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: Qwen2.5-1.5B-Math-R1-GRPO-Three-Reward
+tags:
+- generated_from_trainer
+- trl
+- grpo
+licence: license
+---
+
+# Model Card for Qwen2.5-1.5B-Math-R1-GRPO-Three-Reward
+
+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="Xuchen-Li/Qwen2.5-1.5B-Math-R1-GRPO-Three-Reward", device="cuda")
+output = generator([{"role": "user", "content": question}], max_new_tokens=128, return_full_text=False)[0]
+print(output["generated_text"])
+```
+
+## Training procedure
+
+ 
+
+
+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
+- Transformers: 4.48.2
+- Pytorch: 2.5.1
+- Datasets: 3.3.1
+- 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": 0.01265178020875292,
+    "train_runtime": 42016.9225,
+    "train_samples": 72441,
+    "train_samples_per_second": 1.724,
+    "train_steps_per_second": 0.007
+}

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": null,
+  "tie_word_embeddings": true,
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.48.2",
+  "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": "4.48.2"
+}

model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:41fde4dda35e8d00340c4159fc72a57dac5efbf102d44de1c4b21a4dcaa5d9ac
+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 sha256:5eee858c5123a4279c3e1f7b81247343f356ac767940b2692a928ad929543214
+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        {{- '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": 0.01265178020875292,
+    "train_runtime": 42016.9225,
+    "train_samples": 72441,
+    "train_samples_per_second": 1.724,
+    "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": 622.866654586792,
+      "epoch": 0.0176678445229682,
+      "grad_norm": 0.02329571545124054,
+      "kl": 3.776401281356812e-05,
+      "learning_rate": 3.448275862068966e-06,
+      "loss": 0.0,
+      "reward": 0.3915178759023547,
+      "reward_std": 0.34093408063054087,
+      "rewards/accuracy_reward": 0.3902901979163289,
+      "rewards/format_reward": 0.0012276786263100802,
+      "step": 5
+    },
+    {
+      "completion_length": 634.6379760742187,
+      "epoch": 0.0353356890459364,
+      "grad_norm": 0.0056246621534228325,
+      "kl": 0.0015024781227111817,
+      "learning_rate": 6.896551724137932e-06,
+      "loss": 0.0001,
+      "reward": 0.4254464492201805,
+      "reward_std": 0.3254653673619032,
+      "rewards/accuracy_reward": 0.42421876937150954,
+      "rewards/format_reward": 0.0012276786263100802,
+      "step": 10
+    },
+    {
+      "completion_length": 647.5926582336426,
+      "epoch": 0.053003533568904596,
+      "grad_norm": 0.004786658100783825,
+      "kl": 0.027044105529785156,
+      "learning_rate": 1.0344827586206898e-05,
+      "loss": 0.0011,
+      "reward": 0.5095982380211354,
+      "reward_std": 0.31292589046061037,
+      "rewards/accuracy_reward": 0.509375024586916,
+      "rewards/format_reward": 0.00022321429569274187,
+      "step": 15
+    },
+    {
+      "completion_length": 645.3291557312011,
+      "epoch": 0.0706713780918728,
+      "grad_norm": 0.00284953904338181,
+      "kl": 0.008718490600585938,
+      "learning_rate": 1.3793103448275863e-05,
+      "loss": 0.0003,
+      "reward": 0.5450893104076385,
+      "reward_std": 0.29312025886029003,
+      "rewards/accuracy_reward": 0.5449777029454708,
+      "rewards/format_reward": 0.00011160714784637093,
+      "step": 20
+    },
+    {
+      "completion_length": 645.8677711486816,
+      "epoch": 0.08833922261484099,
+      "grad_norm": 0.002752735046669841,
+      "kl": 0.01061248779296875,
+      "learning_rate": 1.7241379310344828e-05,
+      "loss": 0.0004,
+      "reward": 0.5610491331666708,
+      "reward_std": 0.28372098952531816,
+      "rewards/accuracy_reward": 0.5610491331666708,
+      "rewards/format_reward": 0.0,
+      "step": 25
+    },
+    {
+      "completion_length": 641.2229057312012,
+      "epoch": 0.10600706713780919,
+      "grad_norm": 0.0027106315828859806,
+      "kl": 0.011400985717773437,
+      "learning_rate": 1.999923511388017e-05,
+      "loss": 0.0005,
+      "reward": 0.5921875219792128,
+      "reward_std": 0.263329023681581,
+      "rewards/accuracy_reward": 0.5921875219792128,
+      "rewards/format_reward": 0.0,
+      "step": 30
+    },
+    {
+      "completion_length": 630.1242469787597,
+      "epoch": 0.12367491166077739,
+      "grad_norm": 0.002474562032148242,
+      "kl": 0.022439193725585938,
+      "learning_rate": 1.9972476383747748e-05,
+      "loss": 0.0009,
+      "reward": 0.5995536006987094,
+      "reward_std": 0.2560196537524462,
+      "rewards/accuracy_reward": 0.5995536006987094,
+      "rewards/format_reward": 0.0,
+      "step": 35
+    },
+    {
+      "completion_length": 618.7995788574219,
+      "epoch": 0.1413427561837456,
+      "grad_norm": 0.002974502742290497,
+      "kl": 0.020103073120117186,
+      "learning_rate": 1.9907590277344582e-05,
+      "loss": 0.0008,
+      "reward": 0.6072544913738966,
+      "reward_std": 0.24586861627176404,
+      "rewards/accuracy_reward": 0.6072544913738966,
+      "rewards/format_reward": 0.0,
+      "step": 40
+    },
+    {
+      "completion_length": 628.1982410430908,
+      "epoch": 0.15901060070671377,
+      "grad_norm": 0.004846589639782906,
+      "kl": 0.041345977783203126,
+      "learning_rate": 1.9804824871166254e-05,
+      "loss": 0.0017,
+      "reward": 0.5873884197324515,
+      "reward_std": 0.2790112975984812,
+      "rewards/accuracy_reward": 0.5873884197324515,
+      "rewards/format_reward": 0.0,
+      "step": 45
+    },
+    {
+      "completion_length": 611.4961177825928,
+      "epoch": 0.17667844522968199,
+      "grad_norm": 0.003141767578199506,
+      "kl": 0.0728302001953125,
+      "learning_rate": 1.9664573064143604e-05,
+      "loss": 0.0029,
+      "reward": 0.5887277036905288,
+      "reward_std": 0.27428281418979167,
+      "rewards/accuracy_reward": 0.5887277036905288,
+      "rewards/format_reward": 0.0,
+      "step": 50
+    },
+    {
+      "completion_length": 607.3044876098633,
+      "epoch": 0.19434628975265017,
+      "grad_norm": 0.003312925109639764,
+      "kl": 0.0822174072265625,
+      "learning_rate": 1.948737107548771e-05,
+      "loss": 0.0033,
+      "reward": 0.5859375264495611,
+      "reward_std": 0.24934234898537397,
+      "rewards/accuracy_reward": 0.5859375264495611,
+      "rewards/format_reward": 0.0,
+      "step": 55
+    },
+    {
+      "completion_length": 629.2710105895997,
+      "epoch": 0.21201413427561838,
+      "grad_norm": 0.38133352994918823,
+      "kl": 0.1975006103515625,
+      "learning_rate": 1.9273896394584103e-05,
+      "loss": 0.0079,
+      "reward": 0.550334845110774,
+      "reward_std": 0.27198409736156465,
+      "rewards/accuracy_reward": 0.550334845110774,
+      "rewards/format_reward": 0.0,
+      "step": 60
+    },
+    {
+      "completion_length": 609.5240196228027,
+      "epoch": 0.22968197879858657,
+      "grad_norm": 0.006227842066437006,
+      "kl": 1.1461837768554688,
+      "learning_rate": 1.9024965190774262e-05,
+      "loss": 0.0458,
+      "reward": 0.602678595855832,
+      "reward_std": 0.2419098662212491,
+      "rewards/accuracy_reward": 0.602678595855832,
+      "rewards/format_reward": 0.0,
+      "step": 65
+    },
+    {
+      "completion_length": 607.7145351409912,
+      "epoch": 0.24734982332155478,
+      "grad_norm": 0.0062450082041323185,
+      "kl": 0.0861968994140625,
+      "learning_rate": 1.8741529192927528e-05,
+      "loss": 0.0034,
+      "reward": 0.5861607410013676,
+      "reward_std": 0.24603348365053535,
+      "rewards/accuracy_reward": 0.5861607410013676,
+      "rewards/format_reward": 0.0,
+      "step": 70
+    },
+    {
+      "completion_length": 606.3534866333008,
+      "epoch": 0.26501766784452296,
+      "grad_norm": 0.005692186299711466,
+      "kl": 0.1029083251953125,
+      "learning_rate": 1.8424672050733577e-05,
+      "loss": 0.0041,
+      "reward": 0.5786830626428128,
+      "reward_std": 0.25585202854126693,
+      "rewards/accuracy_reward": 0.5786830626428128,
+      "rewards/format_reward": 0.0,
+      "step": 75
+    },
+    {
+      "completion_length": 621.4455604553223,
+      "epoch": 0.2826855123674912,
+      "grad_norm": 0.007482562679797411,
+      "kl": 0.1179046630859375,
+      "learning_rate": 1.8075605191627242e-05,
+      "loss": 0.0047,
+      "reward": 0.5845982395112514,
+      "reward_std": 0.2532924499362707,
+      "rewards/accuracy_reward": 0.5845982395112514,
+      "rewards/format_reward": 0.0,
+      "step": 80
+    },
+    {
+      "completion_length": 628.4788261413574,
+      "epoch": 0.3003533568904594,
+      "grad_norm": 0.012330959551036358,
+      "kl": 0.1165985107421875,
+      "learning_rate": 1.7695663189185703e-05,
+      "loss": 0.0047,
+      "reward": 0.5766741320490837,
+      "reward_std": 0.2558022084645927,
+      "rewards/accuracy_reward": 0.5766741320490837,
+      "rewards/format_reward": 0.0,
+      "step": 85
+    },
+    {
+      "completion_length": 617.1456726074218,
+      "epoch": 0.31802120141342755,
+      "grad_norm": 0.07277228683233261,
+      "kl": 0.302655029296875,
+      "learning_rate": 1.7286298660705877e-05,
+      "loss": 0.0121,
+      "reward": 0.5973214555531741,
+      "reward_std": 0.23560365894809365,
+      "rewards/accuracy_reward": 0.5973214555531741,
+      "rewards/format_reward": 0.0,
+      "step": 90
+    },
+    {
+      "completion_length": 601.8499160766602,
+      "epoch": 0.33568904593639576,
+      "grad_norm": 0.008128706365823746,
+      "kl": 0.9103607177734375,
+      "learning_rate": 1.6849076713469914e-05,
+      "loss": 0.0364,
+      "reward": 0.5960937775671482,
+      "reward_std": 0.2516948331147432,
+      "rewards/accuracy_reward": 0.5960937775671482,
+      "rewards/format_reward": 0.0,
+      "step": 95
+    },
+    {
+      "completion_length": 613.2512561798096,
+      "epoch": 0.35335689045936397,
+      "grad_norm": 0.010601451620459557,
+      "kl": 0.1064056396484375,
+      "learning_rate": 1.6385668960932143e-05,
+      "loss": 0.0043,
+      "reward": 0.5683035928755998,
+      "reward_std": 0.25203096950426696,
+      "rewards/accuracy_reward": 0.5683035928755998,
+      "rewards/format_reward": 0.0,
+      "step": 100
+    },
+    {
+      "epoch": 0.35335689045936397,
+      "eval_completion_length": 621.3350830078125,
+      "eval_kl": 0.1502685546875,
+      "eval_loss": 0.007704772986471653,
+      "eval_reward": 0.573660746216774,
+      "eval_reward_std": 0.2696436755359173,
+      "eval_rewards/accuracy_reward": 0.573660746216774,
+      "eval_rewards/format_reward": 0.0,
+      "eval_runtime": 56.6658,
+      "eval_samples_per_second": 1.747,
+      "eval_steps_per_second": 0.018,
+      "step": 100
+    },
+    {
+      "completion_length": 605.3703407287597,
+      "epoch": 0.3710247349823322,
+      "grad_norm": 0.016205137595534325,
+      "kl": 0.16383056640625,
+      "learning_rate": 1.5897847131705194e-05,
+      "loss": 0.0066,
+      "reward": 0.5819196697324515,
+      "reward_std": 0.2612144500948489,
+      "rewards/accuracy_reward": 0.5819196697324515,
+      "rewards/format_reward": 0.0,
+      "step": 105
+    },
+    {
+      "completion_length": 602.7196727752686,
+      "epoch": 0.38869257950530034,
+      "grad_norm": 0.03076363541185856,
+      "kl": 0.348724365234375,
+      "learning_rate": 1.5387476295779737e-05,
+      "loss": 0.014,
+      "reward": 0.5632812755182386,
+      "reward_std": 0.25463338308036326,
+      "rewards/accuracy_reward": 0.5632812755182386,
+      "rewards/format_reward": 0.0,
+      "step": 110
+    },
+    {
+      "completion_length": 605.1398712158203,
+      "epoch": 0.40636042402826855,
+      "grad_norm": 0.041805580258369446,
+      "kl": 0.9538330078125,
+      "learning_rate": 1.4856507733875837e-05,
+      "loss": 0.0382,
+      "reward": 0.5839285969734191,
+      "reward_std": 0.24314276818186045,
+      "rewards/accuracy_reward": 0.5839285969734191,
+      "rewards/format_reward": 0.0,
+      "step": 115
+    },
+    {
+      "completion_length": 610.0774848937988,
+      "epoch": 0.42402826855123676,
+      "grad_norm": 0.07552725076675415,
+      "kl": 0.90081787109375,
+      "learning_rate": 1.4306971477188223e-05,
+      "loss": 0.036,
+      "reward": 0.5680803816765547,
+      "reward_std": 0.2557636009529233,
+      "rewards/accuracy_reward": 0.5680803816765547,
+      "rewards/format_reward": 0.0,
+      "step": 120
+    },
+    {
+      "completion_length": 613.6580638885498,
+      "epoch": 0.4416961130742049,
+      "grad_norm": 0.02007584273815155,
+      "kl": 0.47030029296875,
+      "learning_rate": 1.3740968546047935e-05,
+      "loss": 0.0188,
+      "reward": 0.5672991309314966,
+      "reward_std": 0.26936373468488456,
+      "rewards/accuracy_reward": 0.5672991309314966,
+      "rewards/format_reward": 0.0,
+      "step": 125
+    },
+    {
+      "completion_length": 601.1768115997314,
+      "epoch": 0.45936395759717313,
+      "grad_norm": 0.03517204895615578,
+      "kl": 0.3914794921875,
+      "learning_rate": 1.3160662917174045e-05,
+      "loss": 0.0157,
+      "reward": 0.5974330645054579,
+      "reward_std": 0.24408907871693372,
+      "rewards/accuracy_reward": 0.5974330645054579,
+      "rewards/format_reward": 0.0,
+      "step": 130
+    },
+    {
+      "completion_length": 592.4332832336426,
+      "epoch": 0.47703180212014135,
+      "grad_norm": 0.017994452267885208,
+      "kl": 0.313238525390625,
+      "learning_rate": 1.2568273250226681e-05,
+      "loss": 0.0125,
+      "reward": 0.6005580626428128,
+      "reward_std": 0.23314145486801863,
+      "rewards/accuracy_reward": 0.6005580626428128,
+      "rewards/format_reward": 0.0,
+      "step": 135
+    },
+    {
+      "completion_length": 606.3159889221191,
+      "epoch": 0.49469964664310956,
+      "grad_norm": 0.029751351103186607,
+      "kl": 0.370184326171875,
+      "learning_rate": 1.1966064405292887e-05,
+      "loss": 0.0148,
+      "reward": 0.6023437790572643,
+      "reward_std": 0.24306385917589068,
+      "rewards/accuracy_reward": 0.6023437790572643,
+      "rewards/format_reward": 0.0,
+      "step": 140
+    },
+    {
+      "completion_length": 602.2370769500733,
+      "epoch": 0.5123674911660777,
+      "grad_norm": 0.009959203191101551,
+      "kl": 0.3168365478515625,
+      "learning_rate": 1.1356338783736256e-05,
+      "loss": 0.0127,
+      "reward": 0.5839286014437676,
+      "reward_std": 0.2592818034812808,
+      "rewards/accuracy_reward": 0.5839286014437676,
+      "rewards/format_reward": 0.0,
+      "step": 145
+    },
+    {
+      "completion_length": 597.7812774658203,
+      "epoch": 0.5300353356890459,
+      "grad_norm": 0.011643326841294765,
+      "kl": 0.3456634521484375,
+      "learning_rate": 1.0741427525516463e-05,
+      "loss": 0.0138,
+      "reward": 0.6031250298023224,
+      "reward_std": 0.23491650996729732,
+      "rewards/accuracy_reward": 0.6031250298023224,
+      "rewards/format_reward": 0.0,
+      "step": 150
+    },
+    {
+      "completion_length": 599.5233509063721,
+      "epoch": 0.5477031802120141,
+      "grad_norm": 0.007601267658174038,
+      "kl": 0.22601318359375,
+      "learning_rate": 1.012368159663363e-05,
+      "loss": 0.009,
+      "reward": 0.6027902077883482,
+      "reward_std": 0.23668794049881398,
+      "rewards/accuracy_reward": 0.6027902077883482,
+      "rewards/format_reward": 0.0,
+      "step": 155
+    },
+    {
+      "completion_length": 591.527705001831,
+      "epoch": 0.5653710247349824,
+      "grad_norm": 0.03776765987277031,
+      "kl": 0.39940185546875,
+      "learning_rate": 9.505462800772612e-06,
+      "loss": 0.016,
+      "reward": 0.612611635774374,
+      "reward_std": 0.23676070980727673,
+      "rewards/accuracy_reward": 0.612611635774374,
+      "rewards/format_reward": 0.0,
+      "step": 160
+    },
+    {
+      "completion_length": 603.9175487518311,
+      "epoch": 0.5830388692579506,
+      "grad_norm": 0.028675267472863197,
+      "kl": 0.3670166015625,
+      "learning_rate": 8.889134749511956e-06,
+      "loss": 0.0147,
+      "reward": 0.5803571697324514,
+      "reward_std": 0.24291018256917596,
+      "rewards/accuracy_reward": 0.5803571697324514,
+      "rewards/format_reward": 0.0,
+      "step": 165
+    },
+    {
+      "completion_length": 612.8111881256103,
+      "epoch": 0.6007067137809188,
+      "grad_norm": 0.0103794289752841,
+      "kl": 0.315142822265625,
+      "learning_rate": 8.277053825620836e-06,
+      "loss": 0.0126,
+      "reward": 0.5681919891387224,
+      "reward_std": 0.23275800403207542,
+      "rewards/accuracy_reward": 0.5681919891387224,
+      "rewards/format_reward": 0.0,
+      "step": 170
+    },
+    {
+      "completion_length": 595.3726829528808,
+      "epoch": 0.6183745583038869,
+      "grad_norm": 0.020144131034612656,
+      "kl": 0.2862823486328125,
+      "learning_rate": 7.671560173993588e-06,
+      "loss": 0.0115,
+      "reward": 0.5943080618977546,
+      "reward_std": 0.22843202147632838,
+      "rewards/accuracy_reward": 0.5943080618977546,
+      "rewards/format_reward": 0.0,
+      "step": 175
+    },
+    {
+      "completion_length": 591.2825000762939,
+      "epoch": 0.6360424028268551,
+      "grad_norm": 0.010619827546179295,
+      "kl": 0.30369873046875,
+      "learning_rate": 7.07496875466589e-06,
+      "loss": 0.0122,
+      "reward": 0.589397344738245,
+      "reward_std": 0.23018078347668053,
+      "rewards/accuracy_reward": 0.589397344738245,
+      "rewards/format_reward": 0.0,
+      "step": 180
+    },
+    {
+      "completion_length": 593.4808311462402,
+      "epoch": 0.6537102473498233,
+      "grad_norm": 0.009518198668956757,
+      "kl": 0.3361572265625,
+      "learning_rate": 6.489560492119225e-06,
+      "loss": 0.0134,
+      "reward": 0.5937500301748514,
+      "reward_std": 0.2376224915497005,
+      "rewards/accuracy_reward": 0.5937500301748514,
+      "rewards/format_reward": 0.0,
+      "step": 185
+    },
+    {
+      "completion_length": 592.2255844116211,
+      "epoch": 0.6713780918727915,
+      "grad_norm": 0.009495066478848457,
+      "kl": 0.364111328125,
+      "learning_rate": 5.9175735547120975e-06,
+      "loss": 0.0146,
+      "reward": 0.5970982395112514,
+      "reward_std": 0.24990436108782887,
+      "rewards/accuracy_reward": 0.5970982395112514,
+      "rewards/format_reward": 0.0,
+      "step": 190
+    },
+    {
+      "completion_length": 595.8637493133544,
+      "epoch": 0.6890459363957597,
+      "grad_norm": 0.015603470616042614,
+      "kl": 0.38140869140625,
+      "learning_rate": 5.361194797579108e-06,
+      "loss": 0.0153,
+      "reward": 0.5825893163681031,
+      "reward_std": 0.23760683508589864,
+      "rewards/accuracy_reward": 0.5825893163681031,
+      "rewards/format_reward": 0.0,
+      "step": 195
+    },
+    {
+      "completion_length": 591.4626392364502,
+      "epoch": 0.7067137809187279,
+      "grad_norm": 0.01287364587187767,
+      "kl": 0.37381591796875,
+      "learning_rate": 4.8225514017138205e-06,
+      "loss": 0.0149,
+      "reward": 0.6026785977184772,
+      "reward_std": 0.24596040640026332,
+      "rewards/accuracy_reward": 0.6026785977184772,
+      "rewards/format_reward": 0.0,
+      "step": 200
+    },
+    {
+      "epoch": 0.7067137809187279,
+      "eval_completion_length": 583.0498352050781,
+      "eval_kl": 0.4853515625,
+      "eval_loss": 0.022003287449479103,
+      "eval_reward": 0.6037946790456772,
+      "eval_reward_std": 0.2356761023402214,
+      "eval_rewards/accuracy_reward": 0.6037946790456772,
+      "eval_rewards/format_reward": 0.0,
+      "eval_runtime": 55.7696,
+      "eval_samples_per_second": 1.775,
+      "eval_steps_per_second": 0.018,
+      "step": 200
+    },
+    {
+      "completion_length": 603.030270767212,
+      "epoch": 0.7243816254416962,
+      "grad_norm": 0.030599776655435562,
+      "kl": 0.4007110595703125,
+      "learning_rate": 4.303702741201431e-06,
+      "loss": 0.016,
+      "reward": 0.5718750260770321,
+      "reward_std": 0.23691606232896448,
+      "rewards/accuracy_reward": 0.5718750260770321,
+      "rewards/format_reward": 0.0,
+      "step": 205
+    },
+    {
+      "completion_length": 599.5367473602295,
+      "epoch": 0.7420494699646644,
+      "grad_norm": 0.01362893171608448,
+      "kl": 0.3363525390625,
+      "learning_rate": 3.8066325096949153e-06,
+      "loss": 0.0135,
+      "reward": 0.5901785992085934,
+      "reward_std": 0.2412101237103343,
+      "rewards/accuracy_reward": 0.5901785992085934,
+      "rewards/format_reward": 0.0,
+      "step": 210
+    },
+    {
+      "completion_length": 591.5522598266601,
+      "epoch": 0.7597173144876325,
+      "grad_norm": 0.009048458188772202,
+      "kl": 0.44503173828125,
+      "learning_rate": 3.3332411362372063e-06,
+      "loss": 0.0178,
+      "reward": 0.5824777066707612,
+      "reward_std": 0.23562156101688742,
+      "rewards/accuracy_reward": 0.5824777066707612,
+      "rewards/format_reward": 0.0,
+      "step": 215
+    },
+    {
+      "completion_length": 587.6553813934327,
+      "epoch": 0.7773851590106007,
+      "grad_norm": 0.02096477337181568,
+      "kl": 0.331298828125,
+      "learning_rate": 2.8853385194256677e-06,
+      "loss": 0.0133,
+      "reward": 0.5870535984635353,
+      "reward_std": 0.24385144235566258,
+      "rewards/accuracy_reward": 0.5870535984635353,
+      "rewards/format_reward": 0.0,
+      "step": 220
+    },
+    {
+      "completion_length": 597.7669910430908,
+      "epoch": 0.7950530035335689,
+      "grad_norm": 0.06585746258497238,
+      "kl": 0.42650146484375,
+      "learning_rate": 2.464637107698046e-06,
+      "loss": 0.0171,
+      "reward": 0.5850446704775095,
+      "reward_std": 0.24386880425736307,
+      "rewards/accuracy_reward": 0.5850446704775095,
+      "rewards/format_reward": 0.0,
+      "step": 225
+    },
+    {
+      "completion_length": 589.5904312133789,
+      "epoch": 0.8127208480565371,
+      "grad_norm": 0.012867830693721771,
+      "kl": 0.3681884765625,
+      "learning_rate": 2.072745352195794e-06,
+      "loss": 0.0147,
+      "reward": 0.5989955630153417,
+      "reward_std": 0.241961353039369,
+      "rewards/accuracy_reward": 0.5989955630153417,
+      "rewards/format_reward": 0.0,
+      "step": 230
+    },
+    {
+      "completion_length": 587.985852432251,
+      "epoch": 0.8303886925795053,
+      "grad_norm": 0.010206163860857487,
+      "kl": 0.356939697265625,
+      "learning_rate": 1.7111615572361628e-06,
+      "loss": 0.0143,
+      "reward": 0.5984375305473805,
+      "reward_std": 0.2334775916300714,
+      "rewards/accuracy_reward": 0.5984375305473805,
+      "rewards/format_reward": 0.0,
+      "step": 235
+    },
+    {
+      "completion_length": 592.4740253448487,
+      "epoch": 0.8480565371024735,
+      "grad_norm": 0.01556379348039627,
+      "kl": 0.39150390625,
+      "learning_rate": 1.381268151904298e-06,
+      "loss": 0.0157,
+      "reward": 0.5761160954833031,
+      "reward_std": 0.23502092063426971,
+      "rewards/accuracy_reward": 0.5761160954833031,
+      "rewards/format_reward": 0.0,
+      "step": 240
+    },
+    {
+      "completion_length": 592.494669342041,
+      "epoch": 0.8657243816254417,
+      "grad_norm": 0.00771762290969491,
+      "kl": 0.3992919921875,
+      "learning_rate": 1.0843264046665558e-06,
+      "loss": 0.016,
+      "reward": 0.5853794906288385,
+      "reward_std": 0.2298426641151309,
+      "rewards/accuracy_reward": 0.5853794906288385,
+      "rewards/format_reward": 0.0,
+      "step": 245
+    },
+    {
+      "completion_length": 596.9780406951904,
+      "epoch": 0.8833922261484098,
+      "grad_norm": 0.01036030612885952,
+      "kl": 0.379351806640625,
+      "learning_rate": 8.214716012124491e-07,
+      "loss": 0.0152,
+      "reward": 0.5733259174972772,
+      "reward_std": 0.24230954172089697,
+      "rewards/accuracy_reward": 0.5733259174972772,
+      "rewards/format_reward": 0.0,
+      "step": 250
+    },
+    {
+      "completion_length": 598.5827262878418,
+      "epoch": 0.901060070671378,
+      "grad_norm": 0.01498055923730135,
+      "kl": 0.422900390625,
+      "learning_rate": 5.937087039615619e-07,
+      "loss": 0.0169,
+      "reward": 0.5680803801864386,
+      "reward_std": 0.2425838416442275,
+      "rewards/accuracy_reward": 0.5680803801864386,
+      "rewards/format_reward": 0.0,
+      "step": 255
+    },
+    {
+      "completion_length": 586.830269241333,
+      "epoch": 0.9187279151943463,
+      "grad_norm": 0.010074852034449577,
+      "kl": 0.3983642578125,
+      "learning_rate": 4.019085098303077e-07,
+      "loss": 0.0159,
+      "reward": 0.5914062775671483,
+      "reward_std": 0.2368315144442022,
+      "rewards/accuracy_reward": 0.5914062775671483,
+      "rewards/format_reward": 0.0,
+      "step": 260
+    },
+    {
+      "completion_length": 598.5914325714111,
+      "epoch": 0.9363957597173145,
+      "grad_norm": 0.010618208907544613,
+      "kl": 0.42635498046875,
+      "learning_rate": 2.4680432094837394e-07,
+      "loss": 0.0171,
+      "reward": 0.5768973477184772,
+      "reward_std": 0.24664643714204432,
+      "rewards/accuracy_reward": 0.5768973477184772,
+      "rewards/format_reward": 0.0,
+      "step": 265
+    },
+    {
+      "completion_length": 599.6976825714112,
+      "epoch": 0.9540636042402827,
+      "grad_norm": 0.009463882073760033,
+      "kl": 0.42564697265625,
+      "learning_rate": 1.289891410535593e-07,
+      "loss": 0.017,
+      "reward": 0.5652902007102967,
+      "reward_std": 0.25262069487944244,
+      "rewards/accuracy_reward": 0.5652902007102967,
+      "rewards/format_reward": 0.0,
+      "step": 270
+    },
+    {
+      "completion_length": 598.8724590301514,
+      "epoch": 0.9717314487632509,
+      "grad_norm": 0.011376498267054558,
+      "kl": 0.4064453125,
+      "learning_rate": 4.8913408283934874e-08,
+      "loss": 0.0163,
+      "reward": 0.5688616305589675,
+      "reward_std": 0.2400667691603303,
+      "rewards/accuracy_reward": 0.5688616305589675,
+      "rewards/format_reward": 0.0,
+      "step": 275
+    },
+    {
+      "completion_length": 592.385181427002,
+      "epoch": 0.9893992932862191,
+      "grad_norm": 0.014952524565160275,
+      "kl": 0.39835205078125,
+      "learning_rate": 6.883273035447335e-09,
+      "loss": 0.0159,
+      "reward": 0.5964285992085934,
+      "reward_std": 0.24566346323117613,
+      "rewards/accuracy_reward": 0.5964285992085934,
+      "rewards/format_reward": 0.0,
+      "step": 280
+    },
+    {
+      "completion_length": 588.4329948425293,
+      "epoch": 1.0,
+      "kl": 0.4519856770833333,
+      "reward": 0.5911458575477203,
+      "reward_std": 0.2489340848599871,
+      "rewards/accuracy_reward": 0.5911458575477203,
+      "rewards/format_reward": 0.0,
+      "step": 283,
+      "total_flos": 0.0,
+      "train_loss": 0.01265178020875292,
+      "train_runtime": 42016.9225,
+      "train_samples_per_second": 1.724,
+      "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 sha256:5b7a8fbf4b6f2c82ebff0e90f4e05962a98e2fb96763d93547d9ec3005c64c45
+size 7608