- Allow quantize fusion for non per-tensor cases if no shape change #35184
- Allow quantize fusion for non per-tensor cases if no shape change
bdbe061c
per_tensor_values default init to false
3d7601b8
prepare_primitive_fusing_through_test is added.
7abd0c0e
dynamic_shape check and helper method for checking per_tensor_values
7eb64d2c
code cleanup
03608889
p-durandin
added this to the 2026.3 milestone 33 days ago
p-durandin
merged
cdc71a24
into master 32 days ago
Assignees
No one assigned
Labels
category: GPU
ExternalPR
Login to write a write a comment.
Login via GitHub