Commits microsoft/onnxruntime

Clarify pooling ops in BF16 plan Phase 2

Copilot committed 72 days ago

Verified 401bb9e7

Add BFloat16 support analysis and plan doc for CPU/CUDA EPs

Copilot committed 72 days ago

Verified 1db6097a

Initial plan

Copilot committed 72 days ago

Verified 849dba87

[WebGPU] Default non-debug validation mode set to disabled (#28679)

hariharans29 committed 72 days ago

Verified e25750d8

[CPU] Add fix for fastmath corner case with failing tests via sbgemm_neon_kernel (#28394)

JonathanC-ARM committed 72 days ago

Verified e0d99e27

fix the case where the buffer for past_state == present_state (#28753)

Guenther Schmuelling committed 72 days ago

Verified e26ad80e

Skip SetupDi device discovery if Win32k system calls are disabled (#28535)

shiyi9801 committed 73 days ago

Verified a8baf5c4

fix(quantization): validate bias scale in QDQ Conv → QLinearConv fusion (#28229)

Rishi-Dave committed 73 days ago

Verified 5c34495a

Use abseil for readable POSIX stack traces in debug builds (#28405)

tianleiwu committed 73 days ago

Verified 8da5e910

feat(quantization): add opset-21 block_size attribute to QDQ (#28522)

Rishi-Dave committed 73 days ago

Verified 8fcb725d

fix: preserve Q/DQ around Resize when mode is linear or cubic (#28414)

Rishi-Dave committed 73 days ago

Verified 7abc315f

Fix PRelu returning NaN for infinite inputs on CPU EP (#28750)

Copilot committed 73 days ago

Verified 9f81e2df

[MLAS][KleidiAI] Apply fp16 split independent fixes (#28720)

Laan33 committed 73 days ago

Verified 1e2b6394

Zero-copy I/O for plugin EPs with HOST_ACCESSIBLE memory (#28037)

ericcraw committed 73 days ago

Verified a827f9c3

[CoreML EP] Add GatherND builder (#28598)

maxwbuckley committed 73 days ago

Verified 21e9fe39

Fix ARM CPUIDInfo bounds handling for unknown CPU vendors (#28344)

Copilot committed 73 days ago

Verified d1135499

Honor disable_synchronize_execution_providers for CUDA graph replay (#28686)

tianleiwu committed 73 days ago

Verified 312f5249

[MLAS] Fix MatMulNBits CPU LUT GEMM path to apply optional bias (#28742)

hariharans29 committed 73 days ago

Verified 1077c77e

[WebGPU] Support models with opset 24 ops and KV-shared decoder layers (Gemma 4) (#28501)

feich-ms committed 73 days ago

Verified e1f27d15

Add LpNormalization support for CUDA Execution Provider (#28724)

apsonawane committed 74 days ago

Verified 5ddf568f

Unify path validation in EPs, address security gaps (#28725)

yuslepukhin committed 74 days ago

Verified 2e677fea

Add boundary checks for input validation (#28747)

apsonawane committed 74 days ago

Verified a28d46d3

[CoreML EP] Support bool Cast in ML Program (#28595)

maxwbuckley committed 74 days ago

Verified 62567624

[CUDA] Add QMoE tests for standard swiglu and update doc (#28741)

tianleiwu committed 74 days ago

Verified 9e104a1a

Add Implicit inputs for conditional nodes branches part of node fusio… (#28690)

preetha-intel committed 74 days ago

Verified f09204c6

Fix CUDA 13 nuget packaging pipeline failures and enable CUDA 13.3 builds (#28736)

tianleiwu committed 74 days ago

Verified 291311e7

[CUDA] Chunked dequant+GEMM in MatMulNBits to reduce peak GPU memory for large N (#28712)

apsonawane committed 74 days ago

Verified 84b7da78

Fix IsInfTest (T4 GPU does not support BF16) (#28727)

tianleiwu committed 74 days ago

Verified 17f3a3b3

[CUDA Plugin EP] Add provider options: user_compute_stream, do_copy_in_default_stream, use_ep_level_unified_stream, external allocator (#28603)

tianleiwu committed 74 days ago

Verified 895422db

[NvTensorRTRTX EP] Skip GPU JIT during compile-only sessions (#28503)

umangb-09 committed 74 days ago

Verified 6572dd0b