Installation

Requirements

  • Python 3.8 or higher

  • pip package manager

  • Git (for development installation)

Basic Installation

Install MedExplain-Evals using pip:

git clone https://github.com/heilcheng/MedExplain-Evals.git
cd MedExplain-Evals
pip install .

This command uses the setup.py file to install the package and its dependencies correctly. The setup.py file automatically handles all core dependencies defined in the install_requires section.

Development Installation

For development, install in editable mode with additional dependencies:

git clone https://github.com/heilcheng/MedExplain-Evals.git
cd MedExplain-Evals
pip install -e .[dev]

This installs additional tools for development:

  • pytest for testing

  • black for code formatting

  • flake8 for linting

  • mypy for type checking

Optional Dependencies

Some features require additional packages:

Machine Learning Components:

pip install torch transformers sentence-transformers

Natural Language Processing:

pip install spacy scispacy
python -m spacy download en_core_web_sm

Documentation Building:

pip install sphinx sphinx-rtd-theme myst-parser

Configuration

  1. Create logs directory:

mkdir logs

  1. Set up API keys (required for LLM-as-a-judge evaluation):

export OPENAI_API_KEY="your-openai-api-key"
export ANTHROPIC_API_KEY="your-anthropic-api-key"

  1. Verify installation:

python -c "from src.benchmark import MedExplain; print('Installation successful!')"

Docker Installation (Optional)

For a containerized environment:

# Build the Docker image
docker build -t medexplain-evals .

# Run the container
docker run -it medexplain-evals

Troubleshooting

Common Issues:

  1. Import errors: Ensure all dependencies are installed

  2. API key errors: Set environment variables correctly

  3. Permission errors: Use virtual environment or user installation

Getting Help:

  • Check the GitHub Issues page

  • Review the documentation

  • Contact the development team