Add files using upload-large-folder tool

.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: DeepSeek-R1-Distill-Qwen-7B-S
+tags:
+- generated_from_trainer
+- trl
+- grpo
+licence: license
+---
+
+# Model Card for DeepSeek-R1-Distill-Qwen-7B-S
+
+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="leonMW/DeepSeek-R1-Distill-Qwen-7B-S", 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/leonwenderoth-tu-darmstadt/huggingface/runs/zjrhqpsu) 
+
+
+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: 0.23.0
+- Transformers: 4.56.2
+- Pytorch: 2.7.1
+- Datasets: 4.1.1
+- Tokenizers: 0.22.1
+
+## Citations
+
+Cite GRPO as:
+
+```bibtex
+@article{shao2024deepseekmath,
+    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 Gallou{\'e}dec},
+	year         = 2020,
+	journal      = {GitHub repository},
+	publisher    = {GitHub},
+	howpublished = {\url{https://github.com/huggingface/trl}}
+}
+```

all_results.json

@@ -0,0 +1,8 @@
+{
+    "total_flos": 0.0,
+    "train_loss": 0.003984759748293852,
+    "train_runtime": 17571.6441,
+    "train_samples": 2793,
+    "train_samples_per_second": 0.318,
+    "train_steps_per_second": 0.01
+}

chat_template.jinja

@@ -0,0 +1 @@
+{% if not add_generation_prompt is defined %}{% set add_generation_prompt = false %}{% endif %}{% set ns = namespace(is_first=false, is_tool=false, is_output_first=true, system_prompt='') %}{%- for message in messages %}{%- if message['role'] == 'system' %}{% set ns.system_prompt = message['content'] %}{%- endif %}{%- endfor %}{{bos_token}}{{ns.system_prompt}}{%- for message in messages %}{%- if message['role'] == 'user' %}{%- set ns.is_tool = false -%}{{'<｜User｜>' + message['content']}}{%- endif %}{%- if message['role'] == 'assistant' and message['content'] is none %}{%- set ns.is_tool = false -%}{%- for tool in message['tool_calls']%}{%- if not ns.is_first %}{{'<｜Assistant｜><｜tool▁calls▁begin｜><｜tool▁call▁begin｜>' + tool['type'] + '<｜tool▁sep｜>' + tool['function']['name'] + '\n' + '```json' + '\n' + tool['function']['arguments'] + '\n' + '```' + '<｜tool▁call▁end｜>'}}{%- set ns.is_first = true -%}{%- else %}{{'\n' + '<｜tool▁call▁begin｜>' + tool['type'] + '<｜tool▁sep｜>' + tool['function']['name'] + '\n' + '```json' + '\n' + tool['function']['arguments'] + '\n' + '```' + '<｜tool▁call▁end｜>'}}{{'<｜tool▁calls▁end｜><｜end▁of▁sentence｜>'}}{%- endif %}{%- endfor %}{%- endif %}{%- if message['role'] == 'assistant' and message['content'] is not none %}{%- if ns.is_tool %}{{'<｜tool▁outputs▁end｜>' + message['content'] + '<｜end▁of▁sentence｜>'}}{%- set ns.is_tool = false -%}{%- else %}{% set content = message['content'] %}{{'<｜Assistant｜>' + content + '<｜end▁of▁sentence｜>'}}{%- endif %}{%- endif %}{%- if message['role'] == 'tool' %}{%- set ns.is_tool = true -%}{%- if ns.is_output_first %}{{'<｜tool▁outputs▁begin｜><｜tool▁output▁begin｜>' + message['content'] + '<｜tool▁output▁end｜>'}}{%- set ns.is_output_first = false %}{%- else %}{{'\n<｜tool▁output▁begin｜>' + message['content'] + '<｜tool▁output▁end｜>'}}{%- endif %}{%- endif %}{%- endfor -%}{% if ns.is_tool %}{{'<｜tool▁outputs▁end｜>'}}{% endif %}{% if add_generation_prompt and not ns.is_tool %}{{'<｜Assistant｜>'}}{% endif %}

config.json

@@ -0,0 +1,60 @@
+{
+  "architectures": [
+    "Qwen2ForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 151646,
+  "dtype": "bfloat16",
+  "eos_token_id": 151643,
+  "hidden_act": "silu",
+  "hidden_size": 3584,
+  "initializer_range": 0.02,
+  "intermediate_size": 18944,
+  "layer_types": [
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention"
+  ],
+  "max_position_embeddings": 131072,
+  "max_window_layers": 28,
+  "model_type": "qwen2",
+  "num_attention_heads": 28,
+  "num_hidden_layers": 28,
+  "num_key_value_heads": 4,
+  "pad_token_id": 151643,
+  "rms_norm_eps": 1e-06,
+  "rope_scaling": null,
+  "rope_theta": 10000,
+  "sliding_window": null,
+  "tie_word_embeddings": false,
+  "transformers_version": "4.56.2",
+  "use_cache": true,
+  "use_mrope": false,
+  "use_sliding_window": false,
+  "vocab_size": 152064
+}

generation_config.json

@@ -0,0 +1,7 @@
+{
+  "_from_model_config": true,
+  "bos_token_id": 151646,
+  "eos_token_id": 151643,
+  "pad_token_id": 151643,
+  "transformers_version": "4.56.2"
+}

model-00001-of-00004.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c3c0c35afa9726d843a4c40694f542aecca7cfd840fd3f85dc64bad7fa15b980
+size 4877660776

model-00002-of-00004.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3a82a71ef50a20325c758c1cb5e0c093edffda74ee1a7d54ec25a91b95b42a98
+size 4932751008

model-00003-of-00004.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:62bf82c7bad0923f8c63234471b1d2be3de40b3a176251de24a4baeb02239c12
+size 4330865200

model-00004-of-00004.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5e2e1de767800151514f1aebb06062f2cc159fd28cc6fa725743df31a1c99be7
+size 1089994880

model.safetensors.index.json

@@ -0,0 +1,347 @@
+{
+  "metadata": {
+    "total_parameters": 333312,
+    "total_size": 15231233024
+  },
+  "weight_map": {
+    "lm_head.weight": "model-00004-of-00004.safetensors",
+    "model.embed_tokens.weight": "model-00001-of-00004.safetensors",
+    "model.layers.0.input_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.0.mlp.down_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.0.mlp.gate_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.0.mlp.up_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.0.post_attention_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.0.self_attn.k_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.0.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.0.self_attn.o_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.0.self_attn.q_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.0.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.0.self_attn.v_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.0.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.1.input_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.1.mlp.down_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.1.mlp.gate_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.1.mlp.up_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.1.post_attention_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.1.self_attn.k_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.1.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.1.self_attn.o_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.1.self_attn.q_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.1.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.1.self_attn.v_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.1.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.10.input_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.10.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.10.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.10.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.10.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.10.self_attn.k_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.10.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.10.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.10.self_attn.q_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.10.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.10.self_attn.v_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.10.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.11.input_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.11.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.11.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.11.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.11.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.11.self_attn.k_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.11.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.11.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.11.self_attn.q_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.11.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.11.self_attn.v_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.11.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.12.input_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.12.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.12.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.12.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.12.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.12.self_attn.k_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.12.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.12.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.12.self_attn.q_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.12.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.12.self_attn.v_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.12.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.13.input_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.13.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.13.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.13.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.13.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.13.self_attn.k_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.13.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.13.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.13.self_attn.q_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.13.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.13.self_attn.v_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.13.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.14.input_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.14.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.14.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.14.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.14.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.14.self_attn.k_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.14.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.14.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.14.self_attn.q_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.14.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.14.self_attn.v_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.14.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.15.input_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.15.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.15.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.15.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.15.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.15.self_attn.k_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.15.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.15.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.15.self_attn.q_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.15.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.15.self_attn.v_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.15.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.16.input_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.16.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.16.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.16.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.16.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.16.self_attn.k_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.16.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.16.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.16.self_attn.q_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.16.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.16.self_attn.v_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.16.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.17.input_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.17.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.17.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.17.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.17.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.17.self_attn.k_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.17.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.17.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.17.self_attn.q_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.17.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.17.self_attn.v_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.17.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.18.input_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.18.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.18.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.18.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.18.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.18.self_attn.k_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.18.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.18.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.18.self_attn.q_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.18.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.18.self_attn.v_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.18.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.19.input_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.19.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.19.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.19.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.19.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.19.self_attn.k_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.19.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.19.self_attn.o_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.19.self_attn.q_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.19.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.19.self_attn.v_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.19.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.2.input_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.2.mlp.down_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.2.mlp.gate_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.2.mlp.up_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.2.post_attention_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.2.self_attn.k_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.2.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.2.self_attn.o_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.2.self_attn.q_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.2.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.2.self_attn.v_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.2.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.20.input_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.20.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.20.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.20.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.20.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.20.self_attn.k_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.20.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.20.self_attn.o_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.20.self_attn.q_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.20.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.20.self_attn.v_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.20.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.21.input_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.21.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.21.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.21.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.21.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.21.self_attn.k_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.21.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.21.self_attn.o_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.21.self_attn.q_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.21.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.21.self_attn.v_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.21.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.22.input_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.22.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.22.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.22.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.22.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.22.self_attn.k_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.22.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.22.self_attn.o_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.22.self_attn.q_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.22.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.22.self_attn.v_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.22.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.23.input_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.23.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.23.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.23.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.23.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.23.self_attn.k_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.23.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.23.self_attn.o_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.23.self_attn.q_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.23.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.23.self_attn.v_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.23.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.24.input_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.24.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.24.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.24.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.24.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.24.self_attn.k_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.24.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.24.self_attn.o_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.24.self_attn.q_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.24.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.24.self_attn.v_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.24.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.25.input_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.25.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.25.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.25.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.25.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.25.self_attn.k_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.25.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.25.self_attn.o_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.25.self_attn.q_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.25.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.25.self_attn.v_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.25.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.26.input_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.26.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.26.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.26.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.26.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.26.self_attn.k_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.26.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.26.self_attn.o_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.26.self_attn.q_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.26.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.26.self_attn.v_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.26.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.27.input_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.27.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.27.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.27.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.27.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.27.self_attn.k_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.27.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.27.self_attn.o_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.27.self_attn.q_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.27.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.27.self_attn.v_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.27.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.3.input_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.3.mlp.down_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.3.mlp.gate_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.3.mlp.up_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.3.post_attention_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.3.self_attn.k_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.3.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.3.self_attn.o_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.3.self_attn.q_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.3.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.3.self_attn.v_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.3.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.4.input_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.4.mlp.down_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.4.mlp.gate_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.4.mlp.up_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.4.post_attention_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.4.self_attn.k_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.4.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.4.self_attn.o_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.4.self_attn.q_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.4.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.4.self_attn.v_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.4.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.5.input_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.5.mlp.down_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.5.mlp.gate_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.5.mlp.up_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.5.post_attention_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.5.self_attn.k_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.5.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.5.self_attn.o_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.5.self_attn.q_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.5.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.5.self_attn.v_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.5.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.6.input_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.6.mlp.down_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.6.mlp.gate_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.6.mlp.up_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.6.post_attention_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.6.self_attn.k_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.6.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.6.self_attn.o_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.6.self_attn.q_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.6.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.6.self_attn.v_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.6.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.7.input_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.7.mlp.down_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.7.mlp.gate_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.7.mlp.up_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.7.post_attention_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.7.self_attn.k_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.7.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.7.self_attn.o_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.7.self_attn.q_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.7.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.7.self_attn.v_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.7.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.8.input_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.8.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.8.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.8.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.8.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.8.self_attn.k_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.8.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.8.self_attn.o_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.8.self_attn.q_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.8.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.8.self_attn.v_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.8.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.9.input_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.9.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.9.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.9.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.9.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.9.self_attn.k_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.9.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.9.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.9.self_attn.q_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.9.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.9.self_attn.v_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.9.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
+    "model.norm.weight": "model-00003-of-00004.safetensors"
+  }
+}

special_tokens_map.json

@@ -0,0 +1,23 @@
+{
+  "bos_token": {
+    "content": "<｜begin▁of▁sentence｜>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "eos_token": {
+    "content": "<｜end▁of▁sentence｜>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": {
+    "content": "<｜end▁of▁sentence｜>",
+    "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:a4256422650d141f228fe954acee98679da412984c29a569877eefd3af69315a
+size 11422959

tokenizer_config.json

@@ -0,0 +1,198 @@
+{
+  "add_bos_token": true,
+  "add_eos_token": false,
+  "add_prefix_space": null,
+  "added_tokens_decoder": {
+    "151643": {
+      "content": "<｜end▁of▁sentence｜>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151644": {
+      "content": "<｜User｜>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151645": {
+      "content": "<｜Assistant｜>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151646": {
+      "content": "<｜begin▁of▁sentence｜>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151647": {
+      "content": "<|EOT|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151648": {
+      "content": "<think>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151649": {
+      "content": "</think>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "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
+    }
+  },
+  "bos_token": "<｜begin▁of▁sentence｜>",
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "<｜end▁of▁sentence｜>",
+  "extra_special_tokens": {},
+  "legacy": true,
+  "max_length": null,
+  "model_max_length": 16384,
+  "pad_to_multiple_of": null,
+  "pad_token": "<｜end▁of▁sentence｜>",
+  "pad_token_type_id": 0,
+  "padding_side": "left",
+  "sp_model_kwargs": {},
+  "tokenizer_class": "LlamaTokenizerFast",
+  "unk_token": null,
+  "use_default_system_prompt": false
+}

train_results.json

@@ -0,0 +1,8 @@
+{
+    "total_flos": 0.0,
+    "train_loss": 0.003984759748293852,
+    "train_runtime": 17571.6441,
+    "train_samples": 2793,
+    "train_samples_per_second": 0.318,
+    "train_steps_per_second": 0.01
+}

trainer_state.json

@@ -0,0 +1,1862 @@
+{
+  "best_global_step": 84,
+  "best_metric": 0.00022543604427482933,
+  "best_model_checkpoint": "data/DeepSeek-R1-Distill-Qwen-7B-Staged-3/checkpoint-84",
+  "epoch": 2.0,
+  "eval_steps": 500,
+  "global_step": 168,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "clip_ratio/high_max": 0.0,
+      "clip_ratio/high_mean": 0.0,
+      "clip_ratio/low_mean": 0.0,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.0,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 872.0,
+      "completions/max_terminated_length": 872.0,
+      "completions/mean_length": 493.42919921875,
+      "completions/mean_terminated_length": 493.42919921875,
+      "completions/min_length": 271.0,
+      "completions/min_terminated_length": 271.0,
+      "entropy": 0.8375349938869476,
+      "epoch": 0.011904761904761904,
+      "frac_reward_zero_std": 0.0,
+      "grad_norm": 0.10371755808591843,
+      "learning_rate": 1e-05,
+      "loss": -0.0014,
+      "num_tokens": 2778575.0,
+      "reward": 3.663585662841797,
+      "reward_std": 0.316636860370636,
+      "rewards/ngram_repetition2/mean": 0.9156007766723633,
+      "rewards/ngram_repetition2/std": 0.02841464802622795,
+      "rewards/ngram_repetition3/mean": 0.9844592213630676,
+      "rewards/ngram_repetition3/std": 0.010988580994307995,
+      "rewards/symbolic_reward_accuracy/mean": 0.84228515625,
+      "rewards/symbolic_reward_accuracy/std": 0.36456242203712463,
+      "rewards/symbolic_reward_partial_score/mean": 0.9502766728401184,
+      "rewards/symbolic_reward_partial_score/std": 0.13473224639892578,
+      "rewards/tag_count_reward/mean": 1.0,
+      "rewards/tag_count_reward/std": 0.0,
+      "rewards/thinking_answer_ratio_reward/mean": 0.9737867712974548,
+      "rewards/thinking_answer_ratio_reward/std": 0.0084904246032238,
+      "sampling/importance_sampling_ratio/max": 2.0,
+      "sampling/importance_sampling_ratio/mean": 1.2483623027801514,
+      "sampling/importance_sampling_ratio/min": 0.0012419477570801973,
+      "sampling/sampling_logp_difference/max": 6.691074371337891,
+      "sampling/sampling_logp_difference/mean": 0.3304027318954468,
+      "step": 1
+    },
+    {
+      "clip_ratio/high_max": 0.18229166666666666,
+      "clip_ratio/high_mean": 0.06705729166666667,
+      "clip_ratio/low_mean": 0.390625,
+      "clip_ratio/low_min": 0.21354166666666666,
+      "clip_ratio/region_mean": 0.4576822916666667,
+      "entropy": 0.8554559648036957,
+      "epoch": 0.047619047619047616,
+      "grad_norm": 0.0923309475183487,
+      "learning_rate": 1e-05,
+      "loss": 0.0126,
+      "step": 4
+    },
+    {
+      "clip_ratio/high_max": 0.109375,
+      "clip_ratio/high_mean": 0.0390625,
+      "clip_ratio/low_mean": 0.275390625,
+      "clip_ratio/low_min": 0.14453125,
+      "clip_ratio/region_mean": 0.314453125,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 1237.0,
+      "completions/max_terminated_length": 1237.0,
+      "completions/mean_length": 521.73583984375,
+      "completions/mean_terminated_length": 521.73583984375,
+      "completions/min_length": 301.0,
+      "completions/min_terminated_length": 301.0,
+      "entropy": 0.8218244835734367,
+      "epoch": 0.09523809523809523,
+      "frac_reward_zero_std": 0.0,
+      "grad_norm": 0.07603302597999573,
+      "learning_rate": 1e-05,
+      "loss": 0.0111,
+      "num_tokens": 5698738.0,
+      "reward": 3.8017191886901855,
+      "reward_std": 0.2319643646478653,
+      "rewards/ngram_repetition2/mean": 0.909049391746521,
+      "rewards/ngram_repetition2/std": 0.029375825077295303,
+      "rewards/ngram_repetition3/mean": 0.9826949834823608,
+      "rewards/ngram_repetition3/std": 0.011540538631379604,
+      "rewards/symbolic_reward_accuracy/mean": 0.90087890625,
+      "rewards/symbolic_reward_accuracy/std": 0.29889750480651855,
+      "rewards/symbolic_reward_partial_score/mean": 0.9712728261947632,
+      "rewards/symbolic_reward_partial_score/std": 0.09666982293128967,
+      "rewards/tag_count_reward/mean": 1.0,
+      "rewards/tag_count_reward/std": 0.0,
+      "rewards/thinking_answer_ratio_reward/mean": 0.9771104454994202,
+      "rewards/thinking_answer_ratio_reward/std": 0.00513562373816967,
+      "sampling/importance_sampling_ratio/max": 2.0,
+      "sampling/importance_sampling_ratio/mean": 1.2442681789398193,
+      "sampling/importance_sampling_ratio/min": 0.0008342887158505619,
+      "sampling/sampling_logp_difference/max": 7.088931083679199,
+      "sampling/sampling_logp_difference/mean": 0.325713574886322,
+      "step": 8
+    },
+    {
+      "clip_ratio/high_max": 0.19921875,
+      "clip_ratio/high_mean": 0.107421875,
+      "clip_ratio/low_mean": 0.24462890625,
+      "clip_ratio/low_min": 0.13671875,
+      "clip_ratio/region_mean": 0.35205078125,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 907.0,
+      "completions/max_terminated_length": 907.0,
+      "completions/mean_length": 507.666015625,
+      "completions/mean_terminated_length": 507.666015625,
+      "completions/min_length": 287.0,
+      "completions/min_terminated_length": 287.0,
+      "entropy": 0.8540411740541458,
+      "epoch": 0.14285714285714285,
+      "frac_reward_zero_std": 0.0,
+      "grad_norm": 0.08517201989889145,
+      "learning_rate": 1e-05,
+      "loss": 0.0106,
+      "num_tokens": 8529286.0,
+      "reward": 3.751950740814209,
+      "reward_std": 0.2073085606098175,
+      "rewards/ngram_repetition2/mean": 0.9060913920402527,
+      "rewards/ngram_repetition2/std": 0.029160665348172188,
+      "rewards/ngram_repetition3/mean": 0.9818980097770691,
+      "rewards/ngram_repetition3/std": 0.01140812411904335,
+      "rewards/symbolic_reward_accuracy/mean": 0.88134765625,
+      "rewards/symbolic_reward_accuracy/std": 0.3234579563140869,
+      "rewards/symbolic_reward_partial_score/mean": 0.9606119394302368,
+      "rewards/symbolic_reward_partial_score/std": 0.12397422641515732,
+      "rewards/tag_count_reward/mean": 1.0,
+      "rewards/tag_count_reward/std": 0.0,
+      "rewards/thinking_answer_ratio_reward/mean": 0.9763435125350952,
+      "rewards/thinking_answer_ratio_reward/std": 0.006303399335592985,
+      "sampling/importance_sampling_ratio/max": 2.0,
+      "sampling/importance_sampling_ratio/mean": 1.2473976612091064,
+      "sampling/importance_sampling_ratio/min": 0.0019555450417101383,
+      "sampling/sampling_logp_difference/max": 6.237086296081543,
+      "sampling/sampling_logp_difference/mean": 0.3299524486064911,
+      "step": 12
+    },
+    {
+      "clip_ratio/high_max": 0.1953125,
+      "clip_ratio/high_mean": 0.10302734375,
+      "clip_ratio/low_mean": 0.27294921875,
+      "clip_ratio/low_min": 0.125,
+      "clip_ratio/region_mean": 0.3759765625,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 921.0,
+      "completions/max_terminated_length": 921.0,
+      "completions/mean_length": 461.00634765625,
+      "completions/mean_terminated_length": 461.00634765625,
+      "completions/min_length": 267.0,
+      "completions/min_terminated_length": 267.0,
+      "entropy": 0.829069722443819,
+      "epoch": 0.19047619047619047,
+      "frac_reward_zero_std": 0.0,
+      "grad_norm": 0.06742873042821884,
+      "learning_rate": 1e-05,
+      "loss": 0.0084,
+      "num_tokens": 11300691.0,
+      "reward": 3.798084259033203,
+      "reward_std": 0.20729093253612518,
+      "rewards/ngram_repetition2/mean": 0.9188657999038696,
+      "rewards/ngram_repetition2/std": 0.025240786373615265,
+      "rewards/ngram_repetition3/mean": 0.9856483340263367,
+      "rewards/ngram_repetition3/std": 0.009459859691560268,
+      "rewards/symbolic_reward_accuracy/mean": 0.9033203125,
+      "rewards/symbolic_reward_accuracy/std": 0.2955932915210724,
+      "rewards/symbolic_reward_partial_score/mean": 0.962646484375,
+      "rewards/symbolic_reward_partial_score/std": 0.12614500522613525,
+      "rewards/tag_count_reward/mean": 1.0,
+      "rewards/tag_count_reward/std": 0.0,
+      "rewards/thinking_answer_ratio_reward/mean": 0.9751918911933899,
+      "rewards/thinking_answer_ratio_reward/std": 0.004718962591141462,
+      "sampling/importance_sampling_ratio/max": 2.0,
+      "sampling/importance_sampling_ratio/mean": 1.2463021278381348,
+      "sampling/importance_sampling_ratio/min": 0.0014070722972974181,
+      "sampling/sampling_logp_difference/max": 6.566244125366211,
+      "sampling/sampling_logp_difference/mean": 0.3286193013191223,
+      "step": 16
+    },
+    {
+      "clip_ratio/high_max": 0.22265625,
+      "clip_ratio/high_mean": 0.12646484375,
+      "clip_ratio/low_mean": 0.24462890625,
+      "clip_ratio/low_min": 0.14453125,
+      "clip_ratio/region_mean": 0.37109375,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 704.0,
+      "completions/max_terminated_length": 704.0,
+      "completions/mean_length": 416.75146484375,
+      "completions/mean_terminated_length": 416.75146484375,
+      "completions/min_length": 257.0,
+      "completions/min_terminated_length": 257.0,
+      "entropy": 0.8178543969988823,
+      "epoch": 0.23809523809523808,
+      "frac_reward_zero_std": 0.0,
+      "grad_norm": 0.05325091630220413,
+      "learning_rate": 1e-05,
+      "loss": 0.0057,
+      "num_tokens": 13981782.0,
+      "reward": 3.840620517730713,
+      "reward_std": 0.17522145807743073,
+      "rewards/ngram_repetition2/mean": 0.932222306728363,
+      "rewards/ngram_repetition2/std": 0.02194184437394142,
+      "rewards/ngram_repetition3/mean": 0.989234447479248,
+      "rewards/ngram_repetition3/std": 0.008273402228951454,
+      "rewards/symbolic_reward_accuracy/mean": 0.91845703125,
+      "rewards/symbolic_reward_accuracy/std": 0.2737340033054352,
+      "rewards/symbolic_reward_partial_score/mean": 0.9747720956802368,
+      "rewards/symbolic_reward_partial_score/std": 0.09571196138858795,
+      "rewards/tag_count_reward/mean": 1.0,
+      "rewards/tag_count_reward/std": 0.0,
+      "rewards/thinking_answer_ratio_reward/mean": 0.9719721674919128,
+      "rewards/thinking_answer_ratio_reward/std": 0.005963112693279982,
+      "sampling/importance_sampling_ratio/max": 2.0,
+      "sampling/importance_sampling_ratio/mean": 1.2414987087249756,
+      "sampling/importance_sampling_ratio/min": 0.00114686309825629,
+      "sampling/sampling_logp_difference/max": 6.770724773406982,
+      "sampling/sampling_logp_difference/mean": 0.3227406144142151,
+      "step": 20
+    },
+    {
+      "clip_ratio/high_max": 0.21875,
+      "clip_ratio/high_mean": 0.12451171875,
+      "clip_ratio/low_mean": 0.2451171875,
+      "clip_ratio/low_min": 0.1328125,
+      "clip_ratio/region_mean": 0.36962890625,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 583.0,
+      "completions/max_terminated_length": 583.0,
+      "completions/mean_length": 372.75732421875,
+      "completions/mean_terminated_length": 372.75732421875,
+      "completions/min_length": 235.0,
+      "completions/min_terminated_length": 235.0,
+      "entropy": 0.7815876416862011,
+      "epoch": 0.2857142857142857,
+      "frac_reward_zero_std": 0.0,
+      "grad_norm": 0.05320040136575699,
+      "learning_rate": 1e-05,
+      "loss": 0.0041,
+      "num_tokens": 16548549.0,
+      "reward": 3.79876708984375,
+      "reward_std": 0.11160629987716675,
+      "rewards/ngram_repetition2/mean": 0.9457941055297852,
+      "rewards/ngram_repetition2/std": 0.01891067810356617,
+      "rewards/ngram_repetition3/mean": 0.9925487041473389,
+      "rewards/ngram_repetition3/std": 0.006580323446542025,
+      "rewards/symbolic_reward_accuracy/mean": 0.8974609375,
+      "rewards/symbolic_reward_accuracy/std": 0.3034299910068512,
+      "rewards/symbolic_reward_partial_score/mean": 0.9747720956802368,
+      "rewards/symbolic_reward_partial_score/std": 0.09050152450799942,
+      "rewards/tag_count_reward/mean": 1.0,
+      "rewards/tag_count_reward/std": 0.0,
+      "rewards/thinking_answer_ratio_reward/mean": 0.9689682126045227,
+      "rewards/thinking_answer_ratio_reward/std": 0.0063285622745752335,
+      "sampling/importance_sampling_ratio/max": 2.0,
+      "sampling/importance_sampling_ratio/mean": 1.23650324344635,
+      "sampling/importance_sampling_ratio/min": 0.001662293216213584,
+      "sampling/sampling_logp_difference/max": 6.399557113647461,
+      "sampling/sampling_logp_difference/mean": 0.31696027517318726,
+      "step": 24
+    },
+    {
+      "clip_ratio/high_max": 0.2734375,
+      "clip_ratio/high_mean": 0.1572265625,
+      "clip_ratio/low_mean": 0.22412109375,
+      "clip_ratio/low_min": 0.09765625,
+      "clip_ratio/region_mean": 0.38134765625,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 496.0,
+      "completions/max_terminated_length": 496.0,
+      "completions/mean_length": 334.39111328125,
+      "completions/mean_terminated_length": 334.39111328125,
+      "completions/min_length": 192.0,
+      "completions/min_terminated_length": 192.0,
+      "entropy": 0.7627993300557137,
+      "epoch": 0.3333333333333333,
+      "frac_reward_zero_std": 0.0,
+      "grad_norm": 0.05497128516435623,
+      "learning_rate": 1e-05,
+      "loss": 0.0017,
+      "num_tokens": 19030726.0,
+      "reward": 3.8732995986938477,
+      "reward_std": 0.11424858868122101,
+      "rewards/ngram_repetition2/mean": 0.9574997425079346,
+      "rewards/ngram_repetition2/std": 0.015644928440451622,
+      "rewards/ngram_repetition3/mean": 0.9950100183486938,
+      "rewards/ngram_repetition3/std": 0.005439166910946369,
+      "rewards/symbolic_reward_accuracy/mean": 0.93310546875,
+      "rewards/symbolic_reward_accuracy/std": 0.24990026652812958,
+      "rewards/symbolic_reward_partial_score/mean": 0.9779052734375,
+      "rewards/symbolic_reward_partial_score/std": 0.09342914074659348,
+      "rewards/tag_count_reward/mean": 1.0,
+      "rewards/tag_count_reward/std": 0.0,
+      "rewards/thinking_answer_ratio_reward/mean": 0.9658128023147583,
+      "rewards/thinking_answer_ratio_reward/std": 0.006996896583586931,
+      "sampling/importance_sampling_ratio/max": 2.0,
+      "sampling/importance_sampling_ratio/mean": 1.2309629917144775,
+      "sampling/importance_sampling_ratio/min": 0.0011809006100520492,
+      "sampling/sampling_logp_difference/max": 6.741477966308594,
+      "sampling/sampling_logp_difference/mean": 0.31086698174476624,
+      "step": 28
+    },
+    {
+      "clip_ratio/high_max": 0.25390625,
+      "clip_ratio/high_mean": 0.138671875,
+      "clip_ratio/low_mean": 0.21826171875,
+      "clip_ratio/low_min": 0.09375,
+      "clip_ratio/region_mean": 0.35693359375,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 477.0,
+      "completions/max_terminated_length": 477.0,
+      "completions/mean_length": 311.6484375,
+      "completions/mean_terminated_length": 311.6484375,
+      "completions/min_length": 22.0,
+      "completions/min_terminated_length": 22.0,
+      "entropy": 0.7324606403708458,
+      "epoch": 0.38095238095238093,
+      "frac_reward_zero_std": 0.0,
+      "grad_norm": 0.035085611045360565,
+      "learning_rate": 1e-05,
+      "loss": 0.0006,
+      "num_tokens": 21471382.0,
+      "reward": 3.878781795501709,
+      "reward_std": 0.12647078931331635,
+      "rewards/ngram_repetition2/mean": 0.9654513597488403,
+      "rewards/ngram_repetition2/std": 0.013394074514508247,
+      "rewards/ngram_repetition3/mean": 0.9962821006774902,
+      "rewards/ngram_repetition3/std": 0.004647234920412302,
+      "rewards/symbolic_reward_accuracy/mean": 0.935546875,
+      "rewards/symbolic_reward_accuracy/std": 0.24561835825443268,
+      "rewards/symbolic_reward_partial_score/mean": 0.9784342050552368,
+      "rewards/symbolic_reward_partial_score/std": 0.09432914108037949,
+      "rewards/tag_count_reward/mean": 1.0,
+      "rewards/tag_count_reward/std": 0.0,
+      "rewards/thinking_answer_ratio_reward/mean": 0.9636549353599548,
+      "rewards/thinking_answer_ratio_reward/std": 0.021884752437472343,
+      "sampling/importance_sampling_ratio/max": 2.0,
+      "sampling/importance_sampling_ratio/mean": 1.2252447605133057,
+      "sampling/importance_sampling_ratio/min": 0.0016044432995840907,
+      "sampling/sampling_logp_difference/max": 6.434978485107422,
+      "sampling/sampling_logp_difference/mean": 0.30396580696105957,
+      "step": 32
+    },
+    {
+      "clip_ratio/high_max": 0.2734375,
+      "clip_ratio/high_mean": 0.15869140625,
+      "clip_ratio/low_mean": 0.2080078125,
+      "clip_ratio/low_min": 0.09375,
+      "clip_ratio/region_mean": 0.36669921875,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 430.0,
+      "completions/max_terminated_length": 430.0,
+      "completions/mean_length": 290.2744140625,
+      "completions/mean_terminated_length": 290.2744140625,
+      "completions/min_length": 177.0,
+      "completions/min_terminated_length": 177.0,
+      "entropy": 0.721624631434679,
+      "epoch": 0.42857142857142855,
+      "frac_reward_zero_std": 0.0,
+      "grad_norm": 0.03640465438365936,
+      "learning_rate": 1e-05,
+      "loss": -0.0001,
+      "num_tokens": 23887592.0,
+      "reward": 3.8664052486419678,
+      "reward_std": 0.12858426570892334,
+      "rewards/ngram_repetition2/mean": 0.970293402671814,
+      "rewards/ngram_repetition2/std": 0.012160783633589745,
+      "rewards/ngram_repetition3/mean": 0.9971583485603333,
+      "rewards/ngram_repetition3/std": 0.00398310786113143,
+      "rewards/symbolic_reward_accuracy/mean": 0.92822265625,
+      "rewards/symbolic_reward_accuracy/std": 0.2581818997859955,
+      "rewards/symbolic_reward_partial_score/mean": 0.9806721806526184,
+      "rewards/symbolic_reward_partial_score/std": 0.08271697163581848,
+      "rewards/tag_count_reward/mean": 1.0,
+      "rewards/tag_count_reward/std": 0.0,
+      "rewards/thinking_answer_ratio_reward/mean": 0.9613040089607239,
+      "rewards/thinking_answer_ratio_reward/std": 0.007317277602851391,
+      "sampling/importance_sampling_ratio/max": 2.0,
+      "sampling/importance_sampling_ratio/mean": 1.2203993797302246,
+      "sampling/importance_sampling_ratio/min": 0.001269048429094255,
+      "sampling/sampling_logp_difference/max": 6.669487953186035,
+      "sampling/sampling_logp_difference/mean": 0.29926085472106934,
+      "step": 36
+    },
+    {
+      "clip_ratio/high_max": 0.21875,
+      "clip_ratio/high_mean": 0.11572265625,
+      "clip_ratio/low_mean": 0.25537109375,
+      "clip_ratio/low_min": 0.12890625,
+      "clip_ratio/region_mean": 0.37109375,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 427.0,
+      "completions/max_terminated_length": 427.0,
+      "completions/mean_length": 284.6923828125,
+      "completions/mean_terminated_length": 284.6923828125,
+      "completions/min_length": 184.0,
+      "completions/min_terminated_length": 184.0,
+      "entropy": 0.7025646083056927,
+      "epoch": 0.47619047619047616,
+      "frac_reward_zero_std": 0.0,
+      "grad_norm": 0.04123682901263237,
+      "learning_rate": 1e-05,
+      "loss": -0.0005,
+      "num_tokens": 26295538.0,
+      "reward": 3.7722182273864746,
+      "reward_std": 0.1536768078804016,
+      "rewards/ngram_repetition2/mean": 0.9733011722564697,
+      "rewards/ngram_repetition2/std": 0.012304718606173992,
+      "rewards/ngram_repetition3/mean": 0.9972368478775024,
+      "rewards/ngram_repetition3/std": 0.004414981231093407,
+      "rewards/symbolic_reward_accuracy/mean": 0.88818359375,
+      "rewards/symbolic_reward_accuracy/std": 0.31521740555763245,
+      "rewards/symbolic_reward_partial_score/mean": 0.966552734375,
+      "rewards/symbolic_reward_partial_score/std": 0.10999048501253128,
+      "rewards/tag_count_reward/mean": 1.0,
+      "rewards/tag_count_reward/std": 0.0,
+      "rewards/thinking_answer_ratio_reward/mean": 0.9592961668968201,
+      "rewards/thinking_answer_ratio_reward/std": 0.008976742625236511,
+      "sampling/importance_sampling_ratio/max": 2.0,
+      "sampling/importance_sampling_ratio/mean": 1.2174944877624512,
+      "sampling/importance_sampling_ratio/min": 0.0009174988372251391,
+      "sampling/sampling_logp_difference/max": 6.99385929107666,
+      "sampling/sampling_logp_difference/mean": 0.29805290699005127,
+      "step": 40
+    },
+    {
+      "clip_ratio/high_max": 0.25390625,
+      "clip_ratio/high_mean": 0.14892578125,
+      "clip_ratio/low_mean": 0.228515625,
+      "clip_ratio/low_min": 0.1015625,
+      "clip_ratio/region_mean": 0.37744140625,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 417.0,
+      "completions/max_terminated_length": 417.0,
+      "completions/mean_length": 279.68017578125,
+      "completions/mean_terminated_length": 279.68017578125,
+      "completions/min_length": 178.0,
+      "completions/min_terminated_length": 178.0,
+      "entropy": 0.7209025509655476,
+      "epoch": 0.5238095238095238,
+      "frac_reward_zero_std": 0.0,
+      "grad_norm": 0.03904758393764496,
+      "learning_rate": 1e-05,
+      "loss": -0.001,
+      "num_tokens": 28673731.0,
+      "reward": 3.845343589782715,
+      "reward_std": 0.1021677702665329,
+      "rewards/ngram_repetition2/mean": 0.9727880954742432,
+      "rewards/ngram_repetition2/std": 0.014563015662133694,
+      "rewards/ngram_repetition3/mean": 0.9964158535003662,
+      "rewards/ngram_repetition3/std": 0.006320940796285868,
+      "rewards/symbolic_reward_accuracy/mean": 0.91845703125,
+      "rewards/symbolic_reward_accuracy/std": 0.2737340033054352,
+      "rewards/symbolic_reward_partial_score/mean": 0.9791666269302368,
+      "rewards/symbolic_reward_partial_score/std": 0.07696299254894257,
+      "rewards/tag_count_reward/mean": 1.0,
+      "rewards/tag_count_reward/std": 0.0,
+      "rewards/thinking_answer_ratio_reward/mean": 0.9570740461349487,
+      "rewards/thinking_answer_ratio_reward/std": 0.012159617617726326,
+      "sampling/importance_sampling_ratio/max": 2.0,
+      "sampling/importance_sampling_ratio/mean": 1.2212014198303223,
+      "sampling/importance_sampling_ratio/min": 0.0015326839638873935,
+      "sampling/sampling_logp_difference/max": 6.480734825134277,
+      "sampling/sampling_logp_difference/mean": 0.3009487986564636,
+      "step": 44
+    },
+    {
+      "clip_ratio/high_max": 0.22265625,
+      "clip_ratio/high_mean": 0.1083984375,
+      "clip_ratio/low_mean": 0.2265625,
+      "clip_ratio/low_min": 0.11328125,
+      "clip_ratio/region_mean": 0.3349609375,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 420.0,
+      "completions/max_terminated_length": 420.0,
+      "completions/mean_length": 288.65185546875,
+      "completions/mean_terminated_length": 288.65185546875,
+      "completions/min_length": 163.0,
+      "completions/min_terminated_length": 163.0,
+      "entropy": 0.7586110942065716,
+      "epoch": 0.5714285714285714,
+      "frac_reward_zero_std": 0.0,
+      "grad_norm": 0.05547133460640907,
+      "learning_rate": 1e-05,
+      "loss": -0.0005,
+      "num_tokens": 31079642.0,
+      "reward": 3.8869855403900146,
+      "reward_std": 0.0652470514178276,
+      "rewards/ngram_repetition2/mean": 0.9750747680664062,
+      "rewards/ngram_repetition2/std": 0.013366366736590862,
+      "rewards/ngram_repetition3/mean": 0.9970998764038086,
+      "rewards/ngram_repetition3/std": 0.005164336413145065,
+      "rewards/symbolic_reward_accuracy/mean": 0.9384765625,
+      "rewards/symbolic_reward_accuracy/std": 0.24034665524959564,
+      "rewards/symbolic_reward_partial_score/mean": 0.980712890625,
+      "rewards/symbolic_reward_partial_score/std": 0.08254173398017883,
+      "rewards/tag_count_reward/mean": 1.0,
+      "rewards/tag_count_reward/std": 0.0,
+      "rewards/thinking_answer_ratio_reward/mean": 0.9597762823104858,
+      "rewards/thinking_answer_ratio_reward/std": 0.011353440582752228,
+      "sampling/importance_sampling_ratio/max": 2.0,
+      "sampling/importance_sampling_ratio/mean": 1.2331750392913818,
+      "sampling/importance_sampling_ratio/min": 0.0016639016102999449,
+      "sampling/sampling_logp_difference/max": 6.398590087890625,
+      "sampling/sampling_logp_difference/mean": 0.316028892993927,
+      "step": 48
+    },
+    {
+      "clip_ratio/high_max": 0.20703125,
+      "clip_ratio/high_mean": 0.10888671875,
+      "clip_ratio/low_mean": 0.24267578125,
+      "clip_ratio/low_min": 0.10546875,
+      "clip_ratio/region_mean": 0.3515625,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 461.0,
+      "completions/max_terminated_length": 461.0,
+      "completions/mean_length": 291.1162109375,
+      "completions/mean_terminated_length": 291.1162109375,
+      "completions/min_length": 159.0,
+      "completions/min_terminated_length": 159.0,
+      "entropy": 0.783844955265522,
+      "epoch": 0.6190476190476191,
+      "frac_reward_zero_std": 0.0,
+      "grad_norm": 0.055351365357637405,
+      "learning_rate": 1e-05,
+      "loss": -0.0002,
+      "num_tokens": 33518952.0,
+      "reward": 3.874098300933838,
+      "reward_std": 0.0593051016330719,
+      "rewards/ngram_repetition2/mean": 0.9759907722473145,
+      "rewards/ngram_repetition2/std": 0.012779747135937214,
+      "rewards/ngram_repetition3/mean": 0.9973393082618713,
+      "rewards/ngram_repetition3/std": 0.004218540154397488,
+      "rewards/symbolic_reward_accuracy/mean": 0.93310546875,
+      "rewards/symbolic_reward_accuracy/std": 0.24990026652812958,
+      "rewards/symbolic_reward_partial_score/mean": 0.9785562753677368,
+      "rewards/symbolic_reward_partial_score/std": 0.08766603469848633,
+      "rewards/tag_count_reward/mean": 1.0,
+      "rewards/tag_count_reward/std": 0.0,
+      "rewards/thinking_answer_ratio_reward/mean": 0.9597955942153931,
+      "rewards/thinking_answer_ratio_reward/std": 0.010997203178703785,
+      "sampling/importance_sampling_ratio/max": 2.0,
+      "sampling/importance_sampling_ratio/mean": 1.242828130722046,
+      "sampling/importance_sampling_ratio/min": 0.001207949360832572,
+      "sampling/sampling_logp_difference/max": 6.7188310623168945,
+      "sampling/sampling_logp_difference/mean": 0.32469046115875244,
+      "step": 52
+    },
+    {
+      "clip_ratio/high_max": 0.19921875,
+      "clip_ratio/high_mean": 0.11376953125,
+      "clip_ratio/low_mean": 0.23583984375,
+      "clip_ratio/low_min": 0.125,
+      "clip_ratio/region_mean": 0.349609375,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 430.0,
+      "completions/max_terminated_length": 430.0,
+      "completions/mean_length": 287.13720703125,
+      "completions/mean_terminated_length": 287.13720703125,
+      "completions/min_length": 182.0,
+      "completions/min_terminated_length": 182.0,
+      "entropy": 0.8433509171009064,
+      "epoch": 0.6666666666666666,
+      "frac_reward_zero_std": 0.0,
+      "grad_norm": 0.05898959934711456,
+      "learning_rate": 1e-05,
+      "loss": 0.0,
+      "num_tokens": 35955809.0,
+      "reward": 3.7203783988952637,
+      "reward_std": 0.17166642844676971,
+      "rewards/ngram_repetition2/mean": 0.9761325716972351,
+      "rewards/ngram_repetition2/std": 0.011715667322278023,
+      "rewards/ngram_repetition3/mean": 0.9974805116653442,
+      "rewards/ngram_repetition3/std": 0.003850888693705201,
+      "rewards/symbolic_reward_accuracy/mean": 0.8662109375,
+      "rewards/symbolic_reward_accuracy/std": 0.34050869941711426,
+      "rewards/symbolic_reward_partial_score/mean": 0.9586181640625,
+      "rewards/symbolic_reward_partial_score/std": 0.11637341976165771,
+      "rewards/tag_count_reward/mean": 1.0,
+      "rewards/tag_count_reward/std": 0.0,
+      "rewards/thinking_answer_ratio_reward/mean": 0.9602200984954834,
+      "rewards/thinking_answer_ratio_reward/std": 0.009219513274729252,
+      "sampling/importance_sampling_ratio/max": 2.0,
+      "sampling/importance_sampling_ratio/mean": 1.259500503540039,
+      "sampling/importance_sampling_ratio/min": 0.00019177240028511733,
+      "sampling/sampling_logp_difference/max": 8.55920124053955,
+      "sampling/sampling_logp_difference/mean": 0.34136319160461426,
+      "step": 56
+    },
+    {
+      "clip_ratio/high_max": 0.1875,
+      "clip_ratio/high_mean": 0.1015625,
+      "clip_ratio/low_mean": 0.24365234375,
+      "clip_ratio/low_min": 0.12890625,
+      "clip_ratio/region_mean": 0.34521484375,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 469.0,
+      "completions/max_terminated_length": 469.0,
+      "completions/mean_length": 288.99072265625,
+      "completions/mean_terminated_length": 288.99072265625,
+      "completions/min_length": 182.0,
+      "completions/min_terminated_length": 182.0,
+      "entropy": 0.892413392663002,
+      "epoch": 0.7142857142857143,
+      "frac_reward_zero_std": 0.0,
+      "grad_norm": 0.05464347451925278,
+      "learning_rate": 1e-05,
+      "loss": 0.0005,
+      "num_tokens": 38384750.0,
+      "reward": 3.728700637817383,
+      "reward_std": 0.14373056590557098,
+      "rewards/ngram_repetition2/mean": 0.9769801497459412,
+      "rewards/ngram_repetition2/std": 0.011434270069003105,
+      "rewards/ngram_repetition3/mean": 0.9978038668632507,
+      "rewards/ngram_repetition3/std": 0.0036320064682513475,
+      "rewards/symbolic_reward_accuracy/mean": 0.87060546875,
+      "rewards/symbolic_reward_accuracy/std": 0.3357180655002594,
+      "rewards/symbolic_reward_partial_score/mean": 0.9581298828125,
+      "rewards/symbolic_reward_partial_score/std": 0.11856843531131744,
+      "rewards/tag_count_reward/mean": 1.0,
+      "rewards/tag_count_reward/std": 0.0,
+      "rewards/thinking_answer_ratio_reward/mean": 0.9612185955047607,
+      "rewards/thinking_answer_ratio_reward/std": 0.009071115404367447,
+      "sampling/importance_sampling_ratio/max": 2.0,
+      "sampling/importance_sampling_ratio/mean": 1.2778643369674683,
+      "sampling/importance_sampling_ratio/min": 0.001577620510943234,
+      "sampling/sampling_logp_difference/max": 6.451837539672852,
+      "sampling/sampling_logp_difference/mean": 0.3597729802131653,
+      "step": 60
+    },
+    {
+      "clip_ratio/high_max": 0.23046875,
+      "clip_ratio/high_mean": 0.10693359375,
+      "clip_ratio/low_mean": 0.24462890625,
+      "clip_ratio/low_min": 0.11328125,
+      "clip_ratio/region_mean": 0.3515625,
+      "completions/clipped_ratio": 0.00146484375,
+      "completions/max_length": 3072.0,
+      "completions/max_terminated_length": 414.0,
+      "completions/mean_length": 287.26806640625,
+      "completions/mean_terminated_length": 283.1828918457031,
+      "completions/min_length": 180.0,
+      "completions/min_terminated_length": 180.0,
+      "entropy": 0.9810461550951004,
+      "epoch": 0.7619047619047619,
+      "frac_reward_zero_std": 0.0,
+      "grad_norm": 0.06930692493915558,
+      "learning_rate": 1e-05,
+      "loss": 0.0042,
+      "num_tokens": 40800819.0,
+      "reward": 3.598151683807373,
+      "reward_std": 0.17825113236904144,
+      "rewards/ngram_repetition2/mean": 0.9757882952690125,
+      "rewards/ngram_repetition2/std": 0.011011338792741299,
+      "rewards/ngram_repetition3/mean": 0.997752845287323,
+      "rewards/ngram_repetition3/std": 0.0035420055501163006,
+      "rewards/symbolic_reward_accuracy/mean": 0.814453125,
+      "rewards/symbolic_reward_accuracy/std": 0.38883551955223083,
+      "rewards/symbolic_reward_partial_score/mean": 0.9406331181526184,
+      "rewards/symbolic_reward_partial_score/std": 0.14368508756160736,
+      "rewards/tag_count_reward/mean": 0.999267578125,
+      "rewards/tag_count_reward/std": 0.019127286970615387,
+      "rewards/thinking_answer_ratio_reward/mean": 0.9609147310256958,
+      "rewards/thinking_answer_ratio_reward/std": 0.03739883750677109,
+      "sampling/importance_sampling_ratio/max": 2.0,
+      "sampling/importance_sampling_ratio/mean": 1.2963650226593018,
+      "sampling/importance_sampling_ratio/min": 0.0003925914643332362,
+      "sampling/sampling_logp_difference/max": 7.842741012573242,
+      "sampling/sampling_logp_difference/mean": 0.37864184379577637,
+      "step": 64
+    },
+    {
+      "clip_ratio/high_max": 0.140625,
+      "clip_ratio/high_mean": 0.06494140625,
+      "clip_ratio/low_mean": 0.28662109375,
+      "clip_ratio/low_min": 0.1640625,
+      "clip_ratio/region_mean": 0.3515625,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 462.0,
+      "completions/max_terminated_length": 462.0,
+      "completions/mean_length": 286.521484375,
+      "completions/mean_terminated_length": 286.521484375,
+      "completions/min_length": 168.0,
+      "completions/min_terminated_length": 168.0,
+      "entropy": 1.0116581320762634,
+      "epoch": 0.8095238095238095,
+      "frac_reward_zero_std": 0.0,
+      "grad_norm": 0.050855811685323715,
+      "learning_rate": 1e-05,
+      "loss": 0.0003,
+      "num_tokens": 43193663.0,
+      "reward": 3.882843494415283,
+      "reward_std": 0.035406049340963364,
+      "rewards/ngram_repetition2/mean": 0.9764049053192139,
+      "rewards/ngram_repetition2/std": 0.01124592125415802,
+      "rewards/ngram_repetition3/mean": 0.9978073835372925,
+      "rewards/ngram_repetition3/std": 0.0034347528126090765,
+      "rewards/symbolic_reward_accuracy/mean": 0.93408203125,
+      "rewards/symbolic_reward_accuracy/std": 0.2481992542743683,
+      "rewards/symbolic_reward_partial_score/mean": 0.9853108525276184,
+      "rewards/symbolic_reward_partial_score/std": 0.05851730331778526,
+      "rewards/tag_count_reward/mean": 1.0,
+      "rewards/tag_count_reward/std": 0.0,
+      "rewards/thinking_answer_ratio_reward/mean": 0.9626501798629761,
+      "rewards/thinking_answer_ratio_reward/std": 0.006846116855740547,
+      "sampling/importance_sampling_ratio/max": 2.0,
+      "sampling/importance_sampling_ratio/mean": 1.3146123886108398,
+      "sampling/importance_sampling_ratio/min": 0.0005279523902572691,
+      "sampling/sampling_logp_difference/max": 7.546504497528076,
+      "sampling/sampling_logp_difference/mean": 0.39858800172805786,
+      "step": 68
+    },
+    {
+      "clip_ratio/high_max": 0.2265625,
+      "clip_ratio/high_mean": 0.12109375,
+      "clip_ratio/low_mean": 0.24951171875,
+      "clip_ratio/low_min": 0.12890625,
+      "clip_ratio/region_mean": 0.37060546875,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 491.0,
+      "completions/max_terminated_length": 491.0,
+      "completions/mean_length": 294.37060546875,
+      "completions/mean_terminated_length": 294.37060546875,
+      "completions/min_length": 180.0,
+      "completions/min_terminated_length": 180.0,
+      "entropy": 1.132740594446659,
+      "epoch": 0.8571428571428571,
+      "frac_reward_zero_std": 0.0,
+      "grad_norm": 0.066042959690094,
+      "learning_rate": 1e-05,
+      "loss": 0.0021,
+      "num_tokens": 45636310.0,
+      "reward": 3.707709789276123,
+      "reward_std": 0.08106810599565506,
+      "rewards/ngram_repetition2/mean": 0.9696463346481323,
+      "rewards/ngram_repetition2/std": 0.014249228872358799,
+      "rewards/ngram_repetition3/mean": 0.9962953329086304,
+      "rewards/ngram_repetition3/std": 0.005218631587922573,
+      "rewards/symbolic_reward_accuracy/mean": 0.86328125,
+      "rewards/symbolic_reward_accuracy/std": 0.34363409876823425,
+      "rewards/symbolic_reward_partial_score/mean": 0.9518636465072632,
+      "rewards/symbolic_reward_partial_score/std": 0.1330701857805252,
+      "rewards/tag_count_reward/mean": 1.0,
+      "rewards/tag_count_reward/std": 0.0,
+      "rewards/thinking_answer_ratio_reward/mean": 0.9624096751213074,
+      "rewards/thinking_answer_ratio_reward/std": 0.008981175720691681,
+      "sampling/importance_sampling_ratio/max": 2.0,
+      "sampling/importance_sampling_ratio/mean": 1.3442730903625488,
+      "sampling/importance_sampling_ratio/min": 0.0022706298623234034,
+      "sampling/sampling_logp_difference/max": 6.087697982788086,
+      "sampling/sampling_logp_difference/mean": 0.4308692216873169,
+      "step": 72
+    },
+    {
+      "clip_ratio/high_max": 0.171875,
+      "clip_ratio/high_mean": 0.07421875,
+      "clip_ratio/low_mean": 0.24853515625,
+      "clip_ratio/low_min": 0.1171875,
+      "clip_ratio/region_mean": 0.32275390625,
+      "completions/clipped_ratio": 0.0009765625,
+      "completions/max_length": 3072.0,
+      "completions/max_terminated_length": 566.0,
+      "completions/mean_length": 298.833984375,
+      "completions/mean_terminated_length": 296.1231689453125,
+      "completions/min_length": 176.0,
+      "completions/min_terminated_length": 176.0,
+      "entropy": 1.1593250632286072,
+      "epoch": 0.9047619047619048,
+      "frac_reward_zero_std": 0.0,
+      "grad_norm": 0.07082995772361755,
+      "learning_rate": 1e-05,
+      "loss": 0.0055,
+      "num_tokens": 48059586.0,
+      "reward": 3.8371598720550537,
+      "reward_std": 0.05605494976043701,
+      "rewards/ngram_repetition2/mean": 0.9693177938461304,
+      "rewards/ngram_repetition2/std": 0.015271971002221107,
+      "rewards/ngram_repetition3/mean": 0.9960181713104248,
+      "rewards/ngram_repetition3/std": 0.006110348738729954,
+      "rewards/symbolic_reward_accuracy/mean": 0.91796875,
+      "rewards/symbolic_reward_accuracy/std": 0.27447932958602905,
+      "rewards/symbolic_reward_partial_score/mean": 0.9724527597427368,
+      "rewards/symbolic_reward_partial_score/std": 0.10404884815216064,
+      "rewards/tag_count_reward/mean": 0.99951171875,
+      "rewards/tag_count_reward/std": 0.015621182508766651,
+      "rewards/thinking_answer_ratio_reward/mean": 0.9604611992835999,
+      "rewards/thinking_answer_ratio_reward/std": 0.03180825710296631,
+      "sampling/importance_sampling_ratio/max": 2.0,
+      "sampling/importance_sampling_ratio/mean": 1.350536584854126,
+      "sampling/importance_sampling_ratio/min": 0.0011175222462043166,
+      "sampling/sampling_logp_difference/max": 6.7966413497924805,
+      "sampling/sampling_logp_difference/mean": 0.4359755218029022,
+      "step": 76
+    },
+    {
+      "clip_ratio/high_max": 0.22265625,
+      "clip_ratio/high_mean": 0.1142578125,
+      "clip_ratio/low_mean": 0.22119140625,
+      "clip_ratio/low_min": 0.09375,
+      "clip_ratio/region_mean": 0.33544921875,
+      "completions/clipped_ratio": 0.0009765625,
+      "completions/max_length": 3072.0,
+      "completions/max_terminated_length": 490.0,
+      "completions/mean_length": 302.49365234375,
+      "completions/mean_terminated_length": 299.7864074707031,
+      "completions/min_length": 163.0,
+      "completions/min_terminated_length": 163.0,
+      "entropy": 1.2683400809764862,
+      "epoch": 0.9523809523809523,
+      "frac_reward_zero_std": 0.0,
+      "grad_norm": 0.05979188159108162,
+      "learning_rate": 1e-05,
+      "loss": 0.0054,
+      "num_tokens": 50487189.0,
+      "reward": 3.8846912384033203,
+      "reward_std": 0.12184189260005951,
+      "rewards/ngram_repetition2/mean": 0.9656546115875244,
+      "rewards/ngram_repetition2/std": 0.01537134125828743,
+      "rewards/ngram_repetition3/mean": 0.995905876159668,
+      "rewards/ngram_repetition3/std": 0.005492487922310829,
+      "rewards/symbolic_reward_accuracy/mean": 0.93701171875,
+      "rewards/symbolic_reward_accuracy/std": 0.24300122261047363,
+      "rewards/symbolic_reward_partial_score/mean": 0.98193359375,
+      "rewards/symbolic_reward_partial_score/std": 0.0880974605679512,
+      "rewards/tag_count_reward/mean": 0.99951171875,
+      "rewards/tag_count_reward/std": 0.015621182508766651,
+      "rewards/thinking_answer_ratio_reward/mean": 0.9606635570526123,
+      "rewards/thinking_answer_ratio_reward/std": 0.032527752220630646,
+      "sampling/importance_sampling_ratio/max": 2.0,
+      "sampling/importance_sampling_ratio/mean": 1.3736423254013062,
+      "sampling/importance_sampling_ratio/min": 3.4236355070382807e-12,
+      "sampling/sampling_logp_difference/max": 26.400318145751953,
+      "sampling/sampling_logp_difference/mean": 0.46326184272766113,
+      "step": 80
+    },
+    {
+      "clip_ratio/high_max": 0.17578125,
+      "clip_ratio/high_mean": 0.087890625,
+      "clip_ratio/low_mean": 0.2724609375,
+      "clip_ratio/low_min": 0.15625,
+      "clip_ratio/region_mean": 0.3603515625,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 498.0,
+      "completions/max_terminated_length": 498.0,
+      "completions/mean_length": 304.1044921875,
+      "completions/mean_terminated_length": 304.1044921875,
+      "completions/min_length": 173.0,
+      "completions/min_terminated_length": 173.0,
+      "entropy": 1.2831775099039078,
+      "epoch": 1.0,
+      "frac_reward_zero_std": 0.0,
+      "grad_norm": 0.05425618216395378,
+      "learning_rate": 1e-05,
+      "loss": 0.0019,
+      "num_tokens": 52986987.0,
+      "reward": 3.9170305728912354,
+      "reward_std": 0.060697440057992935,
+      "rewards/ngram_repetition2/mean": 0.965947151184082,
+      "rewards/ngram_repetition2/std": 0.014219055883586407,
+      "rewards/ngram_repetition3/mean": 0.9961841702461243,
+      "rewards/ngram_repetition3/std": 0.005091497208923101,
+      "rewards/symbolic_reward_accuracy/mean": 0.95166015625,
+      "rewards/symbolic_reward_accuracy/std": 0.21453572809696198,
+      "rewards/symbolic_reward_partial_score/mean": 0.9844564199447632,
+      "rewards/symbolic_reward_partial_score/std": 0.07733123749494553,
+      "rewards/tag_count_reward/mean": 1.0,
+      "rewards/tag_count_reward/std": 0.0,
+      "rewards/thinking_answer_ratio_reward/mean": 0.9632632732391357,
+      "rewards/thinking_answer_ratio_reward/std": 0.008934850804507732,
+      "sampling/importance_sampling_ratio/max": 2.0,
+      "sampling/importance_sampling_ratio/mean": 1.3828954696655273,
+      "sampling/importance_sampling_ratio/min": 6.9070331676357455e-09,
+      "sampling/sampling_logp_difference/max": 18.790725708007812,
+      "sampling/sampling_logp_difference/mean": 0.46976035833358765,
+      "step": 84
+    },
+    {
+      "epoch": 1.0,
+      "eval_clip_ratio/high_max": 0.0,
+      "eval_clip_ratio/high_mean": 0.0,
+      "eval_clip_ratio/low_mean": 0.0,
+      "eval_clip_ratio/low_min": 0.0,
+      "eval_clip_ratio/region_mean": 0.0,
+      "eval_completions/clipped_ratio": 0.0,
+      "eval_completions/max_length": 437.8421052631579,
+      "eval_completions/max_terminated_length": 437.8421052631579,
+      "eval_completions/mean_length": 312.72286184210526,
+      "eval_completions/mean_terminated_length": 312.72286184210526,
+      "eval_completions/min_length": 207.52631578947367,
+      "eval_completions/min_terminated_length": 207.52631578947367,
+      "eval_entropy": 1.3218948276419389,
+      "eval_frac_reward_zero_std": 0.0,
+      "eval_loss": 0.00022543604427482933,
+      "eval_num_tokens": 52986987.0,
+      "eval_reward": 3.807651005293194,
+      "eval_reward_std": 0.168629709851781,
+      "eval_rewards/ngram_repetition2/mean": 0.961865462754902,
+      "eval_rewards/ngram_repetition2/std": 0.014227556211775854,
+      "eval_rewards/ngram_repetition3/mean": 0.9952163006130018,
+      "eval_rewards/ngram_repetition3/std": 0.004750296277435203,
+      "eval_rewards/symbolic_reward_accuracy/mean": 0.90625,
+      "eval_rewards/symbolic_reward_accuracy/std": 0.2370802113100102,
+      "eval_rewards/symbolic_reward_partial_score/mean": 0.9659745065789473,
+      "eval_rewards/symbolic_reward_partial_score/std": 0.09205205454245995,
+      "eval_rewards/tag_count_reward/mean": 1.0,
+      "eval_rewards/tag_count_reward/std": 0.0,
+      "eval_rewards/thinking_answer_ratio_reward/mean": 0.9605719666731986,
+      "eval_rewards/thinking_answer_ratio_reward/std": 0.01121363438371765,
+      "eval_runtime": 423.2341,
+      "eval_samples_per_second": 0.354,
+      "eval_sampling/importance_sampling_ratio/max": 2.0,
+      "eval_sampling/importance_sampling_ratio/mean": 1.392908353554575,
+      "eval_sampling/importance_sampling_ratio/min": 0.0026607212291585376,
+      "eval_sampling/sampling_logp_difference/max": 7.379827775453267,
+      "eval_sampling/sampling_logp_difference/mean": 0.4819344439004597,
+      "eval_steps_per_second": 0.005,
+      "step": 84
+    },
+    {
+      "clip_ratio/high_max": 0.2421875,
+      "clip_ratio/high_mean": 0.13232421875,
+      "clip_ratio/low_mean": 0.255859375,
+      "clip_ratio/low_min": 0.13671875,
+      "clip_ratio/region_mean": 0.38818359375,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 486.0,
+      "completions/max_terminated_length": 486.0,
+      "completions/mean_length": 311.048828125,
+      "completions/mean_terminated_length": 311.048828125,
+      "completions/min_length": 186.0,
+      "completions/min_terminated_length": 186.0,
+      "entropy": 1.3394631668925285,
+      "epoch": 1.0476190476190477,
+      "frac_reward_zero_std": 0.0,
+      "grad_norm": 0.050926946103572845,
+      "learning_rate": 1e-05,
+      "loss": 0.0015,
+      "num_tokens": 55457935.0,
+      "reward": 3.749323844909668,
+      "reward_std": 0.11920632421970367,
+      "rewards/ngram_repetition2/mean": 0.9640907049179077,
+      "rewards/ngram_repetition2/std": 0.013935322873294353,
+      "rewards/ngram_repetition3/mean": 0.9960398077964783,
+      "rewards/ngram_repetition3/std": 0.0044942288659513,
+      "rewards/symbolic_reward_accuracy/mean": 0.876953125,
+      "rewards/symbolic_reward_accuracy/std": 0.32857123017311096,
+      "rewards/symbolic_reward_partial_score/mean": 0.9661865234375,
+      "rewards/symbolic_reward_partial_score/std": 0.10201738774776459,
+      "rewards/tag_count_reward/mean": 1.0,
+      "rewards/tag_count_reward/std": 0.0,
+      "rewards/thinking_answer_ratio_reward/mean": 0.9629674553871155,
+      "rewards/thinking_answer_ratio_reward/std": 0.009922855533659458,
+      "sampling/importance_sampling_ratio/max": 2.0,
+      "sampling/importance_sampling_ratio/mean": 1.3947930335998535,
+      "sampling/importance_sampling_ratio/min": 1.1835306068519458e-09,
+      "sampling/sampling_logp_difference/max": 20.554763793945312,
+      "sampling/sampling_logp_difference/mean": 0.4833507537841797,
+      "step": 88
+    },
+    {
+      "clip_ratio/high_max": 0.19140625,
+      "clip_ratio/high_mean": 0.09814453125,
+      "clip_ratio/low_mean": 0.27685546875,
+      "clip_ratio/low_min": 0.15234375,
+      "clip_ratio/region_mean": 0.375,
+      "completions/clipped_ratio": 0.0009765625,
+      "completions/max_length": 3072.0,
+      "completions/max_terminated_length": 477.0,
+      "completions/mean_length": 309.70361328125,
+      "completions/mean_terminated_length": 307.00341796875,
+      "completions/min_length": 191.0,
+      "completions/min_terminated_length": 191.0,
+      "entropy": 1.365867592394352,
+      "epoch": 1.0952380952380953,
+      "frac_reward_zero_std": 0.0,
+      "grad_norm": 0.050299327820539474,
+      "learning_rate": 1e-05,
+      "loss": 0.0053,
+      "num_tokens": 57938000.0,
+      "reward": 3.8700785636901855,
+      "reward_std": 0.055703483521938324,
+      "rewards/ngram_repetition2/mean": 0.9664497375488281,
+      "rewards/ngram_repetition2/std": 0.01341775618493557,
+      "rewards/ngram_repetition3/mean": 0.9965437650680542,
+      "rewards/ngram_repetition3/std": 0.004179587122052908,
+      "rewards/symbolic_reward_accuracy/mean": 0.9296875,
+      "rewards/symbolic_reward_accuracy/std": 0.2557353675365448,
+      "rewards/symbolic_reward_partial_score/mean": 0.98193359375,
+      "rewards/symbolic_reward_partial_score/std": 0.07467525452375412,
+      "rewards/tag_count_reward/mean": 0.99951171875,
+      "rewards/tag_count_reward/std": 0.015621182508766651,
+      "rewards/thinking_answer_ratio_reward/mean": 0.9628340601921082,
+      "rewards/thinking_answer_ratio_reward/std": 0.03138146176934242,
+      "sampling/importance_sampling_ratio/max": 2.0,
+      "sampling/importance_sampling_ratio/mean": 1.397605538368225,
+      "sampling/importance_sampling_ratio/min": 3.21872563324721e-11,
+      "sampling/sampling_logp_difference/max": 24.15945053100586,
+      "sampling/sampling_logp_difference/mean": 0.49306315183639526,
+      "step": 92
+    },
+    {
+      "clip_ratio/high_max": 0.17578125,
+      "clip_ratio/high_mean": 0.08984375,
+      "clip_ratio/low_mean": 0.29541015625,
+      "clip_ratio/low_min": 0.171875,
+      "clip_ratio/region_mean": 0.38525390625,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 474.0,
+      "completions/max_terminated_length": 474.0,
+      "completions/mean_length": 324.9814453125,
+      "completions/mean_terminated_length": 324.9814453125,
+      "completions/min_length": 186.0,
+      "completions/min_terminated_length": 186.0,
+      "entropy": 1.4058298990130424,
+      "epoch": 1.1428571428571428,
+      "frac_reward_zero_std": 0.0,
+      "grad_norm": 0.06967155635356903,
+      "learning_rate": 1e-05,
+      "loss": 0.0021,
+      "num_tokens": 60416266.0,
+      "reward": 3.8939476013183594,
+      "reward_std": 0.05917710065841675,
+      "rewards/ngram_repetition2/mean": 0.9656116962432861,
+      "rewards/ngram_repetition2/std": 0.013012575916945934,
+      "rewards/ngram_repetition3/mean": 0.9965729117393494,
+      "rewards/ngram_repetition3/std": 0.0041229212656617165,
+      "rewards/symbolic_reward_accuracy/mean": 0.9423828125,
+      "rewards/symbolic_reward_accuracy/std": 0.23307503759860992,
+      "rewards/symbolic_reward_partial_score/mean": 0.9798990488052368,
+      "rewards/symbolic_reward_partial_score/std": 0.09529523551464081,
+      "rewards/tag_count_reward/mean": 1.0,
+      "rewards/tag_count_reward/std": 0.0,
+      "rewards/thinking_answer_ratio_reward/mean": 0.9661376476287842,
+      "rewards/thinking_answer_ratio_reward/std": 0.007734327111393213,
+      "sampling/importance_sampling_ratio/max": 2.0,
+      "sampling/importance_sampling_ratio/mean": 1.407299518585205,
+      "sampling/importance_sampling_ratio/min": 0.0021590718533843756,
+      "sampling/sampling_logp_difference/max": 6.1380767822265625,
+      "sampling/sampling_logp_difference/mean": 0.5017811059951782,
+      "step": 96
+    },
+    {
+      "clip_ratio/high_max": 0.12890625,
+      "clip_ratio/high_mean": 0.0595703125,
+      "clip_ratio/low_mean": 0.30810546875,
+      "clip_ratio/low_min": 0.1875,
+      "clip_ratio/region_mean": 0.36767578125,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 513.0,
+      "completions/max_terminated_length": 513.0,
+      "completions/mean_length": 332.615234375,
+      "completions/mean_terminated_length": 332.615234375,
+      "completions/min_length": 187.0,
+      "completions/min_terminated_length": 187.0,
+      "entropy": 1.47202467918396,
+      "epoch": 1.1904761904761905,
+      "frac_reward_zero_std": 0.0,
+      "grad_norm": 0.07687471061944962,
+      "learning_rate": 1e-05,
+      "loss": 0.0022,
+      "num_tokens": 62884182.0,
+      "reward": 3.7261898517608643,
+      "reward_std": 0.08603204041719437,
+      "rewards/ngram_repetition2/mean": 0.962227463722229,
+      "rewards/ngram_repetition2/std": 0.013661311008036137,
+      "rewards/ngram_repetition3/mean": 0.9959136247634888,
+      "rewards/ngram_repetition3/std": 0.004227799363434315,
+      "rewards/symbolic_reward_accuracy/mean": 0.8671875,
+      "rewards/symbolic_reward_accuracy/std": 0.33945485949516296,
+      "rewards/symbolic_reward_partial_score/mean": 0.9625650644302368,
+      "rewards/symbolic_reward_partial_score/std": 0.1090090274810791,
+      "rewards/tag_count_reward/mean": 1.0,
+      "rewards/tag_count_reward/std": 0.0,
+      "rewards/thinking_answer_ratio_reward/mean": 0.9668258428573608,
+      "rewards/thinking_answer_ratio_reward/std": 0.007831486873328686,
+      "sampling/importance_sampling_ratio/max": 2.0,
+      "sampling/importance_sampling_ratio/mean": 1.4201319217681885,
+      "sampling/importance_sampling_ratio/min": 0.00016027815581765026,
+      "sampling/sampling_logp_difference/max": 8.73859977722168,
+      "sampling/sampling_logp_difference/mean": 0.5191107392311096,
+      "step": 100
+    },
+    {
+      "clip_ratio/high_max": 0.11328125,
+      "clip_ratio/high_mean": 0.05224609375,
+      "clip_ratio/low_mean": 0.31005859375,
+      "clip_ratio/low_min": 0.15625,
+      "clip_ratio/region_mean": 0.3623046875,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 586.0,
+      "completions/max_terminated_length": 586.0,
+      "completions/mean_length": 352.5576171875,
+      "completions/mean_terminated_length": 352.5576171875,
+      "completions/min_length": 197.0,
+      "completions/min_terminated_length": 197.0,
+      "entropy": 1.5019978284835815,
+      "epoch": 1.2380952380952381,
+      "frac_reward_zero_std": 0.0,
+      "grad_norm": 0.08184507489204407,
+      "learning_rate": 1e-05,
+      "loss": 0.0038,
+      "num_tokens": 65438860.0,
+      "reward": 3.8091237545013428,
+      "reward_std": 0.08657659590244293,
+      "rewards/ngram_repetition2/mean": 0.9608925580978394,
+      "rewards/ngram_repetition2/std": 0.013776077888906002,
+      "rewards/ngram_repetition3/mean": 0.9957015514373779,
+      "rewards/ngram_repetition3/std": 0.004236322361975908,
+      "rewards/symbolic_reward_accuracy/mean": 0.90234375,
+      "rewards/symbolic_reward_accuracy/std": 0.2969217896461487,
+      "rewards/symbolic_reward_partial_score/mean": 0.9751790761947632,
+      "rewards/symbolic_reward_partial_score/std": 0.07810740172863007,
+      "rewards/tag_count_reward/mean": 1.0,
+      "rewards/tag_count_reward/std": 0.0,
+      "rewards/thinking_answer_ratio_reward/mean": 0.9691292643547058,
+      "rewards/thinking_answer_ratio_reward/std": 0.007102734409272671,
+      "sampling/importance_sampling_ratio/max": 2.0,
+      "sampling/importance_sampling_ratio/mean": 1.4305992126464844,
+      "sampling/importance_sampling_ratio/min": 5.72370809095446e-05,
+      "sampling/sampling_logp_difference/max": 9.768308639526367,
+      "sampling/sampling_logp_difference/mean": 0.5295636057853699,
+      "step": 104
+    },
+    {
+      "clip_ratio/high_max": 0.14453125,
+      "clip_ratio/high_mean": 0.068359375,
+      "clip_ratio/low_mean": 0.263671875,
+      "clip_ratio/low_min": 0.13671875,
+      "clip_ratio/region_mean": 0.33203125,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 535.0,
+      "completions/max_terminated_length": 535.0,
+      "completions/mean_length": 354.79345703125,
+      "completions/mean_terminated_length": 354.79345703125,
+      "completions/min_length": 144.0,
+      "completions/min_terminated_length": 144.0,
+      "entropy": 1.5520799160003662,
+      "epoch": 1.2857142857142856,
+      "frac_reward_zero_std": 0.0,
+      "grad_norm": 0.1121075302362442,
+      "learning_rate": 1e-05,
+      "loss": 0.0041,
+      "num_tokens": 68023781.0,
+      "reward": 3.847431182861328,
+      "reward_std": 0.08936133980751038,
+      "rewards/ngram_repetition2/mean": 0.9565550088882446,
+      "rewards/ngram_repetition2/std": 0.015181444585323334,
+      "rewards/ngram_repetition3/mean": 0.9949396848678589,
+      "rewards/ngram_repetition3/std": 0.004793631378561258,
+      "rewards/symbolic_reward_accuracy/mean": 0.92041015625,
+      "rewards/symbolic_reward_accuracy/std": 0.27072325348854065,
+      "rewards/symbolic_reward_partial_score/mean": 0.9776611328125,
+      "rewards/symbolic_reward_partial_score/std": 0.09112773835659027,
+      "rewards/tag_count_reward/mean": 0.999755859375,
+      "rewards/tag_count_reward/std": 0.011048543266952038,
+      "rewards/thinking_answer_ratio_reward/mean": 0.9678950309753418,
+      "rewards/thinking_answer_ratio_reward/std": 0.022913049906492233,
+      "sampling/importance_sampling_ratio/max": 2.0,
+      "sampling/importance_sampling_ratio/mean": 1.4402625560760498,
+      "sampling/importance_sampling_ratio/min": 6.558643690368626e-06,
+      "sampling/sampling_logp_difference/max": 11.93472671508789,
+      "sampling/sampling_logp_difference/mean": 0.541069746017456,
+      "step": 108
+    },
+    {
+      "clip_ratio/high_max": 0.1484375,
+      "clip_ratio/high_mean": 0.068359375,
+      "clip_ratio/low_mean": 0.26953125,
+      "clip_ratio/low_min": 0.15234375,
+      "clip_ratio/region_mean": 0.337890625,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 509.0,
+      "completions/max_terminated_length": 509.0,
+      "completions/mean_length": 334.44384765625,
+      "completions/mean_terminated_length": 334.44384765625,
+      "completions/min_length": 163.0,
+      "completions/min_terminated_length": 163.0,
+      "entropy": 1.5340367630124092,
+      "epoch": 1.3333333333333333,
+      "frac_reward_zero_std": 0.0,
+      "grad_norm": 0.08651846647262573,
+      "learning_rate": 1e-05,
+      "loss": 0.0032,
+      "num_tokens": 70484050.0,
+      "reward": 3.858900785446167,
+      "reward_std": 0.04411323368549347,
+      "rewards/ngram_repetition2/mean": 0.9606438875198364,
+      "rewards/ngram_repetition2/std": 0.01436516921967268,
+      "rewards/ngram_repetition3/mean": 0.995490550994873,
+      "rewards/ngram_repetition3/std": 0.004695704206824303,
+      "rewards/symbolic_reward_accuracy/mean": 0.92333984375,
+      "rewards/symbolic_reward_accuracy/std": 0.2661164402961731,
+      "rewards/symbolic_reward_partial_score/mean": 0.9829915165901184,
+      "rewards/symbolic_reward_partial_score/std": 0.06513792276382446,
+      "rewards/tag_count_reward/mean": 1.0,
+      "rewards/tag_count_reward/std": 0.0,
+      "rewards/thinking_answer_ratio_reward/mean": 0.966830313205719,
+      "rewards/thinking_answer_ratio_reward/std": 0.007882185280323029,
+      "sampling/importance_sampling_ratio/max": 2.0,
+      "sampling/importance_sampling_ratio/mean": 1.4405858516693115,
+      "sampling/importance_sampling_ratio/min": 0.00030129292281344533,
+      "sampling/sampling_logp_difference/max": 8.107427597045898,
+      "sampling/sampling_logp_difference/mean": 0.5405493974685669,
+      "step": 112
+    },
+    {
+      "clip_ratio/high_max": 0.16015625,
+      "clip_ratio/high_mean": 0.072265625,
+      "clip_ratio/low_mean": 0.283203125,
+      "clip_ratio/low_min": 0.15234375,
+      "clip_ratio/region_mean": 0.35546875,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 550.0,
+      "completions/max_terminated_length": 550.0,
+      "completions/mean_length": 347.263671875,
+      "completions/mean_terminated_length": 347.263671875,
+      "completions/min_length": 159.0,
+      "completions/min_terminated_length": 159.0,
+      "entropy": 1.6167370453476906,
+      "epoch": 1.380952380952381,
+      "frac_reward_zero_std": 0.0,
+      "grad_norm": 0.1051953062415123,
+      "learning_rate": 1e-05,
+      "loss": 0.0045,
+      "num_tokens": 73007790.0,
+      "reward": 3.867628574371338,
+      "reward_std": 0.04132593050599098,
+      "rewards/ngram_repetition2/mean": 0.9570583701133728,
+      "rewards/ngram_repetition2/std": 0.014389286749064922,
+      "rewards/ngram_repetition3/mean": 0.9954445362091064,
+      "rewards/ngram_repetition3/std": 0.004455200862139463,
+      "rewards/symbolic_reward_accuracy/mean": 0.9296875,
+      "rewards/symbolic_reward_accuracy/std": 0.2557353675365448,
+      "rewards/symbolic_reward_partial_score/mean": 0.9790445566177368,
+      "rewards/symbolic_reward_partial_score/std": 0.08579050749540329,
+      "rewards/tag_count_reward/mean": 1.0,
+      "rewards/tag_count_reward/std": 0.0,
+      "rewards/thinking_answer_ratio_reward/mean": 0.9683747291564941,
+      "rewards/thinking_answer_ratio_reward/std": 0.007507936097681522,
+      "sampling/importance_sampling_ratio/max": 2.0,
+      "sampling/importance_sampling_ratio/mean": 1.455453872680664,
+      "sampling/importance_sampling_ratio/min": 6.147195108496817e-08,
+      "sampling/sampling_logp_difference/max": 16.604684829711914,
+      "sampling/sampling_logp_difference/mean": 0.561433732509613,
+      "step": 116
+    },
+    {
+      "clip_ratio/high_max": 0.1171875,
+      "clip_ratio/high_mean": 0.05029296875,
+      "clip_ratio/low_mean": 0.306640625,
+      "clip_ratio/low_min": 0.1796875,
+      "clip_ratio/region_mean": 0.35693359375,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 565.0,
+      "completions/max_terminated_length": 565.0,
+      "completions/mean_length": 359.7197265625,
+      "completions/mean_terminated_length": 359.7197265625,
+      "completions/min_length": 181.0,
+      "completions/min_terminated_length": 181.0,
+      "entropy": 1.664498396217823,
+      "epoch": 1.4285714285714286,
+      "frac_reward_zero_std": 0.0,
+      "grad_norm": 0.08962584286928177,
+      "learning_rate": 1e-05,
+      "loss": 0.0057,
+      "num_tokens": 75565264.0,
+      "reward": 3.8777501583099365,
+      "reward_std": 0.04693379998207092,
+      "rewards/ngram_repetition2/mean": 0.9551849961280823,
+      "rewards/ngram_repetition2/std": 0.01518841739743948,
+      "rewards/ngram_repetition3/mean": 0.9952660799026489,
+      "rewards/ngram_repetition3/std": 0.0045869555324316025,
+      "rewards/symbolic_reward_accuracy/mean": 0.93359375,
+      "rewards/symbolic_reward_accuracy/std": 0.24905169010162354,
+      "rewards/symbolic_reward_partial_score/mean": 0.9816080331802368,
+      "rewards/symbolic_reward_partial_score/std": 0.07948367297649384,
+      "rewards/tag_count_reward/mean": 0.999755859375,
+      "rewards/tag_count_reward/std": 0.011048543266952038,
+      "rewards/thinking_answer_ratio_reward/mean": 0.9694327116012573,
+      "rewards/thinking_answer_ratio_reward/std": 0.02234593965113163,
+      "sampling/importance_sampling_ratio/max": 2.0,
+      "sampling/importance_sampling_ratio/mean": 1.4626634120941162,
+      "sampling/importance_sampling_ratio/min": 0.0012032584054395556,
+      "sampling/sampling_logp_difference/max": 6.722722053527832,
+      "sampling/sampling_logp_difference/mean": 0.571152925491333,
+      "step": 120
+    },
+    {
+      "clip_ratio/high_max": 0.15234375,
+      "clip_ratio/high_mean": 0.07080078125,
+      "clip_ratio/low_mean": 0.29541015625,
+      "clip_ratio/low_min": 0.1796875,
+      "clip_ratio/region_mean": 0.3662109375,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 603.0,
+      "completions/max_terminated_length": 603.0,
+      "completions/mean_length": 367.48046875,
+      "completions/mean_terminated_length": 367.48046875,
+      "completions/min_length": 160.0,
+      "completions/min_terminated_length": 160.0,
+      "entropy": 1.703298144042492,
+      "epoch": 1.4761904761904763,
+      "frac_reward_zero_std": 0.0,
+      "grad_norm": 0.114425890147686,
+      "learning_rate": 1e-05,
+      "loss": 0.0068,
+      "num_tokens": 78163976.0,
+      "reward": 3.6791064739227295,
+      "reward_std": 0.11728330701589584,
+      "rewards/ngram_repetition2/mean": 0.9517308473587036,
+      "rewards/ngram_repetition2/std": 0.016181044280529022,
+      "rewards/ngram_repetition3/mean": 0.9945051074028015,
+      "rewards/ngram_repetition3/std": 0.004997564479708672,
+      "rewards/symbolic_reward_accuracy/mean": 0.8505859375,
+      "rewards/symbolic_reward_accuracy/std": 0.35658320784568787,
+      "rewards/symbolic_reward_partial_score/mean": 0.9487711191177368,
+      "rewards/symbolic_reward_partial_score/std": 0.1491239219903946,
+      "rewards/tag_count_reward/mean": 1.0,
+      "rewards/tag_count_reward/std": 0.0,
+      "rewards/thinking_answer_ratio_reward/mean": 0.970108687877655,
+      "rewards/thinking_answer_ratio_reward/std": 0.00691350968554616,
+      "sampling/importance_sampling_ratio/max": 2.0,
+      "sampling/importance_sampling_ratio/mean": 1.47133207321167,
+      "sampling/importance_sampling_ratio/min": 0.0017549983458593488,
+      "sampling/sampling_logp_difference/max": 6.345287322998047,
+      "sampling/sampling_logp_difference/mean": 0.5865588784217834,
+      "step": 124
+    },
+    {
+      "clip_ratio/high_max": 0.0859375,
+      "clip_ratio/high_mean": 0.0400390625,
+      "clip_ratio/low_mean": 0.29833984375,
+      "clip_ratio/low_min": 0.19140625,
+      "clip_ratio/region_mean": 0.33837890625,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 626.0,
+      "completions/max_terminated_length": 626.0,
+      "completions/mean_length": 372.03515625,
+      "completions/mean_terminated_length": 372.03515625,
+      "completions/min_length": 184.0,
+      "completions/min_terminated_length": 184.0,
+      "entropy": 1.7273126989603043,
+      "epoch": 1.5238095238095237,
+      "frac_reward_zero_std": 0.0,
+      "grad_norm": 0.09851213544607162,
+      "learning_rate": 1e-05,
+      "loss": 0.0081,
+      "num_tokens": 80753328.0,
+      "reward": 3.826509475708008,
+      "reward_std": 0.05608430877327919,
+      "rewards/ngram_repetition2/mean": 0.9504142999649048,
+      "rewards/ngram_repetition2/std": 0.01682715304195881,
+      "rewards/ngram_repetition3/mean": 0.9940425753593445,
+      "rewards/ngram_repetition3/std": 0.005228669848293066,
+      "rewards/symbolic_reward_accuracy/mean": 0.91357421875,
+      "rewards/symbolic_reward_accuracy/std": 0.2810603678226471,
+      "rewards/symbolic_reward_partial_score/mean": 0.970458984375,
+      "rewards/symbolic_reward_partial_score/std": 0.1112869530916214,
+      "rewards/tag_count_reward/mean": 0.999755859375,
+      "rewards/tag_count_reward/std": 0.011048543266952038,
+      "rewards/thinking_answer_ratio_reward/mean": 0.9701868295669556,
+      "rewards/thinking_answer_ratio_reward/std": 0.007006238214671612,
+      "sampling/importance_sampling_ratio/max": 2.0,
+      "sampling/importance_sampling_ratio/mean": 1.474413514137268,
+      "sampling/importance_sampling_ratio/min": 0.002049426781013608,
+      "sampling/sampling_logp_difference/max": 6.190195083618164,
+      "sampling/sampling_logp_difference/mean": 0.5897567868232727,
+      "step": 128
+    },
+    {
+      "clip_ratio/high_max": 0.1015625,
+      "clip_ratio/high_mean": 0.03515625,
+      "clip_ratio/low_mean": 0.3115234375,
+      "clip_ratio/low_min": 0.17578125,
+      "clip_ratio/region_mean": 0.3466796875,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 628.0,
+      "completions/max_terminated_length": 628.0,
+      "completions/mean_length": 364.0087890625,
+      "completions/mean_terminated_length": 364.0087890625,
+      "completions/min_length": 181.0,
+      "completions/min_terminated_length": 181.0,
+      "entropy": 1.7212177067995071,
+      "epoch": 1.5714285714285714,
+      "frac_reward_zero_std": 0.0,
+      "grad_norm": 0.10181765258312225,
+      "learning_rate": 1e-05,
+      "loss": 0.0066,
+      "num_tokens": 83292354.0,
+      "reward": 3.813551425933838,
+      "reward_std": 0.07873611152172089,
+      "rewards/ngram_repetition2/mean": 0.9500588178634644,
+      "rewards/ngram_repetition2/std": 0.016752546653151512,
+      "rewards/ngram_repetition3/mean": 0.9937365055084229,
+      "rewards/ngram_repetition3/std": 0.00556135643273592,
+      "rewards/symbolic_reward_accuracy/mean": 0.9072265625,
+      "rewards/symbolic_reward_accuracy/std": 0.2901855409145355,
+      "rewards/symbolic_reward_partial_score/mean": 0.969970703125,
+      "rewards/symbolic_reward_partial_score/std": 0.1166658028960228,
+      "rewards/tag_count_reward/mean": 1.0,
+      "rewards/tag_count_reward/std": 0.0,
+      "rewards/thinking_answer_ratio_reward/mean": 0.9689657688140869,
+      "rewards/thinking_answer_ratio_reward/std": 0.0076606557704508305,
+      "sampling/importance_sampling_ratio/max": 2.0,
+      "sampling/importance_sampling_ratio/mean": 1.4771678447723389,
+      "sampling/importance_sampling_ratio/min": 1.5947607607813552e-05,
+      "sampling/sampling_logp_difference/max": 11.046201705932617,
+      "sampling/sampling_logp_difference/mean": 0.597915768623352,
+      "step": 132
+    },
+    {
+      "clip_ratio/high_max": 0.10546875,
+      "clip_ratio/high_mean": 0.04150390625,
+      "clip_ratio/low_mean": 0.30908203125,
+      "clip_ratio/low_min": 0.19921875,
+      "clip_ratio/region_mean": 0.3505859375,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 647.0,
+      "completions/max_terminated_length": 647.0,
+      "completions/mean_length": 380.59423828125,
+      "completions/mean_terminated_length": 380.59423828125,
+      "completions/min_length": 205.0,
+      "completions/min_terminated_length": 205.0,
+      "entropy": 1.7492975443601608,
+      "epoch": 1.619047619047619,
+      "frac_reward_zero_std": 0.0,
+      "grad_norm": 0.10864487290382385,
+      "learning_rate": 1e-05,
+      "loss": 0.0077,
+      "num_tokens": 85970979.0,
+      "reward": 3.789088249206543,
+      "reward_std": 0.0549209825694561,
+      "rewards/ngram_repetition2/mean": 0.9480582475662231,
+      "rewards/ngram_repetition2/std": 0.016496751457452774,
+      "rewards/ngram_repetition3/mean": 0.9930194616317749,
+      "rewards/ngram_repetition3/std": 0.005672322120517492,
+      "rewards/symbolic_reward_accuracy/mean": 0.8994140625,
+      "rewards/symbolic_reward_accuracy/std": 0.3008531332015991,
+      "rewards/symbolic_reward_partial_score/mean": 0.9611409306526184,
+      "rewards/symbolic_reward_partial_score/std": 0.14185626804828644,
+      "rewards/tag_count_reward/mean": 1.0,
+      "rewards/tag_count_reward/std": 0.0,
+      "rewards/thinking_answer_ratio_reward/mean": 0.9708449244499207,
+      "rewards/thinking_answer_ratio_reward/std": 0.006850030273199081,
+      "sampling/importance_sampling_ratio/max": 2.0,
+      "sampling/importance_sampling_ratio/mean": 1.4822736978530884,
+      "sampling/importance_sampling_ratio/min": 0.00023519470414612442,
+      "sampling/sampling_logp_difference/max": 8.355096817016602,
+      "sampling/sampling_logp_difference/mean": 0.6033132076263428,
+      "step": 136
+    },
+    {
+      "clip_ratio/high_max": 0.0859375,
+      "clip_ratio/high_mean": 0.029296875,
+      "clip_ratio/low_mean": 0.32568359375,
+      "clip_ratio/low_min": 0.21484375,
+      "clip_ratio/region_mean": 0.35498046875,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 600.0,
+      "completions/max_terminated_length": 600.0,
+      "completions/mean_length": 379.244140625,
+      "completions/mean_terminated_length": 379.244140625,
+      "completions/min_length": 195.0,
+      "completions/min_terminated_length": 195.0,
+      "entropy": 1.7215231880545616,
+      "epoch": 1.6666666666666665,
+      "frac_reward_zero_std": 0.0,
+      "grad_norm": 0.14609399437904358,
+      "learning_rate": 1e-05,
+      "loss": 0.0058,
+      "num_tokens": 88531543.0,
+      "reward": 3.8612782955169678,
+      "reward_std": 0.0853937491774559,
+      "rewards/ngram_repetition2/mean": 0.9484306573867798,
+      "rewards/ngram_repetition2/std": 0.015672912821173668,
+      "rewards/ngram_repetition3/mean": 0.9929347038269043,
+      "rewards/ngram_repetition3/std": 0.0058286660350859165,
+      "rewards/symbolic_reward_accuracy/mean": 0.93310546875,
+      "rewards/symbolic_reward_accuracy/std": 0.24990026652812958,
+      "rewards/symbolic_reward_partial_score/mean": 0.9659423828125,
+      "rewards/symbolic_reward_partial_score/std": 0.14673763513565063,
+      "rewards/tag_count_reward/mean": 1.0,
+      "rewards/tag_count_reward/std": 0.0,
+      "rewards/thinking_answer_ratio_reward/mean": 0.9711217880249023,
+      "rewards/thinking_answer_ratio_reward/std": 0.006088613532483578,
+      "sampling/importance_sampling_ratio/max": 2.0,
+      "sampling/importance_sampling_ratio/mean": 1.4805660247802734,
+      "sampling/importance_sampling_ratio/min": 2.7732114631362492e-06,
+      "sampling/sampling_logp_difference/max": 12.795504570007324,
+      "sampling/sampling_logp_difference/mean": 0.5990421772003174,
+      "step": 140
+    },
+    {
+      "clip_ratio/high_max": 0.1484375,
+      "clip_ratio/high_mean": 0.06201171875,
+      "clip_ratio/low_mean": 0.28369140625,
+      "clip_ratio/low_min": 0.16015625,
+      "clip_ratio/region_mean": 0.345703125,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 594.0,
+      "completions/max_terminated_length": 594.0,
+      "completions/mean_length": 367.521484375,
+      "completions/mean_terminated_length": 367.521484375,
+      "completions/min_length": 206.0,
+      "completions/min_terminated_length": 206.0,
+      "entropy": 1.7308146506547928,
+      "epoch": 1.7142857142857144,
+      "frac_reward_zero_std": 0.0,
+      "grad_norm": 0.08718156069517136,
+      "learning_rate": 1e-05,
+      "loss": 0.006,
+      "num_tokens": 91102787.0,
+      "reward": 3.8670616149902344,
+      "reward_std": 0.047201357781887054,
+      "rewards/ngram_repetition2/mean": 0.9467041492462158,
+      "rewards/ngram_repetition2/std": 0.016446510329842567,
+      "rewards/ngram_repetition3/mean": 0.9924731254577637,
+      "rewards/ngram_repetition3/std": 0.006107911933213472,
+      "rewards/symbolic_reward_accuracy/mean": 0.927734375,
+      "rewards/symbolic_reward_accuracy/std": 0.25899046659469604,
+      "rewards/symbolic_reward_partial_score/mean": 0.9825032353401184,
+      "rewards/symbolic_reward_partial_score/std": 0.07203861325979233,
+      "rewards/tag_count_reward/mean": 1.0,
+      "rewards/tag_count_reward/std": 0.0,
+      "rewards/thinking_answer_ratio_reward/mean": 0.9697957634925842,
+      "rewards/thinking_answer_ratio_reward/std": 0.007489646319299936,
+      "sampling/importance_sampling_ratio/max": 2.0,
+      "sampling/importance_sampling_ratio/mean": 1.4847468137741089,
+      "sampling/importance_sampling_ratio/min": 0.001990552293136716,
+      "sampling/sampling_logp_difference/max": 6.219343185424805,
+      "sampling/sampling_logp_difference/mean": 0.6064555048942566,
+      "step": 144
+    },
+    {
+      "clip_ratio/high_max": 0.1640625,
+      "clip_ratio/high_mean": 0.08203125,
+      "clip_ratio/low_mean": 0.26611328125,
+      "clip_ratio/low_min": 0.125,
+      "clip_ratio/region_mean": 0.34814453125,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 605.0,
+      "completions/max_terminated_length": 605.0,
+      "completions/mean_length": 376.55615234375,
+      "completions/mean_terminated_length": 376.55615234375,
+      "completions/min_length": 190.0,
+      "completions/min_terminated_length": 190.0,
+      "entropy": 1.7707895785570145,
+      "epoch": 1.7619047619047619,
+      "frac_reward_zero_std": 0.0,
+      "grad_norm": 0.10629850625991821,
+      "learning_rate": 1e-05,
+      "loss": 0.0042,
+      "num_tokens": 93708054.0,
+      "reward": 3.8587303161621094,
+      "reward_std": 0.16579969227313995,
+      "rewards/ngram_repetition2/mean": 0.9476821422576904,
+      "rewards/ngram_repetition2/std": 0.015965020284056664,
+      "rewards/ngram_repetition3/mean": 0.9930981993675232,
+      "rewards/ngram_repetition3/std": 0.005606223829090595,
+      "rewards/symbolic_reward_accuracy/mean": 0.9296875,
+      "rewards/symbolic_reward_accuracy/std": 0.2557353675365448,
+      "rewards/symbolic_reward_partial_score/mean": 0.9702554941177368,
+      "rewards/symbolic_reward_partial_score/std": 0.1312272697687149,
+      "rewards/tag_count_reward/mean": 1.0,
+      "rewards/tag_count_reward/std": 0.0,
+      "rewards/thinking_answer_ratio_reward/mean": 0.9691884517669678,
+      "rewards/thinking_answer_ratio_reward/std": 0.013018821366131306,
+      "sampling/importance_sampling_ratio/max": 2.0,
+      "sampling/importance_sampling_ratio/mean": 1.4982185363769531,
+      "sampling/importance_sampling_ratio/min": 0.0010142651153728366,
+      "sampling/sampling_logp_difference/max": 6.893590927124023,
+      "sampling/sampling_logp_difference/mean": 0.6270837783813477,
+      "step": 148
+    },
+    {
+      "clip_ratio/high_max": 0.13671875,
+      "clip_ratio/high_mean": 0.05419921875,
+      "clip_ratio/low_mean": 0.30078125,
+      "clip_ratio/low_min": 0.15234375,
+      "clip_ratio/region_mean": 0.35498046875,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 586.0,
+      "completions/max_terminated_length": 586.0,
+      "completions/mean_length": 362.5859375,
+      "completions/mean_terminated_length": 362.5859375,
+      "completions/min_length": 203.0,
+      "completions/min_terminated_length": 203.0,
+      "entropy": 1.767895646393299,
+      "epoch": 1.8095238095238095,
+      "frac_reward_zero_std": 0.0,
+      "grad_norm": 0.09717681258916855,
+      "learning_rate": 1e-05,
+      "loss": 0.0033,
+      "num_tokens": 96295942.0,
+      "reward": 3.8618547916412354,
+      "reward_std": 0.061989087611436844,
+      "rewards/ngram_repetition2/mean": 0.9507569074630737,
+      "rewards/ngram_repetition2/std": 0.015048976056277752,
+      "rewards/ngram_repetition3/mean": 0.993833065032959,
+      "rewards/ngram_repetition3/std": 0.005530180875211954,
+      "rewards/symbolic_reward_accuracy/mean": 0.9267578125,
+      "rewards/symbolic_reward_accuracy/std": 0.26059725880622864,
+      "rewards/symbolic_reward_partial_score/mean": 0.9792073369026184,
+      "rewards/symbolic_reward_partial_score/std": 0.08610634505748749,
+      "rewards/tag_count_reward/mean": 1.0,
+      "rewards/tag_count_reward/std": 0.0,
+      "rewards/thinking_answer_ratio_reward/mean": 0.9685899019241333,
+      "rewards/thinking_answer_ratio_reward/std": 0.006216762121766806,
+      "sampling/importance_sampling_ratio/max": 2.0,
+      "sampling/importance_sampling_ratio/mean": 1.5025081634521484,
+      "sampling/importance_sampling_ratio/min": 5.1178951252950355e-05,
+      "sampling/sampling_logp_difference/max": 9.880182266235352,
+      "sampling/sampling_logp_difference/mean": 0.6323176622390747,
+      "step": 152
+    },
+    {
+      "clip_ratio/high_max": 0.16015625,
+      "clip_ratio/high_mean": 0.0771484375,
+      "clip_ratio/low_mean": 0.2802734375,
+      "clip_ratio/low_min": 0.16015625,
+      "clip_ratio/region_mean": 0.357421875,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 559.0,
+      "completions/max_terminated_length": 559.0,
+      "completions/mean_length": 366.6572265625,
+      "completions/mean_terminated_length": 366.6572265625,
+      "completions/min_length": 83.0,
+      "completions/min_terminated_length": 83.0,
+      "entropy": 1.791026197373867,
+      "epoch": 1.8571428571428572,
+      "frac_reward_zero_std": 0.0,
+      "grad_norm": 0.09605922549962997,
+      "learning_rate": 1e-05,
+      "loss": 0.0037,
+      "num_tokens": 98858760.0,
+      "reward": 3.8958053588867188,
+      "reward_std": 0.14521583914756775,
+      "rewards/ngram_repetition2/mean": 0.9516223669052124,
+      "rewards/ngram_repetition2/std": 0.014798388816416264,
+      "rewards/ngram_repetition3/mean": 0.9940202236175537,
+      "rewards/ngram_repetition3/std": 0.005294241476804018,
+      "rewards/symbolic_reward_accuracy/mean": 0.9443359375,
+      "rewards/symbolic_reward_accuracy/std": 0.22932782769203186,
+      "rewards/symbolic_reward_partial_score/mean": 0.9779866337776184,
+      "rewards/symbolic_reward_partial_score/std": 0.10208848118782043,
+      "rewards/tag_count_reward/mean": 1.0,
+      "rewards/tag_count_reward/std": 0.0,
+      "rewards/thinking_answer_ratio_reward/mean": 0.9690417051315308,
+      "rewards/thinking_answer_ratio_reward/std": 0.006724427919834852,
+      "sampling/importance_sampling_ratio/max": 2.0,
+      "sampling/importance_sampling_ratio/mean": 1.5027318000793457,
+      "sampling/importance_sampling_ratio/min": 0.001545182429254055,
+      "sampling/sampling_logp_difference/max": 6.472613334655762,
+      "sampling/sampling_logp_difference/mean": 0.6329264640808105,
+      "step": 156
+    },
+    {
+      "clip_ratio/high_max": 0.16015625,
+      "clip_ratio/high_mean": 0.06298828125,
+      "clip_ratio/low_mean": 0.29931640625,
+      "clip_ratio/low_min": 0.16796875,
+      "clip_ratio/region_mean": 0.3623046875,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 598.0,
+      "completions/max_terminated_length": 598.0,
+      "completions/mean_length": 392.0107421875,
+      "completions/mean_terminated_length": 392.0107421875,
+      "completions/min_length": 228.0,
+      "completions/min_terminated_length": 228.0,
+      "entropy": 1.7721411362290382,
+      "epoch": 1.9047619047619047,
+      "frac_reward_zero_std": 0.0,
+      "grad_norm": 0.10836534202098846,
+      "learning_rate": 1e-05,
+      "loss": 0.0051,
+      "num_tokens": 101483006.0,
+      "reward": 3.9561607837677,
+      "reward_std": 0.0488056018948555,
+      "rewards/ngram_repetition2/mean": 0.9511752128601074,
+      "rewards/ngram_repetition2/std": 0.014944043941795826,
+      "rewards/ngram_repetition3/mean": 0.9939863085746765,
+      "rewards/ngram_repetition3/std": 0.005151876714080572,
+      "rewards/symbolic_reward_accuracy/mean": 0.96728515625,
+      "rewards/symbolic_reward_accuracy/std": 0.17793269455432892,
+      "rewards/symbolic_reward_partial_score/mean": 0.992431640625,
+      "rewards/symbolic_reward_partial_score/std": 0.04925445467233658,
+      "rewards/tag_count_reward/mean": 1.0,
+      "rewards/tag_count_reward/std": 0.0,
+      "rewards/thinking_answer_ratio_reward/mean": 0.9707227945327759,
+      "rewards/thinking_answer_ratio_reward/std": 0.006023992318660021,
+      "sampling/importance_sampling_ratio/max": 2.0,
+      "sampling/importance_sampling_ratio/mean": 1.5052523612976074,
+      "sampling/importance_sampling_ratio/min": 0.0011821264633908868,
+      "sampling/sampling_logp_difference/max": 6.740440368652344,
+      "sampling/sampling_logp_difference/mean": 0.6356411576271057,
+      "step": 160
+    },
+    {
+      "clip_ratio/high_max": 0.0859375,
+      "clip_ratio/high_mean": 0.0302734375,
+      "clip_ratio/low_mean": 0.31298828125,
+      "clip_ratio/low_min": 0.171875,
+      "clip_ratio/region_mean": 0.34326171875,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 598.0,
+      "completions/max_terminated_length": 598.0,
+      "completions/mean_length": 383.38427734375,
+      "completions/mean_terminated_length": 383.38427734375,
+      "completions/min_length": 46.0,
+      "completions/min_terminated_length": 46.0,
+      "entropy": 1.8137650862336159,
+      "epoch": 1.9523809523809523,
+      "frac_reward_zero_std": 0.0,
+      "grad_norm": 0.21898648142814636,
+      "learning_rate": 1e-05,
+      "loss": 0.0041,
+      "num_tokens": 104089105.0,
+      "reward": 3.929980754852295,
+      "reward_std": 0.06001705303788185,
+      "rewards/ngram_repetition2/mean": 0.9539239406585693,
+      "rewards/ngram_repetition2/std": 0.014461600221693516,
+      "rewards/ngram_repetition3/mean": 0.9946612119674683,
+      "rewards/ngram_repetition3/std": 0.004792888183146715,
+      "rewards/symbolic_reward_accuracy/mean": 0.95654296875,
+      "rewards/symbolic_reward_accuracy/std": 0.20393338799476624,
+      "rewards/symbolic_reward_partial_score/mean": 0.9879556894302368,
+      "rewards/symbolic_reward_partial_score/std": 0.07069577276706696,
+      "rewards/tag_count_reward/mean": 0.999755859375,
+      "rewards/tag_count_reward/std": 0.011048543266952038,
+      "rewards/thinking_answer_ratio_reward/mean": 0.9697635173797607,
+      "rewards/thinking_answer_ratio_reward/std": 0.022018132731318474,
+      "sampling/importance_sampling_ratio/max": 2.0,
+      "sampling/importance_sampling_ratio/mean": 1.5047733783721924,
+      "sampling/importance_sampling_ratio/min": 0.0021326851565390825,
+      "sampling/sampling_logp_difference/max": 6.150373458862305,
+      "sampling/sampling_logp_difference/mean": 0.6389479637145996,
+      "step": 164
+    },
+    {
+      "clip_ratio/high_max": 0.03125,
+      "clip_ratio/high_mean": 0.00732421875,
+      "clip_ratio/low_mean": 0.3271484375,
+      "clip_ratio/low_min": 0.19140625,
+      "clip_ratio/region_mean": 0.33447265625,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 612.0,
+      "completions/max_terminated_length": 612.0,
+      "completions/mean_length": 395.89501953125,
+      "completions/mean_terminated_length": 395.89501953125,
+      "completions/min_length": 224.0,
+      "completions/min_terminated_length": 224.0,
+      "entropy": 1.8114138692617416,
+      "epoch": 2.0,
+      "frac_reward_zero_std": 0.0,
+      "grad_norm": 0.1692064255475998,
+      "learning_rate": 1e-05,
+      "loss": 0.0046,
+      "num_tokens": 106675066.0,
+      "reward": 3.9362592697143555,
+      "reward_std": 0.06057881936430931,
+      "rewards/ngram_repetition2/mean": 0.9569206833839417,
+      "rewards/ngram_repetition2/std": 0.013300254940986633,
+      "rewards/ngram_repetition3/mean": 0.9952021241188049,
+      "rewards/ngram_repetition3/std": 0.004495474975556135,
+      "rewards/symbolic_reward_accuracy/mean": 0.95947265625,
+      "rewards/symbolic_reward_accuracy/std": 0.1972406506538391,
+      "rewards/symbolic_reward_partial_score/mean": 0.9880777597427368,
+      "rewards/symbolic_reward_partial_score/std": 0.07768868654966354,
+      "rewards/tag_count_reward/mean": 1.0,
+      "rewards/tag_count_reward/std": 0.0,
+      "rewards/thinking_answer_ratio_reward/mean": 0.9714963436126709,
+      "rewards/thinking_answer_ratio_reward/std": 0.0055397688411176205,
+      "sampling/importance_sampling_ratio/max": 2.0,
+      "sampling/importance_sampling_ratio/mean": 1.506116509437561,
+      "sampling/importance_sampling_ratio/min": 0.001569847809150815,
+      "sampling/sampling_logp_difference/max": 6.4567766189575195,
+      "sampling/sampling_logp_difference/mean": 0.6404174566268921,
+      "step": 168
+    },
+    {
+      "epoch": 2.0,
+      "eval_clip_ratio/high_max": 0.0,
+      "eval_clip_ratio/high_mean": 0.0,
+      "eval_clip_ratio/low_mean": 0.0,
+      "eval_clip_ratio/low_min": 0.0,
+      "eval_clip_ratio/region_mean": 0.0,
+      "eval_completions/clipped_ratio": 0.0,
+      "eval_completions/max_length": 678.8947368421053,
+      "eval_completions/max_terminated_length": 678.8947368421053,
+      "eval_completions/mean_length": 428.8622532894737,
+      "eval_completions/mean_terminated_length": 428.8622532894737,
+      "eval_completions/min_length": 293.6842105263158,
+      "eval_completions/min_terminated_length": 293.6842105263158,
+      "eval_entropy": 1.8865187983763845,
+      "eval_frac_reward_zero_std": 0.0,
+      "eval_loss": 0.0013121640076860785,
+      "eval_num_tokens": 106675066.0,
+      "eval_reward": 3.922859505603188,
+      "eval_reward_std": 0.12427335997697235,
+      "eval_rewards/ngram_repetition2/mean": 0.9468266618879217,
+      "eval_rewards/ngram_repetition2/std": 0.017048097411660773,
+      "eval_rewards/ngram_repetition3/mean": 0.9923422650287026,
+      "eval_rewards/ngram_repetition3/std": 0.008230763512026323,
+      "eval_rewards/symbolic_reward_accuracy/mean": 0.9547697368421053,
+      "eval_rewards/symbolic_reward_accuracy/std": 0.14237226741878609,
+      "eval_rewards/symbolic_reward_partial_score/mean": 0.9857113455471239,
+      "eval_rewards/symbolic_reward_partial_score/std": 0.059331425465643406,
+      "eval_rewards/tag_count_reward/mean": 0.9985608552631579,
+      "eval_rewards/tag_count_reward/std": 0.01490654286585356,
+      "eval_rewards/thinking_answer_ratio_reward/mean": 0.9656107896252683,
+      "eval_rewards/thinking_answer_ratio_reward/std": 0.042114860623290666,
+      "eval_runtime": 491.5378,
+      "eval_samples_per_second": 0.305,
+      "eval_sampling/importance_sampling_ratio/max": 2.0,
+      "eval_sampling/importance_sampling_ratio/mean": 1.520386018251118,
+      "eval_sampling/importance_sampling_ratio/min": 0.003561224910969797,
+      "eval_sampling/sampling_logp_difference/max": 5.717697494908383,
+      "eval_sampling/sampling_logp_difference/mean": 0.6679018484918695,
+      "eval_steps_per_second": 0.004,
+      "step": 168
+    },
+    {
+      "epoch": 2.0,
+      "step": 168,
+      "total_flos": 0.0,
+      "train_loss": 0.003984759748293852,
+      "train_runtime": 17571.6441,
+      "train_samples_per_second": 0.318,
+      "train_steps_per_second": 0.01
+    }
+  ],
+  "logging_steps": 4,
+  "max_steps": 168,
+  "num_input_tokens_seen": 106675066,
+  "num_train_epochs": 2,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "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:d7d17f6dd86688e6438a2b3ae905cca6fd392558ebb39cd510b6c0de36210110
+size 11665