openvinotoolkit/openvino

Pull Requests Commits

[NPUW] Check router Softmax in MoE LLM test

dylanneve1 committed 31 days ago

44ba08d6

[NPUW] Fix MoE test for expert.router naming

dylanneve1 committed 31 days ago

7f66779b

[NPUW] Fix MoE model builder for NPU compatibility

dylanneve1 committed 31 days ago

203aaddf

[NPUW] Add Mixture of Experts support to model builder

dylanneve1 committed 31 days ago

e401972d

[PyOV][ARM] Skip i16 async mixed-values case with share_inputs (#34384)

aobolensk committed 31 days ago

Verified 052f8039

[NPUW] Model builder: refactor Qwen3 embedding support (#35045)

dylanneve1 committed 31 days ago

Verified 83bc8751

[GPU] Add squeeze/unsqueeze to the compressed conv1x1 transformation (#34957)

mdvoretc-intel committed 31 days ago

Verified 70bcd498

Out-of-bounds Write in ONNX Loop Converter body_inputs Access leads to Crash / DoS (#34762)

bumbosiepsak committed 31 days ago

Verified 728dd933

[NPUW] support for Whisper KV-cache names in precision conversion (#35431)

esmirno committed 31 days ago

Verified b3de8aa1

[CORE] Update GatedDeltaNet to support GQA (#35472)

mmikolajcz committed 31 days ago

Verified f2959dc6

Memory optimization during conversion of onnx model OV IR format (#35207)

susbhere committed 31 days ago

Verified c1bf7550

[TRANSFORMATIONS] Move convert_pagedattn_inputs to paged_attention dir (#35471)

mryzhov committed 31 days ago

Verified 2699faec

Preserve current_iteration element type in UnrollTensorIterator (#35428)

ssam18 committed 31 days ago

Verified 349dcb5c

Update memory tests (#35214)

culhatsker committed 31 days ago

Verified 571863cb

[GPU] Fix implicit narrowing conversion warnings (C4244) in GPU plugin - Step 2 (#35444)

ahnyoung-paul committed 31 days ago

Verified e1dde008

Fix in lto support check (#35354)

olpipi committed 31 days ago

Verified d2f02370

NPUW: Serialization refactoring (#35441)

dmatveev committed 31 days ago

Verified 5ef1deb2

Rtti fix (#35443)

JeevakaPrabu committed 31 days ago

Verified 5f388e7d

[GPU] Add Power(-0.5) pattern to RMSFusion (#35460)

jade-cho committed 31 days ago

Verified e68f6589

[GPU] Fix RMS bfyx_opt kernel producing wrong results with static padding (#35350)

zlma7001 committed 31 days ago

Verified 26c3c8f9

[GPU] minor fix in log message (#35467)

isanghao committed 31 days ago

Verified e0ce2ec5

[NPUW]Optimize pipeline initialize time in online partitioner. (#34958)

intelgaoxiong committed 32 days ago

Verified 6ed00030

[NPU] Create execution context for dynamic graph execution (#35380)

XinWangIntel committed 32 days ago

Verified b62da294

[GHA][Coverage] Add GPU tests code coverage (#35236)

mkruszyx committed 32 days ago

Verified e9187467

[GPU] use parallel ifstream to boost file reading latency (#34679)

riverlijunjie committed 32 days ago

Verified 27e83577

[NPU] Add NPU_COMPILER_VERSION to CompiledModel property list (#35360)

alexandruenache1111 committed 32 days ago

Verified 0c81a247

Update setuptools requirement from <=82.0.0,>=77 to >=77,<=82.0.1 (#35078)

dependabot[bot] committed 32 days ago

Verified 53d3b91e

Introduce new check for CompressFloatConstantsImpl (#35415)

evkotov committed 32 days ago

Verified b344668c

Migrate ONNX frontend InputModel and other to use std::fs::path (#35091)

almilosz committed 32 days ago

Verified 091d0b36

Bump typescript-eslint from 8.58.2 to 8.59.0 in /src/bindings/js/node in the npm-packages group (#35439)

dependabot[bot] committed 32 days ago

Verified c6169ed2

Older