microsoft/onnxruntime

Pull Requests Commits

Address code review: remove intermediate variable, improve platform detection comment

Copilot committed 56 days ago

7584a783

Refactor FP16 softmax buffer size into lightweight testable helper

Copilot committed 56 days ago

1854774d

Copilot committed 56 days ago

3c78b023

expect test overflow in 32-bit builds, update test name and comment

edgchen1 committed 56 days ago

605328f3

increase tolerance for test output

edgchen1 committed 56 days ago

70cd5b14

edgchen1 committed 57 days ago

d30ee590

edgchen1 committed 57 days ago

20332e89

fix overflow issue

edgchen1 committed 57 days ago

1769061a

edgchen1 committed 57 days ago

b72abd5e

add GetTotalPhysicalMemoryBytes helper

edgchen1 committed 57 days ago

c9f71907

Fix QNN SDK version propagation in Linux ort-qnn wheel build (#27800)

derdeljan-msft committed 57 days ago

Verified 36c962fd

[CUDA] Extend Pad support through opset 25 with wrap mode (#27774)

Copilot committed 57 days ago

Verified fde4e032

[CUDA] RoiAlign for opset versions 16 and 22 (#27646)

tianleiwu committed 57 days ago

Verified 45b5900d

Fix non-ASCII Unicode model path crash across session and provider code (#27724)

sagarbhure-msft committed 57 days ago

Verified 3bb032e8

[CPU/CUDA EP] Add DeformConv op support (#27393)

ShirasawaSama committed 58 days ago

Verified 163f6149

Fix tf.function retracing in TensorFlow benchmark (#27665)

Rishi-Dave committed 58 days ago

Verified 727256f8

[CPU] Update the NCHWc transformer to handle more patterns (#27691)

hariharans29 committed 58 days ago

Verified aa6f2e33

[Core] Add SpaceToDepth fusion pattern (#27747)

hariharans29 committed 58 days ago

Verified 3fa42051

Add opset 21/23 CUDA kernel registrations for Flatten, Identity, If, Loop, Scan, ConstantOfShape, Size (#27728)

Rishi-Dave committed 59 days ago

Verified ce3158ac

Bump flatted from 3.2.7 to 3.4.2 in /onnxruntime/test/wasm (#27787)

dependabot[bot] committed 60 days ago

Verified b1c5e044

Fix out of bounds read issue in cross.entropy.cc (#27568)

vraspar committed 60 days ago

Verified 0aa6adb5

Fill CUDA EP opset gap for GRU operator (14 → 22) (#27738)

Copilot committed 60 days ago

Verified 830a29ee

[React Native] Fix float16 tensor crash by mapping to Uint16Array (#27549)

kevinlam92 committed 60 days ago

Verified b10fa545

Dev/erscor/2026/3/19 make react native optional (#27772)

eserscor committed 60 days ago

Verified e3ebf875

[WebGPU] DequantizeLinear op fixes (#27706)

edgchen1 committed 60 days ago

Verified a6857897

Fill GlobalAveragePool and GlobalMaxPool opset gap in CUDA provider (1→22) (#27733)

Copilot committed 61 days ago

Verified f546db86

Add SkipLayerNorm fusion with bias Add (#27765)

kunal-vaishnavi committed 61 days ago

Verified 405dcd72

[CUDA] Refactor CUDA helpers or ops for plugin EP (#27719)

tianleiwu committed 61 days ago

Verified 988d05e4

Enable KleidiAI for asymmetric 4-bit MatMulNBits on ARM64 (#27751)

jambayk committed 61 days ago

Verified d9bb7602

Support softcap and softmax_precision in Attention(CUDA) (#27714)

titaiwangms committed 61 days ago

Verified 93d31cfd

Older