[HOP] Rework Autograd DispatchKey for scan and map #153336
Open
+53
−17
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
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/153336
Note: Links to docs will display an error until the docs builds have been completed. ❌ 1 New Failure, 1 Unrelated FailureAs of commit 9f9adc1 with merge base 47df195 ( NEW FAILURE - The following job has failed:
BROKEN TRUNK - The following job failed but were present on the merge base:👉 Rebase onto the `viable/strict` branch to avoid these failures
This comment was automatically generated by Dr. CI and updates every 15 minutes. |
|
@pytorchbot label "topic: not user facing" |
bohnstingl
commented
May 12, 2025
Comment on lines
+184
to
+191
| # outs = f(xs, *args) | ||
| # if pytree.tree_any(lambda elem: not isinstance(elem, torch.Tensor) if elem is not None else False, outs): | ||
| # raise RuntimeError( | ||
| # "Expect outputs of map only contains tensors or None. " | ||
| # f"Got types {[type(out) for out in pytree.tree_leaves(outs)]}." | ||
| # ) | ||
| # return outs | ||
|
|
There was a problem hiding this comment.
When using the new key @map_impl.py_autograd_impl, the map_dense gets triggered directly and thus create_fw_bw_graph is not used. Therefore, the check of the outputs in create_fw_bw_graph is also not performed and the check can either be done here or below (see comment)
bohnstingl
commented
May 12, 2025
Comment on lines
+269
to
+277
| outs_flatten = pytree.tree_leaves(pytrees) | ||
| if any( | ||
| # pytree.tree_map(lambda elem: not isinstance(elem, torch.Tensor), outs) | ||
| not isinstance(out, torch.Tensor) for out in outs_flatten if out is not None | ||
| ): | ||
| raise RuntimeError( | ||
| "Expect outputs of map only contains tensors or None. " | ||
| f"Got types {[type(out) for out in outs_flatten]}." | ||
| ) |
There was a problem hiding this comment.
or here.
Which option would be preferable?
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 PR introduces the
py_autograd_implinstead of theDispatchKey.Autogradfor some HOPs.cc @voznesenskym @penguinwu @EikanWang @jgong5 @Guobing-Chen @XiaobingSuper @zhuhaozhe @blzheng @wenzhe-nrv @jiayisunx @chenyang78 @kadeng @chauhang @amjames