Open
Conversation
granite-4.0-h-small is a hybrid Mamba/Attention model with GQA group counts that are not divisible by 2, causing an assertion error on HPU when tensor-parallel-size=2. Reduce to 1 since the model is small enough (1.8B params) to run on a single card. Signed-off-by: Willy Hardy <whardy@redhat.com> Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Summary
tensor-parallel-sizefrom 2 to 1 foribm-granite/granite-4.0-h-smallperformance server configassert n_groups % self.tp_size == 0on HPUTest plan
ocp-test.ymlwithconfig_ref=INFERENG-6076/granite-hpu-tp-fixto verify granite-4.0-h-small passes🤖 Generated with Claude Code