InternLM/HuixiangDou
HuixiangDou: Overcoming Group Chat Scenarios with LLM-based Technical Assistance
Implements a three-stage pipeline (preprocess, rejection, response) using hybrid retrieval combining dense embeddings for documents, sparse methods for code, and knowledge graphs to intelligently filter irrelevant group chat messages and generate contextual responses. Integrates with multiple LLM providers (DeepSeek, InternLM, GLM) and chat platforms (WeChat, Lark, Read the Docs) while supporting multimodal inputs including document OCR, image-text retrieval, and coreference resolution for context understanding.
2,481 stars and 12 monthly downloads. Available on PyPI.
Stars
2,481
Forks
182
Language
Python
License
BSD-3-Clause
Category
Last pushed
Nov 24, 2025
Monthly downloads
12
Commits (30d)
0
Dependencies
37
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/rag/InternLM/HuixiangDou"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Related tools
neuml/txtchat
⚡ Local chat assistants with AI superpowers
matterai/DataKitsune
Introducing DataKitsune, an open-source Telegram bot designed to enhance the way you manage and...
minhbtrc/chatbot-template
A modular backend framework for building AI chat applications powered by large language models (LLMs)
Flagro/TelegramRPBot
Add a roleplay AI bot to your group chats
QIN2DIM/telegram-dify-bot
一个为 Dify Workflow “量身定做”的 Telegram 群聊机器人。