MatMulNBits + Add fusion #20587
edgchen1
merged
30 commits into
main
from
edgchen1/matmul_nbits_bias
save work
350f3e20
handle bias in fallback impl
752b7c95
reorder includes
4121ff44
save work - initial impl of optimizer
96fa18b4
Merge remote-tracking branch 'origin/main' into edgchen1/matmul_nbits…
11c3b922
Add test, refine impl.
baab944a
Merge remote-tracking branch 'origin/main' into edgchen1/matmul_nbits…
15687585
fix fusion, update test
6402ea5d
Enable bias in SQNBitGemm benchmark.
6a1feb28
clean up
a6c42e41
move test functions around to avoid unused function warnings
7fa290b7
add /bigobj for graph_tarnsform_test.cc
b5057eb4
lint
823c9db2
Add Neural Speed build to post-merge-jobs.yml.
7555f92f
move graph_utils::IsSupportedProvider() in header so it is available …
10ba1db0
put MatMulNBitsBiasFusion test in contrib ops ifdef
471429c3
update ContribOperators.md
7f07d93c
call RunWithConfig instead of Run
97a15a12
update OperatorKernels.md
905cbd22
add matmul_nbits_fusion files to extended minimal build source
0abc2cd7
Merge remote-tracking branch 'origin/main' into edgchen1/matmul_nbits…
1a4b5a0f
Make DML MatMulNBits input check accomodate new inputs.
50cb266d
expect input arg counts of 0 for missing optional inputs
131141f3
Merge branch 'edgchen1/matmul_nbits_bias' of github.com:microsoft/onn…
2417b802
add runtime optimization test for matmulnbits and add fusion
1d9c2f50
github-advanced-security
commented
on 2024-05-10
github-advanced-security
commented
on 2024-05-10
add test onnx file
afacb57c
lint
55851e33
github-advanced-security
commented
on 2024-05-10
yufenglee
commented
on 2024-05-14
yufenglee
commented
on 2024-05-14
Add !ORT_NEURAL_SPEED ifdefs around adding MatMulNBitsFusion transfor…
badcd0fa
add ifdef around test
b90c8cab
yufenglee
dismissed these changes
on 2024-05-15
Merge branch 'main' into edgchen1/matmul_nbits_bias
92ce6ebb
skottmckay
approved these changes
on 2024-05-16
Reviewers
Assignees
No one assigned
Labels
None yet
Milestone
No milestone
Login to write a write a comment.
Login via GitHub