microsoft/onnxruntime

Pull Requests Commits

Refresh PR on latest main after focused validation

Gopalakrishnan Nallasamy committed 35 days ago

3ba1b177

Add direct mixed FP32 and FP16-helper reporting regression

Gopalakrishnan Nallasamy committed 35 days ago

87da4e5b

Validate CUDA GatherND indices bounds in release mode (#31646)

apsonawane committed 35 days ago

Verified ec1add89

Move standalone FP16 skip into helper without hiding FP32 coverage

Gopalakrishnan Nallasamy committed 35 days ago

766064c8

[WebGPU] Cache sparse indexer QSA visibility and scores (#33017)

sayanshaw24 committed 35 days ago

Verified 98468bff

[WebGPU] Add FP16 GEMV specialization for single-row decode projections (#33070)

aciddelgado committed 35 days ago

Verified b5cb5174

Fix provider bridge initialization failures without exceptions (#33051)

mustjab committed 35 days ago

Verified 4e2e14e4

Fix duplicate node name when QDQ fusion absorbs a redundant Clip/Relu (#33115)

1duo committed 35 days ago

Verified f3c4f31e

Fix WebGPU grouped Conv vectorization across group boundaries (#33129)

xadupre committed 35 days ago

Verified aac791ef

[js/web] Add LoRA adapter support for the standard(non-proxy) run path (#32886)

saravanan-a-r committed 35 days ago

Verified 11922849

Remove windows x86 workflow (#33116)

eserscor committed 35 days ago

Verified 35edaacc

Fix WebGPU DequantizeLinear scalar/vector shader cache collisions (#33122)

gyagp committed 35 days ago

Verified dafbea65

Add JavaScript EPContext data callbacks (#32308)

GopalakrishnanN committed 35 days ago

Verified 99433be6

Add Java EPContext data callbacks (#32305)

GopalakrishnanN committed 35 days ago

Verified e0227cb9

Update Android minimal binary size baseline (#33124)

GopalakrishnanN committed 35 days ago

Verified 402add8d

Add C# EPContext data callbacks (#32270)

GopalakrishnanN committed 35 days ago

Verified 63ec2b0f

Add Objective-C EPContext data callbacks (#32309)

GopalakrishnanN committed 35 days ago

Verified 1ebb4338

Add Python bindings for EPContext data callbacks (#32271)

GopalakrishnanN committed 35 days ago

Verified 86907281

Add Rust EPContext data callbacks (#32310)

GopalakrishnanN committed 35 days ago

Verified 642c0da9

[CUDA] [WebGPU] Optimize sparse attention indexers and add packed QSA state updates (#32656)

Copilot committed 35 days ago

Verified 9dde2edb

Add CUDA QMoE block-scaled FP8 support with routed-expert compaction (#32887)

apsonawane committed 35 days ago

Verified 51f98622

Validate optional graph inputs and Constant node shape during model load (#32940)

GopalakrishnanN committed 35 days ago

Verified 8aa1feca

Bind external data validation to opened files (#33063)

apsonawane committed 35 days ago

Verified 780c0c01

Add reference examples and guidance (especially for multi-device scenarios) for ValidateCompiledModelCompatibilityInfo (#29168)

adrastogi committed 36 days ago

Verified dba3629f

Add WebGPU INT2 and mixed-width QMoE support (#32878)

sayanshaw24 committed 36 days ago

Verified ec1d573c

Fix Python WHL builds for older CUDA version. Filter out __grid_constant__ from tinygemm2 (#33111)

eserscor committed 36 days ago

Verified fc8f71cf

Promote EPContext data callbacks to stable API (#32265)

GopalakrishnanN committed 36 days ago

Verified 772123e9

[CUDA][WebGPU] Add packed sparse attention indexer for continuous batching (#32618)

Copilot committed 36 days ago

Verified 3188a2ee

Validate RestorePadding token offsets (#33099)

apsonawane committed 36 days ago

Verified 0ea42b2b

Raise ir_version to 10 when MatMulNBitsQuantizer emits INT4/UINT4 tensors (#33068)

andife committed 36 days ago

Verified f152edba

Older