CSC-chatbot

A chatbot that tells you about the Centro di Sonologia Computazionale (CSC) of the University of Padova and its people.

Installation

Requirements:

• Python 3.10 or higher
• Poetry for managing dependencies or pip
• Makefile to run commands (optional)

To install the project from source you can download the repository:

git clone https://github.com/matteospanio/csc-chatbot.git
cd csc-chatbot

Then you can install the dependencies with Poetry:

poetry install

Without Poetry (not recommended)

If you don't have Poetry installed, you can install the project with pip by running the following command in the project root directory:

pip install -e .

Otherwise you can install just the dependencies with pip:

pip install -r requirements.txt

Configuration

Configuration is done through the config.yml to set the models parameters. In addition you need to set some environment variables:

• OPENAI_API_KEY to use the OpenAI API
• PDF_PATH, the path to the PDF files to load
• DATA_PATH, the path to the web links to load
• CHROMA_PATH the folder name where to store the chroma database files
• CHAT_MEMORY the file name where to store the chat messages history

Usage

with poetry

If you have Poetry installed, you can run the chatbot with the following command:

poetry run chatbot --help

or enter the virtual environment with:

poetry shell

and then run the chatbot with:

chatbot --help

without poetry

If you installed the project with pip, you can run the chatbot with:

chatbot --help

If you installed only the dependencies with pip, you can run the chatbot with:

python src/chatbot/ --help

Features/Commands

The CLI chatbot has the following commands:

• --help: show the help message
• --version: show the version of the chatbot
• chat: start the chatbot
• ingest: setup the chatbot

chat starts the chatbot and you can ask questions about the CSC and its people.

setup is used to setup the chatbot memory. It will read the data from the data directory and store it in the chatbot memory. It accepts two flags:

• --with-pdf: Load the PDF files in the data/pdf directory, extract the text and store it in the chatbot memory
• --with-web: Load the web pages in the data/csc.yml file, extract the text and store it in the chatbot memory

License

This project is licensed under the MIT License - see the LICENSE file for details.