microsoft/onnxruntime

Pull Requests Commits

Address pr review comments

baijumeswani committed 6 hours ago

4b3e8910

Support LoRA adapters with plugin EP allocators

baijumeswani committed 6 hours ago

8f679d3e

Fix C API docs Doxygen download (#32210)

Wayne-Ch committed 1 day ago

Verified df012c2f

[ARM] MLAS: SVE i8mm (svmmla) int8 QGEMM kernels, portable machine code (#31146)

mirounga committed 1 day ago

Verified a4156be6

[MLAS] Add a NEON fused kernel for LinearAttention (#32178)

mirounga committed 1 day ago

Verified 27f212c4

[WebGPU] Regenerate WGSL in-tree golden fixtures to match current templates (#32115)

4n4ny4 committed 1 day ago

Verified 775526b8

[WebGPU] Make MatMulNaiveProgram's pipeline cache key cover everything it bakes into WGSL (#32048)

4n4ny4 committed 1 day ago

Verified 8772f770

Pin GitHub Actions to full-length commit SHAs (#32176)

danfiedler-msft committed 1 day ago

Verified 5fd8e199

Resolve possible onnx CVE in python docs (#32190)

eserscor committed 1 day ago

Verified 1fbac8f3

[MLAS] AVX-512 16-wide Erf kernel and NCHWc reorder transpose for MobileClip-S0 model (#31958)

swetha097 committed 2 days ago

Verified a88c1168

[WebGPU] Optimized PagedAttention implementation (2/n) (#31727)

hariharans29 committed 2 days ago

Verified ebd7bad1

[MLAS] Hardswish Fusion implement for Mobilenetv3 models (#31957)

swetha097 committed 2 days ago

Verified 4db33198

Rework agent code review guidance (#31969)

edgchen1 committed 2 days ago

Verified 5d9676ea

Harden Crop operator validation (#32157)

edgchen1 committed 2 days ago

Verified 71f7c781

Validate per-element split sizes on the input-tensor path to prevent OOB read (#29461)

chilo-ms committed 2 days ago

Verified abc74dc0

Retain Python async run resources (#32041)

apsonawane committed 2 days ago

Verified 1cb4307d

Validate ScatterND index depth (#32034)

apsonawane committed 2 days ago

Verified d63396d9

Validate QEmbed segment inputs (#32144)

apsonawane committed 2 days ago

Verified 924b4d1d

Validate empty reduction axes (#32156)

apsonawane committed 2 days ago

Verified 113375ed

Validate Conv bias size (#32160)

apsonawane committed 2 days ago

Verified d9872d8e

Reject scalar Normalizer inputs (#32161)

apsonawane committed 2 days ago

Verified c6a81e4b

Skip overridable initializer fusion (#32143)

apsonawane committed 2 days ago

Verified 8805b399

Handle empty initializer axis scaling (#32138)

apsonawane committed 2 days ago

Verified 597e4a6a

Canonicalize external data locations (#32135)

apsonawane committed 2 days ago

Verified d3b86e7c

Avoid overflow in CPU TensorScatter indices (#32012)

apsonawane committed 2 days ago

Verified f703c528

Fix #31573, prevent ARM64 SymmQgemm int16 overflow (#32057)

sylvesterkaczmarek committed 3 days ago

Verified f73b9ef8

[Build] Update cuda plugin linux aarch64 parallel to 8 (#32165)

tianleiwu committed 3 days ago

Verified 00b6cdb5

Use commit timestamps for plugin EP dev versions (#32095)

edgchen1 committed 3 days ago

Verified 84e87b4b

[CUDA] upgrade cutlass 4.7 and cudnn-frontend 1.27 (#32111)

tianleiwu committed 3 days ago

Verified fccfa10c

[WebGPU] Enable GeluFusion and BiasGeluFusion for the WebGPU EP (#32053)

fanchenkong1 committed 3 days ago

Verified 10ea381c

Older