Tensor learning compiler binary distribution package.
We use github action to build wheel and conda packages nightly.
Checkout .github/workflows
- Build docker images
./docker/build_docker <CONTAINER_TYPE>
CONTAINER_NAME: Type of the docker container used to build wheels, e.g., (cpu|cu100|cu101|cu102)
- Build tlcpack PIP wheels.
To build wheels for all Python versions (3.6, 3.7, 3.8) with CPU and all CUDA versions (10.0, 10.1, 10.2), run
./scripts/build_pip_wheel.sh
To build wheels for a specific CUDA version, for example, CUDA 10.1, run
./scripts/build_pip_wheel.sh --cuda 10.1
Or, to build wheels for CPU only, run
./scripts/build_pip_wheel.sh --cuda none
Check ./scripts/build_pip_wheel.sh --help
for other options.