hiyouga/EasyR1
EasyR1: An Efficient, Scalable, Multi-Modality RL Training Framework based on veRL
This project helps AI researchers and machine learning engineers efficiently train powerful, large-scale language and vision-language models using reinforcement learning. It takes large datasets of text or text-image pairs and various models like Llama3 or Qwen2-VL, then applies advanced training algorithms to produce highly optimized models ready for deployment. This is for professionals working on developing and fine-tuning cutting-edge AI.
4,721 stars. Actively maintained with 2 commits in the last 30 days.
Use this if you need to fine-tune very large language models or vision-language models with reinforcement learning on large datasets, and require a highly efficient and scalable training framework.
Not ideal if you are looking for a simple tool for basic model training or if you do not have significant computational resources (multiple high-end GPUs) available.
Stars
4,721
Forks
362
Language
Python
License
Apache-2.0
Category
Last pushed
Mar 10, 2026
Commits (30d)
2
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/llm-tools/hiyouga/EasyR1"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Related tools
hud-evals/hud-python
OSS RL environment + evals toolkit
OpenRL-Lab/openrl
Unified Reinforcement Learning Framework
sail-sg/oat
🌾 OAT: A research-friendly framework for LLM online alignment, including reinforcement learning,...
opendilab/awesome-RLHF
A curated list of reinforcement learning with human feedback resources (continually updated)
NVlabs/GDPO
Official implementation of GDPO: Group reward-Decoupled Normalization Policy Optimization for...