[rocm] Fix current torch.linalg.lstsq limitation in InverseMelScale module#3939
Draft
amd-sriram wants to merge 1 commit intopytorch:mainfrom
Draft
[rocm] Fix current torch.linalg.lstsq limitation in InverseMelScale module#3939amd-sriram wants to merge 1 commit intopytorch:mainfrom
amd-sriram wants to merge 1 commit intopytorch:mainfrom
Conversation
RuntimeError: torch.linalg.lstsq: only overdetermined systems (input.size(-2) >= input.size(-1)) are allowed on CUDA. Please rebuild with cuSOLVER Checking the pytorch code aten/src/ATen/native/cuda/linalg/BatchLinearAlgebra.cpp, it calls magma. According to magma documentation, least squares method - magma_sgels_gpu does supports only overdetermined systems https://icl.utk.edu/projectsfiles/magma/doxygen/group__magma__gels.html The proposed temporary solution uses https://docs.pytorch.org/docs/stable/generated/torch.linalg.pinv.html as a replacement for linear least squares. This change can be reverted once torch.linalg.lstsq uses hipSolver for underconstrained systems. Tested on rocm with pytest test/torchaudio_unittest/transforms/transforms_cuda_test.py -k test_inverse_melscale
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/audio/3939
Note: Links to docs will display an error until the docs builds have been completed. This comment was automatically generated by Dr. CI and updates every 15 minutes. |
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
This PR fixes the following error:
RuntimeError: torch.linalg.lstsq: only overdetermined systems (input.size(-2) >= input.size(-1)) are allowed on CUDA. Please rebuild with cuSOLVERChecking the pytorch code https://github.com/ROCm/pytorch/blob/main/aten/src/ATen/native/cuda/linalg/BatchLinearAlgebra.cpp, it calls magma. According to magma documentation, least squares method - magma_sgels_gpu supports only overdetermined systems https://icl.utk.edu/projectsfiles/magma/doxygen/group__magma__gels.html
The proposed temporary solution uses https://docs.pytorch.org/docs/stable/generated/torch.linalg.pinv.html as a replacement for linear least squares. This change can be reverted once torch.linalg.lstsq uses hipSolver for underconstrained systems.
Tested on rocm with
pytest test/torchaudio_unittest/transforms/transforms_cuda_test.py -k test_inverse_melscale