microsoft/onnxruntime

Pull Requests Commits

Address CUDA GQA validation feedback

Copilot committed 23 days ago

Verified 4a14d910

Fix CUDA GQA graph replay test

Copilot committed 23 days ago

Verified 8af50060

Copilot committed 23 days ago

Verified 5817f62e

[CUDA] Make GatherND capture-safe (#32120)

tianleiwu committed 23 days ago

Verified 71b33460

Specify GroupQueryAttention windowed KV cache layout contract (#32419)

tianleiwu committed 23 days ago

Verified cd9da083

[CUDA] Add cuDNN paged SDPA decode tier for PagedAttention (#32493)

hariharans29 committed 23 days ago

Verified 4725a200

Add FP8/FP4 block-quantized support to GatherBlockQuantized contrib op (#32480)

Copilot committed 23 days ago

Verified aad07ef2

Suppress preload errors for optional CUDA libraries (#32112)

Copilot committed 23 days ago

Verified 84d758f8

Skip compiling Flash Attention .cu sources when onnxruntime_USE_FLASH_ATTENTION=OFF (#32580)

Copilot committed 23 days ago

Verified e8787e28

Add Qwen4-Exp EngramGate and NGramHashMapping support (#32285)

Copilot committed 23 days ago

Verified d71d2789

[MLAS] Fix Inf/NaN comparison and add fp16 extreme-value coverage for ARM64 HGEMM NEON (#31761)

kjg0724 committed 23 days ago

Verified 6f6e8815

Add GQA XQA and Flash workspace recipes (#32453)

titaiwangms committed 23 days ago

Verified 8ed3bea2

Start transitioning pools to unified Windows pool (#32612)

eserscor committed 23 days ago

Verified 1474dfc6

Merge upstream main for #32239 maintenance

justinchuby committed 23 days ago

ab9e50ac

[WebGPU] Fix OOB store in 8x16x16 subgroup-matrix MatMulNBits (#32582)

jchen10 committed 23 days ago

Verified 3578d771

Add contributor sorting option to compile_contributors.py (#32549)

tianleiwu committed 24 days ago

Verified c96cf8de

Enable WebGPU GQA local-window flash attention (#32462)

feich-ms committed 24 days ago

Verified fed7f01a

[WebGPU] Vectorize NHWC depthwise convolution across channels (#32498)

linzuojian committed 24 days ago

Verified 73534a26

[WebGPU EP] implements VarlenCausalConvWithState operator (#32477)

prathikr committed 24 days ago

Verified d6750b39

[CUDA] Fix SM120 QMoE MoE GEMM TMA build failure with GCC (#32584)

bheu committed 24 days ago

Verified 5f44b5b5

[Telemetry] Reduce 1DS event volume (#32425)

bmehta001 committed 24 days ago

Verified 0a82f2ca

Deprecate ACL EP (#32564)

edgchen1 committed 24 days ago

Verified 719b9fa3

Bump core mltools: 7.1 -> 9.0 (#32586)

eserscor committed 24 days ago

Verified 154c8218

Move kCutlassSafeMaskFilterValue outside USE_MEMORY_EFFICIENT_ATTENTION guard (#32588)

SIDDARTHAREDDY8 committed 24 days ago

Verified bf8b4f1a

[CUDA] Preserve graph replay test includes after merge

justinchuby committed 24 days ago

3341e351

[CUDA] Fix BNSH indexing in FlashDecode graph replay test

justinchuby committed 24 days ago

8e2d9dbc

Merge latest main into CUDA graph-safe FlashDecode branch

justinchuby committed 24 days ago

0f2d109f

Add GQA workspace preparation recipes (#32446)

titaiwangms committed 24 days ago

Verified b03fb522

Auto-generated baselines by 1ES Pipeline Templates (#31136)

microsoft-github-policy-service[bot] committed 24 days ago

Verified dfd2228b

[WebGPU] Support QuickGelu activation in the im2col-matmul path (#32560)

xhcao committed 25 days ago

Verified 8d85527a

Older