openvinotoolkit/openvino

Pull Requests Commits

Fix clang formatting and tidy up

dylanneve1 committed 34 days ago

ae97b394

Add profiling to identify first-inference NPU overhead root cause

dylanneve1 committed 34 days ago

a8f643de

Simplify 1/prefill:3.infer profiling

dylanneve1 committed 34 days ago

e5720d6f

Fix clang formatting

dylanneve1 committed 34 days ago

9834854d

Merge N/generate:(1.prepare/2.prepare)

dylanneve1 committed 34 days ago

aef5514f

Rename to 1/prefill:1.prepare_for_new_conversation

dylanneve1 committed 34 days ago

feb48dc8

Remove unused namespace alias in prepare_for_new_conversation

dylanneve1 committed 34 days ago

597ff5b1

Fix clang-format issue in llm_infer_request.cpp

dylanneve1 committed 34 days ago

dadfaa7f

Cleanup profiling

dylanneve1 committed 34 days ago

44e08727

[NPUW] Add LLM-level profiling for 1st token generation analysis

dylanneve1 committed 34 days ago

7ef31aa5

[GPU] Add fully_connected_bf_tiled_dyn_b kernel for small-batch INT4 FC (#34525)

jade-cho committed 34 days ago

Verified 4589d335

Refactor frontends' supported & load impl (#34466)

almilosz committed 34 days ago

Verified 53ef535a

[NPU] Use ZeroTensor instead of ZeroHostTensor for weightless graph (#34592)

pereanub committed 34 days ago

Verified 8306c8e7

[GPU] Select new impl when propagate constant makes dynamic shape layer to constant (#34115)

wilson-seok committed 34 days ago

Verified 9d8ca139

[GPU] Enable segment_max-16 (#34037)

wilson-seok committed 34 days ago

Verified aebbbdb7

[NPU] Acknowledge OV model compression in weightless compilation (#34609)

razvanapetroaie committed 34 days ago

Verified 9e83aea1

[TRANSFORMATIONS] Add Qwen 2.5 VL M-RoPE handling to SDPAToPA (#34365)

CuriousPanCake committed 34 days ago

Verified 68322ce9

Migrate to std::fs::path dev and auto plugin (#34545)

almilosz committed 34 days ago

Verified e3195385

Add compute runtime license (#34541)

jkasprza committed 34 days ago

Verified 6f79c848

[core] Create single file multiple blobs storage (#34191)

t-jankowski committed 34 days ago

Verified 5625ab2f

[NPU] Fix when run set_tensors after set_tensor (#34578)

pereanub committed 34 days ago

Verified f65dc110

[NPU] Blacklist legacy NPU device (#34544)

sbutnari committed 34 days ago

Verified 0d1317de

[GPU] Fix OOB memory access in gemm_tiled_opt kernel for non-aligned tile dimensions (#34482)

wilson-seok committed 35 days ago

Verified c4ce6e56

[OVC] Introduce torch.export in convert_model as non-default (#34471)

mvafin committed 35 days ago

Verified 2908f218

[CPU][DEBUG_CAPS] Include model name to IR dump path (#34397)

EgorDuplensky committed 35 days ago

Verified a3c8ffea

[CPU] fix zero size reorder (#34569)

azhai219 committed 35 days ago

Verified a89b1315

[IR FE] Weight share ir fe mark weights (#34454)

praasz committed 35 days ago

Verified f3c663ee

[core]Feature/weight sharing properties (#34282)

praasz committed 35 days ago

Verified 26e55ee6

[OV JS] Bump NPM package verstion up to 2026.1 (#34583)

Retribution98 committed 35 days ago

Verified f24b8b5e

Specify GatedDeltaNet operation for Stateful mode (#34529)

rkazants committed 35 days ago

Verified 4402dee0

Older