# chatglm-finetune

**Repository Path**: aizpy/chatglm-finetune

## Basic Information

- **Project Name**: chatglm-finetune
- **Description**: 对 ChatGLM-6B 做 Fine-tuning
- **Primary Language**: Python
- **License**: BSD-3-Clause
- **Default Branch**: main
- **Homepage**: None
- **GVP Project**: No

## Statistics

- **Stars**: 4
- **Forks**: 3
- **Created**: 2023-04-03
- **Last Updated**: 2023-06-06

## Categories & Tags

**Categories**: Uncategorized

**Tags**: None

## README

# chatglm-finetune
Fine-tuning based on [THUDM/chatglm-6b](https://huggingface.co/THUDM/chatglm-6b). 
The first Fine-tuning method is [LoRA](https://arxiv.org/abs/2106.09685), and more methods will be made available if possible.

## LoRA
Please checkout the [Tutorial](https://aizpy.com/2023/03/30/chatglm-6b-lora/) if you can read Chinese.

## Dataset
An AI-generated QA dataset about AI explorer ([AI探险家](https://aizpy.com/about/)) is provided. The trained model saved in the output folder is also based on this dataset.

## Infer test
![image](https://user-images.githubusercontent.com/127382813/229401551-3819c3f5-0795-4927-b88e-8fc554cf5ed2.png)