ChatLearn

Introduction

  • ChatLearn: A flexible and efficient reinforcement learning framework for large language models(LLMs)

Installation

  • Environment and Code Preparation

Tutorial

  • End-to-End GRPO Training Tutorial with FSDP
  • End-to-End GRPO Training Tutorial with Mcore
  • Multi-Node Distributed Training
  • Resume Training and Fault Tolerance
  • Performance Tuning Guide
  • Profile

Customized Task

  • Dataset Preparation
  • Customize Reward Function

Configuration

  • Config Explanation

FAQ

  • FAQ
  • Common Errors
ChatLearn
  • Search

© Copyright 2024, Alibaba Cloud.

Built with Sphinx using a theme provided by Read the Docs.