microsoft/onnxruntime

Pull Requests Commits

Gate RISC-V FP16 kernels on acceleration

Copilot committed 7 hours ago

Verified f2349d0d

Fix CPU provider test include

Copilot committed 11 hours ago

Verified fb743f93

Gate CPU FP16 Gemm on acceleration

Copilot committed 11 hours ago

Verified a933add2

Copilot committed 11 hours ago

Verified 3e9bd871

[CUDA] Add FP16-cache paged XQA decode for head size 256 (#32263)

baijumeswani committed 1 day ago

Verified b1f76d58

Fix ort-ci/SKILL.md yaml frontmatter syntax. (#32248)

edgchen1 committed 1 day ago

Verified 05491530

Harden head-size-256 paged XQA coverage (#32262)

baijumeswani committed 2 days ago

Verified d67cc269

Run wgsl_template Python tests in CI (#32214)

edgchen1 committed 2 days ago

Verified dc999734

[MLAS] Reject Arm® KleidiAI™ Q4 prepack with dynamic scales (#32068)

martin-klacer-arm committed 3 days ago

Verified 9cbf665b

[MLAS] Refactor QNBit KleidiAI integration into dedicated layer (#31720)

martin-klacer-arm committed 3 days ago

Verified 5cb241b5

Validate MatMulNBits 8-bit g_idx bounds on CUDA (#31643)

apsonawane committed 3 days ago

Verified 066b90ef

Fix Windows CUDA 12.9 SM120 compilation (#32114)

Copilot committed 3 days ago

Verified 939e180a

pin review dog (#32208)

eserscor committed 3 days ago

Verified ddc8f0c3

Upgrade Protobuf to v33.6 (#29906)

Copilot committed 3 days ago

Verified 993cd135

Extend memory importer with host pointer support (#29726)

javier-intel committed 3 days ago

Verified 710f0ce3

[Telemetry] Prevent crashes in shell-less containers (#32226)

bmehta001 committed 3 days ago

Verified 9aa8fb06

[CUDA] Add VarlenCausalConvWithState for continuous batching (#32168)

baijumeswani committed 3 days ago

Verified 2a9bf1e9

Support LoRA adapters with plugin EP allocators (#32221)

baijumeswani committed 3 days ago

Verified 4f7d0f5b

[CUDA] Extend paged XQA decode to head size 256 (#32229)

baijumeswani committed 3 days ago

Verified 15920d08

Move ort-release-notes skill to new .github/skills location. (#32193)

edgchen1 committed 3 days ago

Verified ddbcef7d

[WebGPU] Supply fused activation parameters to Conv/MatMul as uniforms (#32116)

4n4ny4 committed 5 days ago

Verified 4d308dac

[WebGPU] Scale Dawn pipeline compilation workers with CPU count (#29820)

xiaofeihan1 committed 5 days ago

Verified 89b01b40

Fix C API docs Doxygen download (#32210)

Wayne-Ch committed 6 days ago

Verified df012c2f

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

mirounga committed 6 days ago

Verified a4156be6

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

mirounga committed 6 days ago

Verified 27f212c4

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

4n4ny4 committed 6 days ago

Verified 775526b8

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

4n4ny4 committed 6 days ago

Verified 8772f770

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

danfiedler-msft committed 6 days ago

Verified 5fd8e199

Resolve possible onnx CVE in python docs (#32190)

eserscor committed 6 days ago

Verified 1fbac8f3

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

swetha097 committed 7 days ago

Verified a88c1168

Older