Let wrap() preserve metadata for custom TVTensor subclasses#9490
Merged
Conversation
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/vision/9490
Note: Links to docs will display an error until the docs builds have been completed. ❗ 1 Active SEVsThere are 1 currently active SEVs. If your PR is affected, please view them below: This comment was automatically generated by Dr. CI and updates every 15 minutes. |
|
Hey @NicolasHug! You merged this PR, but no labels were added. |
yeetypete
added a commit
to yeetypete/vision3d
that referenced
this pull request
May 21, 2026
- Move CameraIntrinsics torchvision-op kernels (resize/crop/center_crop/ pad/resized_crop) from tensors/_camera.py to transforms/functional/ _camera.py, matching torchvision's layout (kernels next to dispatchers, not next to tv_tensor classes). Rename them to public functional names (resize_camera_intrinsics, etc.), mirroring torchvision's resize_keypoints precedent. - Add public wrap() classmethods to BoundingBoxes3D and CameraIntrinsics; rewrite vision3d.tensors.wrap() to dispatch via getattr(type(like), "wrap", None). Matches the dispatch introduced in pytorch/vision#9490 (torchvision 0.28). Vision3d's wrap() can collapse to a re-export once the pin bumps to torchvision>=0.28. - Import transforms from vision3d/__init__.py so importing any vision3d submodule registers the v2 kernels, matching torchvision/__init__.py.
3 tasks
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Closes #9398
Closes #9333
This PR lets custom classes passed to
tv_tensors.wrap()preserve their metadata by allowing each subclass to define how it wants to be wrapped by defining a .wrap() method.cc @vfdev-5