HuanjinYao/R1-ShareVL-7B

Mulberry

R1-ShareVL-7b is a reasoning MLLM trained by ShareGRPO.

Paper: https://arxiv.org/abs/2505.16673

Code: https://github.com/HJYao00/R1-ShareVL

More Details

Base Model: https://huggingface.co/Qwen/Qwen2.5-VL-7B-Instruct

Training Framework: EasyR1

Hardware: 8x NVIDIA H100