Yi-Hong Lyu 02a0be3599 Optimize Transpose around QLinearSoftmax (#22849) ... ### Description <!-- Describe your changes. --> - Improved Transpose around QLinearSoftmax in Level 3 NHWC Transformer. - Removed redundant code HandleQLinearConcat, HandleQLinearBinaryOp. ### Motivation and Context <!-- - Why is this change required? What problem does it solve? - If it fixes an open issue, please link to the issue here. --> By merging and eliminating redundant transpose , the Image Segmentation i8 model (MobileNetv2 + DeepLabv3) achieves a 2.34X speedup.		2024-11-18 06:58:21 -08:00
..
contrib_ops	Refactor SkipLayerNorm and handle beta properly (#22862)	2024-11-17 14:51:16 -08:00
core	Optimize Transpose around QLinearSoftmax (#22849)	2024-11-18 06:58:21 -08:00
lora	Accomodate BE platforms. Make sure we always write flatbuffers LE (#22375)	2024-10-11 09:14:44 -07:00
python	[Quant tool] Handle input models with pre-quantized weights (#22633)	2024-11-14 13:48:46 -08:00
test	Optimize Transpose around QLinearSoftmax (#22849)	2024-11-18 06:58:21 -08:00
tool/etw
wasm	[WebNN] Fixed WebNN Module undefined issue (#22795)	2024-11-11 21:31:24 -08:00
__init__.py	bumps up version in main from 1.20 -> 1.21 (#22482)	2024-10-17 12:32:35 -07:00
ReformatSource.ps1
ReformatSourcePython.bat
VSCodeCoverage.runsettings