[MPS] Convert output back to ChannelsLast for MaxPool2D #94877
Closed
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
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/94877
Note: Links to docs will display an error until the docs builds have been completed. ✅ No FailuresAs of commit 5c8c4df: This comment was automatically generated by Dr. CI and updates every 15 minutes. |
pytorch-bot
bot
added
ciflow/mps
kulinseth
approved these changes
Feb 15, 2023
|
@pytorchbot merge -g "All tests are green." |
|
❌ 🤖 pytorchbot command failed: Try |
|
@pytorchbot merge -f "All tests are green." |
Merge startedYour change will be merged immediately since you used the force (-f) flag, bypassing any CI checks (ETA: 1-5 minutes). Learn more about merging in the wiki. Questions? Feedback? Please reach out to the PyTorch DevX Team |
drisspg
pushed a commit
to drisspg/pytorch
that referenced
this pull request
Feb 16, 2023
Since we re-stride the indices and output in MPS pooling from ChannelsLast to Contiguous, we need to convert the results back to ChannelsLast. This will fix the failure with test_memory_format with MaxPool2D in test_modules.py. Pull Request resolved: pytorch#94877 Approved by: https://github.com/kulinseth, https://github.com/DenisVieriu97
kulinseth
pushed a commit
to kulinseth/pytorch
that referenced
this pull request
Feb 22, 2023
Since we re-stride the indices and output in MPS pooling from ChannelsLast to Contiguous, we need to convert the results back to ChannelsLast. This will fix the failure with test_memory_format with MaxPool2D in test_modules.py. Pull Request resolved: pytorch#94877 Approved by: https://github.com/kulinseth, https://github.com/DenisVieriu97
This was referenced Feb 22, 2023
atalman
pushed a commit
that referenced
this pull request
Feb 24, 2023
* [MPS] Fix upsample for NHWC output (#94963) Fixes huggingface/diffusers#941 **Before**: <img width="1144" alt="Screenshot 2023-02-15 at 8 11 53 PM" src="https://user-images.githubusercontent.com/104024078/219266709-6a77636a-2fc0-4802-b130-85069b95953f.png"> **After**: <img width="1144" alt="Screenshot 2023-02-15 at 8 12 02 PM" src="https://user-images.githubusercontent.com/104024078/219266694-ea743c02-fb55-44f1-b7d6-5946106527c3.png"> Pull Request resolved: #94963 Approved by: https://github.com/razarmehr * [MPS] Move max_pool2d to mps dispatch key (#90772) Related issue: #77394 This PR also modifies some assertions in the codegen, an explanatory comment for it has been added. Pull Request resolved: #90772 Approved by: https://github.com/albanD * [MPS] Convert output back to ChannelsLast for MaxPool2D (#94877) Since we re-stride the indices and output in MPS pooling from ChannelsLast to Contiguous, we need to convert the results back to ChannelsLast. This will fix the failure with test_memory_format with MaxPool2D in test_modules.py. Pull Request resolved: #94877 Approved by: https://github.com/kulinseth, https://github.com/DenisVieriu97 --------- Co-authored-by: Denis Vieriu <104024078+DenisVieriu97@users.noreply.github.com> Co-authored-by: Li-Huai (Allan) Lin <qqaatw@gmail.com> Co-authored-by: Ramin Azarmehr <razarmehr@apple.com>
pruthvistony
added a commit
to ROCm/pytorch
that referenced
this pull request
May 2, 2023
…rch#94877)" This reverts commit 7dd7dde.
pruthvistony
pushed a commit
to ROCm/pytorch
that referenced
this pull request
May 3, 2023
* [MPS] Fix upsample for NHWC output (pytorch#94963) Fixes huggingface/diffusers#941 **Before**: <img width="1144" alt="Screenshot 2023-02-15 at 8 11 53 PM" src="https://user-images.githubusercontent.com/104024078/219266709-6a77636a-2fc0-4802-b130-85069b95953f.png"> **After**: <img width="1144" alt="Screenshot 2023-02-15 at 8 12 02 PM" src="https://user-images.githubusercontent.com/104024078/219266694-ea743c02-fb55-44f1-b7d6-5946106527c3.png"> Pull Request resolved: pytorch#94963 Approved by: https://github.com/razarmehr * [MPS] Move max_pool2d to mps dispatch key (pytorch#90772) Related issue: pytorch#77394 This PR also modifies some assertions in the codegen, an explanatory comment for it has been added. Pull Request resolved: pytorch#90772 Approved by: https://github.com/albanD * [MPS] Convert output back to ChannelsLast for MaxPool2D (pytorch#94877) Since we re-stride the indices and output in MPS pooling from ChannelsLast to Contiguous, we need to convert the results back to ChannelsLast. This will fix the failure with test_memory_format with MaxPool2D in test_modules.py. Pull Request resolved: pytorch#94877 Approved by: https://github.com/kulinseth, https://github.com/DenisVieriu97 --------- Co-authored-by: Denis Vieriu <104024078+DenisVieriu97@users.noreply.github.com> Co-authored-by: Li-Huai (Allan) Lin <qqaatw@gmail.com> Co-authored-by: Ramin Azarmehr <razarmehr@apple.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.
Since we re-stride the indices and output in MPS pooling from ChannelsLast to Contiguous, we need to convert the results back to ChannelsLast.
This will fix the failure with test_memory_format with MaxPool2D in test_modules.py.