This repository provides a Chinese dialogue system built using the Rasa framework. It is intended for developers and researchers working with conversational AI in Chinese, offering a practical implementation and guidance on overcoming common challenges.

How It Works

The system leverages Rasa Core and Rasa NLU for building conversational agents. The project focuses on a specific version of Rasa (1.10.18) and includes a custom pipeline optimized for Chinese language understanding, with mentions of BERT component support for improved performance.

Quick Start & Requirements

• Install packages: pip install -r requirements.txt
• Python version: >= 3.6
• Train model: make train
• Run model: make run
• Test via shell: make shell
• Test via HTTP server: http://localhost:5005/webhooks/rest/webhook (POST request with {"sender": "0001", "message": "你好"})

Highlighted Details

• Based on Rasa version 1.10.18 with a Chinese-optimized pipeline.
• Includes a series of articles detailing the development and troubleshooting process ("rasa对话系统踩坑记").
• Mentions potential integration with BERT components for enhanced Chinese NLP.

Maintenance & Community

• The project mentions a WeChat group for Rasa Chinese casual chat. Contact coffee199029 via WeChat to join.
• The README indicates that the code is from an earlier version and that the master branch will be updated with newer Rasa versions.

Licensing & Compatibility

• The repository does not explicitly state a license.

Limitations & Caveats

The code is based on an older Rasa version (1.10.18), and the README notes significant changes in Rasa 2.0 and beyond, suggesting that the current master branch might not reflect the latest Rasa features or best practices.