Ecosystem & Stack: megatron-core
Reinforcement Learning Library for LLMs
Ray
3.1k
alibaba/ROLL
An efficient and user-friendly scaling library designed to optimize Reinforcement Learning with Large Language Models, enhancing performance in complex AI tasks.