This repository contains the data and resources to build an AI assistant on public service using the Retrieval Augmented Generation (RAG) approach. The RAG method combines information retrieval with powerful generative models to provide accurate, context-aware answers..
- Public service website https://service-public.bj/
- Build a RAG system using Langchain: https://learn.deeplearning.ai/courses/langchain-chat-with-your-data
- Retrieval with Langchain https://python.langchain.com/v0.1/docs/modules/data_connection/
- Retrievers evaluation metrics https://hal.science/hal-03467001/document
- Embeddings models leaderboard https://huggingface.co/spaces/mteb/leaderboard
- LLM Chatbot leaderboard https://huggingface.co/spaces/lmsys/chatbot-arena-leaderboard
- LLM finetuning https://www.leewayhertz.com/parameter-efficient-fine-tuning/
- LLM evaluation metrics https://fabianofalcao.medium.com/metrics-for-evaluating-summarization-of-texts-performed-by-transformers-how-to-evaluate-the-b3ce68a309c3