Commits microsoft/onnxruntime

Add clarifying notes for compensated ReduceSum fix

Copilot committed 103 days ago

Verified 1d0f58f4

Fix float32 ReduceSum stability for large full reductions

Copilot committed 103 days ago

Verified deb25252

Initial plan

Copilot committed 104 days ago

Verified 270f80ea

Derive ORT_TEST_VERBOSE from System.Debug via runtime coalesce in plugin test pipelines (#28517)

edgchen1 committed 104 days ago

Verified 13630176

Add quantized KV cache support in CPU GroupQueryAttention (#28576)

tianleiwu committed 104 days ago

Verified 6fbda6d8

add more ORT_ENFORCE when building the tree structure of TreeEnsemble operators (#27677)

xadupre committed 104 days ago

Verified f6e8b0b8

QMoE CUDA EP — FP4/FP8/WFP4AFP8 Quantized Mixture-of-Experts + MoE GEMM Refactor (#28467)

tianleiwu committed 104 days ago

Verified 548ab6e5

[MLAS] Correcting Compilation Errors: onnxruntime/onnxruntime/core/mlas/lib/riscv64/sconv_depthwise_kernel_rvv.cpp:138:18: error: unused variable 'pad_bottom' (#28506)

zejianzhang1982 committed 104 days ago

Verified 3a071a67

Fix Cargo registry auth for network-isolated Linux Python wheel builds (#28556)

kpkbandi committed 104 days ago

Verified a48b4982

Fix protoc error when cross compiling on a linux aarch64 host (#28567)

jocelyn-stericker committed 104 days ago

Verified 192ef6b8

Add EP and hardware device type to Windows ML telemetry (#28477)

angelser committed 104 days ago

Verified 4d1dce86

[MLAS] test: accept canonical NaN in activation NaN round-trip check (#28538)

qiurui144 committed 104 days ago

Verified b38731bf

Validate sparse tensor external file paths (#28408)

adrianlizarraga committed 104 days ago

Verified 6a8f021b

Fix XNNPACK Gemm SIGSEGV on missing and scalar 'C' bias (#28546)

dparikh79 committed 104 days ago

Verified 8c650d81

Fix int32 overflow in CUDA Cast and UnaryElementWise kernels for tensors with >2^31 elements (#28386)

Copilot committed 104 days ago

Verified cd05ef40

[Security] Validate CPU cache_indirection beam indices (#28486)

tianleiwu committed 104 days ago

Verified 2d9f35fc

Skip file path validation in CreateEpContextModel when the output is not a file (#28431)

shiyi9801 committed 105 days ago

Verified ce6aaeec

Add unit test for fix of plugin EP library handle leak in PR 28396 (#28430)

adrianlizarraga committed 105 days ago

Verified 40e30488

Support kquant mixed for whisper export (#28515)

jiafatom committed 105 days ago

Verified d2453bc3

contrib: add mask/input shape consistency checks in MaxpoolWithMask::Compute (#28223)

Copilot committed 105 days ago

Verified 9883e93d

Reject recursive local function definitions during model load (#28187)

tianleiwu committed 105 days ago

Verified 22e537b8

[WebGPU plugin EP] Package improvements (#28525)

edgchen1 committed 105 days ago

Verified 770da7dd

webgpu: Generalize FlashAttention prefill shared-memory path (#28520)

qjia7 committed 106 days ago

Verified aa416f5c

fix: avoid vector copies in CheckIfSubtreesAreEqual (#27854)

lhrios committed 106 days ago

Verified 4f620c38

Fix Inverse kernel rank underflow before indexing trailing dims (#28400)

Copilot committed 108 days ago

Verified b429dff3

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

hariharans29 committed 108 days ago

Verified eb63cc6b

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

rvandermeulen committed 108 days ago

Verified 45c06632

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

qti-ashwshan committed 108 days ago

Verified ad6d34c6

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

vraspar committed 108 days ago

Verified 00e9ef3d

Fix WeaklyCanonicalPath ERROR_ACCESS_DENIED in Windows AppContainers (#28509)

bsosnader committed 108 days ago

Verified ddd1eae2