Skip to content

Pull requests: microsoft/onnxruntime

New pull request New
661 Open 19,195 Closed
661 Open 19,195 Closed
Author
Filter by author
Loading
Label
Filter by label
Loading
Use alt + click/return to exclude labels
or ⇧ + click/return for logical OR
Projects
Filter by project
Loading
Milestones
Filter by milestone
Loading
Reviews
Filter by reviews
No reviews Review required Approved review Changes requested
Assignee
Filter by who’s assigned
Assigned to nobody Loading
Sort
Sort by
Newest Oldest Most commented Least commented Recently updated Least recently updated Best match
Most reactions
👍 👎 😄 🎉 😕 ❤️ 🚀 👀

Pull requests list

upgrades adm-zip for onnxruntime-node
#31192 opened Jul 30, 2026 by prathikr Contributor Loading…
[CUDA] Port ShardedMoE to the new MoE GEMM backend
#31191 opened Jul 30, 2026 by tianleiwu Contributor • Draft
Enforce K/V shape consistency in 3D Attention [CPU]
#31190 opened Jul 30, 2026 by nenad1002 Contributor Loading…
3
[XNNPACK EP] Read Gemm M from the input tensor at Compute time
#31189 opened Jul 30, 2026 by guptaishaan Loading…
1
1
Fix shape inference failure with file-backed external data
#31169 opened Jul 30, 2026 by yoonseok-kim Loading…
1
2
[CUDA] Speed up the NVFP4 QMoE decode GEMV and enable it for MTP verify
#31159 opened Jul 30, 2026 by tianleiwu Contributor • Draft
3
[CUDA] Add LinearAttentionGate and GatedRMSNorm contrib operators
#31158 opened Jul 30, 2026 by tianleiwu Contributor • Draft
2
[CUDA/CPU/WebGPU] Add state_window to LinearAttention and CausalConvWithState
#31157 opened Jul 30, 2026 by tianleiwu Contributor • Draft
[CUDA] Cut the MoE routing prologue/epilogue cost for small-batch decode
#31156 opened Jul 30, 2026 by tianleiwu Contributor Loading…
2
[CUDA] Block-scaled MatMul decode GEMV: tensor cores, packed FP4 decode and M-tiling
#31155 opened Jul 30, 2026 by tianleiwu Contributor Loading…
3
[CUDA] QMoE: release raw MXFP4 weight initializers after PrePack
#31154 opened Jul 30, 2026 by tianleiwu Contributor Loading…
1
[WebNN EP] Fix MatMulNBits when K is not a multiple of block_size
#31152 opened Jul 30, 2026 by Honry Contributor Loading…
2
[WebNN EP] Reuse shared WASM loader for Blob-backed external data
#31151 opened Jul 30, 2026 by Honry Contributor Loading…
2
Skip redundant present_key/value copy when aliased to external KV cache (follow-up to #29715) ep:CUDA issues related to the CUDA execution provider
#31150 opened Jul 30, 2026 by titaiwangms Contributor Loading…
2
Throw error on negative Split tensor axis [CPU]
#31149 opened Jul 30, 2026 by nenad1002 Contributor Loading…
2
Fix LayerNorm fusion selecting the Div output as scale or bias
#31148 opened Jul 29, 2026 by Sammy-Dabbas Contributor Loading…
1
1
[ARM] MLAS: SVE i8mm (svmmla) int8 QGEMM kernels, portable machine code
#31146 opened Jul 29, 2026 by mirounga Contributor • Draft
3
[ARM] MLAS: portable machine-code SVE elementwise kernels, FEXPA exp
#31145 opened Jul 29, 2026 by mirounga Contributor • Draft
Do not fuse LayerNorm when ReduceMean has keepdims=0
#31144 opened Jul 29, 2026 by guptaishaan Loading…
1
4
[ARM] MLAS: route fp32 SGEMM to the KleidiAI SVE ukernel on SVE-only
#31143 opened Jul 29, 2026 by mirounga Contributor • Draft
Fix segfault in MatMulNBits prepacking when weights/scales are parent-graph initializers in subgraph
#31141 opened Jul 29, 2026 by Copilot AI Loading…
1
@tianleiwu
10
[OpenVINO EP] Fix get/set_float_initializer_data for raw_data backed float initializers
#31138 opened Jul 29, 2026 by wangw-1991 Contributor Loading…
14
Auto-generated baselines by 1ES Pipeline Templates
#31136 opened Jul 29, 2026 by microsoft-github-policy-service Bot Loading…
1
[WebGPU EP] Support int64 for Tile and Concat ep:WebGPU ort-web webgpu provider
#31049 opened Jul 29, 2026 by miaobin Contributor Loading…
4
Build onnxruntime_providers as an object library for shared-lib builds
#30511 opened Jul 28, 2026 by rvandermeulen Contributor Loading…
1
ProTip! Add no:assignee to see everything that’s not assigned.