rag qna

RAG QnA ?

In this project we use a Vector Database (Chroma DB) backed LLM Inference using LangChain to create a QnA system.

While running in colab use real OpenAI credentials for LLM inference.

Running it

git clone https://github.com/ketanip/VectorDB-LangChain-LLM.git

Open src/code.ipynb and execute cells one by one.

Tools used in this project:

1. ? LangChain
2. Vector Database (Chroma DB)
3. LLM (Zephyr 7B β - Mistral Architecture) powered by LM Studio.

Features:

1. Loads data from the raw_data folder .
2. Generates vector embeddings for the data and stores it in Chroma DB.
3. Searches relevant data based on the query ?.
4. Combines that data with the input question in a custom prompt template using LangChain.
5. Infers it using LLM and displays the results ?.