Set up environment

Pre-requisites :

1. Create virtual environment (I'm using venv)
2. Run pip install -r requirements.txt
3. Download and install Ollama

Ollama After installing, open Ollama and run command ollama run llama3.2 on your terminal to download the pretrained LLM model. In this project, I'm using llama3.2-3B because it's lightweight and has good performance.

Deployment

I'm using Streamlit and Docker to deploy the chatbot.

Step 1 : Build Docker Image

  docker build -t ecommerce-chatbot-app .

Step 2 : Run Docker Container

  docker run -d --restart always --gpus all --name ecommerce-chatbot-app -p 8501:8501 ecommerce-chatbot-app

To stop and delete the container run

  docker stop ecommerce-chatbot-app && docker rm ecommerce-chatbot-app

Chatbot Overview

Example of Questions

Retrieved data from API