Building conda binaries

• Change the BUILD_VERSION and BUILD_NUMBER variables in build_pytorch.sh as appropriate
• If pytorch-$BUILD_VERSION folder doesn't exist, copy over the last version and change the meta.yaml if necessary (if tests change etc.)
  • cp -r pytorch-0.1.3 pytorch-$BUILD_VERSION
  • git add pytorch-$BUILD_VERSION
• Run ./build_pytorch.sh on an OSX machine and a Linux machine

build base docker images

conda/build_all_docker.sh
# Will push all of the images
docker push pytorch/conda-builder

building pytorch / torchvision etc.

# building pytorch
docker run --rm -it \
    -e PACKAGE_TYPE=conda \
    -e DESIRED_CUDA=cu92 \
    -e DESIRED_PYTHON=3.8 \
    -e PYTORCH_BUILD_VERSION=1.5.0 \
    -e PYTORCH_BUILD_NUMBER=1 \
    -e TORCH_CONDA_BUILD_FOLDER=pytorch-nightly \
    -v /path/to/pytorch:/pytorch \
    -v /path/to/builder:/builder \
    -v "$(pwd):/final_pkgs" \
    -u "$(id -u):$(id -g)"  \
    pytorch/conda-cuda \
    /builder/conda/build_pytorch.sh

building magma

MOVED TO magma/README.md