Simplify exclude_from_weight_decay
implementation
#2676
Merged
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Outdated suggestions cannot be applied.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.
Description
This PR changes the handling of
exclude_from_weight_decay
to ensure that the regex is only evaluated once per step. Currently it gets called within a distributedmerge_call
which might lead to problems.I implemented this when trying to debug a performance regression which turned out to be unrelated to this, but I think this still is a nice cleanup to have since it consolidates the handling of
decay_var_list
andexclude_from_weight_decay
.Type of change
Checklist:
How Has This Been Tested?
I adapted the unittests that where directly relying on the internal functions that had been modified.