Skip to content

ducanhho2296/LangChain-RAG

Repository files navigation

LangChain-RAG Experiments and Embedding Model Testing

📌 Project Overview

In this project, we:

Experiment with different embedding models via API calls using LangChain.

Create and manage LLM agents for retrieval tasks.

Test vector databases for efficient storage and querying.

Build, refine, and benchmark a custom RAG pipeline.

This repository serves as a playground for testing and evaluating the performance of RAG systems under various conditions and configurations.

🚀 Key Features

Embedding Model Testing: Evaluate performance of different embedding models using APIs (e.g., OpenAI, Hugging Face, Cohere).

Agent Creation: Build custom agents with LangChain for intelligent querying and response generation.

Vector Database Integration: Experiment with ChromaDB, FAISS, Pinecone, and Milvus.

Dynamic RAG Pipelines: Implement RAG pipelines with different configurations to handle unstructured data retrieval and response generation.

Benchmarking and Analysis: Compare accuracy, retrieval speed, and resource usage across various setups.

About

This repository contains a series of experiments exploring Retrieval-Augmented Generation (RAG) using various embedding models and LLM agents with LangChain. The goal is to evaluate and compare different embeddings, databases, and agent behaviors to optimize RAG pipelines for diverse applications.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors