PR #16196 [ROCm] Add ROCm Triton TunableOp for GroupNorm

[ROCm] Add ROCm Triton TunableOp for GroupNorm #16196

mindest merged 12 commits into main from linmin/triton_groupnorm

Update existing Triton Softmax code

c3c434ca

some more refactor on Triton, Softmax

1df0ecd2

Add Triton implementation in GroupNorm TunableOp

59be5ace

github-advanced-security commented on 2023-06-01

tianleiwu commented on 2023-06-01

Fix CI errors/warnings

5e7b1174

Merge branch 'main' into linmin/triton_groupnorm

8e88a34a

fix CI

f221c26f

mindest force pushed from 47ba1b12 to f221c26f 3 years ago

Improve Triton implementation; fix CI.

d29adb87

Merge branch 'main' into linmin/triton_groupnorm

dd342abc

minor

90dd00a8

Improve load logic, update configs

77ac2424

Merge branch 'main' into linmin/triton_groupnorm

3957e3a8

mindest marked this pull request as ready for review 3 years ago

mindest requested a review from

kailums 3 years ago

mindest changed the title ~~[WIP] Add ROCm Triton TunableOp for GroupNorm~~ [ROCm] Add ROCm Triton TunableOp for GroupNorm 3 years ago

mindest requested a review from

zhangyaobit 3 years ago

mindest requested a review from

codego7250 3 years ago

kailums dismissed these changes on 2023-06-26

mindest dismissed their stale review via e9c49550 3 years ago

mindest force pushed from 5e83a14d to 48c8eeea 3 years ago

reduce all axes; minor refactor.

3216a047

mindest force pushed from 48c8eeea to 3216a047 3 years ago

kailums approved these changes on 2023-06-30

mindest merged 347c963d into main 3 years ago

mindest deleted the linmin/triton_groupnorm branch 3 years ago

Reviewers

kailums

tianleiwu

github-advanced-security

zhangyaobit

codego7250

Assignees

No one assigned

Labels

None yet

Milestone

No milestone

onnxruntime [ROCm] Add ROCm Triton TunableOp for GroupNorm #16196 Merged

[ROCm] Add ROCm Triton TunableOp for GroupNorm #16196

onnxruntime
[ROCm] Add ROCm Triton TunableOp for GroupNorm
#16196

Merged