Skip to content

Latest commit

 

History

History
150 lines (117 loc) · 4.07 KB

BUILD.md

File metadata and controls

150 lines (117 loc) · 4.07 KB

Build Instructions

Supported Python Versions

This project supports Python 3.9, 3.10, and 3.11. Ensure you have one of these versions installed on your system before proceeding.

Setting Up the Environment

1. Create a Virtual Environment

Using a virtual environment is recommended to isolate project dependencies and avoid conflicts with other Python projects.

On Linux/macOS:

python -m venv .venv
source .venv/bin/activate

On Windows:

python -m venv .venv
.venv\Scripts\activate

2. Update pip

Before installing the project dependencies, ensure that pip is up to date:

pip install --upgrade pip

3. Install Dependencies

Basic Installation:

Install the required dependencies from the requirements.txt file:

pip install -r requirements.txt

GPU Support:

  • macOS (darwin):

    pip install -r requirements-gpu.txt

  • Windows with CUDA Support: If you are using GPU acceleration on Windows, you'll need to specify the appropriate CUDA version. For example, for CUDA 12.4:

    pip install --extra-index-url https://download.pytorch.org/whl/cu124 -r requirements-gpu.txt

    Adjust the CUDA version (cu124) in the URL as necessary for your setup.

Using Astral UV for Advanced Dependency Management

The project uses Astral UV for advanced dependency management, including custom indices and conditional dependencies.

Install Astral UV

Astral UV can be installed using pip:

pip install uv

Install Dependencies with Astral UV

Use the uv CLI to install dependencies as per the configurations in pyproject.toml:

uv install

This will resolve dependencies, including those with custom indices or conditional markers.

Using the pyproject.toml

This project uses a pyproject.toml file for dependency and build configuration. To interact with this framework, you can use tools such as pip, poetry, or Astral UV:

Install Dependencies with pip

If not already done, dependencies listed in pyproject.toml can be installed with:

pip install .

Using Poetry (Optional)

If you prefer using poetry for dependency management and builds:

  1. Install Poetry:

    pip install poetry

  2. Install dependencies:

    poetry install

  3. Activate the Poetry shell (optional):

    poetry shell

Additional Notes

  • The project uses the astral-uv library for dependency management. Ensure it is properly installed as part of the dependencies.
  • For advanced builds or custom configurations, refer to the comments in pyproject.toml and the respective tool documentation.

Verifying the Installation

To verify that everything is set up correctly:

  1. Activate the virtual environment:

    source .venv/bin/activate  # or .venv\Scripts\activate on Windows

  2. Run the project's test suite or a basic command to confirm:

    python -m unittest discover

    or

    python run.py

You should now be ready to develop and run the project

Upgrading dependencies

uv lock --upgrade

Regenerating the requirements.txt files

uv export --no-hashes > .\requirements.txt
uv export --no-hashes --extra cuda_gpu > .\requirements-gpu.txt

On windows, make sure the output is in utf-8, not utf-16 is it would be by default:

uv export --no-hashes | Out-File -FilePath .\requirements.txt -Encoding utf8
uv export --no-hashes --extra cuda_gpu | Out-File -FilePath .\requirements-gpu.txt -Encoding utf8

You can add --no-dev to uv export to ignore the dev dependencies.

Before submitting a change request

# pip install ruff
ruff check verbatim tests # try `ruff check --fix verbatim tests`if you see some failures
# pip install pylint
pylint verbatim $(git ls-files 'tests/*.py')
# pip install flake8
flake8 --count --select=E9,F63,F7,F82 --show-source --statistics verbatim tests
flake8 --count --exit-zero --max-complexity=10 --max-line-length=127 --statistics verbatim tests
# pip install bandit
bandit -r verbatim tests