Add Qwen2-VL #1542
Open
Add Qwen2-VL #1542
+812
−0
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
nngokhale
requested review from
ssarkar2,
bhargaveede,
vivekgoe and
regisss
as code owners
ssarkar2
reviewed
Dec 3, 2024
| @@ -67,6 +68,7 @@ def gaudi_MaxTimeCriteria_call( | |||
| def gaudi_EosTokenCriteria_call( | |||
| self, input_ids: torch.LongTensor, scores: torch.FloatTensor, **kwargs | |||
| ) -> Union[torch.BoolTensor, bool]: | |||
| htcore.mark_step() | |||
There was a problem hiding this comment.
this is a general change, and may affect performance in other models.
Can you add some numbers for the compile times u have with and without this change?
There was a problem hiding this comment.
is it possible to add teh markstep at teh end of the qwen model or something like that so the effect of this is isolated or qwen only?
otherwise we'd have to do a larger study to make sure it doesnt have any bad effect on any other model/usecase
There was a problem hiding this comment.
mark_step at the end of the model doesn't have the same effect.
There is 2x improvement in warmup time.
I agree this should be independent PR with more tests.
I will remove this particular change.
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Outdated suggestions cannot be applied.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?
Adds optimized Qwen2-VL to Optimum habana.
Uses HF static cache.
Adds OH static shapes.
Fixes graph recompilations.
Adds FusedSDPA with use_flash_attention kwarg
Removed "Adds mark_step in eos stopping criterion for reduced graph recompilation time"
Fixes # (issue)
Before submitting