[MLAS/CPU EP] Improve performance of Silu activation path within the QuickGelu CPU kernel #26753
Add Silu implementation in MLAS
7cb74751
Fix
db4fd886
Hook-up CPU EP op with MLAS code
abfd8293
Fix
17144ec9
hariharans29
changed the title WIP: [MLAS] [DO NOT REVIEW] Implement vectorized Silu operation WIP: [MLAS] [DO NOT REVIEW] Implement vectorized fused Silu operation 8 days ago
Expt
6c5ebd1c
Revert separate Silu opration
0836eac3
hariharans29
changed the title WIP: [MLAS] [DO NOT REVIEW] Implement vectorized fused Silu operation WIP: [MLAS] Improve performance of Silu activation path within the QuickGelu CPU kernel 7 days ago
Add vectorized elementwise operation
4736b8fb
hook up the kernel
5f3478ee
Remove unnecessary lines of code
0d2c03b7
hariharans29
changed the title WIP: [MLAS] Improve performance of Silu activation path within the QuickGelu CPU kernel WIP: [MLAS/CPU EP] Improve performance of Silu activation path within the QuickGelu CPU kernel 7 days ago
Bug fix
a89867d8
Merge remote-tracking branch 'origin/main' into hari/mlas_silu
667f22b9
Update activations.h
9f801d33
Update activations.h
52e2b9d6
hariharans29
changed the title WIP: [MLAS/CPU EP] Improve performance of Silu activation path within the QuickGelu CPU kernel [MLAS/CPU EP] Improve performance of Silu activation path within the QuickGelu CPU kernel 7 days ago
Update onnxruntime/core/mlas/lib/eltwise.cpp
9883d055
Add MLAS test
b6397172
Adjust
bdeaf672
Assignees
No one assigned
Login to write a write a comment.
Login via GitHub