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="susumuota/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/llm-m_wandb-weblab/Qwen2.5-1.5B-GRPO/runs/8mhfzw0w) 
+
+
+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": 0.02524358952622005,
+    "train_runtime": 29066.0481,
+    "train_samples": 72441,
+    "train_samples_per_second": 2.492,
+    "train_steps_per_second": 0.01
+}

config.json

@@ -0,0 +1,29 @@
+{
+  "_name_or_path": "Qwen/Qwen2.5-1.5B-Instruct",
+  "architectures": [
+    "Qwen2ForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 151643,
+  "eos_token_id": 151645,
+  "hidden_act": "silu",
+  "hidden_size": 1536,
+  "initializer_range": 0.02,
+  "intermediate_size": 8960,
+  "max_position_embeddings": 32768,
+  "max_window_layers": 21,
+  "model_type": "qwen2",
+  "num_attention_heads": 12,
+  "num_hidden_layers": 28,
+  "num_key_value_heads": 2,
+  "rms_norm_eps": 1e-06,
+  "rope_scaling": null,
+  "rope_theta": 1000000.0,
+  "sliding_window": null,
+  "tie_word_embeddings": true,
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.49.0.dev0",
+  "use_cache": false,
+  "use_sliding_window": false,
+  "vocab_size": 151936
+}

generation_config.json

@@ -0,0 +1,14 @@
+{
+  "bos_token_id": 151643,
+  "do_sample": true,
+  "eos_token_id": [
+    151645,
+    151643
+  ],
+  "pad_token_id": 151643,
+  "repetition_penalty": 1.1,
+  "temperature": 0.7,
+  "top_k": 20,
+  "top_p": 0.8,
+  "transformers_version": "4.49.0.dev0"
+}

model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7dad40ae883e71ba0a6c04423de48515c4485bbb4c7d4cc81d7245715f1938c9
+size 3087467144

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": 0.02524358952622005,
+    "train_runtime": 29066.0481,
+    "train_samples": 72441,
+    "train_samples_per_second": 2.492,
+    "train_steps_per_second": 0.01
+}

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": 384.0463357925415,
+      "epoch": 0.0176678445229682,
+      "grad_norm": 0.6529819382356065,
+      "kl": 0.0007791996002197265,
+      "learning_rate": 3.448275862068966e-06,
+      "loss": 0.0,
+      "reward": 0.640290205180645,
+      "reward_std": 0.43957694210112097,
+      "rewards/accuracy_reward": 0.14720982844009994,
+      "rewards/format_reward": 0.4930803783237934,
+      "step": 5
+    },
+    {
+      "completion_length": 211.52143774032592,
+      "epoch": 0.0353356890459364,
+      "grad_norm": 8.069932901761003,
+      "kl": 0.6061183929443359,
+      "learning_rate": 6.896551724137932e-06,
+      "loss": 0.0242,
+      "reward": 0.9496652163565159,
+      "reward_std": 0.264773327531293,
+      "rewards/accuracy_reward": 0.07377232476137578,
+      "rewards/format_reward": 0.8758928958326578,
+      "step": 10
+    },
+    {
+      "completion_length": 146.1306981086731,
+      "epoch": 0.053003533568904596,
+      "grad_norm": 0.43113546878909836,
+      "kl": 0.077935791015625,
+      "learning_rate": 1.0344827586206898e-05,
+      "loss": 0.0031,
+      "reward": 1.0457589790225028,
+      "reward_std": 0.20272804964333774,
+      "rewards/accuracy_reward": 0.08671875351574272,
+      "rewards/format_reward": 0.9590402156114578,
+      "step": 15
+    },
+    {
+      "completion_length": 99.81663384437562,
+      "epoch": 0.0706713780918728,
+      "grad_norm": 0.4019533843400798,
+      "kl": 0.19752197265625,
+      "learning_rate": 1.3793103448275863e-05,
+      "loss": 0.0079,
+      "reward": 1.0904018372297286,
+      "reward_std": 0.21298125991597772,
+      "rewards/accuracy_reward": 0.12154018429573625,
+      "rewards/format_reward": 0.968861648440361,
+      "step": 20
+    },
+    {
+      "completion_length": 152.5420826435089,
+      "epoch": 0.08833922261484099,
+      "grad_norm": 2409.3058850659404,
+      "kl": 4.647747802734375,
+      "learning_rate": 1.7241379310344828e-05,
+      "loss": 0.187,
+      "reward": 1.1687500491738319,
+      "reward_std": 0.28142741243354974,
+      "rewards/accuracy_reward": 0.20290179601870478,
+      "rewards/format_reward": 0.9658482521772385,
+      "step": 25
+    },
+    {
+      "completion_length": 214.08862552642822,
+      "epoch": 0.10600706713780919,
+      "grad_norm": 0.2986315582789042,
+      "kl": 0.159442138671875,
+      "learning_rate": 1.999923511388017e-05,
+      "loss": 0.0064,
+      "reward": 1.1717634417116642,
+      "reward_std": 0.3338914422318339,
+      "rewards/accuracy_reward": 0.23303572479635476,
+      "rewards/format_reward": 0.9387277208268643,
+      "step": 30
+    },
+    {
+      "completion_length": 149.01953802108764,
+      "epoch": 0.12367491166077739,
+      "grad_norm": 0.2289227294262673,
+      "kl": 0.17880859375,
+      "learning_rate": 1.9972476383747748e-05,
+      "loss": 0.0072,
+      "reward": 1.2422991648316384,
+      "reward_std": 0.2768098399043083,
+      "rewards/accuracy_reward": 0.26517858244478704,
+      "rewards/format_reward": 0.9771205797791481,
+      "step": 35
+    },
+    {
+      "completion_length": 247.47110538482667,
+      "epoch": 0.1413427561837456,
+      "grad_norm": 370.97596271434315,
+      "kl": 5.921478271484375,
+      "learning_rate": 1.9907590277344582e-05,
+      "loss": 0.2362,
+      "reward": 1.2344866633415221,
+      "reward_std": 0.320804962515831,
+      "rewards/accuracy_reward": 0.27901786835864184,
+      "rewards/format_reward": 0.9554687894880771,
+      "step": 40
+    },
+    {
+      "completion_length": 251.62902946472167,
+      "epoch": 0.15901060070671377,
+      "grad_norm": 58.23088108276416,
+      "kl": 1.127880859375,
+      "learning_rate": 1.9804824871166254e-05,
+      "loss": 0.045,
+      "reward": 1.1912946954369545,
+      "reward_std": 0.31524865701794624,
+      "rewards/accuracy_reward": 0.2500000115483999,
+      "rewards/format_reward": 0.9412946835160255,
+      "step": 45
+    },
+    {
+      "completion_length": 176.58136978149415,
+      "epoch": 0.17667844522968199,
+      "grad_norm": 1.2616830529166947,
+      "kl": 0.3770751953125,
+      "learning_rate": 1.9664573064143604e-05,
+      "loss": 0.0151,
+      "reward": 1.1544643394649028,
+      "reward_std": 0.35408149342983963,
+      "rewards/accuracy_reward": 0.24129465334117411,
+      "rewards/format_reward": 0.9131696835160256,
+      "step": 50
+    },
+    {
+      "completion_length": 122.38192529678345,
+      "epoch": 0.19434628975265017,
+      "grad_norm": 0.2806316603874717,
+      "kl": 0.38321533203125,
+      "learning_rate": 1.948737107548771e-05,
+      "loss": 0.0153,
+      "reward": 1.2000000566244124,
+      "reward_std": 0.2845297777093947,
+      "rewards/accuracy_reward": 0.24564733263105154,
+      "rewards/format_reward": 0.9543527215719223,
+      "step": 55
+    },
+    {
+      "completion_length": 134.8651843070984,
+      "epoch": 0.21201413427561838,
+      "grad_norm": 2.6512614670552606,
+      "kl": 0.39910888671875,
+      "learning_rate": 1.9273896394584103e-05,
+      "loss": 0.016,
+      "reward": 1.2407366573810577,
+      "reward_std": 0.2907262988388538,
+      "rewards/accuracy_reward": 0.2858259055763483,
+      "rewards/format_reward": 0.9549107551574707,
+      "step": 60
+    },
+    {
+      "completion_length": 136.180140209198,
+      "epoch": 0.22968197879858657,
+      "grad_norm": 1.338921430387086,
+      "kl": 2.08179931640625,
+      "learning_rate": 1.9024965190774262e-05,
+      "loss": 0.0833,
+      "reward": 1.275223271548748,
+      "reward_std": 0.26917654536664487,
+      "rewards/accuracy_reward": 0.3123884086497128,
+      "rewards/format_reward": 0.9628348648548126,
+      "step": 65
+    },
+    {
+      "completion_length": 305.0924233436584,
+      "epoch": 0.24734982332155478,
+      "grad_norm": 2.4885869901417887,
+      "kl": 2.4799072265625,
+      "learning_rate": 1.8741529192927528e-05,
+      "loss": 0.0992,
+      "reward": 0.9402902211993933,
+      "reward_std": 0.42617332013323905,
+      "rewards/accuracy_reward": 0.2555803691968322,
+      "rewards/format_reward": 0.6847098521888256,
+      "step": 70
+    },
+    {
+      "completion_length": 273.31150884628295,
+      "epoch": 0.26501766784452296,
+      "grad_norm": 0.42249549223947436,
+      "kl": 0.40208740234375,
+      "learning_rate": 1.8424672050733577e-05,
+      "loss": 0.0161,
+      "reward": 1.0387277275323867,
+      "reward_std": 0.46621278002858163,
+      "rewards/accuracy_reward": 0.2751116196624935,
+      "rewards/format_reward": 0.7636161059141159,
+      "step": 75
+    },
+    {
+      "completion_length": 122.03739376068116,
+      "epoch": 0.2826855123674912,
+      "grad_norm": 0.5350338358000201,
+      "kl": 0.6455322265625,
+      "learning_rate": 1.8075605191627242e-05,
+      "loss": 0.0258,
+      "reward": 1.2546875521540641,
+      "reward_std": 0.26898800805211065,
+      "rewards/accuracy_reward": 0.2853794780559838,
+      "rewards/format_reward": 0.9693080767989158,
+      "step": 80
+    },
+    {
+      "completion_length": 152.50893516540526,
+      "epoch": 0.3003533568904594,
+      "grad_norm": 1.6918463013914555,
+      "kl": 0.884033203125,
+      "learning_rate": 1.7695663189185703e-05,
+      "loss": 0.0354,
+      "reward": 1.260156300663948,
+      "reward_std": 0.2852775551378727,
+      "rewards/accuracy_reward": 0.3037946572527289,
+      "rewards/format_reward": 0.956361646950245,
+      "step": 85
+    },
+    {
+      "completion_length": 186.635165309906,
+      "epoch": 0.31802120141342755,
+      "grad_norm": 3.0673822516554643,
+      "kl": 1.114306640625,
+      "learning_rate": 1.7286298660705877e-05,
+      "loss": 0.0446,
+      "reward": 1.2143973752856254,
+      "reward_std": 0.3637195309624076,
+      "rewards/accuracy_reward": 0.301004477776587,
+      "rewards/format_reward": 0.9133929021656513,
+      "step": 90
+    },
+    {
+      "completion_length": 184.8044722557068,
+      "epoch": 0.33568904593639576,
+      "grad_norm": 0.8799900343158491,
+      "kl": 1.20673828125,
+      "learning_rate": 1.6849076713469914e-05,
+      "loss": 0.0483,
+      "reward": 1.1883929029107094,
+      "reward_std": 0.352879635989666,
+      "rewards/accuracy_reward": 0.2754464400932193,
+      "rewards/format_reward": 0.9129464730620385,
+      "step": 95
+    },
+    {
+      "completion_length": 459.28539962768554,
+      "epoch": 0.35335689045936397,
+      "grad_norm": 1.4660921102611524,
+      "kl": 2.501123046875,
+      "learning_rate": 1.6385668960932143e-05,
+      "loss": 0.1,
+      "reward": 0.7824777118861675,
+      "reward_std": 0.5532245114445686,
+      "rewards/accuracy_reward": 0.1621651851804927,
+      "rewards/format_reward": 0.6203125301748514,
+      "step": 100
+    },
+    {
+      "epoch": 0.35335689045936397,
+      "eval_completion_length": 471.5725402832031,
+      "eval_kl": 2.90625,
+      "eval_loss": 0.11904626339673996,
+      "eval_reward": 0.690848246216774,
+      "eval_reward_std": 0.6071737855672836,
+      "eval_rewards/accuracy_reward": 0.1261160746216774,
+      "eval_rewards/format_reward": 0.564732164144516,
+      "eval_runtime": 46.7851,
+      "eval_samples_per_second": 2.116,
+      "eval_steps_per_second": 0.021,
+      "step": 100
+    },
+    {
+      "completion_length": 432.415310382843,
+      "epoch": 0.3710247349823322,
+      "grad_norm": 0.48111406831320014,
+      "kl": 1.6375,
+      "learning_rate": 1.5897847131705194e-05,
+      "loss": 0.0655,
+      "reward": 0.7103794921189547,
+      "reward_std": 0.4915090943686664,
+      "rewards/accuracy_reward": 0.1224330407101661,
+      "rewards/format_reward": 0.5879464514553547,
+      "step": 105
+    },
+    {
+      "completion_length": 44.55658693313599,
+      "epoch": 0.38869257950530034,
+      "grad_norm": 0.4829508094562178,
+      "kl": 1.22216796875,
+      "learning_rate": 1.5387476295779737e-05,
+      "loss": 0.0489,
+      "reward": 1.173437552154064,
+      "reward_std": 0.16121841138228774,
+      "rewards/accuracy_reward": 0.19732143701985477,
+      "rewards/format_reward": 0.9761160992085933,
+      "step": 110
+    },
+    {
+      "completion_length": 102.14576349258422,
+      "epoch": 0.40636042402826855,
+      "grad_norm": 13.346352830691668,
+      "kl": 0.862548828125,
+      "learning_rate": 1.4856507733875837e-05,
+      "loss": 0.0345,
+      "reward": 1.1206473730504514,
+      "reward_std": 0.30396630289033055,
+      "rewards/accuracy_reward": 0.20636161593720317,
+      "rewards/format_reward": 0.9142857596278191,
+      "step": 115
+    },
+    {
+      "completion_length": 102.4949821472168,
+      "epoch": 0.42402826855123676,
+      "grad_norm": 0.2561020453511286,
+      "kl": 0.561376953125,
+      "learning_rate": 1.4306971477188223e-05,
+      "loss": 0.0224,
+      "reward": 1.201450951397419,
+      "reward_std": 0.22256765561178327,
+      "rewards/accuracy_reward": 0.22254465399309992,
+      "rewards/format_reward": 0.9789062902331352,
+      "step": 120
+    },
+    {
+      "completion_length": 238.2740068435669,
+      "epoch": 0.4416961130742049,
+      "grad_norm": 0.265018502031295,
+      "kl": 0.1888427734375,
+      "learning_rate": 1.3740968546047935e-05,
+      "loss": 0.0076,
+      "reward": 1.191964340209961,
+      "reward_std": 0.323291926831007,
+      "rewards/accuracy_reward": 0.2588169751688838,
+      "rewards/format_reward": 0.9331473611295223,
+      "step": 125
+    },
+    {
+      "completion_length": 310.94075145721433,
+      "epoch": 0.45936395759717313,
+      "grad_norm": 0.17518365102004282,
+      "kl": 0.142608642578125,
+      "learning_rate": 1.3160662917174045e-05,
+      "loss": 0.0057,
+      "reward": 1.2094866700470448,
+      "reward_std": 0.41736746579408646,
+      "rewards/accuracy_reward": 0.33258930016309024,
+      "rewards/format_reward": 0.8768973611295223,
+      "step": 130
+    },
+    {
+      "completion_length": 316.3401933670044,
+      "epoch": 0.47703180212014135,
+      "grad_norm": 0.17370338211422123,
+      "kl": 0.132843017578125,
+      "learning_rate": 1.2568273250226681e-05,
+      "loss": 0.0053,
+      "reward": 1.191852729022503,
+      "reward_std": 0.43695004992187025,
+      "rewards/accuracy_reward": 0.34654019474983216,
+      "rewards/format_reward": 0.8453125394880772,
+      "step": 135
+    },
+    {
+      "completion_length": 416.6068260192871,
+      "epoch": 0.49469964664310956,
+      "grad_norm": 0.20545441542706755,
+      "kl": 0.15596923828125,
+      "learning_rate": 1.1966064405292887e-05,
+      "loss": 0.0062,
+      "reward": 1.046428620815277,
+      "reward_std": 0.525279750674963,
+      "rewards/accuracy_reward": 0.3165178705006838,
+      "rewards/format_reward": 0.7299107514321804,
+      "step": 140
+    },
+    {
+      "completion_length": 242.95715389251708,
+      "epoch": 0.5123674911660777,
+      "grad_norm": 0.3073638444949766,
+      "kl": 0.17847900390625,
+      "learning_rate": 1.1356338783736256e-05,
+      "loss": 0.0071,
+      "reward": 1.2570313096046448,
+      "reward_std": 0.36572824846953156,
+      "rewards/accuracy_reward": 0.350446443259716,
+      "rewards/format_reward": 0.9065848618745804,
+      "step": 145
+    },
+    {
+      "completion_length": 210.5078224182129,
+      "epoch": 0.5300353356890459,
+      "grad_norm": 0.8165939885649788,
+      "kl": 0.1881591796875,
+      "learning_rate": 1.0741427525516463e-05,
+      "loss": 0.0075,
+      "reward": 1.293192020058632,
+      "reward_std": 0.3428498702123761,
+      "rewards/accuracy_reward": 0.3569196606054902,
+      "rewards/format_reward": 0.9362723641097546,
+      "step": 150
+    },
+    {
+      "completion_length": 261.776350402832,
+      "epoch": 0.5477031802120141,
+      "grad_norm": 0.1582208660243009,
+      "kl": 0.1717041015625,
+      "learning_rate": 1.012368159663363e-05,
+      "loss": 0.0069,
+      "reward": 1.243973270058632,
+      "reward_std": 0.39117210209369657,
+      "rewards/accuracy_reward": 0.3363839453086257,
+      "rewards/format_reward": 0.9075893312692642,
+      "step": 155
+    },
+    {
+      "completion_length": 239.85179653167725,
+      "epoch": 0.5653710247349824,
+      "grad_norm": 0.1704685274719256,
+      "kl": 0.14776611328125,
+      "learning_rate": 9.505462800772612e-06,
+      "loss": 0.0059,
+      "reward": 1.2680804081261159,
+      "reward_std": 0.35588377732783555,
+      "rewards/accuracy_reward": 0.335825908370316,
+      "rewards/format_reward": 0.9322545066475868,
+      "step": 160
+    },
+    {
+      "completion_length": 229.55492095947267,
+      "epoch": 0.5830388692579506,
+      "grad_norm": 0.15800892653159143,
+      "kl": 0.138787841796875,
+      "learning_rate": 8.889134749511956e-06,
+      "loss": 0.0055,
+      "reward": 1.2492188036441803,
+      "reward_std": 0.3586545692756772,
+      "rewards/accuracy_reward": 0.31674108635634185,
+      "rewards/format_reward": 0.9324777208268642,
+      "step": 165
+    },
+    {
+      "completion_length": 249.42702083587648,
+      "epoch": 0.6007067137809188,
+      "grad_norm": 0.15989747781862085,
+      "kl": 0.13619384765625,
+      "learning_rate": 8.277053825620836e-06,
+      "loss": 0.0054,
+      "reward": 1.234040232002735,
+      "reward_std": 0.3690490124747157,
+      "rewards/accuracy_reward": 0.315625012665987,
+      "rewards/format_reward": 0.9184152215719223,
+      "step": 170
+    },
+    {
+      "completion_length": 266.64677543640136,
+      "epoch": 0.6183745583038869,
+      "grad_norm": 0.1728727033658703,
+      "kl": 0.143048095703125,
+      "learning_rate": 7.671560173993588e-06,
+      "loss": 0.0057,
+      "reward": 1.2549107685685157,
+      "reward_std": 0.3902056057006121,
+      "rewards/accuracy_reward": 0.3504464441910386,
+      "rewards/format_reward": 0.90446432903409,
+      "step": 175
+    },
+    {
+      "completion_length": 267.99766731262207,
+      "epoch": 0.6360424028268551,
+      "grad_norm": 0.16828899543606332,
+      "kl": 0.157781982421875,
+      "learning_rate": 7.07496875466589e-06,
+      "loss": 0.0063,
+      "reward": 1.2652902334928513,
+      "reward_std": 0.394162866845727,
+      "rewards/accuracy_reward": 0.36138394605368374,
+      "rewards/format_reward": 0.9039062932133675,
+      "step": 180
+    },
+    {
+      "completion_length": 251.41809253692628,
+      "epoch": 0.6537102473498233,
+      "grad_norm": 0.21026602400413624,
+      "kl": 0.2133056640625,
+      "learning_rate": 6.489560492119225e-06,
+      "loss": 0.0085,
+      "reward": 1.237165230512619,
+      "reward_std": 0.3938104841858149,
+      "rewards/accuracy_reward": 0.3395089440047741,
+      "rewards/format_reward": 0.8976562954485416,
+      "step": 185
+    },
+    {
+      "completion_length": 214.81686115264893,
+      "epoch": 0.6713780918727915,
+      "grad_norm": 0.18932563174791797,
+      "kl": 0.2186767578125,
+      "learning_rate": 5.9175735547120975e-06,
+      "loss": 0.0087,
+      "reward": 1.2207589730620385,
+      "reward_std": 0.41313072480261326,
+      "rewards/accuracy_reward": 0.3296875163912773,
+      "rewards/format_reward": 0.8910714708268642,
+      "step": 190
+    },
+    {
+      "completion_length": 244.51027965545654,
+      "epoch": 0.6890459363957597,
+      "grad_norm": 0.2257367270272996,
+      "kl": 0.21044921875,
+      "learning_rate": 5.361194797579108e-06,
+      "loss": 0.0084,
+      "reward": 1.2325893357396125,
+      "reward_std": 0.40438184086233375,
+      "rewards/accuracy_reward": 0.34832590743899344,
+      "rewards/format_reward": 0.8842634335160255,
+      "step": 195
+    },
+    {
+      "completion_length": 229.27824668884278,
+      "epoch": 0.7067137809187279,
+      "grad_norm": 0.19283868985883293,
+      "kl": 0.20889892578125,
+      "learning_rate": 4.8225514017138205e-06,
+      "loss": 0.0084,
+      "reward": 1.2724330857396127,
+      "reward_std": 0.37237234245985745,
+      "rewards/accuracy_reward": 0.358370553702116,
+      "rewards/format_reward": 0.9140625432133674,
+      "step": 200
+    },
+    {
+      "epoch": 0.7067137809187279,
+      "eval_completion_length": 220.90455627441406,
+      "eval_kl": 0.200927734375,
+      "eval_loss": 0.008105829358100891,
+      "eval_reward": 1.3314732909202576,
+      "eval_reward_std": 0.33083294332027435,
+      "eval_rewards/accuracy_reward": 0.3984375223517418,
+      "eval_rewards/format_reward": 0.933035746216774,
+      "eval_runtime": 31.5112,
+      "eval_samples_per_second": 3.142,
+      "eval_steps_per_second": 0.032,
+      "step": 200
+    },
+    {
+      "completion_length": 248.72456493377686,
+      "epoch": 0.7243816254416962,
+      "grad_norm": 0.1784321065416981,
+      "kl": 0.2092529296875,
+      "learning_rate": 4.303702741201431e-06,
+      "loss": 0.0084,
+      "reward": 1.2589286297559739,
+      "reward_std": 0.38014698009938,
+      "rewards/accuracy_reward": 0.34765626564621926,
+      "rewards/format_reward": 0.9112723655998707,
+      "step": 205
+    },
+    {
+      "completion_length": 260.51217727661134,
+      "epoch": 0.7420494699646644,
+      "grad_norm": 0.1683054026863793,
+      "kl": 0.2004150390625,
+      "learning_rate": 3.8066325096949153e-06,
+      "loss": 0.008,
+      "reward": 1.255580408871174,
+      "reward_std": 0.40174516644328834,
+      "rewards/accuracy_reward": 0.35680805090814827,
+      "rewards/format_reward": 0.8987723641097546,
+      "step": 210
+    },
+    {
+      "completion_length": 268.3051458358765,
+      "epoch": 0.7597173144876325,
+      "grad_norm": 0.19491423500924795,
+      "kl": 0.21324462890625,
+      "learning_rate": 3.3332411362372063e-06,
+      "loss": 0.0085,
+      "reward": 1.222433079779148,
+      "reward_std": 0.4209536049515009,
+      "rewards/accuracy_reward": 0.33761162161827085,
+      "rewards/format_reward": 0.8848214723169804,
+      "step": 215
+    },
+    {
+      "completion_length": 242.99431762695312,
+      "epoch": 0.7773851590106007,
+      "grad_norm": 0.1827115506822177,
+      "kl": 0.2093017578125,
+      "learning_rate": 2.8853385194256677e-06,
+      "loss": 0.0084,
+      "reward": 1.256361658871174,
+      "reward_std": 0.3803428867831826,
+      "rewards/accuracy_reward": 0.34787948057055473,
+      "rewards/format_reward": 0.9084821864962578,
+      "step": 220
+    },
+    {
+      "completion_length": 229.23405113220215,
+      "epoch": 0.7950530035335689,
+      "grad_norm": 0.19240194787621448,
+      "kl": 0.21724853515625,
+      "learning_rate": 2.464637107698046e-06,
+      "loss": 0.0087,
+      "reward": 1.2588170126080513,
+      "reward_std": 0.37087023109197614,
+      "rewards/accuracy_reward": 0.3417410867288709,
+      "rewards/format_reward": 0.917075938731432,
+      "step": 225
+    },
+    {
+      "completion_length": 243.72713241577148,
+      "epoch": 0.8127208480565371,
+      "grad_norm": 0.1741043475951591,
+      "kl": 0.2067626953125,
+      "learning_rate": 2.072745352195794e-06,
+      "loss": 0.0083,
+      "reward": 1.243750050663948,
+      "reward_std": 0.39028936978429557,
+      "rewards/accuracy_reward": 0.3424107315018773,
+      "rewards/format_reward": 0.901339328289032,
+      "step": 230
+    },
+    {
+      "completion_length": 245.68751201629638,
+      "epoch": 0.8303886925795053,
+      "grad_norm": 0.24505417166133048,
+      "kl": 0.21961669921875,
+      "learning_rate": 1.7111615572361628e-06,
+      "loss": 0.0088,
+      "reward": 1.2446429148316382,
+      "reward_std": 0.4114550109952688,
+      "rewards/accuracy_reward": 0.35022323187440635,
+      "rewards/format_reward": 0.8944196850061417,
+      "step": 235
+    },
+    {
+      "completion_length": 241.52579193115236,
+      "epoch": 0.8480565371024735,
+      "grad_norm": 0.20757665190714725,
+      "kl": 0.21859130859375,
+      "learning_rate": 1.381268151904298e-06,
+      "loss": 0.0087,
+      "reward": 1.2525670200586319,
+      "reward_std": 0.3901967225596309,
+      "rewards/accuracy_reward": 0.3515625149011612,
+      "rewards/format_reward": 0.9010045081377029,
+      "step": 240
+    },
+    {
+      "completion_length": 244.1225549697876,
+      "epoch": 0.8657243816254417,
+      "grad_norm": 0.17183432547369637,
+      "kl": 0.21815185546875,
+      "learning_rate": 1.0843264046665558e-06,
+      "loss": 0.0087,
+      "reward": 1.249665230512619,
+      "reward_std": 0.3750341447070241,
+      "rewards/accuracy_reward": 0.3444196585565805,
+      "rewards/format_reward": 0.9052455805242061,
+      "step": 245
+    },
+    {
+      "completion_length": 243.7822650909424,
+      "epoch": 0.8833922261484098,
+      "grad_norm": 0.1916692184784476,
+      "kl": 0.21383056640625,
+      "learning_rate": 8.214716012124491e-07,
+      "loss": 0.0086,
+      "reward": 1.24776791036129,
+      "reward_std": 0.3710861327126622,
+      "rewards/accuracy_reward": 0.33995537385344504,
+      "rewards/format_reward": 0.9078125409781933,
+      "step": 250
+    },
+    {
+      "completion_length": 235.8559259414673,
+      "epoch": 0.901060070671378,
+      "grad_norm": 0.23346599184542174,
+      "kl": 0.22025146484375,
+      "learning_rate": 5.937087039615619e-07,
+      "loss": 0.0088,
+      "reward": 1.2597098812460898,
+      "reward_std": 0.3739591669291258,
+      "rewards/accuracy_reward": 0.34542412087321284,
+      "rewards/format_reward": 0.9142857551574707,
+      "step": 255
+    },
+    {
+      "completion_length": 251.72690830230712,
+      "epoch": 0.9187279151943463,
+      "grad_norm": 0.20314456248803323,
+      "kl": 0.2180908203125,
+      "learning_rate": 4.019085098303077e-07,
+      "loss": 0.0087,
+      "reward": 1.2440848767757415,
+      "reward_std": 0.39818168375641105,
+      "rewards/accuracy_reward": 0.3501116232946515,
+      "rewards/format_reward": 0.8939732573926449,
+      "step": 260
+    },
+    {
+      "completion_length": 248.0124002456665,
+      "epoch": 0.9363957597173145,
+      "grad_norm": 0.1993341399262892,
+      "kl": 0.22598876953125,
+      "learning_rate": 2.4680432094837394e-07,
+      "loss": 0.009,
+      "reward": 1.2393973782658576,
+      "reward_std": 0.3911540800705552,
+      "rewards/accuracy_reward": 0.34441965762525795,
+      "rewards/format_reward": 0.8949777200818062,
+      "step": 265
+    },
+    {
+      "completion_length": 249.1703239440918,
+      "epoch": 0.9540636042402827,
+      "grad_norm": 0.2591538960467843,
+      "kl": 0.2145263671875,
+      "learning_rate": 1.289891410535593e-07,
+      "loss": 0.0086,
+      "reward": 1.247767911851406,
+      "reward_std": 0.40619117505848407,
+      "rewards/accuracy_reward": 0.34966519437730315,
+      "rewards/format_reward": 0.8981027238070964,
+      "step": 270
+    },
+    {
+      "completion_length": 245.0108373641968,
+      "epoch": 0.9717314487632509,
+      "grad_norm": 0.18045259609614844,
+      "kl": 0.22490234375,
+      "learning_rate": 4.8913408283934874e-08,
+      "loss": 0.009,
+      "reward": 1.24698666036129,
+      "reward_std": 0.3904744828119874,
+      "rewards/accuracy_reward": 0.3506696572527289,
+      "rewards/format_reward": 0.8963170073926449,
+      "step": 275
+    },
+    {
+      "completion_length": 248.61061344146728,
+      "epoch": 0.9893992932862191,
+      "grad_norm": 0.1910344572252932,
+      "kl": 0.2626220703125,
+      "learning_rate": 6.883273035447335e-09,
+      "loss": 0.0105,
+      "reward": 1.2524554118514062,
+      "reward_std": 0.4037581391632557,
+      "rewards/accuracy_reward": 0.3559151943773031,
+      "rewards/format_reward": 0.8965402223169804,
+      "step": 280
+    },
+    {
+      "completion_length": 249.9445120493571,
+      "epoch": 1.0,
+      "kl": 0.21577962239583334,
+      "reward": 1.2481399315098922,
+      "reward_std": 0.4004522568235795,
+      "rewards/accuracy_reward": 0.35435269710918266,
+      "rewards/format_reward": 0.8937872474392256,
+      "step": 283,
+      "total_flos": 0.0,
+      "train_loss": 0.02524358952622005,
+      "train_runtime": 29066.0481,
+      "train_samples_per_second": 2.492,
+      "train_steps_per_second": 0.01
+    }
+  ],
+  "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 sha256:461246ac0b79296cdc78f3c64b765a1779adbd870145dbd953883e336a0f7d67
+size 7480