EasyR1: An Efficient, Scalable, Multi-Modality RL Training Framework based on veRL

A powerful tool for creating fine-tuning datasets for LLM

Unify Efficient Fine-Tuning of 100+ LLMs

LLaMA Factory Document

Official code repo for our work "Native Visual Understanding: Resolving Resolution Dilemmas in Vision-Language Models"