This project is a Retrieval-Augmented Generation (RAG) application that generates responses to queries using Langchain and foundation models in JavaScript. The application leverages vector stores and embeddings to enhance the quality of the responses.
RAG (Retrieval-Augmented Generation) is a technique that combines retrieval-based and generation-based approaches to provide more accurate and contextually relevant responses to user queries. This application uses Langchain for managing the language models and embeddings, and foundation models to generate high-quality responses.
- Document Loading: Load documents from a directory to create a knowledge base.
- Text Splitting: Split documents into manageable chunks for better processing.
- Embedding Creation: Generate embeddings for document chunks using foundation models.
- Query Handling: Handle user queries and provide accurate responses using retrieval-augmented generation.
- Chroma Vector Store: Store and manage document embeddings using Chroma.
Before running the application, ensure you have the following installed:
- Node.js (>= 14.x)
- npm (Node Package Manager)
- chromadb
-
Clone the Repository
git clone https://github.com/Gooner12/rag-js.git cd rag-application -
Install Dependencies
npm install
-
Run ChromaDB Server
You can run the chromadb locally using Docker or using chromadb cli.
- Using Docker:
git clone [email protected]:chroma-core/chroma.git cd chroma docker-compose up -d --build
-
Using ChromaDB CLI:
To use chromadb cli, you need to install the chromadb using:
pip install chromadb chroma run --host localhost --port 8000 --path ./my_chroma_data
For more information on running the chromadb server, check out this link.
-
Configure Environment Variables
Lastly, create a .env file with environment variables containing your secret key and access key.