microsoft/onnxruntime

Pull Requests Commits

add flush and wait for completion support for webgpu (#33181)

aciddelgado committed 4 days ago

Verified 2a8e1f0c

Bump reviewdog/action-cpplint from 1.11.0 to 1.12.0 (#33190)

dependabot[bot] committed 4 days ago

Verified 62a06e40

Bump gradle/actions/wrapper-validation from 5.0.2 to 6.4.0 (#33191)

dependabot[bot] committed 4 days ago

Verified 4a43f7ca

Bump next from 15.5.24 to 15.5.27 in /js/web/test/e2e/exports/testcases/nextjs-default (#33194)

dependabot[bot] committed 4 days ago

Verified f7cc86ce

Bump version to 1.32.0 (#33197)

tianleiwu committed 4 days ago

Verified 768987fe

[Telemetry] Default native builds to 1DS (#32384)

bmehta001 committed 4 days ago

Verified b131c79b

[CUDA] Add opt-in wave-aware fpA_intB GEMV dispatch (#33186)

tianleiwu committed 4 days ago

Verified cd50f9d0

[WebGPU] Add block-scaled FP8 QMoE (#32794)

sayanshaw24 committed 4 days ago

Verified a571b726

Promote model package API to stable (#33166)

xiaoyu-work committed 4 days ago

Verified 5df65067

Publish CUDA13 jars, remove unnecessary parameter, add jni tests to testing package. (#33131)

eserscor committed 4 days ago

Verified 3ddd2d87

Skip CUDA MoE counting tests on pre-SM80 GPUs (#33180)

Copilot committed 4 days ago

Verified 2e076d52

[WebGPU EP] implements chunked gated delta net optimization (#33143)

prathikr committed 4 days ago

Verified ce9bb081

Preserve ONNX element types when loading node tests (#33168)

apsonawane committed 4 days ago

Verified 9f6baac7

Fix optional OpenVINO provider probe without exceptions (#33112)

mustjab committed 4 days ago

Verified 6ce927e6

Fix broken link to ONNX Runtime technical reference docs (#33030)

Zapakio6 committed 4 days ago

Verified 397657de

[CUDA] Fix GQA build regression after eligibility/workspace integration (#33179)

chilo-ms committed 4 days ago

Verified 05517e12

Update Android minimal binary size baseline (#33164)

xadupre committed 4 days ago

Verified 3d9d664a

[CUDA] Plumb GQA workspace KV-length envelope knob into the L1 estimator (reader-only) (#33014)

chilo-ms committed 5 days ago

Verified f83bf57d

[CUDA] Consolidate GQA scratch sizing (MEA, unfused, XQA, Flash) into shared workspace recipes (#32945)

chilo-ms committed 5 days ago

Verified e3f71e01

[CUDA] Extract GQA seq-free route eligibility into shared helpers (#33001)

chilo-ms committed 5 days ago

Verified ec82e2a3

[CUDA] Suppress CUDA 13 MSVC build warnings (#33149)

tianleiwu committed 5 days ago

Verified 0eb8a047

Fix dead links in docs and CONTRIBUTING (#32880)

pratikgx committed 5 days ago

Verified 9a7c1939

Validate ORT format input argument counts (#32865)

danielsongmicrosoft committed 5 days ago

Verified 958927fb

Account for string payloads in initializer size limits (#32867)

danielsongmicrosoft committed 5 days ago

Verified 0d427f8c

Handle missing ConstantOfShape value during float16 conversion (#33089)

Copilot committed 5 days ago

Verified ae245d03

Fix ReshapeFusion crash when fused Reshape's shape input is node-produced (#33091)

anubhavagr committed 5 days ago

Verified 5652d3fa

fix(quantization): avoid intermediate overflow when calculating quantization ranges (#33104)

galleonli committed 5 days ago

Verified b40377d9

Optimize CPU BatchNormalization for NC inputs (#33113)

artyomrabosh committed 5 days ago

Verified 8478e365

Bump sharp from 0.35.4 to 0.35.5 in /js/web/test/e2e/exports/testcases/nextjs-default (#33138)

dependabot[bot] committed 5 days ago

Verified 43c6f17e

Fix Windows clang/no-exceptions portability (#33140)

mustjab committed 5 days ago

Verified 84207093

Newer Older