onnxruntime
[webgpu] Optimize matmulnbits with M > 1
#23102
Merged

[webgpu] Optimize matmulnbits with M > 1 #23102

guschmue merged 6 commits into microsoft:main from qjia7:webgpu_matmulnbits
qjia7
qjia7 [webgpu] Optimize matmulnbits with M > 1
d30cf803
qjia7
sushraja-msft
guschmue
guschmue
guschmue
azure-pipelines
guschmue
guschmue
azure-pipelines
azure-pipelines
azure-pipelines
guschmue guschmue added ep:WebGPU
github-advanced-security
github-advanced-security commented on 2024-12-14
qjia7 Remove MatMulNBitsProgramPrefill
a349ad4f
qjia7 remove components_a limitation
a7a7d9b7
qjia7 make tile_m as class member
be81377e
qjia7 merge MatMulNBitsWithLargeMProgram to MatMulNBitsProgram
d6277ea1
qjia7 set tile M threshold
ca8ef7ab
qjia7
guschmue
guschmue
guschmue
azure-pipelines
guschmue
azure-pipelines
azure-pipelines
azure-pipelines
guschmue
azure-pipelines
guschmue
guschmue approved these changes on 2024-12-17
guschmue guschmue merged 0981bbf4 into main 1 year ago

Login to write a write a comment.

Login via GitHub

Assignees
No one assigned
Labels
Milestone