microsoft/onnxruntime

Pull Requests Commits

MatMul: clarify inline-batch comment for DISABLE_ABSEIL build

Gopalakrishnan Nallasamy committed 18 hours ago

37c9dcd1

MatMul: use inlined_containers_fwd.h and shared constexpr for inline batch cutoff

Gopalakrishnan Nallasamy committed 19 hours ago

6fa54e4c

Avoid small MatMul batch parameter heap allocations

Gopalakrishnan Nallasamy committed 20 days ago

5a6196f1

[WebGPU] Make max_num_pending_dispatches configurable (#28894)

Reranko05 committed 20 days ago

Verified 57f6079d

Remove redundant ORT FlatBuffer table offset validation (#29068)

tianleiwu committed 20 days ago

Verified 964cf4d7

Integrate ONNX 1.22.0 (opset 27) — issue #28752 (#28754)

titaiwangms committed 20 days ago

Verified 43fd961a

Single-source OrtErrorCode values to keep StatusCode in sync (#29065)

edgchen1 committed 20 days ago

Verified cbb74576

Add python implementation for wgsl-gen (#28355)

danielsongmicrosoft committed 20 days ago

Verified 9c12e9d0

Fix build failure from direct ONNX schema include in layernorm optimizer tests (#29073)

JonathanC-ARM committed 20 days ago

Verified 291bd33a

Encapsulate global random seed in a function (#28979)

raagrawal committed 20 days ago

Verified 6f691bfe

Bump protobufjs from 7.6.0 to 7.6.3 in /js/web (#29061)

dependabot[bot] committed 20 days ago

Verified c2126deb

Bump tar from 7.5.15 to 7.5.16 in /js/node (#29063)

dependabot[bot] committed 20 days ago

Verified 84be9344

Bump js-yaml from 4.1.1 to 4.2.0 in /js (#29062)

dependabot[bot] committed 20 days ago

Verified 6ae00279

webgpu: use naive reduction (#28174)

xhcao committed 21 days ago

Verified a534347f

Harden ORT FlatBuffer model loader against malformed buffers (#28186)

tianleiwu committed 21 days ago

Verified 9cacffa6

Bump tmp from 0.2.6 to 0.2.7 in /onnxruntime/test/wasm (#29055)

dependabot[bot] committed 21 days ago

Verified 693e090e

Add 2-bit quantization support to WebGPU GatherBlockQuantized operator (#29054)

Shivani767 committed 21 days ago

Verified 5c3c2b70

Fix OOB pointer in TensorAt for sub-byte packed types (#28973)

neilmsft committed 21 days ago

Verified f490d892

Bump WebGPU plugin EP version number to 0.3.0 (#29056)

edgchen1 committed 21 days ago

Verified 981b4757

Fix WebGPU GatherBlockQuantized dispatch failure for empty indices (#29030)

tairenpiao committed 21 days ago

Verified f67f7482

[WebGPU] Fix nearest half-tie handling for round_prefer_ceil/floor (#28757)

haoxli committed 21 days ago

Verified eabc40d9

[WebGPU] Enable Cast to int64 by default (#28804)

fanchenkong1 committed 21 days ago

Verified f0da3c4e

[WebNN EP] Support 8-bit MatMulNBits (#29025)

Honry committed 21 days ago

Verified c5aa86e1

Bump esbuild from 0.25.12 to 0.28.1 in /js (#29044)

dependabot[bot] committed 21 days ago

Verified bc226539

[CUDA] QMoE GEMV fast path for batch-1 decode (#29038)

tianleiwu committed 21 days ago

Verified 56f6fee6

Update protobuf to mitigate CVE-2026-0994 (#28967)

sushraja-msft committed 21 days ago

Verified 7e5c63ff

[WebNN EP] Remove unnecessary cast around normalization ops (#29024)

Honry committed 23 days ago

Verified bf7e27a4

Lower CUDA plugin EP minimum ORT version to 1.24.4 with version-gated callbacks (#28824)

tianleiwu committed 24 days ago

Verified 33b389aa

Add conditional logic to improve functionality (#29015)

apsonawane committed 24 days ago

Verified ed7a29b5

fix: NodeJS pkging stage needs to use CFS (#29006)

sanaa-hamel-microsoft committed 24 days ago

Verified d6e4a2d2

Older