Doggo RAG Chatbot 🐶

A simple streamlit chatbot to showcase functioning of Retrieval Augmented Generation (RAG) and how to integrate it with a public LLM to create custom chatbot with a specific knowledge.

I write about this code in my LinkedIn article https://www.linkedin.com/pulse/creating-custom-chatbot-based-rag-llm-ivan-magdolen-splfe

• 1st article
  • branch "part_1"
• 2nd article
  • branch "part_2"

Features

• You can chat with custom chatbot who knows data about Zaira (my dog)

Requirements

• Python 3.8 or higher
• see requirements.txt

Usage

Clone and run virtual environment

$ git clone https://github.com/doggo-rag-chatbot.git
$ cd doggo-rag-chatbot
$ python -m venv .
$ ./Scripts/activate
$ pip install -r requirements.txt

Set you API key in environment variables

OPEN_AI_KEY='mykey'

Create embeddings database

$ python embeddings_init.py

Run the chatbot

$ streamlit run streamlit-app.py

And go to

http://localhost:8501