WikiChat and multi-mediawiki-rag
About WikiChat
stanford-oval/WikiChat
WikiChat is an improved RAG. It stops the hallucination of large language models by retrieving data from a corpus.
This project helps anyone who uses large language model chatbots (like ChatGPT) to get accurate, factual answers to their questions, especially on recent events or niche topics. It takes your question and an existing language model, then searches Wikipedia or other documents to provide a fact-checked response, complete with citations. This is ideal for researchers, journalists, or anyone who relies on accurate information from AI tools.
About multi-mediawiki-rag
tylertitsworth/multi-mediawiki-rag
A simple RAG chatbot that can retrieve from a mediawiki data dump
Leverages Langchain to orchestrate multi-source MediaWiki XML dumps through Chroma vector embeddings and Ollama-hosted LLMs, with configurable Huggingface sentence transformers for semantic search. Includes a Chainlit web interface and Discord bot integration for chatbot deployment. Supports custom model quantization and tuning via Modelfiles for optimized document retrieval.
Scores updated daily from GitHub, PyPI, and npm data. How scores work