KEMBAR78
[ROCm] miopen benchmark behavior now better aligns with cudnn by jeffdaily · Pull Request #145294 · pytorch/pytorch · GitHub
Skip to content

Conversation

@jeffdaily
Copy link
Collaborator

@jeffdaily jeffdaily commented Jan 21, 2025

The default benchmark setting is now false. The new miopen behavior means when benchmarking is disabled, for any shape that doesn't have a find hit, then it will do a quick search (same behavior as the prior default), and use that result. Now when benchmark is enabled, it will perform an exhaustive search and update any DBs. miopen immediate mode is still available and is used when deterministic is true and benchmark is false.

cc @sunway513 @jithunnair-amd @pruthvistony @ROCmSupport @dllehr-amd @jataylo @hongxiayang @naromero77amd

@pytorch-bot
Copy link

pytorch-bot bot commented Jan 21, 2025

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/145294

Note: Links to docs will display an error until the docs builds have been completed.

❗ 1 Active SEVs

There are 1 currently active SEVs. If your PR is affected, please view them below:

⏳ No Failures, 40 Pending

As of commit f0b12f0 with merge base 93d98ac (image):
💚 Looks good so far! There are no failures yet. 💚

This comment was automatically generated by Dr. CI and updates every 15 minutes.

@pytorch-bot pytorch-bot bot added ciflow/rocm Trigger "default" config CI on ROCm module: rocm AMD GPU support for Pytorch labels Jan 21, 2025
@pytorch-bot pytorch-bot bot had a problem deploying to upload-benchmark-results January 21, 2025 21:56 Failure
@pytorch-bot pytorch-bot bot temporarily deployed to upload-benchmark-results January 21, 2025 21:56 Inactive
@pytorch-bot pytorch-bot bot temporarily deployed to upload-benchmark-results January 21, 2025 21:56 Inactive
@pytorch-bot pytorch-bot bot temporarily deployed to upload-benchmark-results January 21, 2025 21:56 Inactive
@pytorch-bot pytorch-bot bot temporarily deployed to upload-benchmark-results January 21, 2025 21:56 Inactive
@pytorch-bot pytorch-bot bot temporarily deployed to upload-benchmark-results January 21, 2025 21:56 Inactive
Copy link

@BrianHarrisonAMD BrianHarrisonAMD left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

LGTM

@jeffdaily jeffdaily marked this pull request as ready for review January 31, 2025 05:40
@colesbury colesbury added the triaged This issue has been looked at a team member, and triaged and prioritized into an appropriate module label Jan 31, 2025
@jeffdaily jeffdaily added release notes: rocm mandatorylabel release notes: nn release notes category labels Feb 5, 2025
@jeffdaily
Copy link
Collaborator Author

@pytorchbot rebase

@jeffdaily jeffdaily requested a review from malfet February 5, 2025 15:52
@pytorchmergebot
Copy link
Collaborator

@pytorchbot started a rebase job onto refs/remotes/origin/viable/strict. Check the current status here

The default benchmark setting is now false. The new miopen behavior
means when benchmarking is disabled, for any shape that doesn't have a
find hit, then it will do a quick search (same behavior as the prior
default), and use that result. Now when benchmark is enabled, it will
perform an exhaustive search and update any DBs. miopen immediate mode
is still available and is used when deterministic is true and benchmark
is false.
@pytorchmergebot
Copy link
Collaborator

Successfully rebased miopen_benchmark_false onto refs/remotes/origin/viable/strict, please pull locally before adding more changes (for example, via git checkout miopen_benchmark_false && git pull --rebase)

Copy link
Contributor

@malfet malfet left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Looks good to me

@malfet
Copy link
Contributor

malfet commented Feb 5, 2025

@pytorchbot merge -f "ROCM builds are green?"

@pytorchmergebot
Copy link
Collaborator

Merge started

Your change will be merged immediately since you used the force (-f) flag, bypassing any CI checks (ETA: 1-5 minutes). Please use -f as last resort and instead consider -i/--ignore-current to continue the merge ignoring current failures. This will allow currently pending tests to finish and report signal before the merge.

Learn more about merging in the wiki.

Questions? Feedback? Please reach out to the PyTorch DevX Team

Advanced Debugging
Check the merge workflow status
here

pytorchmergebot pushed a commit that referenced this pull request Jul 25, 2025
The MIOpen integration has changed over the years.  In the past, the MIOpen default for benchmark was True and if it were set to False it would use MIOpen Immediate Mode.  But with #145294 the MIOpen benchmark default changed to False and to activate immediate mode you would set the deterministic flag to True.  This has proved too restrictive because benchmark and deterministic flags are independent from immediate mode.  Thus, immediate mode needs its own flag.  Though MIOpen still masquerades behind torch.backends.cudnn and its flags, it seemed inappropriate to add an miopen-exclusive flag to the set of cudnn flags.  This PR adds the first miopen-only flag to control its immediate mode.

Pull Request resolved: #158951
Approved by: https://github.com/jeffdaily

Co-authored-by: Jeff Daily <jeff.daily@amd.com>
yangw-dev pushed a commit that referenced this pull request Aug 1, 2025
The MIOpen integration has changed over the years.  In the past, the MIOpen default for benchmark was True and if it were set to False it would use MIOpen Immediate Mode.  But with #145294 the MIOpen benchmark default changed to False and to activate immediate mode you would set the deterministic flag to True.  This has proved too restrictive because benchmark and deterministic flags are independent from immediate mode.  Thus, immediate mode needs its own flag.  Though MIOpen still masquerades behind torch.backends.cudnn and its flags, it seemed inappropriate to add an miopen-exclusive flag to the set of cudnn flags.  This PR adds the first miopen-only flag to control its immediate mode.

Pull Request resolved: #158951
Approved by: https://github.com/jeffdaily

Co-authored-by: Jeff Daily <jeff.daily@amd.com>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

ciflow/rocm Trigger "default" config CI on ROCm Merged module: rocm AMD GPU support for Pytorch open source release notes: nn release notes category release notes: rocm mandatorylabel triaged This issue has been looked at a team member, and triaged and prioritized into an appropriate module

Projects

None yet

Development

Successfully merging this pull request may close these issues.

6 participants