This repository provides a training pipeline for fine-tuning the ChatGLM-6B large language model using LoRA (Low-Rank Adaptation), a parameter-efficient fine-tuning method. It processes datasets such as Alpaca, tokenizes them into a structured format, and runs distributed training with configurable batch sizes, gradient accumulation, and rank settings. The project also offers a Google Colab notebook for users to run fine-tuning without local GPU infrastructure.