---
base_model:
- Qwen/Qwen2.5-VL-7B-Instruct
datasets:
- Code2Logic/GameQA-140K
- Code2Logic/GameQA-5K
license: apache-2.0
pipeline_tag: image-text-to-text
library_name: transformers
---
***This model (GameQA-Qwen2.5-VL-7B) results from training Qwen2.5-VL-7B with GRPO solely on our [GameQA-5K](https://huggingface.co/datasets/Code2Logic/GameQA-5K) (sampled from the full [GameQA-140K](https://huggingface.co/datasets/Gabriel166/GameQA-140K) dataset).***
# Evaluation Results on General Vision BenchMarks
***(The inference and evaluation configurations were unified across both the original open-source models and our trained models.)***
It's also found that getting trained on 5k samples from our GameQA dataset can lead to better results than on [multimodal-open-r1-8k-verified](https://huggingface.co/datasets/lmms-lab/multimodal-open-r1-8k-verified).
# Code2Logic: Game-Code-Driven Data Synthesis for Enhancing VLMs General Reasoning
This is the first work, to the best of our knowledge, that leverages ***game code*** to synthesize multimodal reasoning data for ***training*** VLMs. Furthermore, when trained with a GRPO strategy solely on **GameQA** (synthesized via our proposed **Code2Logic** approach), multiple cutting-edge open-source models exhibit significantly enhanced out-of-domain generalization.
[[📖 Paper](https://arxiv.org/abs/2505.13886)] [[🤗 GameQA-140K Dataset](https://huggingface.co/datasets/Gabriel166/GameQA-140K)] [[🤗 GameQA-5K Dataset](https://huggingface.co/datasets/Code2Logic/GameQA-5K)] [[🤗 GameQA-InternVL3-8B](https://huggingface.co/Code2Logic/GameQA-InternVL3-8B) ] [[🤗 GameQA-Qwen2.5-VL-7B](https://huggingface.co/Code2Logic/GameQA-Qwen2.5-VL-7B)] [[🤗 GameQA-LLaVA-OV-7B](https://huggingface.co/Code2Logic/GameQA-llava-onevision-qwen2-7b-ov-hf) ]
Code: https://github.com/tongjingqi/Code2Logic
## News
* We've open-sourced the ***three*** models trained with GRPO on GameQA on [Huggingface](https://huggingface.co/Code2Logic).
