[NLLB-MoE] Fix NLLB MoE 4bit inference
#27012
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 documentation is not available anymore as the PR was closed or merged. |
ArthurZucker
approved these changes
Oct 23, 2023
|
Thanks! |
staghado
pushed a commit
to staghado/transformers
that referenced
this pull request
Oct 24, 2023
fix NLLB MoE 4bit
EduardoPach
pushed a commit
to EduardoPach/transformers
that referenced
this pull request
Nov 19, 2023
fix NLLB MoE 4bit
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Outdated suggestions cannot be applied.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?
Fixes: #26898
The hidden states gets silenty casted in
unint8leading to the error described in #26898The check
and self.fc2.weight.dtype != torch.int8is not sufficient in order to cover 4bit models, for these models the weights are inuint8, hence adding an extra condition to cover 4bit models fixes the inference issuecc @ArthurZucker