[TRANSFORMATIONS] SDPAToPagedAttention transformation: support decompression case in the Qwen-7b-Chat pattern #28514
+59
−18
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Outdated suggestions cannot be applied.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.
Qwen-7b-Chat has a decompression if the model is executed in lower
precision resulting into the model having additional Converts (i.e. FP16
to FP32).
Handle this case of optional Convert in PositionIDsReplacerQwen
Added a unit test for it.
Tickets:
Signed-off-by: Andrii Staikov [email protected]
Signed-off-by: Ivan Tikhonov [email protected]