Add perturb() to allow capturing intermediate gradients #2476
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
|
Looks good! Should we add a simple test? |
Test added! |
levskaya
approved these changes
Sep 23, 2022
There was a problem hiding this comment.
Looks good as a start. I still think we should think about whether there are ways to simplify using "perturb" in practice w.o. incurring too much extra memory burden. We may wish to reserve the right to modify this API as we understand how people will use it - perhaps it's worth marking as "experimental" in docstrings to be clear that we may need to tweak it in the future.
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Outdated suggestions cannot be applied.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.
Add a perturb() function in nn.Module, which could be used to capture the intermediate gradients inside a module's call.
See example in the docstring. There will be a future Colab/notebook with better examples.