XU-YIJIE/grpo-flat

XU-YIJIE/grpo-flat

GitHubGitHub
GitHubgithub.com
Releases0
Stars71
Train your grpo with zero dataset and low resources, 8bit/4bit/lora/qlora supported, multi-gpu supported ...
Log in to subscribe

Collections containing this project

Showing collections based on your access.

This project is not in any collections you can view.