Enable CUDA 12.4 builds

GHA results show this is needed to fix errors in pytorch/pytorch#121684 Reference: pytorch#1374
nWEIdia · Apr 11, 2024 · bf63b1a · bf63b1a
1 parent 87cdc8c
commit bf63b1a
Show file tree

Hide file tree

Showing 3 changed files with 11 additions and 3 deletions.
diff --git a/conda/build_pytorch.sh b/conda/build_pytorch.sh
@@ -268,7 +268,10 @@ else
  . ./switch_cuda_version.sh "$desired_cuda"
  # TODO, simplify after anaconda fixes their cudatoolkit versioning inconsistency.
  # see: https:/conda-forge/conda-forge.github.io/issues/687#issuecomment-460086164
- if [[ "$desired_cuda" == "12.1" ]]; then
+ if [[ "$desired_cuda" == "12.4" ]]; then
+ export CONDA_CUDATOOLKIT_CONSTRAINT=" - pytorch-cuda >=12.4,<12.5 # [not osx]"
+ export MAGMA_PACKAGE=" - magma-cuda124 # [not osx and not win]"
+ elif [[ "$desired_cuda" == "12.1" ]]; then
  export CONDA_CUDATOOLKIT_CONSTRAINT=" - pytorch-cuda >=12.1,<12.2 # [not osx]"
  export MAGMA_PACKAGE=" - magma-cuda121 # [not osx and not win]"
  elif [[ "$desired_cuda" == "11.8" ]]; then

diff --git a/conda/pytorch-nightly/build.sh b/conda/pytorch-nightly/build.sh
@@ -60,10 +60,11 @@ if [[ -n "$build_with_cuda" ]]; then
  TORCH_CUDA_ARCH_LIST="$TORCH_CUDA_ARCH_LIST;3.7+PTX;9.0"
  #for cuda 11.8 include all dynamic loading libraries
  DEPS_LIST=(/usr/local/cuda/lib64/libcudnn*.so.8 /usr/local/cuda-11.8/extras/CUPTI/lib64/libcupti.so.11.8 /usr/local/cuda/lib64/libcusparseLt.so.0)
- elif [[ $CUDA_VERSION == 12.1* ]]; then
+ elif [[ $CUDA_VERSION == 12.1* || $CUDA_VERSION == 12.4* ]]; then
  # cuda 12 does not support sm_3x
  TORCH_CUDA_ARCH_LIST="$TORCH_CUDA_ARCH_LIST;9.0"
  # for cuda 12.1 we use cudnn 8.8 and include all dynamic loading libraries
+ # for cuda 12.4 we use cudnn 8.9
  DEPS_LIST=(/usr/local/cuda/lib64/libcudnn*.so.8 /usr/local/cuda-12.1/extras/CUPTI/lib64/libcupti.so.12 /usr/local/cuda/lib64/libcusparseLt.so.0)
  fi
  if [[ -n "$OVERRIDE_TORCH_CUDA_ARCH_LIST" ]]; then

diff --git a/manywheel/build_cuda.sh b/manywheel/build_cuda.sh
@@ -59,6 +59,10 @@ cuda_version_nodot=$(echo $CUDA_VERSION | tr -d '.')
 
 TORCH_CUDA_ARCH_LIST="5.0;6.0;7.0;7.5;8.0;8.6"
 case ${CUDA_VERSION} in
+ 12.4)
+ TORCH_CUDA_ARCH_LIST="${TORCH_CUDA_ARCH_LIST};9.0"
+ EXTRA_CAFFE2_CMAKE_FLAGS+=("-DATEN_NO_TEST=ON")
+ ;;
  12.1)
  TORCH_CUDA_ARCH_LIST="${TORCH_CUDA_ARCH_LIST};9.0"
  EXTRA_CAFFE2_CMAKE_FLAGS+=("-DATEN_NO_TEST=ON")
@@ -131,7 +135,7 @@ if [[ $USE_CUSPARSELT == "1" ]]; then
  )
 fi
 
-if [[ $CUDA_VERSION == "12.1" ]]; then
+if [[ $CUDA_VERSION == "12.1" || $CUDA_VERSION == "12.4" ]]; then
  export USE_STATIC_CUDNN=0
  # Try parallelizing nvcc as well
  export TORCH_NVCC_FLAGS="-Xfatbin -compress-all --threads 2"