Cache OpenBLAS to docker image for SBSA builds #1842
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
tinglvv
changed the title
tinglvv
changed the title
|
hi @tinglvv looks good, but please open test PY in pytorch/pytorch to test this change with aarch64 cpu builds, to make sure we are not failing |
atalman
requested changes
May 31, 2024
atalman
approved these changes
May 31, 2024
atalman
added a commit
that referenced
this pull request
Jun 17, 2024
* Remove triton constraint for py312 (#1846) * Cache OpenBLAS to docker image for SBSA builds (#1842) * apply openblas cache for cpu-aarch64 * reapply for cuda-aarch64 * [MacOS] Don't build wheel while building libtorch Not sure why this was ever done twice * Allow validate doker images to be called from different workflow (#1850) * Allow validate doker images to be called from different workflow * Revert "[MacOS] Don't build wheel while building libtorch" This reverts commit d88495a. * [MacOS] Don't build libtorch twice (take 2) By not invoking `tools/build_libtorch.py` as as it's not done on Linux * [MacOs][LibTorch] Copy libomp.dylib into libtorch package * Update cudnn from v8 to v9 across CUDA versions and x86/arm (#1847) * Update cudnn to v9.1.0.70 for cuda11.8, cuda12.1, and cuda12.4 * Add CUDNN_VERSION variable * Remove 2 spaces for install_cu124 * trivial fix * Fix DEPS_LIST and DEPS_SONAME for x86 Update cudnn to v9 for arm cuda binary as well * libcudnn_adv_infer/libcudnn_adv_train becomes libcudnn_adv * Change DEPS due to cudnn v9 libraries name changes (and additions) * Fix lint * Add missing changes to cu121/cu124 * Change OpenSSL URL (#1854) * Change OpenSSL URL * Change to use openssl URL (but no longer ftp!) * Update build-manywheel-images.yml - Add a note about manylinux_2_28 state * Revert "Update cudnn from v8 to v9 across CUDA versions and x86/arm" (#1855) This reverts commit 5783bcc. * Don't run torch.compile on runtime images in docker validations (#1858) * Don't run torch.compile on runtime images * test * Don't run torch.compile on runtime images in docker validations * Update cudnn from v8 to v9 across CUDA versions and x86/arm (#1857) * Update cudnn to v9.1.0.70 for cuda11.8, cuda12.1, and cuda12.4 * Add CUDNN_VERSION variable * Remove 2 spaces for install_cu124 * trivial fix * Fix DEPS_LIST and DEPS_SONAME for x86 Update cudnn to v9 for arm cuda binary as well * libcudnn_adv_infer/libcudnn_adv_train becomes libcudnn_adv * Change DEPS due to cudnn v9 libraries name changes (and additions) * Fix lint * Add missing changes to cu121/cu124 * Fix aarch64 cuda typos * Update validate-docker-images.yml - disable runtime error check for now * Update validate-docker-images.yml - use validation_runner rather then hardcoded one * Update validate-docker-images.yml - fix MATRIX_GPU_ARCH_TYPE setting for cpu only workflows * [aarch64 cuda cudnn] Add RUNPATH to libcudnn_graph.so.9 (#1859) * Add executorch to pypi prep, promotion and validation scripts (#1860) * Add AOTriton install step for ROCm manylinux images (#1862) * Add AOTriton install step for ROCm * No common_utils.sh needed * temporary disable runtime error check * Add python 3.13 builder (#1845) --------- Co-authored-by: Ting Lu <92425201+tinglvv@users.noreply.github.com> Co-authored-by: Nikita Shulga <nikita.shulga@gmail.com> Co-authored-by: Wei Wang <143543872+nWEIdia@users.noreply.github.com> Co-authored-by: Jithun Nair <37884920+jithunnair-amd@users.noreply.github.com>
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
follow up on #1833
applying it for cpu aarch64 as build is failing with
to reduce build time
Caching OpenBLAS to docker image.
Combining fixes for cpu-aarch64 and cuda-aarch64.
relevant issue: pytorch/pytorch#126980
cc @atalman @ptrblck @nWEIdia @Aidyn-A