onnxruntime
Add batched and strided batched gemm as TunableOp
#13841
Merged
Overview
Commits
18
Changes
View On GitHub

Add batched and strided batched gemm as TunableOp #13841

cloudhan merged 18 commits into main from guangyunhan/more-tunable-gemm
cloudhan
cloudhan cloudhan marked this pull request as ready for review 3 years ago
cloudhan cloudhan requested a review from zhangyaobit zhangyaobit 3 years ago
cloudhan cloudhan requested a review from PeixuanZuo PeixuanZuo 3 years ago
cloudhan cloudhan requested a review from abudup abudup 3 years ago
cloudhan
cloudhan commented on 2022-12-06
abudup
abudup commented on 2022-12-06
cloudhan cloudhan force pushed from e77b2e7d to 5a18f264 3 years ago
abudup
abudup commented on 2022-12-14
cloudhan cloudhan force pushed from 5a18f264 to e4206e3b 3 years ago
PeixuanZuo
PeixuanZuo commented on 2022-12-20
cloudhan cloudhan force pushed from be7ef039 to 5ebd02ed 3 years ago
cloudhan cloudhan force pushed from 5ebd02ed to cb1d1b89 3 years ago
cloudhan
abudup
abudup dismissed these changes on 2023-01-04
cloudhan cloudhan dismissed their stale review via f46305a3 3 years ago
cloudhan cloudhan force pushed from cb1d1b89 to f46305a3 3 years ago
cloudhan cloudhan requested a review from abudup abudup 3 years ago
abudup
abudup dismissed these changes on 2023-01-05
cloudhan Add tunable strided batch gemm composed from rocblas and ck
28b52abe
cloudhan Adjust gemm_test.py to be unified with strided_batched_gemm_test.py
438bc889
cloudhan Add batched gemm tunable op
b24de039
cloudhan Remove staled TODOs
51e339a9
cloudhan Use tunable StridedBatchedGemm for MatMul
d488f104
cloudhan Switch remaining rocblasGemmHelper usage in Gemm to tunable Gemm
4ee27326
cloudhan Switch contrib ops to use tunable gemms
c412f5eb
cloudhan Fix batched gemm for ke
94b45323
cloudhan Add tests for alpha and beta
ec6bc368
cloudhan Update after rebase onto main
c52bafdc
cloudhan Fix size calculation
433f90b3
cloudhan Fix typo
09c7e594
cloudhan Factor out the `constexpr if`s
9bace1eb
cloudhan Fix
af875bdc
cloudhan Minor
eb34e43c
cloudhan Use new API after rebase
7a6bf14f
cloudhan Enable all ck tests after ck update
5b439ab5
cloudhan Improve test speed for all gemms
19107853
cloudhan cloudhan dismissed their stale review via 19107853 3 years ago
cloudhan cloudhan force pushed from f46305a3 to 19107853 3 years ago
cloudhan cloudhan requested a review from PeixuanZuo PeixuanZuo 3 years ago
PeixuanZuo
PeixuanZuo approved these changes on 2023-01-06
cloudhan cloudhan merged be879c11 into main 3 years ago
cloudhan cloudhan deleted the guangyunhan/more-tunable-gemm branch 3 years ago

Login to write a write a comment.

Login via GitHub

Reviewers
Assignees
No one assigned
Labels
None yet
Milestone
No milestone
Terms Of Use Privacy Policy Refund Policy Impressum