openvinotoolkit/openvino

Pull Requests Commits

Refactor pagedAttention transpose

mangguo321 committed 156 days ago

9ef60bcd

NPUW: Disable V-tensors transpose for Whisper models (#33053)

eshiryae committed 159 days ago

Verified 24654ef2

Fix ConvTranspose with output_shape containing batch and channel dimensions (#32854)

evkotov committed 159 days ago

Verified 2e27a0d0

[PT FE] Support axis=None for aten::mean operation (#32909)

mvafin committed 159 days ago

Verified 6129c50a

[NPUW] Share kvcache between prefill and generate when chunking is enabled (#32642)

smirnov-alexey committed 159 days ago

Verified 3797c559

[GPU] Enable BY_CHANNEL de-quantize/re-quantize support in KV_CACHE_R… (#32783)

Prithviraj-R committed 159 days ago

Verified 1687ab91

[GHA] Return back the normal timeout for C++ unit on Win Debug (#33052)

akladiev committed 159 days ago

Verified b237dd6a

Updated External Benchmarking details for OV-2025.4 (#33029)

mfhansen committed 159 days ago

Verified cb5138e4

[CPU] Implement X-Attention for intel CPU (#32086)

mangguo321 committed 159 days ago

Verified 3ac45ad5

[core] reduce test iteration to fix time degradation (#33039)

walidbarakat committed 159 days ago

Verified 3d46a4b3

[ONNX FE] Add operations in the documentation page - SequenceAt, SplitToSequence (#32990)

bumbosiepsak committed 159 days ago

Verified b8c79f98

[core] Remove `cache_model_path` from config pass to device (#33037)

praasz committed 159 days ago

Verified a4e91411

[GPU] Choose byxf for dynamic-batch case of OneDNN (#32874)

hyunback committed 160 days ago

Verified 0e3f77a7

Switch Debian 10 repositories to archive.debian.org (#33047)

ababushk committed 160 days ago

Verified d7eb2588

[NPU][Properties] Remove INTERNAL_SUPPORTED_PROPERTIES property (#32841)

sbutnari committed 160 days ago

Verified b1074859

[CPU] Update xbyak_riscv to v1.00 (#33045)

aobolensk committed 160 days ago

Verified 5046eff7

[CPU] Remove all unused header files (#33042)

aobolensk committed 160 days ago

Verified 923c5503

[TRANSFORMATIONS] Do not convert Range inputs (#32985)

mryzhov committed 160 days ago

Verified 45e7c3de

[Snippets] SplitLoops: inner loops splitting support (#32008)

v-Golubev committed 160 days ago

Verified 6d02ccc1

[PyOV] Move `packaging` from runtime dependencies to frontend dependencies (#32951)

p-wysocki committed 160 days ago

Verified b96c23a3

[NPU] Adding support for importing raw pointer into zero memory through remote tensor (#32908)

pereanub committed 160 days ago

Verified 7bf1918b

[GPU] Add 1d_conv flag in convolution primitive (#32864)

wilson-seok committed 160 days ago

Verified bee83079

[CPU] Disable deconvolution Prelu post-ops fusion (#32832)

mangguo321 committed 160 days ago

Verified e626f670

[GHA] Don't run pipelines by default for draft but allow manual restart (#31996)

akladiev committed 160 days ago

Verified 5161b6cd

[core] Throw error if data type is not recognized (#32955)

dnkurek committed 160 days ago

Verified 425c863e

[CPU] Fix u8 -> bool conversion in cpu_convert.cpp (#32919)

aobolensk committed 160 days ago

Verified 6b5d5d74

[GPU][Wiki] Update GPU document: runtime_skippable and KV cache (#32943)

yeonbok committed 160 days ago

Verified 132820e9

[`NPU`] [`icompiler_adapter`] Extend `icompiler_adapter` interface to accept optional `value` (#32823)

MirceaDan99 committed 160 days ago

Verified 9d1f078e

[TF Lite FE] Embedding Lookup implementation (#32820)

akodanka committed 160 days ago

Verified 4291e7a5

[GPU] init enable_large_allocations to silent valgrind (#33022)

michal-miotk committed 160 days ago

Verified 806eb8be

Older