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="kppkkp/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/hasueryu-project/open-r1/runs/p2n44gck) 
+
+
+This model was trained with GRPO, a method introduced in [DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models](https://huggingface.co/papers/2402.03300).
+
+### Framework versions
+
+- TRL: 0.16.0.dev0
+- Transformers: 4.49.0.dev0
+- 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": 0.040519028099076065,
+    "train_runtime": 54178.7039,
+    "train_samples": 72441,
+    "train_samples_per_second": 1.337,
+    "train_steps_per_second": 0.005
+}

config.json

@@ -0,0 +1,29 @@
+{
+  "_name_or_path": "/nlp_group/yulinhao/models/Qwen/Qwen2.5-1.5B-Instruct",
+  "architectures": [
+    "Qwen2ForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 151643,
+  "eos_token_id": 151645,
+  "hidden_act": "silu",
+  "hidden_size": 1536,
+  "initializer_range": 0.02,
+  "intermediate_size": 8960,
+  "max_position_embeddings": 32768,
+  "max_window_layers": 21,
+  "model_type": "qwen2",
+  "num_attention_heads": 12,
+  "num_hidden_layers": 28,
+  "num_key_value_heads": 2,
+  "rms_norm_eps": 1e-06,
+  "rope_scaling": null,
+  "rope_theta": 1000000.0,
+  "sliding_window": 32768,
+  "tie_word_embeddings": true,
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.49.0.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 sha256:42ece6821a1bb28efba4bf7d1e004ec669e01e49c61f517e05e85bbab83c0e88
+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": 0.040519028099076065,
+    "train_runtime": 54178.7039,
+    "train_samples": 72441,
+    "train_samples_per_second": 1.337,
+    "train_steps_per_second": 0.005
+}

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": 385.89532871246337,
+      "epoch": 0.0176678445229682,
+      "grad_norm": 0.5257675647735596,
+      "kl": 0.0012482523918151856,
+      "learning_rate": 3.448275862068966e-06,
+      "loss": 0.0001,
+      "reward": 0.6444196742027998,
+      "reward_std": 0.4479222524911165,
+      "rewards/accuracy_reward": 0.1539062581025064,
+      "rewards/format_reward": 0.49051341526210307,
+      "step": 5
+    },
+    {
+      "completion_length": 229.56596965789794,
+      "epoch": 0.0353356890459364,
+      "grad_norm": 0.5334790945053101,
+      "kl": 0.3332817077636719,
+      "learning_rate": 6.896551724137932e-06,
+      "loss": 0.0133,
+      "reward": 0.9689732536673545,
+      "reward_std": 0.26624082447960973,
+      "rewards/accuracy_reward": 0.08214286140864716,
+      "rewards/format_reward": 0.8868303976953029,
+      "step": 10
+    },
+    {
+      "completion_length": 157.9798059463501,
+      "epoch": 0.053003533568904596,
+      "grad_norm": 0.287689745426178,
+      "kl": 0.04854583740234375,
+      "learning_rate": 1.0344827586206898e-05,
+      "loss": 0.0019,
+      "reward": 1.062388438731432,
+      "reward_std": 0.20184952337294818,
+      "rewards/accuracy_reward": 0.10022321890573949,
+      "rewards/format_reward": 0.962165217846632,
+      "step": 15
+    },
+    {
+      "completion_length": 171.60927028656005,
+      "epoch": 0.0706713780918728,
+      "grad_norm": 0.2567259669303894,
+      "kl": 0.078216552734375,
+      "learning_rate": 1.3793103448275863e-05,
+      "loss": 0.0031,
+      "reward": 1.1186384424567222,
+      "reward_std": 0.24078293070197104,
+      "rewards/accuracy_reward": 0.15345982832368463,
+      "rewards/format_reward": 0.9651786126196384,
+      "step": 20
+    },
+    {
+      "completion_length": 210.81998748779296,
+      "epoch": 0.08833922261484099,
+      "grad_norm": 0.5349090695381165,
+      "kl": 0.112786865234375,
+      "learning_rate": 1.7241379310344828e-05,
+      "loss": 0.0045,
+      "reward": 1.1583705946803093,
+      "reward_std": 0.3293624483048916,
+      "rewards/accuracy_reward": 0.21551340306177735,
+      "rewards/format_reward": 0.9428571835160255,
+      "step": 25
+    },
+    {
+      "completion_length": 293.2053703308105,
+      "epoch": 0.10600706713780919,
+      "grad_norm": 0.19419077038764954,
+      "kl": 0.09522705078125,
+      "learning_rate": 1.999923511388017e-05,
+      "loss": 0.0038,
+      "reward": 1.2208705872297287,
+      "reward_std": 0.36669483222067356,
+      "rewards/accuracy_reward": 0.2871651930734515,
+      "rewards/format_reward": 0.9337054029107094,
+      "step": 30
+    },
+    {
+      "completion_length": 323.5448799133301,
+      "epoch": 0.12367491166077739,
+      "grad_norm": 0.22766603529453278,
+      "kl": 0.39053955078125,
+      "learning_rate": 1.9972476383747748e-05,
+      "loss": 0.0157,
+      "reward": 1.2860491678118706,
+      "reward_std": 0.30488016121089456,
+      "rewards/accuracy_reward": 0.3202009078115225,
+      "rewards/format_reward": 0.9658482551574707,
+      "step": 35
+    },
+    {
+      "completion_length": 285.7590543746948,
+      "epoch": 0.1413427561837456,
+      "grad_norm": 0.3550393283367157,
+      "kl": 0.118524169921875,
+      "learning_rate": 1.9907590277344582e-05,
+      "loss": 0.0047,
+      "reward": 1.2512277334928512,
+      "reward_std": 0.34426755234599116,
+      "rewards/accuracy_reward": 0.30758929941803215,
+      "rewards/format_reward": 0.9436384335160255,
+      "step": 40
+    },
+    {
+      "completion_length": 269.9738945007324,
+      "epoch": 0.15901060070671377,
+      "grad_norm": 2.2805335521698,
+      "kl": 1.9856201171875,
+      "learning_rate": 1.9804824871166254e-05,
+      "loss": 0.0796,
+      "reward": 1.1523438021540642,
+      "reward_std": 0.46314122155308723,
+      "rewards/accuracy_reward": 0.316294657997787,
+      "rewards/format_reward": 0.8360491469502449,
+      "step": 45
+    },
+    {
+      "completion_length": 249.81585788726807,
+      "epoch": 0.17667844522968199,
+      "grad_norm": 6.213081359863281,
+      "kl": 3.11435546875,
+      "learning_rate": 1.9664573064143604e-05,
+      "loss": 0.1247,
+      "reward": 0.8168527133762836,
+      "reward_std": 0.42027441747486594,
+      "rewards/accuracy_reward": 0.22991072395816445,
+      "rewards/format_reward": 0.5869419902563096,
+      "step": 50
+    },
+    {
+      "completion_length": 837.4478942871094,
+      "epoch": 0.19434628975265017,
+      "grad_norm": 0.1708061844110489,
+      "kl": 0.7300537109375,
+      "learning_rate": 1.948737107548771e-05,
+      "loss": 0.0292,
+      "reward": 0.0875000040512532,
+      "reward_std": 0.16735761975869537,
+      "rewards/accuracy_reward": 0.03895089457510039,
+      "rewards/format_reward": 0.048549109499435875,
+      "step": 55
+    },
+    {
+      "completion_length": 1024.0,
+      "epoch": 0.21201413427561838,
+      "grad_norm": 0.16706973314285278,
+      "kl": 0.364111328125,
+      "learning_rate": 1.9273896394584103e-05,
+      "loss": 0.0146,
+      "reward": 0.09107143309665844,
+      "reward_std": 0.14619714424479752,
+      "rewards/accuracy_reward": 0.09084821873111651,
+      "rewards/format_reward": 0.00022321429569274187,
+      "step": 60
+    },
+    {
+      "completion_length": 1024.0,
+      "epoch": 0.22968197879858657,
+      "grad_norm": 0.10041823238134384,
+      "kl": 0.2697998046875,
+      "learning_rate": 1.9024965190774262e-05,
+      "loss": 0.0108,
+      "reward": 0.17421875786967575,
+      "reward_std": 0.23079481534659863,
+      "rewards/accuracy_reward": 0.17421875786967575,
+      "rewards/format_reward": 0.0,
+      "step": 65
+    },
+    {
+      "completion_length": 1024.0,
+      "epoch": 0.24734982332155478,
+      "grad_norm": 0.3122642934322357,
+      "kl": 0.28310546875,
+      "learning_rate": 1.8741529192927528e-05,
+      "loss": 0.0113,
+      "reward": 0.19218751024454833,
+      "reward_std": 0.23374766409397124,
+      "rewards/accuracy_reward": 0.19218751024454833,
+      "rewards/format_reward": 0.0,
+      "step": 70
+    },
+    {
+      "completion_length": 1024.0,
+      "epoch": 0.26501766784452296,
+      "grad_norm": 0.3012256622314453,
+      "kl": 0.3281494140625,
+      "learning_rate": 1.8424672050733577e-05,
+      "loss": 0.0131,
+      "reward": 0.20256697395816445,
+      "reward_std": 0.23207057397812605,
+      "rewards/accuracy_reward": 0.20122768832370638,
+      "rewards/format_reward": 0.0013392857741564511,
+      "step": 75
+    },
+    {
+      "completion_length": 1024.0,
+      "epoch": 0.2826855123674912,
+      "grad_norm": 0.2279616743326187,
+      "kl": 0.569970703125,
+      "learning_rate": 1.8075605191627242e-05,
+      "loss": 0.0228,
+      "reward": 0.1684151873923838,
+      "reward_std": 0.1943290094844997,
+      "rewards/accuracy_reward": 0.16808036593720316,
+      "rewards/format_reward": 0.0003348214435391128,
+      "step": 80
+    },
+    {
+      "completion_length": 1023.7799125671387,
+      "epoch": 0.3003533568904594,
+      "grad_norm": 0.1516251116991043,
+      "kl": 0.51630859375,
+      "learning_rate": 1.7695663189185703e-05,
+      "loss": 0.0207,
+      "reward": 0.21607143906876444,
+      "reward_std": 0.20886625591665506,
+      "rewards/accuracy_reward": 0.21607143906876444,
+      "rewards/format_reward": 0.0,
+      "step": 85
+    },
+    {
+      "completion_length": 1023.1735549926758,
+      "epoch": 0.31802120141342755,
+      "grad_norm": 0.09773323684930801,
+      "kl": 0.39691162109375,
+      "learning_rate": 1.7286298660705877e-05,
+      "loss": 0.0159,
+      "reward": 0.2613839410245419,
+      "reward_std": 0.2178319870494306,
+      "rewards/accuracy_reward": 0.2613839410245419,
+      "rewards/format_reward": 0.0,
+      "step": 90
+    },
+    {
+      "completion_length": 1022.9333808898925,
+      "epoch": 0.33568904593639576,
+      "grad_norm": 0.2660332918167114,
+      "kl": 0.3833740234375,
+      "learning_rate": 1.6849076713469914e-05,
+      "loss": 0.0153,
+      "reward": 0.2964285839349031,
+      "reward_std": 0.23299281364306806,
+      "rewards/accuracy_reward": 0.2964285839349031,
+      "rewards/format_reward": 0.0,
+      "step": 95
+    },
+    {
+      "completion_length": 1022.3206634521484,
+      "epoch": 0.35335689045936397,
+      "grad_norm": 1.3615121841430664,
+      "kl": 0.835400390625,
+      "learning_rate": 1.6385668960932143e-05,
+      "loss": 0.0334,
+      "reward": 0.1241071482654661,
+      "reward_std": 0.16077298847958446,
+      "rewards/accuracy_reward": 0.1241071482654661,
+      "rewards/format_reward": 0.0,
+      "step": 100
+    },
+    {
+      "epoch": 0.35335689045936397,
+      "eval_completion_length": 1021.8020172119141,
+      "eval_kl": 0.794921875,
+      "eval_loss": 0.03294466808438301,
+      "eval_reward": 0.1272321492433548,
+      "eval_reward_std": 0.20546478778123856,
+      "eval_rewards/accuracy_reward": 0.1272321492433548,
+      "eval_rewards/format_reward": 0.0,
+      "eval_runtime": 83.4283,
+      "eval_samples_per_second": 1.187,
+      "eval_steps_per_second": 0.012,
+      "step": 100
+    },
+    {
+      "completion_length": 1023.4450950622559,
+      "epoch": 0.3710247349823322,
+      "grad_norm": 1.5146944522857666,
+      "kl": 1.085205078125,
+      "learning_rate": 1.5897847131705194e-05,
+      "loss": 0.0434,
+      "reward": 0.12544643497094513,
+      "reward_std": 0.1771216381341219,
+      "rewards/accuracy_reward": 0.12544643497094513,
+      "rewards/format_reward": 0.0,
+      "step": 105
+    },
+    {
+      "completion_length": 1023.4677505493164,
+      "epoch": 0.38869257950530034,
+      "grad_norm": 0.6001514196395874,
+      "kl": 1.1712890625,
+      "learning_rate": 1.5387476295779737e-05,
+      "loss": 0.0469,
+      "reward": 0.16138393647270277,
+      "reward_std": 0.20901122770737857,
+      "rewards/accuracy_reward": 0.15982143628643825,
+      "rewards/format_reward": 0.0015625000698491931,
+      "step": 110
+    },
+    {
+      "completion_length": 1023.904689025879,
+      "epoch": 0.40636042402826855,
+      "grad_norm": 0.357604444026947,
+      "kl": 0.9640380859375,
+      "learning_rate": 1.4856507733875837e-05,
+      "loss": 0.0386,
+      "reward": 0.2774553706869483,
+      "reward_std": 0.22977914968505503,
+      "rewards/accuracy_reward": 0.27712054904550315,
+      "rewards/format_reward": 0.0003348214435391128,
+      "step": 115
+    },
+    {
+      "completion_length": 1023.7766761779785,
+      "epoch": 0.42402826855123676,
+      "grad_norm": 2.9171788692474365,
+      "kl": 1.54794921875,
+      "learning_rate": 1.4306971477188223e-05,
+      "loss": 0.0619,
+      "reward": 0.29587054867297413,
+      "reward_std": 0.22431441079825162,
+      "rewards/accuracy_reward": 0.2954241203144193,
+      "rewards/format_reward": 0.00044642859138548373,
+      "step": 120
+    },
+    {
+      "completion_length": 1023.5507850646973,
+      "epoch": 0.4416961130742049,
+      "grad_norm": 1.7767579555511475,
+      "kl": 1.08740234375,
+      "learning_rate": 1.3740968546047935e-05,
+      "loss": 0.0435,
+      "reward": 0.26272322572767737,
+      "reward_std": 0.2170619947835803,
+      "rewards/accuracy_reward": 0.2623884044587612,
+      "rewards/format_reward": 0.0003348214435391128,
+      "step": 125
+    },
+    {
+      "completion_length": 1023.353914642334,
+      "epoch": 0.45936395759717313,
+      "grad_norm": 0.626575231552124,
+      "kl": 1.624072265625,
+      "learning_rate": 1.3160662917174045e-05,
+      "loss": 0.065,
+      "reward": 0.2738839427009225,
+      "reward_std": 0.21952001163735985,
+      "rewards/accuracy_reward": 0.27332590725272893,
+      "rewards/format_reward": 0.0005580357392318547,
+      "step": 130
+    },
+    {
+      "completion_length": 1023.6706512451171,
+      "epoch": 0.47703180212014135,
+      "grad_norm": 2.2120277881622314,
+      "kl": 1.8474609375,
+      "learning_rate": 1.2568273250226681e-05,
+      "loss": 0.0739,
+      "reward": 0.27935269083827735,
+      "reward_std": 0.2362092829309404,
+      "rewards/accuracy_reward": 0.2791294766589999,
+      "rewards/format_reward": 0.00022321429569274187,
+      "step": 135
+    },
+    {
+      "completion_length": 1023.9494422912597,
+      "epoch": 0.49469964664310956,
+      "grad_norm": 3.246752977371216,
+      "kl": 2.728662109375,
+      "learning_rate": 1.1966064405292887e-05,
+      "loss": 0.1091,
+      "reward": 0.2922991219907999,
+      "reward_std": 0.23428730978630483,
+      "rewards/accuracy_reward": 0.2919643005356193,
+      "rewards/format_reward": 0.0003348214435391128,
+      "step": 140
+    },
+    {
+      "completion_length": 1023.9381698608398,
+      "epoch": 0.5123674911660777,
+      "grad_norm": 11.377167701721191,
+      "kl": 2.4001953125,
+      "learning_rate": 1.1356338783736256e-05,
+      "loss": 0.096,
+      "reward": 0.31104912236332893,
+      "reward_std": 0.2248768277466297,
+      "rewards/accuracy_reward": 0.31093751527369023,
+      "rewards/format_reward": 0.00011160714784637093,
+      "step": 145
+    },
+    {
+      "completion_length": 1023.8702018737793,
+      "epoch": 0.5300353356890459,
+      "grad_norm": 4.7356181144714355,
+      "kl": 0.99365234375,
+      "learning_rate": 1.0741427525516463e-05,
+      "loss": 0.0398,
+      "reward": 0.2786830499768257,
+      "reward_std": 0.22981408620253205,
+      "rewards/accuracy_reward": 0.2775669790804386,
+      "rewards/format_reward": 0.0011160714784637094,
+      "step": 150
+    },
+    {
+      "completion_length": 1023.8892868041992,
+      "epoch": 0.5477031802120141,
+      "grad_norm": 0.48026323318481445,
+      "kl": 1.191552734375,
+      "learning_rate": 1.012368159663363e-05,
+      "loss": 0.0477,
+      "reward": 0.2672991200350225,
+      "reward_std": 0.2149678454734385,
+      "rewards/accuracy_reward": 0.2670759057626128,
+      "rewards/format_reward": 0.00022321429569274187,
+      "step": 155
+    },
+    {
+      "completion_length": 1023.9833709716797,
+      "epoch": 0.5653710247349824,
+      "grad_norm": 0.2914314866065979,
+      "kl": 0.31871337890625,
+      "learning_rate": 9.505462800772612e-06,
+      "loss": 0.0127,
+      "reward": 0.2546875115483999,
+      "reward_std": 0.21248297598212956,
+      "rewards/accuracy_reward": 0.2534598330967128,
+      "rewards/format_reward": 0.0012276786263100802,
+      "step": 160
+    },
+    {
+      "completion_length": 1023.8994430541992,
+      "epoch": 0.5830388692579506,
+      "grad_norm": 0.403013676404953,
+      "kl": 0.38603515625,
+      "learning_rate": 8.889134749511956e-06,
+      "loss": 0.0154,
+      "reward": 0.23883929578587412,
+      "reward_std": 0.21736905500292777,
+      "rewards/accuracy_reward": 0.23627233160659672,
+      "rewards/format_reward": 0.0025669644004665316,
+      "step": 165
+    },
+    {
+      "completion_length": 1023.8919662475586,
+      "epoch": 0.6007067137809188,
+      "grad_norm": 2.4790608882904053,
+      "kl": 0.843505859375,
+      "learning_rate": 8.277053825620836e-06,
+      "loss": 0.0337,
+      "reward": 0.22020090287551283,
+      "reward_std": 0.225645115878433,
+      "rewards/accuracy_reward": 0.21886161714792252,
+      "rewards/format_reward": 0.0013392857741564511,
+      "step": 170
+    },
+    {
+      "completion_length": 1023.5271240234375,
+      "epoch": 0.6183745583038869,
+      "grad_norm": 1.517407774925232,
+      "kl": 1.15859375,
+      "learning_rate": 7.671560173993588e-06,
+      "loss": 0.0463,
+      "reward": 0.2196428684517741,
+      "reward_std": 0.22536969408392907,
+      "rewards/accuracy_reward": 0.21886161863803863,
+      "rewards/format_reward": 0.0007812500349245966,
+      "step": 175
+    },
+    {
+      "completion_length": 1023.7799140930176,
+      "epoch": 0.6360424028268551,
+      "grad_norm": 3.477236747741699,
+      "kl": 1.853125,
+      "learning_rate": 7.07496875466589e-06,
+      "loss": 0.0742,
+      "reward": 0.18906251017469913,
+      "reward_std": 0.21059290650300683,
+      "rewards/accuracy_reward": 0.18861608181614428,
+      "rewards/format_reward": 0.00044642859138548373,
+      "step": 180
+    },
+    {
+      "completion_length": 1023.8558059692383,
+      "epoch": 0.6537102473498233,
+      "grad_norm": 4.594732284545898,
+      "kl": 1.274169921875,
+      "learning_rate": 6.489560492119225e-06,
+      "loss": 0.051,
+      "reward": 0.22857143869623542,
+      "reward_std": 0.23066243380308152,
+      "rewards/accuracy_reward": 0.22745536724105478,
+      "rewards/format_reward": 0.0011160714784637094,
+      "step": 185
+    },
+    {
+      "completion_length": 1023.9906257629394,
+      "epoch": 0.6713780918727915,
+      "grad_norm": 5.503405570983887,
+      "kl": 1.36650390625,
+      "learning_rate": 5.9175735547120975e-06,
+      "loss": 0.0547,
+      "reward": 0.2881696566008031,
+      "reward_std": 0.24815111914649607,
+      "rewards/accuracy_reward": 0.28437501201406123,
+      "rewards/format_reward": 0.0037946430500596763,
+      "step": 190
+    },
+    {
+      "completion_length": 1023.9542419433594,
+      "epoch": 0.6890459363957597,
+      "grad_norm": 1.3981853723526,
+      "kl": 2.093359375,
+      "learning_rate": 5.361194797579108e-06,
+      "loss": 0.0838,
+      "reward": 0.29575894251465795,
+      "reward_std": 0.2531373543664813,
+      "rewards/accuracy_reward": 0.28939733523875477,
+      "rewards/format_reward": 0.0063616074505262075,
+      "step": 195
+    },
+    {
+      "completion_length": 1023.9152908325195,
+      "epoch": 0.7067137809187279,
+      "grad_norm": 7.645321369171143,
+      "kl": 1.69521484375,
+      "learning_rate": 4.8225514017138205e-06,
+      "loss": 0.0678,
+      "reward": 0.26662947684526445,
+      "reward_std": 0.2479257956147194,
+      "rewards/accuracy_reward": 0.25881697619333865,
+      "rewards/format_reward": 0.007812500360887497,
+      "step": 200
+    },
+    {
+      "epoch": 0.7067137809187279,
+      "eval_completion_length": 1024.0,
+      "eval_kl": 2.50390625,
+      "eval_loss": 0.10209327191114426,
+      "eval_reward": 0.310267873108387,
+      "eval_reward_std": 0.22309495136141777,
+      "eval_rewards/accuracy_reward": 0.305803582072258,
+      "eval_rewards/format_reward": 0.004464285913854837,
+      "eval_runtime": 91.3405,
+      "eval_samples_per_second": 1.084,
+      "eval_steps_per_second": 0.011,
+      "step": 200
+    },
+    {
+      "completion_length": 1024.0,
+      "epoch": 0.7243816254416962,
+      "grad_norm": 4.966090679168701,
+      "kl": 2.449951171875,
+      "learning_rate": 4.303702741201431e-06,
+      "loss": 0.098,
+      "reward": 0.24832590399309992,
+      "reward_std": 0.24168844958767294,
+      "rewards/accuracy_reward": 0.24118304727599024,
+      "rewards/format_reward": 0.007142857485450804,
+      "step": 205
+    },
+    {
+      "completion_length": 1023.8795768737793,
+      "epoch": 0.7420494699646644,
+      "grad_norm": 7.903662204742432,
+      "kl": 1.901416015625,
+      "learning_rate": 3.8066325096949153e-06,
+      "loss": 0.076,
+      "reward": 0.25558037031441927,
+      "reward_std": 0.2529288594610989,
+      "rewards/accuracy_reward": 0.2431919751688838,
+      "rewards/format_reward": 0.012388393504079432,
+      "step": 210
+    },
+    {
+      "completion_length": 1023.8687515258789,
+      "epoch": 0.7597173144876325,
+      "grad_norm": 2.2350668907165527,
+      "kl": 1.812060546875,
+      "learning_rate": 3.3332411362372063e-06,
+      "loss": 0.0725,
+      "reward": 0.258482154738158,
+      "reward_std": 0.26162059921771286,
+      "rewards/accuracy_reward": 0.24118304559960962,
+      "rewards/format_reward": 0.01729910804424435,
+      "step": 215
+    },
+    {
+      "completion_length": 1023.8536842346191,
+      "epoch": 0.7773851590106007,
+      "grad_norm": 2.8165674209594727,
+      "kl": 1.880078125,
+      "learning_rate": 2.8853385194256677e-06,
+      "loss": 0.0752,
+      "reward": 0.24375001061707735,
+      "reward_std": 0.25496505089104177,
+      "rewards/accuracy_reward": 0.23002233263105154,
+      "rewards/format_reward": 0.013727679383009672,
+      "step": 220
+    },
+    {
+      "completion_length": 1024.0,
+      "epoch": 0.7950530035335689,
+      "grad_norm": 3.0542807579040527,
+      "kl": 1.645703125,
+      "learning_rate": 2.464637107698046e-06,
+      "loss": 0.0659,
+      "reward": 0.26216519055888055,
+      "reward_std": 0.27050148248672484,
+      "rewards/accuracy_reward": 0.23950894009321927,
+      "rewards/format_reward": 0.022656251245643945,
+      "step": 225
+    },
+    {
+      "completion_length": 1023.835604095459,
+      "epoch": 0.8127208480565371,
+      "grad_norm": 1.1989268064498901,
+      "kl": 1.440283203125,
+      "learning_rate": 2.072745352195794e-06,
+      "loss": 0.0576,
+      "reward": 0.2896205481141806,
+      "reward_std": 0.28690752685070037,
+      "rewards/accuracy_reward": 0.2671875134110451,
+      "rewards/format_reward": 0.0224330369848758,
+      "step": 230
+    },
+    {
+      "completion_length": 1023.9050231933594,
+      "epoch": 0.8303886925795053,
+      "grad_norm": 0.9027553200721741,
+      "kl": 1.21279296875,
+      "learning_rate": 1.7111615572361628e-06,
+      "loss": 0.0485,
+      "reward": 0.2984375137835741,
+      "reward_std": 0.29320847503840924,
+      "rewards/accuracy_reward": 0.27187501136213543,
+      "rewards/format_reward": 0.026562501350417732,
+      "step": 235
+    },
+    {
+      "completion_length": 1023.9849334716797,
+      "epoch": 0.8480565371024735,
+      "grad_norm": 1.0079853534698486,
+      "kl": 0.9890380859375,
+      "learning_rate": 1.381268151904298e-06,
+      "loss": 0.0396,
+      "reward": 0.3024553697556257,
+      "reward_std": 0.3057010589167476,
+      "rewards/accuracy_reward": 0.2669642990455031,
+      "rewards/format_reward": 0.03549107307335362,
+      "step": 240
+    },
+    {
+      "completion_length": 1023.9882820129394,
+      "epoch": 0.8657243816254417,
+      "grad_norm": 0.6375559568405151,
+      "kl": 0.80595703125,
+      "learning_rate": 1.0843264046665558e-06,
+      "loss": 0.0322,
+      "reward": 0.28783483654260633,
+      "reward_std": 0.2901507246308029,
+      "rewards/accuracy_reward": 0.256250012293458,
+      "rewards/format_reward": 0.0315848228870891,
+      "step": 245
+    },
+    {
+      "completion_length": 1023.8103805541992,
+      "epoch": 0.8833922261484098,
+      "grad_norm": 0.3038390576839447,
+      "kl": 0.7016357421875,
+      "learning_rate": 8.214716012124491e-07,
+      "loss": 0.0281,
+      "reward": 0.28537947684526443,
+      "reward_std": 0.2940663579851389,
+      "rewards/accuracy_reward": 0.2517857262864709,
+      "rewards/format_reward": 0.033593751536682245,
+      "step": 250
+    },
+    {
+      "completion_length": 1023.9936386108399,
+      "epoch": 0.901060070671378,
+      "grad_norm": 0.7605869174003601,
+      "kl": 0.6392333984375,
+      "learning_rate": 5.937087039615619e-07,
+      "loss": 0.0256,
+      "reward": 0.3013393010944128,
+      "reward_std": 0.2973380209878087,
+      "rewards/accuracy_reward": 0.26551340594887735,
+      "rewards/format_reward": 0.035825894423760475,
+      "step": 255
+    },
+    {
+      "completion_length": 1024.0,
+      "epoch": 0.9187279151943463,
+      "grad_norm": 0.4661541283130646,
+      "kl": 0.6113037109375,
+      "learning_rate": 4.019085098303077e-07,
+      "loss": 0.0245,
+      "reward": 0.30870537031441925,
+      "reward_std": 0.301556083932519,
+      "rewards/accuracy_reward": 0.27533483430743216,
+      "rewards/format_reward": 0.03337053736904636,
+      "step": 260
+    },
+    {
+      "completion_length": 1023.9744422912597,
+      "epoch": 0.9363957597173145,
+      "grad_norm": 0.5394344925880432,
+      "kl": 0.59677734375,
+      "learning_rate": 2.4680432094837394e-07,
+      "loss": 0.0239,
+      "reward": 0.2986607299186289,
+      "reward_std": 0.3167744716629386,
+      "rewards/accuracy_reward": 0.2603794766589999,
+      "rewards/format_reward": 0.03828125168802217,
+      "step": 265
+    },
+    {
+      "completion_length": 1023.7898460388184,
+      "epoch": 0.9540636042402827,
+      "grad_norm": 0.35241708159446716,
+      "kl": 0.6021240234375,
+      "learning_rate": 1.289891410535593e-07,
+      "loss": 0.0241,
+      "reward": 0.30212054755538703,
+      "reward_std": 0.30877320375293493,
+      "rewards/accuracy_reward": 0.2652901913970709,
+      "rewards/format_reward": 0.036830358975566926,
+      "step": 270
+    },
+    {
+      "completion_length": 1023.873885345459,
+      "epoch": 0.9717314487632509,
+      "grad_norm": 0.3670821189880371,
+      "kl": 0.5589111328125,
+      "learning_rate": 4.8913408283934874e-08,
+      "loss": 0.0224,
+      "reward": 0.30368304885923864,
+      "reward_std": 0.30286577958613636,
+      "rewards/accuracy_reward": 0.2606026901863515,
+      "rewards/format_reward": 0.04308035911526531,
+      "step": 275
+    },
+    {
+      "completion_length": 1023.9140632629394,
+      "epoch": 0.9893992932862191,
+      "grad_norm": 0.4401331841945648,
+      "kl": 0.5879638671875,
+      "learning_rate": 6.883273035447335e-09,
+      "loss": 0.0235,
+      "reward": 0.31294644335284827,
+      "reward_std": 0.3160593772307038,
+      "rewards/accuracy_reward": 0.2747767997905612,
+      "rewards/format_reward": 0.038169644912704824,
+      "step": 280
+    },
+    {
+      "completion_length": 1023.8789800008138,
+      "epoch": 1.0,
+      "kl": 0.8118896484375,
+      "reward": 0.30822173940638703,
+      "reward_std": 0.3040016482894619,
+      "rewards/accuracy_reward": 0.27101935756703216,
+      "rewards/format_reward": 0.03720238265426209,
+      "step": 283,
+      "total_flos": 0.0,
+      "train_loss": 0.040519028099076065,
+      "train_runtime": 54178.7039,
+      "train_samples_per_second": 1.337,
+      "train_steps_per_second": 0.005
+    }
+  ],
+  "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 sha256:951891df8dddf769f5f5706f5ea995ef82af3f17c11f27d4ae40dc27fddb76c4
+size 7672