Decoder: fix performance regression #755
Merged
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
|
The docs for this PR live here. All of your documentation changes will be reflected on that endpoint. The docs are available until 30 days after the last update. |
The logits warpers used for sampling are now included in the list of processors obtained when calling GenerationMixin._get_logits_processor. We need to remove them explicitly because we use instead a fused logits warper (which is 10x faster).
dacorvo
force-pushed
the
fix_perf_regression
branch
from
cd6ebb6 to
d4084a3
Compare
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Outdated suggestions cannot be applied.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.
What does this PR do?
With
transformers >= 4.46.0the top-k, top-p and temperature logits processors have been included in the common logits processors lists, where they were previously included in a logits warpers list used only when sampling.This defeated the optimization previously introduced in
optimum-neuronto replace the logits warpers by a single fused warper. This led to a huge increase of the sampling latency.This pull-request removes the superfluous
transformerssampling warpers and restores the optimized processing.