ai-parrot

AI Parrot: Python package for creating Chatbots

This is an open-source Python package for creating Chatbots based on Langchain and Navigator. This README provides instructions for installation, development, testing, and releasing Parrot.

Installation

Creating a virtual environment:

This is recommended for development and isolation from system-wide libraries. Run the following command in your terminal:

Debian-based systems installation:

sudo apt install gcc python3.11-venv python3.11-full python3.11-dev libmemcached-dev zlib1g-dev build-essential libffi-dev unixodbc unixodbc-dev libsqliteodbc libev4 libev-dev

For Qdrant installation:

docker pull qdrant/qdrant
docker run -d -p 6333:6333 -p 6334:6334 --name qdrant -v $(pwd)/qdrant_storage:/qdrant/storage:z qdrant/qdrant

For VertexAI, creates a folder on "env" called "google" and copy the JSON credentials file into it.

make venv

This will create a virtual environment named .venv. To activate it, run:

source .venv/bin/activate  # Linux/macOS

Once activated, install Parrot within the virtual environment:

make install

The output will remind you to activate the virtual environment before development.

Optional (for developers):

pip install -e .

Start http server

python run.py

Development Setup

This section explains how to set up your development environment:

1. Install development requirements:

   make setup
   

   This installs development dependencies like linters and test runners mentioned in the docs/requirements-dev.txt file.

2. Install Parrot in editable mode:

   This allows you to make changes to the code and test them without reinstalling:

   make dev
   

   This uses flit to install Parrot in editable mode.

Usage (Replace with actual usage instructions)

Once you have set up your development environment, you can start using Parrot.

Test with Code ChatBOT

• Set environment variables for: [google] GOOGLE_API_KEY=apikey GOOGLE_CREDENTIALS_FILE=.../credentials.json VERTEX_PROJECT_ID=vertex-project VERTEX_REGION=region

• Run the chatbot:

  python examples/test_agent.py
  

Testing

To run the test suite:

make test

This will run tests using coverage to report on code coverage.

Code Formatting

To format the code with black:

make format

Linting

To lint the code for style and potential errors:

make lint

This uses pylint and black to check for issues.

Releasing a New Version

This section outlines the steps for releasing a new version of Parrot:

1. Ensure everything is clean and tested:

   make release
   

   This runs lint, test, and clean tasks before proceeding.

2. Publish the package:

   make release
   

   This uses flit to publish the package to a repository like PyPI. You'll need to have publishing credentials configured for flit.

Cleaning Up

To remove the virtual environment:

make distclean

Contributing

We welcome contributions to Parrot! Please refer to the CONTRIBUTING.md file for guidelines on how to contribute.