CoolFace
Datasetpublic

OpenMOSS-Team/GameQA-text

Here we provide a pure-text version of GameQA, encompassing some appropriate games. (See https://github.com/tongjingqi/Code2Logic/issues/2) 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… See the full description on the dataset page: https://huggingface.co/datasets/OpenMOSS-Team/GameQA-text.

sourceHugging Facemitupdated 1y agoView on Hugging Face
3likes51downloads
5 commits on main
25be2e61y ago

Update README.md

lkdhy
f97dd441y ago

Update README.md

lkdhy
b40c25c1y ago

Upload 2 files

lkdhy
6b3e1491y ago

Create README.md

lkdhy
7bfad441y ago

initial commit

lkdhy