onnxruntime
Add CPU QMoE 2-bit support and LUT GEMM fast path
#28185
Merged

Add CPU QMoE 2-bit support and LUT GEMM fast path #28185

tianleiwu merged 25 commits into main from tlwu/qmoe_2bit_cpu
tianleiwu
tianleiwu Add 2 bit QMoE
b7e9fd80
tianleiwu Add LuT GEMM for 2 bits
ae555a3e
tianleiwu Add doc
16ca1f98
tianleiwu tianleiwu marked this pull request as draft 123 days ago
tianleiwu Add doc gen
620d057f
tianleiwu upload docs
c4597096
tianleiwu Merge remote-tracking branch 'origin/main' into tlwu/win_gpu_doc_gen_…
c0596535
tianleiwu Merge remote-tracking branch 'origin/main' into tlwu/qmoe_2bit_cpu
e1f24368
tianleiwu Merge remote-tracking branch 'origin/tlwu/win_gpu_doc_gen_workflow' i…
9eeb2775
tianleiwu Merge remote-tracking branch 'origin/main' into tlwu/qmoe_2bit_cpu
1825452d
tianleiwu tianleiwu requested a review from copilot-pull-request-reviewer copilot-pull-request-reviewer 112 days ago
tianleiwu tianleiwu requested a review from apsonawane apsonawane 112 days ago
copilot-pull-request-reviewer
copilot-pull-request-reviewer commented on 2026-05-03
tianleiwu fix: address review feedback for QMoE 2-bit CPU support
dcb3a68f
github-actions
github-actions commented on 2026-05-03
tianleiwu lintrunner
e7f814fa
tianleiwu tianleiwu requested a review from copilot-pull-request-reviewer copilot-pull-request-reviewer 112 days ago
copilot-pull-request-reviewer
copilot-pull-request-reviewer commented on 2026-05-03
tianleiwu address feedbacks
07d4d43a
tianleiwu tianleiwu requested a review from copilot-pull-request-reviewer copilot-pull-request-reviewer 112 days ago
copilot-pull-request-reviewer
copilot-pull-request-reviewer commented on 2026-05-03
tianleiwu update comment
ecc98f01
tianleiwu tianleiwu marked this pull request as ready for review 112 days ago
tianleiwu tianleiwu requested a review from vraspar vraspar 111 days ago
apsonawane
apsonawane commented on 2026-05-05
vraspar
vraspar commented on 2026-05-05
vraspar
vraspar commented on 2026-05-05
vraspar
vraspar commented on 2026-05-05
vraspar
vraspar commented on 2026-05-05
vraspar
vraspar commented on 2026-05-05
vraspar
vraspar commented on 2026-05-05
tianleiwu fix: address review feedback for 2-bit QMoE CPU support
6acd4a6e
tianleiwu tianleiwu requested a review from copilot-pull-request-reviewer copilot-pull-request-reviewer 110 days ago
tianleiwu tianleiwu requested a review from vraspar vraspar 110 days ago
copilot-pull-request-reviewer
copilot-pull-request-reviewer commented on 2026-05-05
tianleiwu fix: align dequant block size to zp_pack_size for row-wise asymmetric…
3185da90
tianleiwu test: add regression test for 2-bit row-wise asymmetric ZP alignment
c14f41da
tianleiwu tianleiwu requested a review from copilot-pull-request-reviewer copilot-pull-request-reviewer 110 days ago
tianleiwu tianleiwu requested a review from apsonawane apsonawane 110 days ago
tianleiwu test: fix QMoETest_CPU_Int2_MLAS missing swiglu_fusion attribute
bfd89717
copilot-pull-request-reviewer
copilot-pull-request-reviewer commented on 2026-05-06
tianleiwu fix: correct test comment and doc block-wise shape formulas
5da077fe
tianleiwu Merge remote-tracking branch 'origin/main' into tlwu/qmoe_2bit_cpu
3511b177
tianleiwu update doc
10370dcf
tianleiwu tianleiwu requested a review from copilot-pull-request-reviewer copilot-pull-request-reviewer 109 days ago
copilot-pull-request-reviewer
copilot-pull-request-reviewer commented on 2026-05-06
vraspar
vraspar commented on 2026-05-06
tianleiwu fix(cpu): remove redundant MlasInitLutGemmKernelConfig in TryRunLutGemm
b380eb3c
tianleiwu tianleiwu requested a review from vraspar vraspar 108 days ago
tianleiwu update doc
ba731ad0
vraspar
vraspar commented on 2026-05-11
vraspar
vraspar commented on 2026-05-11
vraspar
vraspar commented on 2026-05-11
vraspar
vraspar commented on 2026-05-11
vraspar
vraspar dismissed these changes on 2026-05-11
tianleiwu address feedbacks
2be976a2
tianleiwu tianleiwu dismissed their stale review via 2be976a2 104 days ago
tianleiwu tianleiwu requested a review from vraspar vraspar 104 days ago
tianleiwu tianleiwu requested a review from copilot-pull-request-reviewer copilot-pull-request-reviewer 104 days ago
copilot-pull-request-reviewer
copilot-pull-request-reviewer commented on 2026-05-11
tianleiwu Merge remote-tracking branch 'origin/main' into tlwu/qmoe_2bit_cpu
b0de83db
tianleiwu Address feedbacks
7830ec38
vraspar
vraspar approved these changes on 2026-05-12
tianleiwu tianleiwu enabled auto-merge (squash) 103 days ago
tianleiwu tianleiwu merged dac5a143 into main 103 days ago
tianleiwu tianleiwu deleted the tlwu/qmoe_2bit_cpu branch 103 days ago

Login to write a write a comment.

Login via GitHub

Assignees
No one assigned
Labels
Milestone