Skip to content

Installation

greyhound requires Python >= 3.11, PyTorch >= 2.7.1, and a CUDA-capable GPU.

greyhound-kernels is available on PyPI. It provides the greyhound Python import package:

Terminal window
pip install greyhound-kernels

Or with uv:

Terminal window
uv add greyhound-kernels

Clone the repository:

Terminal window
git clone https://github.com/tyler-romero/greyhound.git
cd greyhound

Then install in editable mode with development dependencies:

Terminal window
uv sync --all-extras
Terminal window
pip install -e ".[dev]"

For the Quack dependency used by bonus Newton-Schulz utilities and Quack comparison benchmarks:

Terminal window
uv sync --extra quack
Terminal window
pip install -e ".[quack]"

For the full benchmark provider set:

Terminal window
uv sync --extra dev --group thirdparty

This uv-only dependency group installs comparison providers used by benchmark scripts, such as Dion, liger-kernel, gram-newton-schulz, flash-linear-attention, and causal-conv1d. The group is intentionally not published as a Greyhound package extra because PyPI rejects direct Git dependencies in uploaded package metadata.

For Modal benchmark execution:

Terminal window
uv sync --extra modal
Terminal window
pip install -e ".[modal]"