fixes non-granite model instantiation with Liger Kernel #476
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
We change the model forward method to use a different loss function. Liger Kernel's `FusedLinearCrossEntropy` kernel doesn't materialize these logits, so our custom forward was broken for Qwen2.5, Llama3, Gemma3, etc. This switches `FusedLinearCrossEntropy` for `FusedCrossEntropy` as the default kernel used. Signed-off-by: James Kunstle <[email protected]>
cdoern
approved these changes
Apr 16, 2025
Maxusmusti
approved these changes
Apr 16, 2025
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
4 participants
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Outdated suggestions cannot be applied.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.
We change the model forward method to use a different loss function. Liger Kernel's
FusedLinearCrossEntropykernel doesn't materialize these logits, so our custom forward was broken for Qwen2.5, Llama3, Gemma3, etc.This switches
FusedLinearCrossEntropyforFusedCrossEntropyas the default kernel used.