adding blackwell support for precompiled wheels #1285
Conversation
This file contains hidden or 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
Signed-off-by: LosCrossos <[email protected]>
facebook-github-bot
added
the
CLA Signed
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
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Outdated suggestions cannot be applied.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.
the current github wheel build process uses TORCH_CUDA_ARCH_LIST = 7.5 8.0+PTX 9.0a
this should be forwards compatible but on in some cases for blackwell it isnt (see #1251).
Currently the compilation code already supports blackwell. Building with TORCH_CUDA_ARCH_LIST=12.0 solves the issue on #1251 but the pre-built wheels are not being compiled with that flag. Since the build process is done with CudaTK12.8 already i added capability 100 and 120 to the workflow. I already submitted a PR for the FA2 code path (#1254).
this PR adds support for blackwell on the main build workflow for the pipy builds.