onnxruntime
Improvements to the INT8 GEMM portion of the code for Power
#20595
Merged
Merged
Improvements to the INT8 GEMM portion of the code for Power #20595
yufenglee
merged
13 commits into
microsoft:main
from
ChipKerchner:powerImprovements
Faster S8 Packing routines.
12b6219f
Use vectors instead of individual element in MlasGemmQuantCopyPackB8x…
6b4f4ccd
Remove comment and fix tab.
263bfe41
Merge branch 'main' into powerImprovements
e25ea1b8
Merge branch 'main' into powerImprovements
d887654f
Fix lint warnings.
a65477d9
Merge branch 'main' into powerImprovements
6c4c9326
Merge remote-tracking branch 'origin_chip/powerImprovements' into pow…
3fd91665
Fix commenters suggestions and other lint warnings.
89d63c22
Unroll MlasQgemmStoreVectorMMA and vectorize other variables.
1b6553a3
ChipKerchner
changed the title
Merge remote-tracking branch 'origin/main' into powerImprovements
9be49a0a
Merge branch 'powerImprovements' of github.ibm.com:PowerAppLibs/onnx …
3e576277
More lint fixes.
84725ba9
yufenglee
approved these changes
on 2024-06-05
Reviewers
Assignees
No one assigned
Labels
None yet
Milestone
No milestone
Login to write a write a comment.
Login via GitHub