This chatbot application allows users to upload PDF documents, process their content, and interact with an AI-powered chatbot that answers questions based on the uploaded documents. The chatbot utilizes Phi-2, a small but powerful NLP model, and employs semantic search using embeddings and FAISS.
- Upload PDFs – Users can upload single or multiple PDF documents.
- Automatic Text Extraction – Extracts text from uploaded PDFs.
- Vector Store for Semantic Search – Uses embeddings to store and retrieve relevant document information.
- AI-Powered Chatbot – Answers questions based on the extracted knowledge using Phi-2.
- FastAPI-based API – A lightweight backend for seamless chatbot interactions.
- Docker Support – Easily deploy the application in a Docker container.
git clone https://github.com/aliai22/ChatBot.git
cd chatbot
python -m venv venv
source venv/bin/activate # On macOS/Linux
venv\Scripts\activate # On Windows
pip install -r requirements.txt
uvicorn app.main:app --reload
This chatbot application is powered by FastAPI, which provides an interactive API documentation. You can access it at:
http://127.0.0.1:8000/docs
- Build the Docker Image:
docker build -t chatbot-app .
- Run the Container:
docker run -p 8000:8000 chatbot-app
- Access the Chatbot:
http://127.0.0.1:8000/