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 +1,67 @@
+---
+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 [None](https://huggingface.co/None).
+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="taozihuahua/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
+
+ 
+
+
+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: 4.48.3
+- Pytorch: 2.5.1
+- Datasets: 3.3.0
+- Tokenizers: 0.21.0
+
+## Citations
+
+Cite GRPO as:
+
+```bibtex
+@article{zhihong2024deepseekmath,
+    title        = {{DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models}},
+    author       = {Zhihong Shao and Peiyi Wang and Qihao Zhu and Runxin Xu and Junxiao Song and Mingchuan Zhang and Y. K. Li and Y. Wu and Daya Guo},
+    year         = 2024,
+    eprint       = {arXiv:2402.03300},
+}
+
+```
+
+Cite TRL as:
+    
+```bibtex
+@misc{vonwerra2022trl,
+	title        = {{TRL: Transformer Reinforcement Learning}},
+	author       = {Leandro von Werra and Younes Belkada and Lewis Tunstall and Edward Beeching and Tristan Thrush and Nathan Lambert and Shengyi Huang and Kashif Rasul and Quentin Gallouédec},
+	year         = 2020,
+	journal      = {GitHub repository},
+	publisher    = {GitHub},
+	howpublished = {\url{https://github.com/huggingface/trl}}
+}
+```

added_tokens.json

@@ -0,0 +1,24 @@
+{
+  "</tool_call>": 151658,
+  "<tool_call>": 151657,
+  "<|box_end|>": 151649,
+  "<|box_start|>": 151648,
+  "<|endoftext|>": 151643,
+  "<|file_sep|>": 151664,
+  "<|fim_middle|>": 151660,
+  "<|fim_pad|>": 151662,
+  "<|fim_prefix|>": 151659,
+  "<|fim_suffix|>": 151661,
+  "<|im_end|>": 151645,
+  "<|im_start|>": 151644,
+  "<|image_pad|>": 151655,
+  "<|object_ref_end|>": 151647,
+  "<|object_ref_start|>": 151646,
+  "<|quad_end|>": 151651,
+  "<|quad_start|>": 151650,
+  "<|repo_name|>": 151663,
+  "<|video_pad|>": 151656,
+  "<|vision_end|>": 151653,
+  "<|vision_pad|>": 151654,
+  "<|vision_start|>": 151652
+}

all_results.json

@@ -0,0 +1,8 @@
+{
+    "total_flos": 0.0,
+    "train_loss": 0.7804505577981542,
+    "train_runtime": 45227.2818,
+    "train_samples": 72441,
+    "train_samples_per_second": 1.602,
+    "train_steps_per_second": 0.006
+}

config.json

@@ -0,0 +1,29 @@
+{
+  "_name_or_path": "/jfs/taozhihua/projects/training_code/deepseek/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": "4.48.3",
+  "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": "4.48.3"
+}

model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2c83c3079630168e485249618fe4546a14cab3c5ddfb5c95cbb9f22f16937ceb
+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": 0.7804505577981542,
+    "train_runtime": 45227.2818,
+    "train_samples": 72441,
+    "train_samples_per_second": 1.602,
+    "train_steps_per_second": 0.006
+}

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": 384.2656421661377,
+      "epoch": 0.0176678445229682,
+      "grad_norm": 0.03746618330478668,
+      "kl": 0.0010451078414916992,
+      "learning_rate": 3.448275862068966e-06,
+      "loss": 0.0,
+      "reward": 0.6437500260770321,
+      "reward_std": 0.4394187767058611,
+      "rewards/accuracy_reward": 0.15212054261937738,
+      "rewards/format_reward": 0.49162948355078695,
+      "step": 5
+    },
+    {
+      "completion_length": 261.34967670440676,
+      "epoch": 0.0353356890459364,
+      "grad_norm": 0.0510084368288517,
+      "kl": 0.41996231079101565,
+      "learning_rate": 6.896551724137932e-06,
+      "loss": 0.0168,
+      "reward": 0.9005580745637417,
+      "reward_std": 0.3328772446140647,
+      "rewards/accuracy_reward": 0.08348214707802981,
+      "rewards/format_reward": 0.8170759327709675,
+      "step": 10
+    },
+    {
+      "completion_length": 160.25893526077272,
+      "epoch": 0.053003533568904596,
+      "grad_norm": 0.03750701621174812,
+      "kl": 3.9105926513671876,
+      "learning_rate": 1.0344827586206898e-05,
+      "loss": 0.156,
+      "reward": 0.9283482581377029,
+      "reward_std": 0.28247026447206736,
+      "rewards/accuracy_reward": 0.07600446732249111,
+      "rewards/format_reward": 0.8523437850177288,
+      "step": 15
+    },
+    {
+      "completion_length": 78.71875367164611,
+      "epoch": 0.0706713780918728,
+      "grad_norm": 0.026410209015011787,
+      "kl": 0.2958984375,
+      "learning_rate": 1.3793103448275863e-05,
+      "loss": 0.0118,
+      "reward": 1.0590402320027352,
+      "reward_std": 0.2208268778398633,
+      "rewards/accuracy_reward": 0.11250000515719875,
+      "rewards/format_reward": 0.94654021859169,
+      "step": 20
+    },
+    {
+      "completion_length": 99.18750429153442,
+      "epoch": 0.08833922261484099,
+      "grad_norm": 0.06202094256877899,
+      "kl": 0.2493408203125,
+      "learning_rate": 1.7241379310344828e-05,
+      "loss": 0.01,
+      "reward": 1.1320313036441803,
+      "reward_std": 0.23392355674877763,
+      "rewards/accuracy_reward": 0.15993304285220802,
+      "rewards/format_reward": 0.9720982521772384,
+      "step": 25
+    },
+    {
+      "completion_length": 101.18650045394898,
+      "epoch": 0.10600706713780919,
+      "grad_norm": 0.022969426587224007,
+      "kl": 0.3260009765625,
+      "learning_rate": 1.999923511388017e-05,
+      "loss": 0.013,
+      "reward": 1.170312550663948,
+      "reward_std": 0.2822354474104941,
+      "rewards/accuracy_reward": 0.21841519000008702,
+      "rewards/format_reward": 0.9518973603844643,
+      "step": 30
+    },
+    {
+      "completion_length": 98.8357186794281,
+      "epoch": 0.12367491166077739,
+      "grad_norm": 0.02503274939954281,
+      "kl": 0.78983154296875,
+      "learning_rate": 1.9972476383747748e-05,
+      "loss": 0.0317,
+      "reward": 1.2123884364962578,
+      "reward_std": 0.22811748776584864,
+      "rewards/accuracy_reward": 0.2267857262864709,
+      "rewards/format_reward": 0.9856027126312256,
+      "step": 35
+    },
+    {
+      "completion_length": 254.09867248535156,
+      "epoch": 0.1413427561837456,
+      "grad_norm": 0.04796220734715462,
+      "kl": 0.21435546875,
+      "learning_rate": 1.9907590277344582e-05,
+      "loss": 0.0086,
+      "reward": 1.1683036170899868,
+      "reward_std": 0.2912572468165308,
+      "rewards/accuracy_reward": 0.20970983013976366,
+      "rewards/format_reward": 0.9585937902331352,
+      "step": 40
+    },
+    {
+      "completion_length": 253.43382778167725,
+      "epoch": 0.15901060070671377,
+      "grad_norm": 0.017703380435705185,
+      "kl": 0.24912109375,
+      "learning_rate": 1.9804824871166254e-05,
+      "loss": 0.01,
+      "reward": 1.1513393327593804,
+      "reward_std": 0.2743695043027401,
+      "rewards/accuracy_reward": 0.19397322395816446,
+      "rewards/format_reward": 0.9573661118745804,
+      "step": 45
+    },
+    {
+      "completion_length": 200.55023231506348,
+      "epoch": 0.17667844522968199,
+      "grad_norm": 0.0570429302752018,
+      "kl": 0.2423828125,
+      "learning_rate": 1.9664573064143604e-05,
+      "loss": 0.0097,
+      "reward": 1.1926339820027352,
+      "reward_std": 0.25827550180256365,
+      "rewards/accuracy_reward": 0.21953126033768058,
+      "rewards/format_reward": 0.9731027163565159,
+      "step": 50
+    },
+    {
+      "completion_length": 196.50491943359376,
+      "epoch": 0.19434628975265017,
+      "grad_norm": 0.013810385018587112,
+      "kl": 0.22283935546875,
+      "learning_rate": 1.948737107548771e-05,
+      "loss": 0.0089,
+      "reward": 1.211383979022503,
+      "reward_std": 0.2685432176105678,
+      "rewards/accuracy_reward": 0.23638393906876445,
+      "rewards/format_reward": 0.9750000379979611,
+      "step": 55
+    },
+    {
+      "completion_length": 176.04018640518188,
+      "epoch": 0.21201413427561838,
+      "grad_norm": 2.573392629623413,
+      "kl": 0.5369873046875,
+      "learning_rate": 1.9273896394584103e-05,
+      "loss": 0.0215,
+      "reward": 1.220870591700077,
+      "reward_std": 0.25569039974361657,
+      "rewards/accuracy_reward": 0.24185269009321927,
+      "rewards/format_reward": 0.9790179051458836,
+      "step": 60
+    },
+    {
+      "completion_length": 218.72221965789794,
+      "epoch": 0.22968197879858657,
+      "grad_norm": 0.03989923745393753,
+      "kl": 0.237548828125,
+      "learning_rate": 1.9024965190774262e-05,
+      "loss": 0.0095,
+      "reward": 1.229910773038864,
+      "reward_std": 0.3029924972914159,
+      "rewards/accuracy_reward": 0.26774554727599026,
+      "rewards/format_reward": 0.9621652208268643,
+      "step": 65
+    },
+    {
+      "completion_length": 180.19543266296387,
+      "epoch": 0.24734982332155478,
+      "grad_norm": 0.06838759034872055,
+      "kl": 0.45032958984375,
+      "learning_rate": 1.8741529192927528e-05,
+      "loss": 0.018,
+      "reward": 1.204799158871174,
+      "reward_std": 0.30533219743520024,
+      "rewards/accuracy_reward": 0.25468751080334184,
+      "rewards/format_reward": 0.9501116439700127,
+      "step": 70
+    },
+    {
+      "completion_length": 140.01819801330566,
+      "epoch": 0.26501766784452296,
+      "grad_norm": 0.04944866895675659,
+      "kl": 0.508447265625,
+      "learning_rate": 1.8424672050733577e-05,
+      "loss": 0.0204,
+      "reward": 1.1905134454369546,
+      "reward_std": 0.33636278491467236,
+      "rewards/accuracy_reward": 0.2647321538999677,
+      "rewards/format_reward": 0.9257812954485416,
+      "step": 75
+    },
+    {
+      "completion_length": 80.20279369354247,
+      "epoch": 0.2826855123674912,
+      "grad_norm": 0.13146764039993286,
+      "kl": 0.6988525390625,
+      "learning_rate": 1.8075605191627242e-05,
+      "loss": 0.028,
+      "reward": 1.122544700652361,
+      "reward_std": 0.31514625139534475,
+      "rewards/accuracy_reward": 0.20680804452858864,
+      "rewards/format_reward": 0.9157366476953029,
+      "step": 80
+    },
+    {
+      "completion_length": 49.505471158027646,
+      "epoch": 0.3003533568904594,
+      "grad_norm": 0.34890076518058777,
+      "kl": 3.626953125,
+      "learning_rate": 1.7695663189185703e-05,
+      "loss": 0.1454,
+      "reward": 1.056361659616232,
+      "reward_std": 0.3539266254752874,
+      "rewards/accuracy_reward": 0.18928572302684188,
+      "rewards/format_reward": 0.8670759327709675,
+      "step": 85
+    },
+    {
+      "completion_length": 117.59018392562866,
+      "epoch": 0.31802120141342755,
+      "grad_norm": 154.60848999023438,
+      "kl": 66.207421875,
+      "learning_rate": 1.7286298660705877e-05,
+      "loss": 2.6486,
+      "reward": 0.873437537252903,
+      "reward_std": 0.39913321323692796,
+      "rewards/accuracy_reward": 0.13069197018630802,
+      "rewards/format_reward": 0.7427455738186837,
+      "step": 90
+    },
+    {
+      "completion_length": 304.4634063720703,
+      "epoch": 0.33568904593639576,
+      "grad_norm": 341.2224426269531,
+      "kl": 143.1623046875,
+      "learning_rate": 1.6849076713469914e-05,
+      "loss": 5.7314,
+      "reward": 0.8466518253087998,
+      "reward_std": 0.4182041293010116,
+      "rewards/accuracy_reward": 0.11473214884754271,
+      "rewards/format_reward": 0.731919676810503,
+      "step": 95
+    },
+    {
+      "completion_length": 744.7376457214356,
+      "epoch": 0.35335689045936397,
+      "grad_norm": 0.5984399318695068,
+      "kl": 863.16181640625,
+      "learning_rate": 1.6385668960932143e-05,
+      "loss": 34.5017,
+      "reward": 0.3026785840280354,
+      "reward_std": 0.4176575379446149,
+      "rewards/accuracy_reward": 0.08136161119909957,
+      "rewards/format_reward": 0.22131697526201605,
+      "step": 100
+    },
+    {
+      "epoch": 0.35335689045936397,
+      "eval_completion_length": 456.6484375,
+      "eval_kl": 2.5625,
+      "eval_loss": 0.0743500366806984,
+      "eval_reward": 0.7232142984867096,
+      "eval_reward_std": 0.5973740816116333,
+      "eval_rewards/accuracy_reward": 0.2109375111758709,
+      "eval_rewards/format_reward": 0.5122768133878708,
+      "eval_runtime": 68.5229,
+      "eval_samples_per_second": 1.445,
+      "eval_steps_per_second": 0.015,
+      "step": 100
+    },
+    {
+      "completion_length": 293.03695430755613,
+      "epoch": 0.3710247349823322,
+      "grad_norm": 0.015755295753479004,
+      "kl": 1.394775390625,
+      "learning_rate": 1.5897847131705194e-05,
+      "loss": 0.0559,
+      "reward": 1.0056920126080513,
+      "reward_std": 0.4020020180381835,
+      "rewards/accuracy_reward": 0.2078125095926225,
+      "rewards/format_reward": 0.797879496589303,
+      "step": 105
+    },
+    {
+      "completion_length": 200.97579021453856,
+      "epoch": 0.38869257950530034,
+      "grad_norm": 0.32614070177078247,
+      "kl": 0.37606201171875,
+      "learning_rate": 1.5387476295779737e-05,
+      "loss": 0.0151,
+      "reward": 1.2574777364730836,
+      "reward_std": 0.2175285722129047,
+      "rewards/accuracy_reward": 0.26316965660080316,
+      "rewards/format_reward": 0.9943080559372902,
+      "step": 110
+    },
+    {
+      "completion_length": 248.8611722946167,
+      "epoch": 0.40636042402826855,
+      "grad_norm": 0.012247471138834953,
+      "kl": 0.25589599609375,
+      "learning_rate": 1.4856507733875837e-05,
+      "loss": 0.0102,
+      "reward": 1.264843797683716,
+      "reward_std": 0.2261615320108831,
+      "rewards/accuracy_reward": 0.2718750134110451,
+      "rewards/format_reward": 0.9929687723517417,
+      "step": 115
+    },
+    {
+      "completion_length": 288.8208837509155,
+      "epoch": 0.42402826855123676,
+      "grad_norm": 0.0323672890663147,
+      "kl": 0.3291748046875,
+      "learning_rate": 1.4306971477188223e-05,
+      "loss": 0.0132,
+      "reward": 1.2723214834928513,
+      "reward_std": 0.25960605759173633,
+      "rewards/accuracy_reward": 0.2868303697556257,
+      "rewards/format_reward": 0.9854911111295224,
+      "step": 120
+    },
+    {
+      "completion_length": 294.0188756942749,
+      "epoch": 0.4416961130742049,
+      "grad_norm": 0.053948838263750076,
+      "kl": 0.231494140625,
+      "learning_rate": 1.3740968546047935e-05,
+      "loss": 0.0093,
+      "reward": 1.2631696969270707,
+      "reward_std": 0.2896059764549136,
+      "rewards/accuracy_reward": 0.2962053693830967,
+      "rewards/format_reward": 0.9669643238186836,
+      "step": 125
+    },
+    {
+      "completion_length": 294.04755783081055,
+      "epoch": 0.45936395759717313,
+      "grad_norm": 0.03586830198764801,
+      "kl": 0.40794677734375,
+      "learning_rate": 1.3160662917174045e-05,
+      "loss": 0.0163,
+      "reward": 1.2707589879631995,
+      "reward_std": 0.32381807956844566,
+      "rewards/accuracy_reward": 0.3318080514669418,
+      "rewards/format_reward": 0.9389509350061417,
+      "step": 130
+    },
+    {
+      "completion_length": 328.0676498413086,
+      "epoch": 0.47703180212014135,
+      "grad_norm": 0.01469908282160759,
+      "kl": 0.6115966796875,
+      "learning_rate": 1.2568273250226681e-05,
+      "loss": 0.0245,
+      "reward": 1.2416295170783997,
+      "reward_std": 0.3711954679340124,
+      "rewards/accuracy_reward": 0.34330358598381283,
+      "rewards/format_reward": 0.8983259335160255,
+      "step": 135
+    },
+    {
+      "completion_length": 360.57222690582273,
+      "epoch": 0.49469964664310956,
+      "grad_norm": 0.04768484830856323,
+      "kl": 0.79281005859375,
+      "learning_rate": 1.1966064405292887e-05,
+      "loss": 0.0317,
+      "reward": 1.2180804163217545,
+      "reward_std": 0.408589731156826,
+      "rewards/accuracy_reward": 0.33995537236332896,
+      "rewards/format_reward": 0.8781250402331352,
+      "step": 140
+    },
+    {
+      "completion_length": 432.28651466369627,
+      "epoch": 0.5123674911660777,
+      "grad_norm": 0.04908701032400131,
+      "kl": 0.44599609375,
+      "learning_rate": 1.1356338783736256e-05,
+      "loss": 0.0178,
+      "reward": 1.1873884446918965,
+      "reward_std": 0.4431950330734253,
+      "rewards/accuracy_reward": 0.34799108523875477,
+      "rewards/format_reward": 0.8393973596394062,
+      "step": 145
+    },
+    {
+      "completion_length": 751.7043853759766,
+      "epoch": 0.5300353356890459,
+      "grad_norm": 0.09671700745820999,
+      "kl": 0.7987060546875,
+      "learning_rate": 1.0741427525516463e-05,
+      "loss": 0.0319,
+      "reward": 0.7703125398606062,
+      "reward_std": 0.5760972078889608,
+      "rewards/accuracy_reward": 0.28314733477309345,
+      "rewards/format_reward": 0.4871651985682547,
+      "step": 150
+    },
+    {
+      "completion_length": 997.5923439025879,
+      "epoch": 0.5477031802120141,
+      "grad_norm": 0.06775002181529999,
+      "kl": 0.5507080078125,
+      "learning_rate": 1.012368159663363e-05,
+      "loss": 0.022,
+      "reward": 0.19564733072184026,
+      "reward_std": 0.28081326140090823,
+      "rewards/accuracy_reward": 0.1513392934575677,
+      "rewards/format_reward": 0.04430803757859394,
+      "step": 155
+    },
+    {
+      "completion_length": 936.0987037658691,
+      "epoch": 0.5653710247349824,
+      "grad_norm": 0.010140431113541126,
+      "kl": 0.304296875,
+      "learning_rate": 9.505462800772612e-06,
+      "loss": 0.0122,
+      "reward": 0.11752232649596409,
+      "reward_std": 0.17188074269797654,
+      "rewards/accuracy_reward": 0.11283482633298263,
+      "rewards/format_reward": 0.004687500221189111,
+      "step": 160
+    },
+    {
+      "completion_length": 572.9960075378418,
+      "epoch": 0.5830388692579506,
+      "grad_norm": 0.01122345682233572,
+      "kl": 0.3446533203125,
+      "learning_rate": 8.889134749511956e-06,
+      "loss": 0.0138,
+      "reward": 0.07767857551807537,
+      "reward_std": 0.1298783538164571,
+      "rewards/accuracy_reward": 0.07700893274741247,
+      "rewards/format_reward": 0.0006696428870782256,
+      "step": 165
+    },
+    {
+      "completion_length": 445.03171615600587,
+      "epoch": 0.6007067137809188,
+      "grad_norm": 0.03271189704537392,
+      "kl": 0.393603515625,
+      "learning_rate": 8.277053825620836e-06,
+      "loss": 0.0157,
+      "reward": 0.07667411079164595,
+      "reward_std": 0.1326357247773558,
+      "rewards/accuracy_reward": 0.07566964647267013,
+      "rewards/format_reward": 0.0010044643306173385,
+      "step": 170
+    },
+    {
+      "completion_length": 444.26196403503417,
+      "epoch": 0.6183745583038869,
+      "grad_norm": 0.04383590444922447,
+      "kl": 0.4544677734375,
+      "learning_rate": 7.671560173993588e-06,
+      "loss": 0.0182,
+      "reward": 0.11026786249130964,
+      "reward_std": 0.1732286940328777,
+      "rewards/accuracy_reward": 0.109933041036129,
+      "rewards/format_reward": 0.0003348214435391128,
+      "step": 175
+    },
+    {
+      "completion_length": 441.78182754516604,
+      "epoch": 0.6360424028268551,
+      "grad_norm": 0.015269740484654903,
+      "kl": 0.4688232421875,
+      "learning_rate": 7.07496875466589e-06,
+      "loss": 0.0188,
+      "reward": 0.18348215110599994,
+      "reward_std": 0.22411425951868297,
+      "rewards/accuracy_reward": 0.1827009011991322,
+      "rewards/format_reward": 0.0007812500349245966,
+      "step": 180
+    },
+    {
+      "completion_length": 441.2130783081055,
+      "epoch": 0.6537102473498233,
+      "grad_norm": 0.09357426315546036,
+      "kl": 0.55419921875,
+      "learning_rate": 6.489560492119225e-06,
+      "loss": 0.0222,
+      "reward": 0.2385044758208096,
+      "reward_std": 0.24473562026396395,
+      "rewards/accuracy_reward": 0.23794644037261606,
+      "rewards/format_reward": 0.0005580357392318547,
+      "step": 185
+    },
+    {
+      "completion_length": 419.17278785705565,
+      "epoch": 0.6713780918727915,
+      "grad_norm": 0.020369287580251694,
+      "kl": 0.49603271484375,
+      "learning_rate": 5.9175735547120975e-06,
+      "loss": 0.0198,
+      "reward": 0.26886161994189023,
+      "reward_std": 0.2588370142504573,
+      "rewards/accuracy_reward": 0.2681919766589999,
+      "rewards/format_reward": 0.0006696428870782256,
+      "step": 190
+    },
+    {
+      "completion_length": 404.33562088012695,
+      "epoch": 0.6890459363957597,
+      "grad_norm": 0.012796707451343536,
+      "kl": 0.49130859375,
+      "learning_rate": 5.361194797579108e-06,
+      "loss": 0.0197,
+      "reward": 0.3069196566939354,
+      "reward_std": 0.2584508447907865,
+      "rewards/accuracy_reward": 0.3060267990455031,
+      "rewards/format_reward": 0.0008928571827709675,
+      "step": 195
+    },
+    {
+      "completion_length": 399.09253997802733,
+      "epoch": 0.7067137809187279,
+      "grad_norm": 0.029935084283351898,
+      "kl": 0.470458984375,
+      "learning_rate": 4.8225514017138205e-06,
+      "loss": 0.0188,
+      "reward": 0.29910715743899347,
+      "reward_std": 0.255822978913784,
+      "rewards/accuracy_reward": 0.2979910857975483,
+      "rewards/format_reward": 0.0011160714784637094,
+      "step": 200
+    },
+    {
+      "epoch": 0.7067137809187279,
+      "eval_completion_length": 405.41170501708984,
+      "eval_kl": 0.54638671875,
+      "eval_loss": 0.02145611308515072,
+      "eval_reward": 0.3470982313156128,
+      "eval_reward_std": 0.2499225027859211,
+      "eval_rewards/accuracy_reward": 0.3470982313156128,
+      "eval_rewards/format_reward": 0.0,
+      "eval_runtime": 62.8803,
+      "eval_samples_per_second": 1.574,
+      "eval_steps_per_second": 0.016,
+      "step": 200
+    },
+    {
+      "completion_length": 406.24164657592775,
+      "epoch": 0.7243816254416962,
+      "grad_norm": 0.040414728224277496,
+      "kl": 0.4975341796875,
+      "learning_rate": 4.303702741201431e-06,
+      "loss": 0.0199,
+      "reward": 0.29620537012815473,
+      "reward_std": 0.2518943578004837,
+      "rewards/accuracy_reward": 0.29564733505249025,
+      "rewards/format_reward": 0.0005580357392318547,
+      "step": 205
+    },
+    {
+      "completion_length": 403.4329418182373,
+      "epoch": 0.7420494699646644,
+      "grad_norm": 0.030647961422801018,
+      "kl": 0.47685546875,
+      "learning_rate": 3.8066325096949153e-06,
+      "loss": 0.0191,
+      "reward": 0.3157366219907999,
+      "reward_std": 0.25109751569107175,
+      "rewards/accuracy_reward": 0.31484376527369023,
+      "rewards/format_reward": 0.0008928571827709675,
+      "step": 210
+    },
+    {
+      "completion_length": 406.2974487304688,
+      "epoch": 0.7597173144876325,
+      "grad_norm": 0.02896731346845627,
+      "kl": 0.5581298828125,
+      "learning_rate": 3.3332411362372063e-06,
+      "loss": 0.0223,
+      "reward": 0.3091517997905612,
+      "reward_std": 0.2563653405755758,
+      "rewards/accuracy_reward": 0.3085937643423676,
+      "rewards/format_reward": 0.0005580357392318547,
+      "step": 215
+    },
+    {
+      "completion_length": 395.1335018157959,
+      "epoch": 0.7773851590106007,
+      "grad_norm": 0.013939591124653816,
+      "kl": 0.58065185546875,
+      "learning_rate": 2.8853385194256677e-06,
+      "loss": 0.0232,
+      "reward": 0.31049108412116766,
+      "reward_std": 0.253873204998672,
+      "rewards/accuracy_reward": 0.3099330486729741,
+      "rewards/format_reward": 0.0005580357392318547,
+      "step": 220
+    },
+    {
+      "completion_length": 406.0737899780273,
+      "epoch": 0.7950530035335689,
+      "grad_norm": 0.03492565453052521,
+      "kl": 0.5447265625,
+      "learning_rate": 2.464637107698046e-06,
+      "loss": 0.0218,
+      "reward": 0.3137276933528483,
+      "reward_std": 0.2610215944238007,
+      "rewards/accuracy_reward": 0.3131696577183902,
+      "rewards/format_reward": 0.0005580357392318547,
+      "step": 225
+    },
+    {
+      "completion_length": 398.26832046508787,
+      "epoch": 0.8127208480565371,
+      "grad_norm": 0.014227672480046749,
+      "kl": 0.52509765625,
+      "learning_rate": 2.072745352195794e-06,
+      "loss": 0.021,
+      "reward": 0.31406251452863215,
+      "reward_std": 0.25744989719241856,
+      "rewards/accuracy_reward": 0.31328126452863214,
+      "rewards/format_reward": 0.0007812500349245966,
+      "step": 230
+    },
+    {
+      "completion_length": 403.2072723388672,
+      "epoch": 0.8303886925795053,
+      "grad_norm": 0.061746783554553986,
+      "kl": 0.47880859375,
+      "learning_rate": 1.7111615572361628e-06,
+      "loss": 0.0192,
+      "reward": 0.3147321568801999,
+      "reward_std": 0.2717663310468197,
+      "rewards/accuracy_reward": 0.31417412105947734,
+      "rewards/format_reward": 0.0005580357392318547,
+      "step": 235
+    },
+    {
+      "completion_length": 406.78651428222656,
+      "epoch": 0.8480565371024735,
+      "grad_norm": 0.05263536050915718,
+      "kl": 0.5779541015625,
+      "learning_rate": 1.381268151904298e-06,
+      "loss": 0.0231,
+      "reward": 0.31506697898730635,
+      "reward_std": 0.2468398400582373,
+      "rewards/accuracy_reward": 0.3146205506287515,
+      "rewards/format_reward": 0.00044642859138548373,
+      "step": 240
+    },
+    {
+      "completion_length": 415.2616268157959,
+      "epoch": 0.8657243816254417,
+      "grad_norm": 0.01796867325901985,
+      "kl": 0.5435302734375,
+      "learning_rate": 1.0843264046665558e-06,
+      "loss": 0.0217,
+      "reward": 0.3033482272177935,
+      "reward_std": 0.24437342984601856,
+      "rewards/accuracy_reward": 0.3030134059488773,
+      "rewards/format_reward": 0.0003348214435391128,
+      "step": 245
+    },
+    {
+      "completion_length": 420.59242935180663,
+      "epoch": 0.8833922261484098,
+      "grad_norm": 0.014822212047874928,
+      "kl": 0.6258056640625,
+      "learning_rate": 8.214716012124491e-07,
+      "loss": 0.025,
+      "reward": 0.29631697880104185,
+      "reward_std": 0.2492463287897408,
+      "rewards/accuracy_reward": 0.29575894298031924,
+      "rewards/format_reward": 0.0005580357392318547,
+      "step": 250
+    },
+    {
+      "completion_length": 424.70124740600585,
+      "epoch": 0.901060070671378,
+      "grad_norm": 0.0215270034968853,
+      "kl": 0.586328125,
+      "learning_rate": 5.937087039615619e-07,
+      "loss": 0.0234,
+      "reward": 0.29787947833538053,
+      "reward_std": 0.24997795289382338,
+      "rewards/accuracy_reward": 0.29765626415610313,
+      "rewards/format_reward": 0.00022321429569274187,
+      "step": 255
+    },
+    {
+      "completion_length": 419.2236808776855,
+      "epoch": 0.9187279151943463,
+      "grad_norm": 0.01607878878712654,
+      "kl": 0.5501953125,
+      "learning_rate": 4.019085098303077e-07,
+      "loss": 0.022,
+      "reward": 0.3072544798254967,
+      "reward_std": 0.25280203791335226,
+      "rewards/accuracy_reward": 0.30691965818405154,
+      "rewards/format_reward": 0.0003348214435391128,
+      "step": 260
+    },
+    {
+      "completion_length": 424.92613525390624,
+      "epoch": 0.9363957597173145,
+      "grad_norm": 0.03978170454502106,
+      "kl": 0.533837890625,
+      "learning_rate": 2.4680432094837394e-07,
+      "loss": 0.0214,
+      "reward": 0.2920759066939354,
+      "reward_std": 0.24946068711578845,
+      "rewards/accuracy_reward": 0.29151787124574186,
+      "rewards/format_reward": 0.0005580357392318547,
+      "step": 265
+    },
+    {
+      "completion_length": 426.96754188537597,
+      "epoch": 0.9540636042402827,
+      "grad_norm": 0.027451133355498314,
+      "kl": 0.5416015625,
+      "learning_rate": 1.289891410535593e-07,
+      "loss": 0.0217,
+      "reward": 0.29977679867297413,
+      "reward_std": 0.2573161173611879,
+      "rewards/accuracy_reward": 0.2993303703144193,
+      "rewards/format_reward": 0.00044642859138548373,
+      "step": 270
+    },
+    {
+      "completion_length": 420.65091171264646,
+      "epoch": 0.9717314487632509,
+      "grad_norm": 0.026210250332951546,
+      "kl": 0.568017578125,
+      "learning_rate": 4.8913408283934874e-08,
+      "loss": 0.0227,
+      "reward": 0.2981026927009225,
+      "reward_std": 0.25082770437002183,
+      "rewards/accuracy_reward": 0.29754465706646444,
+      "rewards/format_reward": 0.0005580357392318547,
+      "step": 275
+    },
+    {
+      "completion_length": 420.31932601928713,
+      "epoch": 0.9893992932862191,
+      "grad_norm": 0.023707764223217964,
+      "kl": 0.5349365234375,
+      "learning_rate": 6.883273035447335e-09,
+      "loss": 0.0214,
+      "reward": 0.31428572740405797,
+      "reward_std": 0.267278019990772,
+      "rewards/accuracy_reward": 0.31328126322478056,
+      "rewards/format_reward": 0.0010044643306173385,
+      "step": 280
+    },
+    {
+      "completion_length": 417.32022921244305,
+      "epoch": 1.0,
+      "kl": 0.5493570963541666,
+      "reward": 0.30375745457907516,
+      "reward_std": 0.2610333391154806,
+      "rewards/accuracy_reward": 0.30319941850999993,
+      "rewards/format_reward": 0.0005580357392318547,
+      "step": 283,
+      "total_flos": 0.0,
+      "train_loss": 0.7804505577981542,
+      "train_runtime": 45227.2818,
+      "train_samples_per_second": 1.602,
+      "train_steps_per_second": 0.006
+    }
+  ],
+  "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 sha256:51dd48c8006e74811d7f9d39bf6efb6ed77db2db4113c477666ae2f41aa0d9eb
+size 7480