Put back previous tolerance for test_classification and test_video #7202
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
NicolasHug
changed the title
pmeier
approved these changes
Feb 9, 2023
There was a problem hiding this comment.
LGTM if CI is green-ish again. Thanks Nicolas! Just to make sure: #6380 changed the precision on three tests, but here we are only reverting two of them. Meaning, test_quantized_classification_model is fine with the stricter tolerance?
|
Yes, I'm not seeing failures for the quantized models, so it looks like tightening that tolerance was OK |
NicolasHug
added
module: tests
revert(ed)
|
Hey @NicolasHug! You merged this PR, but no labels were added. The list of valid labels is available at https://github.com/pytorch/vision/blob/main/.github/process_commit.py |
facebook-github-bot
pushed a commit
that referenced
this pull request
Mar 28, 2023
…_video (#7202) Reviewed By: vmoens Differential Revision: D44416276 fbshipit-source-id: 40ce41797c05765010a401abaaf46d845207750a
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Outdated suggestions cannot be applied.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.
This reverts #6380 which was identified to be the cause of various recent failures #7114 (comment).
This PR fixes the current model failures for video and classification (except for this weird cuda-vit_h_14 failure which is tracked in #7143.
EDIT: there's still a failure for resnet101 as well
I'll submit another PR for the detection model and mark them as flaky (like what #7130 already does)
Like all of those quick-fix PRs, this one is less than ideal. It's just a patch over an already messy situation. In the (hopefully near) future, we should focus on more robust changes like the one proposed in #7130
cc @pmeier