onnxruntime
[CPU EP] Add blocked quantization to QuantizeLinear op kernel
#20977

Merged

[CPU EP] Add blocked quantization to QuantizeLinear op kernel #20977

fajin-corp merged 10 commits into main from fajin/qblockedquantize

yufenglee commented on 2024-06-11

added baseline Q op kernel

8e2b6a62

finalized thread blocking approach

c3fa1cb4

passed build

abe14263

basedline passed UT

da9763aa

expanded UT to cover multithreading cases

4f7edfff

finished perf

7ac407d9

reset task block size

0b0075b6

resolve CI build failures

227386c2

change namings

c8afdbc8

fix ci warning

81d32fcf

fajin-corp force pushed from 2034cbd8 to 81d32fcf 1 year ago

yufenglee approved these changes on 2024-06-12

fajin-corp merged 9be30348 into main 1 year ago

fajin-corp deleted the fajin/qblockedquantize branch 1 year ago

Reviewers

yufenglee

Assignees

No one assigned

Labels

None yet

Milestone

No milestone