Commits microsoft/onnxruntime

Add ms-gsl version>=4.2.1 constraint in cmake/vcpkg.json

Copilot committed 2 days ago

Verified 4f190849

Update microsoft_gsl from v4.0.0 to v4.2.1 and fix GSL_SUPPRESS C4875 warning

Copilot committed 2 days ago

Verified 1d55d854

Initial plan

Copilot committed 2 days ago

Verified a6020fd5

[MLAS] Fix scratch buffer over-allocation in batched/grouped NCHW Conv (#28482)

hariharans29 committed 2 days ago

Verified eb63cc6b

Fix mlasi_sve.h preprocessor guards to allow clang compilation (#28507)

rvandermeulen committed 2 days ago

Verified 45c06632

[QNN EP] ETW log level rule change (#27593)

qti-ashwshan committed 2 days ago

Verified ad6d34c6

Fix input validation and null-pointer dereference in STFTDecomposition graph transformer (#28465)

vraspar committed 2 days ago

Verified 00e9ef3d

Fix WeaklyCanonicalPath ERROR_ACCESS_DENIED in Windows AppContainers (#28509)

bsosnader committed 2 days ago

Verified ddd1eae2

Build issues on AIX for POWER10 and POWER11 (#26704)

KV2773 committed 3 days ago

Verified 127faf0c

Fill LSTM CUDA operator opset gap: extend coverage from opset 14 to opset 22 (#27737)

Copilot committed 3 days ago

Verified c0b3212f

Validate per-column weight_scale/weight_zero_point shape in CPU QAttention; harden integer arithmetic in QAttention and AttentionBase (#28480)

edgchen1 committed 3 days ago

Verified 18ae6da8

enable dynamic max_k_step in FA for nvidia (#28511)

guschmue committed 3 days ago

Verified b36d8dbd

[MLAS] Add an NHWC implementation of convolution to avoid transposes (#26834)

orlmon01 committed 3 days ago

Verified a8ba94a1

Improve error reporting for pre-allocated outputs with a wrong shape (#28481)

yuslepukhin committed 3 days ago

Verified 70dcc0e2

fix(security): add SafeInt overflow protection in Expand and constant folding output size limit (#28055)

tianleiwu committed 3 days ago

Verified 04f74403

fix: support N-D weights with unit leading dims in MatMulNBitsQuantizer (#28178)

Rishi-Dave committed 3 days ago

Verified 2979bab1

Fix out-of-bounds read in CropBase: validate scale attribute length (#28399)

Copilot committed 3 days ago

Verified 50d04b92

CUDA Plugin Packaging Pipeline Fix TSAUpload and Timeout (#28497)

tianleiwu committed 3 days ago

Verified dcf87757

fix: auto-upgrade model opset to 21 for int16/uint16 QDQ quantization (#28202)

Rishi-Dave committed 3 days ago

Verified 0ebd8cc9

Bump next from 16.1.5 to 16.1.7 in /js/web/test/e2e/exports/testcases/nextjs-default (#27705)

dependabot[bot] committed 3 days ago

Verified 4d9aa47a

Address string attribute issue (#28494)

yuslepukhin committed 4 days ago

Verified 4bc20949

Optimize FlashAttention for M4 Max (20x speedup) (#27780)

xenova committed 4 days ago

Verified 938b6075

Address latent bug in LabelEncoder (#28496)

yuslepukhin committed 4 days ago

Verified 5f729560

Fill opset gap for RandomNormal, RandomNormalLike, RandomUniform, RandomUniformLike CUDA ops with BFloat16 support (#27759)

Copilot committed 4 days ago

Verified 0139499a

Fixes for unicode path handling (#28390)

jnagi-intel committed 4 days ago

Verified 9b3fb24a

Add float zero point support for 2-bit LUT GEMM in MatMulNBits (#28354)

vraspar committed 4 days ago

Verified e8ae6ce9

Bound total output allocation size in Tile kernel (#28070)

GopalakrishnanN committed 4 days ago

Verified a363e895

Fix OOB reads in SoftmaxCrossEntropyLoss via label bounds validation (#28004)

vraspar committed 4 days ago

Verified ad11f7cf

Use weights_only=True for remaining torch.load() calls (#28421)

adrianlizarraga committed 4 days ago

Verified 4e377a13

Add LabelEncoder_4 fast path for session creation (#27526)

SamuelLess committed 5 days ago

Verified 73d54165