This repository provides efficient fine-tuning methods for ChatGLM-6B, a Chinese large language model developed by Tsinghua University. It implements PEFT (Parameter-Efficient Fine-Tuning) techniques including LoRA and QLoRA, enabling users to adapt the model with limited computational resources. The project includes a Web UI for training, evaluation and inference, and aligns its API with OpenAI’s format for easy integration into ChatGPT-based applications.