Supporting Multi-modality

[kaka-Cao]

Feature request

支持多模态PRM模型的训练，以及使用多模态PRM与多模态大模型联合进行强化学习训练

Motivation

目前没有github 项目支持多模态大模型的 类O1 强化学习训练

Your contribution

感谢您在LLM方向上的贡献，基于您较为完善的代码，实现多模态的 类O1 训练变得可实现，目前我在尝试在qwen2-vl上训练一个PRM模型

[YanSong97]

Supports for inference with Qwen2-VL will be available soon—stay tuned for updates!

Regarding multi-modal PRM training, we have added it to our to-do list and plan to investigate it in the near future. We also welcome and encourage any contributions in this area.