Fix torch_logs tutorial to support CPU-only environments (#137285)#3924
Open
yyouretoast wants to merge 1 commit into
Open
Fix torch_logs tutorial to support CPU-only environments (#137285)#3924yyouretoast wants to merge 1 commit into
yyouretoast wants to merge 1 commit into
Conversation
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/tutorials/3924
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. |
yyouretoast
force-pushed
the
fix-torch-logs-137285
branch
from
038eaa5 to
16e0b7f
Compare
Remove the hard CUDA gate that caused the tutorial to render as 'Skipped' on the PyTorch docs website. torch.compile works on both CUDA (via Triton) and CPU (via C++ Inductor), so no guard or early exit is needed. Changes: - Use CUDA if available and compute capability >= 7.0, else fall back to CPU (prevents crashes on older CUDA devices while supporting CPU execution). - Remove try/except compilation probe (unnecessary). - Remove sys.exit(0) early exit (dangerous in Sphinx-Gallery subprocess mode). - De-indent all tutorial code from the original if/else block. Fixes pytorch/pytorch#137285
yyouretoast
force-pushed
the
fix-torch-logs-137285
branch
from
16e0b7f to
2f00fa4
Compare
yyouretoast
changed the title
Author
|
Hi! Just bumping this PR. Could a maintainer please approve the |
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
1 participant
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Fixes pytorch/pytorch#137285
The
torch_logstutorial was gated behind a hard CUDA capability check (get_device_capability() < (7, 0)) that printed a skipping message and exited. This caused the tutorial to be skipped during CPU-only documentation builds, resulting in the page rendering as "skipped" on the official PyTorch website.Since
torch.compileis fully supported on CPU (using the C++ Inductor backend) and generates meaningful compilation logging outputs, this PR refactors the tutorial to allow CPU execution. It dynamically selects CUDA if available and the compute capability is>= 7.0(to avoid compile crashes on older GPUs), otherwise falling back to CPU.Changes