Register torch.return_types.* as pytree nodes #75915
Closed
Conversation
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
Contributor
🔗 Helpful links
💊 CI failures summary and remediationsAs of commit 4b8657a (more details on the Dr. CI page): 💚 💚 Looks good so far! There are no failures yet. 💚 💚 This comment was automatically generated by Dr. CI (expand for details).Please report bugs/suggestions to the (internal) Dr. CI Users group. |
zou3519
force-pushed
the
tree_map_return_types
branch
from
518ffd6 to
037dc1e
Compare
ezyang
reviewed
Apr 18, 2022
| _register_pytree_node(tuple, _tuple_flatten, _tuple_unflatten) | ||
| _register_pytree_node(namedtuple, _namedtuple_flatten, _namedtuple_unflatten) | ||
|
|
||
|
|
ezyang
approved these changes
Apr 18, 2022
Contributor
|
I would like to know if there's a more proper way to check structseq but I wouldn't block this on it |
All of the torch.return_types.* are these special things "structseq" that subclass tuple but have a different constructor from tuple :(. This PR iterates through all of torch.return_types.* and adds a pytree registration for them. Test Plan: - add tests for max and min which return torch.return_types.max, and torch.return_types.min, respectively. There's not an easy way to "get all torch ops that return a return_types object".
zou3519
force-pushed
the
tree_map_return_types
branch
from
037dc1e to
4b8657a
Compare
Contributor
Author
I'll file a followup issue |
kshitij12345
approved these changes
Apr 18, 2022
Contributor
Author
|
@pytorchbot merge this please |
Contributor
|
Hey @zou3519. |
This was referenced Apr 19, 2022
facebook-github-bot
pushed a commit
that referenced
this pull request
Apr 19, 2022
Summary: All of the torch.return_types.* are these special things "structseq" that subclass tuple but have a different constructor from tuple :(. This PR iterates through all of torch.return_types.* and adds a pytree registration for them. Pull Request resolved: #75915 Approved by: https://github.com/ezyang, https://github.com/kshitij12345 Test Plan: contbuild & OSS CI, see https://hud.pytorch.org/commit/pytorch/pytorch/18c74d10bbec7728c22e7868caa880c5d4189c1a Test plan from GitHub: - add tests for max and min which return torch.return_types.max, and torch.return_types.min, respectively. There's not an easy way to "get all torch ops that return a return_types object". Fixes #75218 Reviewed By: seemethere Differential Revision: D35751456 Pulled By: zou3519 fbshipit-source-id: 88a5289019eb4f5642f8e9ece6791581a0f2b7a9
malfet
pushed a commit
that referenced
this pull request
Apr 20, 2022
All of the torch.return_types.* are these special things "structseq" that subclass tuple but have a different constructor from tuple :(. This PR iterates through all of torch.return_types.* and adds a pytree registration for them. Test Plan: - add tests for max and min which return torch.return_types.max, and torch.return_types.min, respectively. There's not an easy way to "get all torch ops that return a return_types object". Fixes #75218 Pull Request resolved: #75915 Approved by: https://github.com/ezyang, https://github.com/kshitij12345 (cherry picked from commit 18c74d1)
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.
All of the torch.return_types.* are these special things "structseq"
that subclass tuple but have a different constructor from tuple :(.
This PR iterates through all of torch.return_types.* and adds a pytree
registration for them.
Test Plan:
torch.return_types.min, respectively. There's not an easy way to
"get all torch ops that return a return_types object".
Fixes #75218