0.2594,
+      "reward": 1.265178633481264,
+      "reward_std": 0.32337176529690626,
+      "rewards/accuracy_reward": 0.32433037273585796,
+      "rewards/format_reward": 0.9408482559025287,
+      "step": 140
+    },
+    {
+      "completion_length": 210.5714376449585,
+      "epoch": 0.5123674911660777,
+      "grad_norm": 2.63114333152771,
+      "kl": 0.3202392578125,
+      "learning_rate": 1.1356338783736256e-05,
+      "loss": 0.0326,
+      "reward": 1.3486607745289803,
+      "reward_std": 0.23780102925375105,
+      "rewards/accuracy_reward": 0.3533482296392322,
+      "rewards/format_reward": 0.9953125156462193,
+      "step": 145
+    },
+    {
+      "completion_length": 214.0686477661133,
+      "epoch": 0.5300353356890459,
+      "grad_norm": 1.1995089054107666,
+      "kl": 0.38431396484375,
+      "learning_rate": 1.0741427525516463e-05,
+      "loss": 0.114,
+      "reward": 1.2697545185685157,
+      "reward_std": 0.3020464600995183,
+      "rewards/accuracy_reward": 0.3006696555763483,
+      "rewards/format_reward": 0.9690848618745804,
+      "step": 150
+    },
+    {
+      "completion_length": 368.81084098815916,
+      "epoch": 0.5477031802120141,
+      "grad_norm": 6.58673095703125,
+      "kl": 1.377197265625,
+      "learning_rate": 1.012368159663363e-05,
+      "loss": 0.6451,
+      "reward": 0.9731027223169804,
+      "reward_std": 0.5320311084389686,
+      "rewards/accuracy_reward": 0.20669643813744187,
+      "rewards/format_reward": 0.7664062835276126,
+      "step": 155
+    },
+    {
+      "completion_length": 288.0265756607056,
+      "epoch": 0.5653710247349824,
+      "grad_norm": 1.0641926527023315,
+      "kl": 0.6106201171875,
+      "learning_rate": 9.505462800772612e-06,
+      "loss": 0.478,
+      "reward": 1.054352729767561,
+      "reward_std": 0.4532489124685526,
+      "rewards/accuracy_reward": 0.20569197335280479,
+      "rewards/format_reward": 0.8486607544124126,
+      "step": 160
+    },
+    {
+      "completion_length": 286.00916519165037,
+      "epoch": 0.5830388692579506,
+      "grad_norm": 0.44122907519340515,
+      "kl": 0.3580810546875,
+      "learning_rate": 8.889134749511956e-06,
+      "loss": 0.5047,
+      "reward": 1.1090402357280253,
+      "reward_std": 0.45567981265485286,
+      "rewards/accuracy_reward": 0.24263394130393862,
+      "rewards/format_reward": 0.8664062894880772,
+      "step": 165
+    },
+    {
+      "completion_length": 248.03639526367186,
+      "epoch": 0.6007067137809188,
+      "grad_norm": 0.31353357434272766,
+      "kl": 0.3409423828125,
+      "learning_rate": 8.277053825620836e-06,
+      "loss": 0.4022,
+      "reward": 1.20323666036129,
+      "reward_std": 0.39312053602188823,
+      "rewards/accuracy_reward": 0.2875000137835741,
+      "rewards/format_reward": 0.9157366491854191,
+      "step": 170
+    },
+    {
+      "completion_length": 187.5467725753784,
+      "epoch": 0.6183745583038869,
+      "grad_norm": 33.15998458862305,
+      "kl": 0.42301025390625,
+      "learning_rate": 7.671560173993588e-06,
+      "loss": 0.1912,
+      "reward": 1.291406311094761,
+      "reward_std": 0.2932738855481148,
+      "rewards/accuracy_reward": 0.3219866223633289,
+      "rewards/format_reward": 0.9694196812808513,
+      "step": 175
+    },
+    {
+      "completion_length": 185.00536499023437,
+      "epoch": 0.6360424028268551,
+      "grad_norm": 0.5558907985687256,
+      "kl": 0.245166015625,
+      "learning_rate": 7.07496875466589e-06,
+      "loss": 0.1924,
+      "reward": 1.2816964894533158,
+      "reward_std": 0.2955435009673238,
+      "rewards/accuracy_reward": 0.31372769344598056,
+      "rewards/format_reward": 0.9679687887430191,
+      "step": 180
+    },
+    {
+      "completion_length": 263.3887405395508,
+      "epoch": 0.6537102473498233,
+      "grad_norm": 0.36499133706092834,
+      "kl": 0.28150634765625,
+      "learning_rate": 6.489560492119225e-06,
+      "loss": 0.412,
+      "reward": 1.1849330902099608,
+      "reward_std": 0.40096613820642235,
+      "rewards/accuracy_reward": 0.29218751322478054,
+      "rewards/format_reward": 0.8927455753087997,
+      "step": 185
+    },
+    {
+      "completion_length": 304.0499008178711,
+      "epoch": 0.6713780918727915,
+      "grad_norm": 0.24227161705493927,
+      "kl": 0.26263427734375,
+      "learning_rate": 5.9175735547120975e-06,
+      "loss": 0.4397,
+      "reward": 1.183147370815277,
+      "reward_std": 0.44423861615359783,
+      "rewards/accuracy_reward": 0.3106026917695999,
+      "rewards/format_reward": 0.8725446827709675,
+      "step": 190
+    },
+    {
+      "completion_length": 281.2084943771362,
+      "epoch": 0.6890459363957597,
+      "grad_norm": 0.9491950273513794,
+      "kl": 0.2338623046875,
+      "learning_rate": 5.361194797579108e-06,
+      "loss": 0.2874,
+      "reward": 1.2875000551342963,
+      "reward_std": 0.3527799446135759,
+      "rewards/accuracy_reward": 0.3587053744122386,
+      "rewards/format_reward": 0.9287946857511997,
+      "step": 195
+    },
+    {
+      "completion_length": 279.403472328186,
+      "epoch": 0.7067137809187279,
+      "grad_norm": 0.24906866252422333,
+      "kl": 0.18758544921875,
+      "learning_rate": 4.8225514017138205e-06,
+      "loss": 0.226,
+      "reward": 1.3204241558909415,
+      "reward_std": 0.3341758308932185,
+      "rewards/accuracy_reward": 0.37712055407464506,
+      "rewards/format_reward": 0.9433036096394062,
+      "step": 200
+    },
+    {
+      "epoch": 0.7067137809187279,
+      "eval_completion_length": 290.95755767822266,
+      "eval_kl": 0.213623046875,
+      "eval_loss": 0.34684792160987854,
+      "eval_reward": 1.3035714626312256,
+      "eval_reward_std": 0.37863466888666153,
+      "eval_rewards/accuracy_reward": 0.3816964477300644,
+      "eval_rewards/format_reward": 0.9218750447034836,
+      "eval_runtime": 52.6337,
+      "eval_samples_per_second": 1.881,
+      "eval_steps_per_second": 0.019,
+      "step": 200
+    },
+    {
+      "completion_length": 314.0254592895508,
+      "epoch": 0.7243816254416962,
+      "grad_norm": 0.4518482983112335,
+      "kl": 0.26624755859375,
+      "learning_rate": 4.303702741201431e-06,
+      "loss": 0.3849,
+      "reward": 1.212611661851406,
+      "reward_std": 0.4183976389467716,
+      "rewards/accuracy_reward": 0.32354912124574187,
+      "rewards/format_reward": 0.8890625379979611,
+      "step": 205
+    },
+    {
+      "completion_length": 338.1247905731201,
+      "epoch": 0.7420494699646644,
+      "grad_norm": 0.29115650057792664,
+      "kl": 0.3578857421875,
+      "learning_rate": 3.8066325096949153e-06,
+      "loss": 0.4956,
+      "reward": 1.148772369325161,
+      "reward_std": 0.4888919793069363,
+      "rewards/accuracy_reward": 0.31272322945296765,
+      "rewards/format_reward": 0.8360491462051869,
+      "step": 210
+    },
+    {
+      "completion_length": 337.40280590057375,
+      "epoch": 0.7597173144876325,
+      "grad_norm": 0.4650420546531677,
+      "kl": 0.34840087890625,
+      "learning_rate": 3.3332411362372063e-06,
+      "loss": 0.5048,
+      "reward": 1.1222098745405673,
+      "reward_std": 0.4806524086743593,
+      "rewards/accuracy_reward": 0.29397322703152895,
+      "rewards/format_reward": 0.8282366462051869,
+      "step": 215
+    },
+    {
+      "completion_length": 303.6270210266113,
+      "epoch": 0.7773851590106007,
+      "grad_norm": 0.475253701210022,
+      "kl": 0.33944091796875,
+      "learning_rate": 2.8853385194256677e-06,
+      "loss": 0.4854,
+      "reward": 1.1648438036441804,
+      "reward_std": 0.4518149187788367,
+      "rewards/accuracy_reward": 0.3044643012806773,
+      "rewards/format_reward": 0.8603795073926449,
+      "step": 220
+    },
+    {
+      "completion_length": 294.54075050354004,
+      "epoch": 0.7950530035335689,
+      "grad_norm": 1.6670526266098022,
+      "kl": 0.4580810546875,
+      "learning_rate": 2.464637107698046e-06,
+      "loss": 0.4903,
+      "reward": 1.1580357693135739,
+      "reward_std": 0.44512549713253974,
+      "rewards/accuracy_reward": 0.2974330489523709,
+      "rewards/format_reward": 0.860602717846632,
+      "step": 225
+    },
+    {
+      "completion_length": 275.87791481018064,
+      "epoch": 0.8127208480565371,
+      "grad_norm": 1.4973307847976685,
+      "kl": 0.37625732421875,
+      "learning_rate": 2.072745352195794e-06,
+      "loss": 0.4606,
+      "reward": 1.1925223790109158,
+      "reward_std": 0.4279437720775604,
+      "rewards/accuracy_reward": 0.31104912217706443,
+      "rewards/format_reward": 0.8814732596278191,
+      "step": 230
+    },
+    {
+      "completion_length": 251.60960941314698,
+      "epoch": 0.8303886925795053,
+      "grad_norm": 30.014623641967773,
+      "kl": 0.34725341796875,
+      "learning_rate": 1.7111615572361628e-06,
+      "loss": 0.3987,
+      "reward": 1.243750061094761,
+      "reward_std": 0.41131251212209463,
+      "rewards/accuracy_reward": 0.33627233747392893,
+      "rewards/format_reward": 0.9074777193367481,
+      "step": 235
+    },
+    {
+      "completion_length": 248.69844818115234,
+      "epoch": 0.8480565371024735,
+      "grad_norm": 54.57441711425781,
+      "kl": 1.3016845703125,
+      "learning_rate": 1.381268151904298e-06,
+      "loss": 0.5182,
+      "reward": 1.236607201397419,
+      "reward_std": 0.40282299667596816,
+      "rewards/accuracy_reward": 0.3315848369151354,
+      "rewards/format_reward": 0.9050223655998707,
+      "step": 240
+    },
+    {
+      "completion_length": 247.95547924041747,
+      "epoch": 0.8657243816254417,
+      "grad_norm": 167.22845458984375,
+      "kl": 6.20146484375,
+      "learning_rate": 1.0843264046665558e-06,
+      "loss": 1.1488,
+      "reward": 1.2164063028991223,
+      "reward_std": 0.3988880608230829,
+      "rewards/accuracy_reward": 0.31004465725272895,
+      "rewards/format_reward": 0.9063616514205932,
+      "step": 245
+    },
+    {
+      "completion_length": 246.49398460388184,
+      "epoch": 0.8833922261484098,
+      "grad_norm": 48.688541412353516,
+      "kl": 1.783447265625,
+      "learning_rate": 8.214716012124491e-07,
+      "loss": 0.5749,
+      "reward": 1.2148438021540642,
+      "reward_std": 0.38329584859311583,
+      "rewards/accuracy_reward": 0.30535715632140636,
+      "rewards/format_reward": 0.9094866514205933,
+      "step": 250
+    },
+    {
+      "completion_length": 239.5714382171631,
+      "epoch": 0.901060070671378,
+      "grad_norm": 47.49518966674805,
+      "kl": 0.770556640625,
+      "learning_rate": 5.937087039615619e-07,
+      "loss": 0.4138,
+      "reward": 1.2323661297559738,
+      "reward_std": 0.3774874061346054,
+      "rewards/accuracy_reward": 0.3184151927940547,
+      "rewards/format_reward": 0.9139509372413158,
+      "step": 255
+    },
+    {
+      "completion_length": 249.4630693435669,
+      "epoch": 0.9187279151943463,
+      "grad_norm": 25.54094886779785,
+      "kl": 1.382958984375,
+      "learning_rate": 4.019085098303077e-07,
+      "loss": 0.5233,
+      "reward": 1.21964291036129,
+      "reward_std": 0.39937015902251005,
+      "rewards/accuracy_reward": 0.31763394493609665,
+      "rewards/format_reward": 0.9020089708268643,
+      "step": 260
+    },
+    {
+      "completion_length": 246.59643955230712,
+      "epoch": 0.9363957597173145,
+      "grad_norm": 25.444677352905273,
+      "kl": 2.867236328125,
+      "learning_rate": 2.4680432094837394e-07,
+      "loss": 0.6682,
+      "reward": 1.2183036252856254,
+      "reward_std": 0.40516539234668014,
+      "rewards/accuracy_reward": 0.31696430016309024,
+      "rewards/format_reward": 0.901339328289032,
+      "step": 265
+    },
+    {
+      "completion_length": 249.1561502456665,
+      "epoch": 0.9540636042402827,
+      "grad_norm": 5.572141647338867,
+      "kl": 2.645947265625,
+      "learning_rate": 1.289891410535593e-07,
+      "loss": 0.6572,
+      "reward": 1.2252232685685158,
+      "reward_std": 0.40847705118358135,
+      "rewards/accuracy_reward": 0.32165180034935476,
+      "rewards/format_reward": 0.9035714715719223,
+      "step": 270
+    },
+    {
+      "completion_length": 238.4615062713623,
+      "epoch": 0.9717314487632509,
+      "grad_norm": 17.332447052001953,
+      "kl": 1.982080078125,
+      "learning_rate": 4.8913408283934874e-08,
+      "loss": 0.5867,
+      "reward": 1.2314732655882836,
+      "reward_std": 0.38298603780567647,
+      "rewards/accuracy_reward": 0.32265626490116117,
+      "rewards/format_reward": 0.9088170036673546,
+      "step": 275
+    },
+    {
+      "completion_length": 242.24800186157228,
+      "epoch": 0.9893992932862191,
+      "grad_norm": 13.555401802062988,
+      "kl": 1.88583984375,
+      "learning_rate": 6.883273035447335e-09,
+      "loss": 0.5522,
+      "reward": 1.2324777357280254,
+      "reward_std": 0.3977079253643751,
+      "rewards/accuracy_reward": 0.32622769437730315,
+      "rewards/format_reward": 0.9062500409781933,
+      "step": 280
+    },
+    {
+      "completion_length": 240.61751333872476,
+      "epoch": 1.0,
+      "kl": 1.9344075520833333,
+      "reward": 1.2202381441990535,
+      "reward_std": 0.40573689217368764,
+      "rewards/accuracy_reward": 0.31733632429192465,
+      "rewards/format_reward": 0.9029018295307955,
+      "step": 283,
+      "total_flos": 0.0,
+      "train_loss": 0.40895458704104165,
+      "train_runtime": 40740.9033,
+      "train_samples_per_second": 1.778,
+      "train_steps_per_second": 0.007
+    }
+  ],
+  "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": 16,
+  "trial_name": null,
+  "trial_params": null
+}

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:def16a7ff18f84fbf8e4c85cb30a177896a17fd1c0a5d2120cc9ca3b7605d8cc
+size 7608