This repository provides a Chinese instruction-following dataset designed for fine-tuning large language models, inspired by the Stanford Alpaca dataset. It aims to facilitate the development of Chinese-capable instruction-following models for researchers and developers working with Chinese NLP.

How It Works

The dataset consists of instruction-output pairs, with optional input fields, formatted in JSON. The generation methodology is based on the self-instruct approach, leveraging machine translation and data cleaning techniques to create diverse and high-quality instruction-following examples in Chinese.

Highlighted Details

• Dataset format is identical to the original Alpaca data (JSON).
• Includes examples for various tasks like information retrieval, list generation, and hidden message extraction.
• Generation methods include machine translation and self-instruct.

Maintenance & Community

Information regarding contributors, community channels, or roadmap is not available in the provided README.

Licensing & Compatibility

The license is not specified in the README. Compatibility for commercial use or closed-source linking is undetermined.

Limitations & Caveats

The README indicates that the generation method for each data entry and the specific keyword/rule cleaning processes are pending further documentation.