From 63929e8cbb9138ab51036a611aa0467f3f07bbbf27c2de43f0e008368f9ffea7 Mon Sep 17 00:00:00 2001 From: Alessandro de Oliveira Faria Date: Sun, 7 Sep 2025 04:23:58 +0000 Subject: [PATCH 1/3] =?UTF-8?q?-=20Update=20to=202025.3.0=20-=20More=20Gen?= =?UTF-8?q?AI=20coverage=20and=20framework=20integrations=20to=20minimize?= =?UTF-8?q?=20code=20=20=20changes=20=20=20*=20New=20models=20supported:?= =?UTF-8?q?=20Phi-4-mini-reasoning,=20AFM-4.5B,=20=20=20=20=20Gemma-3-1B-i?= =?UTF-8?q?t,=20Gemma-3-4B-it,=20and=20Gemma-3-12B,=20=20=20*=20NPU=20supp?= =?UTF-8?q?ort=20added=20for:=20Qwen3-1.7B,=20Qwen3-4B,=20and=20Qwen3-8B.?= =?UTF-8?q?=20=20=20*=20LLMs=20optimized=20for=20NPU=20now=20available=20o?= =?UTF-8?q?n=20OpenVINO=20Hugging=20=20=20=20=20Face=20collection.=20-=20B?= =?UTF-8?q?roader=20LLM=20model=20support=20and=20more=20model=20compressi?= =?UTF-8?q?on=20techniques=20=20=20*=20The=20NPU=20plug-in=20adds=20suppor?= =?UTF-8?q?t=20for=20longer=20contexts=20of=20up=20to=20=20=20=20=208K=20t?= =?UTF-8?q?okens,=20dynamic=20prompts,=20and=20dynamic=20LoRA=20for=20impr?= =?UTF-8?q?oved=20=20=20=20=20LLM=20performance.=20=20=20*=20The=20NPU=20p?= =?UTF-8?q?lug-in=20now=20supports=20dynamic=20batch=20sizes=20by=20reshap?= =?UTF-8?q?ing=20=20=20=20=20the=20model=20to=20a=20batch=20size=20of=201?= =?UTF-8?q?=20and=20concurrently=20managing=20=20=20=20=20multiple=20infer?= =?UTF-8?q?ence=20requests,=20enhancing=20performance=20and=20=20=20=20=20?= =?UTF-8?q?optimizing=20memory=20utilization.=20=20=20*=20Accuracy=20impro?= =?UTF-8?q?vements=20for=20GenAI=20models=20on=20both=20built-in=20=20=20?= =?UTF-8?q?=20=20and=20discrete=20graphics=20achieved=20through=20the=20im?= =?UTF-8?q?plementation=20=20=20=20=20of=20the=20key=20cache=20compression?= =?UTF-8?q?=20per=20channel=20technique,=20in=20=20=20=20=20addition=20to?= =?UTF-8?q?=20the=20existing=20KV=20cache=20per-token=20compression=20=20?= =?UTF-8?q?=20=20=20method.=20=20=20*=20OpenVINO=E2=84=A2=20GenAI=20introd?= =?UTF-8?q?uces=20TextRerankPipeline=20for=20improved=20=20=20=20=20retrie?= =?UTF-8?q?val=20relevance=20and=20RAG=20pipeline=20accuracy,=20plus=20=20?= =?UTF-8?q?=20=20=20Structured=20Output=20for=20enhanced=20response=20reli?= =?UTF-8?q?ability=20and=20=20=20=20=20function=20calling=20while=20ensuri?= =?UTF-8?q?ng=20adherence=20to=20predefined=20=20=20=20=20formats.=20-=20M?= =?UTF-8?q?ore=20portability=20and=20performance=20to=20run=20AI=20at=20th?= =?UTF-8?q?e=20edge,=20=20=20in=20the=20cloud,=20or=20locally.=20=20=20*?= =?UTF-8?q?=20Announcing=20support=20for=20Intel=C2=AE=20Arc=E2=84=A2=20Pr?= =?UTF-8?q?o=20B-Series=20=20=20=20=20(B50=20and=20B60).=20=20=20*=20Previ?= =?UTF-8?q?ew:=20Hugging=20Face=20models=20that=20are=20GGUF-enabled=20for?= =?UTF-8?q?=20=20=20=20=20OpenVINO=20GenAI=20are=20now=20supported=20by=20?= =?UTF-8?q?the=20OpenVINO=E2=84=A2=20Model=20=20=20=20=20Server=20for=20po?= =?UTF-8?q?pular=20LLM=20model=20architectures=20such=20as=20=20=20=20=20D?= =?UTF-8?q?eepSeek=20Distill,=20Qwen2,=20Qwen2.5,=20and=20Llama=203.=20=20?= =?UTF-8?q?=20=20=20This=20functionality=20reduces=20memory=20footprint=20?= =?UTF-8?q?and=20=20=20=20=20simplifies=20integration=20for=20GenAI=20work?= =?UTF-8?q?loads.=20=20=20*=20With=20improved=20reliability=20and=20tool?= =?UTF-8?q?=20call=20accuracy,=20=20=20=20=20the=20OpenVINO=E2=84=A2=20Mod?= =?UTF-8?q?el=20Server=20boosts=20support=20for=20=20=20=20=20agentic=20AI?= =?UTF-8?q?=20use=20cases=20on=20AI=20PCs,=20while=20enhancing=20=20=20=20?= =?UTF-8?q?=20performance=20on=20Intel=20CPUs,=20built-in=20GPUs,=20and=20?= =?UTF-8?q?NPUs.=20=20=20*=20int4=20data-aware=20weights=20compression,=20?= =?UTF-8?q?now=20supported=20in=20the=20=20=20=20=20Neural=20Network=20Com?= =?UTF-8?q?pression=20Framework=20(NNCF)=20for=20ONNX=20=20=20=20=20models?= =?UTF-8?q?,=20reduces=20memory=20footprint=20while=20maintaining=20=20=20?= =?UTF-8?q?=20=20accuracy=20and=20enables=20efficient=20deployment=20in=20?= =?UTF-8?q?=20=20=20=20resource-constrained=20environments.?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit OBS-URL: https://build.opensuse.org/package/show/science:machinelearning/openvino?expand=0&rev=42 --- .gitattributes | 23 + .gitignore | 1 + _constraints | 11 + _service | 16 + openvino-2025.3.0.obscpio | 3 + openvino-ComputeLibrary-include-string.patch | 11 + openvino-fix-build-sample-path.patch | 12 + openvino-fix-install-paths.patch | 87 ++ openvino-rpmlintrc | 4 + openvino.changes | 820 +++++++++++++++++++ openvino.obsinfo | 4 + openvino.spec | 441 ++++++++++ 12 files changed, 1433 insertions(+) create mode 100644 .gitattributes create mode 100644 .gitignore create mode 100644 _constraints create mode 100644 _service create mode 100644 openvino-2025.3.0.obscpio create mode 100644 openvino-ComputeLibrary-include-string.patch create mode 100644 openvino-fix-build-sample-path.patch create mode 100644 openvino-fix-install-paths.patch create mode 100644 openvino-rpmlintrc create mode 100644 openvino.changes create mode 100644 openvino.obsinfo create mode 100644 openvino.spec diff --git a/.gitattributes b/.gitattributes new file mode 100644 index 0000000..9b03811 --- /dev/null +++ b/.gitattributes @@ -0,0 +1,23 @@ +## Default LFS +*.7z filter=lfs diff=lfs merge=lfs -text +*.bsp filter=lfs diff=lfs merge=lfs -text +*.bz2 filter=lfs diff=lfs merge=lfs -text +*.gem filter=lfs diff=lfs merge=lfs -text +*.gz filter=lfs diff=lfs merge=lfs -text +*.jar filter=lfs diff=lfs merge=lfs -text +*.lz filter=lfs diff=lfs merge=lfs -text +*.lzma filter=lfs diff=lfs merge=lfs -text +*.obscpio filter=lfs diff=lfs merge=lfs -text +*.oxt filter=lfs diff=lfs merge=lfs -text +*.pdf filter=lfs diff=lfs merge=lfs -text +*.png filter=lfs diff=lfs merge=lfs -text +*.rpm filter=lfs diff=lfs merge=lfs -text +*.tbz filter=lfs diff=lfs merge=lfs -text +*.tbz2 filter=lfs diff=lfs merge=lfs -text +*.tgz filter=lfs diff=lfs merge=lfs -text +*.ttf filter=lfs diff=lfs merge=lfs -text +*.txz filter=lfs diff=lfs merge=lfs -text +*.whl filter=lfs diff=lfs merge=lfs -text +*.xz filter=lfs diff=lfs merge=lfs -text +*.zip filter=lfs diff=lfs merge=lfs -text +*.zst filter=lfs diff=lfs merge=lfs -text diff --git a/.gitignore b/.gitignore new file mode 100644 index 0000000..57affb6 --- /dev/null +++ b/.gitignore @@ -0,0 +1 @@ +.osc diff --git a/_constraints b/_constraints new file mode 100644 index 0000000..69ac422 --- /dev/null +++ b/_constraints @@ -0,0 +1,11 @@ + + + + + 20 + + + 8 + + + diff --git a/_service b/_service new file mode 100644 index 0000000..00184c0 --- /dev/null +++ b/_service @@ -0,0 +1,16 @@ + + + https://github.com/openvinotoolkit/openvino.git + git + 2025.3.0 + 2025.3.0 + enable + openvino + .git + + + + *.tar + zstd + + diff --git a/openvino-2025.3.0.obscpio b/openvino-2025.3.0.obscpio new file mode 100644 index 0000000..dcac4b0 --- /dev/null +++ b/openvino-2025.3.0.obscpio @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:64424d07d5017ed8773c86d9bfa80af8d436d70aa55cb6c4b0a26ca1b2804b1e +size 744031247 diff --git a/openvino-ComputeLibrary-include-string.patch b/openvino-ComputeLibrary-include-string.patch new file mode 100644 index 0000000..44006ce --- /dev/null +++ b/openvino-ComputeLibrary-include-string.patch @@ -0,0 +1,11 @@ +diff -uNr openvino-2025.2.0.orig/src/plugins/intel_cpu/thirdparty/ComputeLibrary/arm_compute/core/utils/logging/IPrinter.h openvino-2025.2.0/src/plugins/intel_cpu/thirdparty/ComputeLibrary/arm_compute/core/utils/logging/IPrinter.h +--- openvino-2025.2.0.orig/src/plugins/intel_cpu/thirdparty/ComputeLibrary/arm_compute/core/utils/logging/IPrinter.h 2025-06-22 18:14:56.561471325 -0300 ++++ openvino-2025.2.0/src/plugins/intel_cpu/thirdparty/ComputeLibrary/arm_compute/core/utils/logging/IPrinter.h 2025-06-22 18:15:23.466678704 -0300 +@@ -29,6 +29,7 @@ + */ + + #include "support/Mutex.h" ++#include + + namespace arm_compute + { diff --git a/openvino-fix-build-sample-path.patch b/openvino-fix-build-sample-path.patch new file mode 100644 index 0000000..f0742b1 --- /dev/null +++ b/openvino-fix-build-sample-path.patch @@ -0,0 +1,12 @@ +diff -uNr openvino.orig/samples/cpp/build_samples.sh openvino/samples/cpp/build_samples.sh +--- openvino.orig/samples/cpp/build_samples.sh 2024-04-25 01:04:42.451868881 -0300 ++++ openvino/samples/cpp/build_samples.sh 2024-04-25 01:05:04.678342617 -0300 +@@ -59,7 +59,7 @@ + printf "\nSetting environment variables for building samples...\n" + + if [ -z "$INTEL_OPENVINO_DIR" ]; then +- if [[ "$SAMPLES_SOURCE_DIR" = "/usr/share/openvino"* ]]; then ++ if [[ "$SAMPLES_SOURCE_DIR" = "/usr/share/OpenVINO"* ]]; then + true + elif [ -e "$SAMPLES_SOURCE_DIR/../../setupvars.sh" ]; then + setupvars_path="$SAMPLES_SOURCE_DIR/../../setupvars.sh" diff --git a/openvino-fix-install-paths.patch b/openvino-fix-install-paths.patch new file mode 100644 index 0000000..c18c137 --- /dev/null +++ b/openvino-fix-install-paths.patch @@ -0,0 +1,87 @@ +diff -uNr openvino-2024.6.0.orig/cmake/developer_package/packaging/archive.cmake openvino-2024.6.0/cmake/developer_package/packaging/archive.cmake +--- openvino-2024.6.0.orig/cmake/developer_package/packaging/archive.cmake 2024-12-27 17:04:54.520685198 -0300 ++++ openvino-2024.6.0/cmake/developer_package/packaging/archive.cmake 2024-12-27 17:02:57.644273948 -0300 +@@ -25,14 +25,18 @@ + macro(ov_archive_cpack_set_dirs) + # common "archive" package locations + # TODO: move current variables to OpenVINO specific locations +- set(OV_CPACK_INCLUDEDIR runtime/include) +- set(OV_CPACK_OPENVINO_CMAKEDIR runtime/cmake) +- set(OV_CPACK_DOCDIR docs) +- set(OV_CPACK_LICENSESDIR licenses) +- set(OV_CPACK_SAMPLESDIR samples) +- set(OV_CPACK_WHEELSDIR wheels) +- set(OV_CPACK_DEVREQDIR tools) +- set(OV_CPACK_PYTHONDIR python) ++ set(OV_CPACK_INCLUDEDIR include) ++ set(OV_CPACK_OPENVINO_CMAKEDIR ${CMAKE_INSTALL_LIBDIR}/cmake/${PROJECT_NAME}) ++ set(OV_CPACK_DOCDIR ${CMAKE_INSTALL_DOCDIR}) ++ set(OV_CPACK_LICENSESDIR ${CMAKE_INSTALL_DATAROOTDIR}/licenses/${PROJECT_NAME}) ++ set(OV_CPACK_SAMPLESDIR ${CMAKE_INSTALL_DATAROOTDIR}/${PROJECT_NAME}/samples) ++ if (ENABLE_PYTHON) ++ find_package(Python3 QUIET COMPONENTS Interpreter) ++ file(RELATIVE_PATH OV_PYTHON_MODPATH ${CMAKE_INSTALL_PREFIX} ${Python3_SITEARCH}) ++ set(OV_CPACK_WHEELSDIR tools) ++ set(OV_CPACK_DEVREQDIR tools) ++ set(OV_CPACK_PYTHONDIR ${OV_PYTHON_MODPATH}) ++ endif() + + if(USE_BUILD_TYPE_SUBFOLDER) + set(build_type ${CMAKE_BUILD_TYPE}) +@@ -49,11 +53,12 @@ + set(OV_CPACK_RUNTIMEDIR runtime/lib/${ARCH_FOLDER}/${build_type}) + set(OV_CPACK_ARCHIVEDIR runtime/lib/${ARCH_FOLDER}/${build_type}) + else() +- set(OV_CPACK_LIBRARYDIR runtime/lib/${ARCH_FOLDER}) +- set(OV_CPACK_RUNTIMEDIR runtime/lib/${ARCH_FOLDER}) +- set(OV_CPACK_ARCHIVEDIR runtime/lib/${ARCH_FOLDER}) ++ set(OV_CPACK_LIBRARYDIR ${CMAKE_INSTALL_LIBDIR}) ++ set(OV_CPACK_RUNTIMEDIR ${CMAKE_INSTALL_LIBDIR}) ++ set(OV_CPACK_ARCHIVEDIR ${CMAKE_INSTALL_LIBDIR}) + endif() +- set(OV_CPACK_PLUGINSDIR ${OV_CPACK_RUNTIMEDIR}) ++ set(OV_CPACK_PLUGINSDIR ${OV_CPACK_RUNTIMEDIR}/${PROJECT_NAME}) ++ + endmacro() + + ov_archive_cpack_set_dirs() +diff -uNr openvino-2024.6.0.orig/src/cmake/openvino.cmake openvino-2024.6.0/src/cmake/openvino.cmake +--- openvino-2024.6.0.orig/src/cmake/openvino.cmake 2024-12-27 17:04:55.240687724 -0300 ++++ openvino-2024.6.0/src/cmake/openvino.cmake 2024-12-27 17:03:50.176459053 -0300 +@@ -267,6 +267,7 @@ + + # define relative paths + file(RELATIVE_PATH PKGCONFIG_OpenVINO_PREFIX "/${OV_CPACK_RUNTIMEDIR}/pkgconfig" "/") ++ cmake_path(NORMAL_PATH PKGCONFIG_OpenVINO_PREFIX) + + set(pkgconfig_in "${OpenVINO_SOURCE_DIR}/cmake/templates/openvino.pc.in") + if(CMAKE_VERSION VERSION_GREATER_EQUAL 3.20 AND OV_GENERATOR_MULTI_CONFIG) +diff -uNr openvino-2024.6.0.orig/src/plugins/intel_npu/tools/compile_tool/cmake/standalone.cmake openvino-2024.6.0/src/plugins/intel_npu/tools/compile_tool/cmake/standalone.cmake +--- openvino-2024.6.0.orig/src/plugins/intel_npu/tools/compile_tool/cmake/standalone.cmake 2024-12-27 17:04:56.868693438 -0300 ++++ openvino-2024.6.0/src/plugins/intel_npu/tools/compile_tool/cmake/standalone.cmake 2024-12-28 00:55:18.661614722 -0300 +@@ -43,5 +43,5 @@ + endif() + + install(TARGETS ${TARGET_NAME} +- DESTINATION "tools/${TARGET_NAME}" ++ DESTINATION "${CMAKE_INSTALL_DATAROOTDIR}/${PROJECT_NAME}/tools/${TARGET_NAME}" + COMPONENT npu_tools) +diff -uNr openvino-2024.6.0.orig/src/plugins/intel_npu/tools/compile_tool/CMakeLists.txt openvino-2024.6.0/src/plugins/intel_npu/tools/compile_tool/CMakeLists.txt +--- openvino-2024.6.0.orig/src/plugins/intel_npu/tools/compile_tool/CMakeLists.txt 2024-12-27 17:04:56.868693438 -0300 ++++ openvino-2024.6.0/src/plugins/intel_npu/tools/compile_tool/CMakeLists.txt 2024-12-28 02:18:52.768816190 -0300 +@@ -41,13 +41,13 @@ + # + + install(TARGETS ${TARGET_NAME} +- RUNTIME DESTINATION "tools/${TARGET_NAME}" ++ RUNTIME DESTINATION "${CMAKE_INSTALL_DATAROOTDIR}/${PROJECT_NAME}/tools/${TARGET_NAME}" + COMPONENT ${NPU_INTERNAL_COMPONENT} + ${OV_CPACK_COMP_NPU_INTERNAL_EXCLUDE_ALL}) + + if(EXISTS "${CMAKE_CURRENT_SOURCE_DIR}/README.md") + install(FILES "${CMAKE_CURRENT_SOURCE_DIR}/README.md" +- DESTINATION "tools/${TARGET_NAME}" ++ DESTINATION "${CMAKE_INSTALL_DATAROOTDIR}/${PROJECT_NAME}/tools/${TARGET_NAME}" + COMPONENT ${NPU_INTERNAL_COMPONENT} + ${OV_CPACK_COMP_NPU_INTERNAL_EXCLUDE_ALL}) + endif() diff --git a/openvino-rpmlintrc b/openvino-rpmlintrc new file mode 100644 index 0000000..4ca21df --- /dev/null +++ b/openvino-rpmlintrc @@ -0,0 +1,4 @@ +addFilter("openvino-sample.*: E: devel-file-in-non-devel-package") + +# These files are part of samples, meant for the user to copy and re-use, so env based hashbangs are preferred +addFilter("openvino-sample.*: E: env-script-interpreter") diff --git a/openvino.changes b/openvino.changes new file mode 100644 index 0000000..1cdbf81 --- /dev/null +++ b/openvino.changes @@ -0,0 +1,820 @@ +------------------------------------------------------------------- +Sun Sep 7 01:21:19 UTC 2025 - Alessandro de Oliveira Faria + +- Update to 2025.3.0 +- More GenAI coverage and framework integrations to minimize code + changes + * New models supported: Phi-4-mini-reasoning, AFM-4.5B, + Gemma-3-1B-it, Gemma-3-4B-it, and Gemma-3-12B, + * NPU support added for: Qwen3-1.7B, Qwen3-4B, and Qwen3-8B. + * LLMs optimized for NPU now available on OpenVINO Hugging + Face collection. +- Broader LLM model support and more model compression techniques + * The NPU plug-in adds support for longer contexts of up to + 8K tokens, dynamic prompts, and dynamic LoRA for improved + LLM performance. + * The NPU plug-in now supports dynamic batch sizes by reshaping + the model to a batch size of 1 and concurrently managing + multiple inference requests, enhancing performance and + optimizing memory utilization. + * Accuracy improvements for GenAI models on both built-in + and discrete graphics achieved through the implementation + of the key cache compression per channel technique, in + addition to the existing KV cache per-token compression + method. + * OpenVINO™ GenAI introduces TextRerankPipeline for improved + retrieval relevance and RAG pipeline accuracy, plus + Structured Output for enhanced response reliability and + function calling while ensuring adherence to predefined + formats. +- More portability and performance to run AI at the edge, + in the cloud, or locally. + * Announcing support for Intel® Arc™ Pro B-Series + (B50 and B60). + * Preview: Hugging Face models that are GGUF-enabled for + OpenVINO GenAI are now supported by the OpenVINO™ Model + Server for popular LLM model architectures such as + DeepSeek Distill, Qwen2, Qwen2.5, and Llama 3. + This functionality reduces memory footprint and + simplifies integration for GenAI workloads. + * With improved reliability and tool call accuracy, + the OpenVINO™ Model Server boosts support for + agentic AI use cases on AI PCs, while enhancing + performance on Intel CPUs, built-in GPUs, and NPUs. + * int4 data-aware weights compression, now supported in the + Neural Network Compression Framework (NNCF) for ONNX + models, reduces memory footprint while maintaining + accuracy and enables efficient deployment in + resource-constrained environments. + +------------------------------------------------------------------- +Wed Jun 25 01:09:14 UTC 2025 - Alessandro de Oliveira Faria + +- openSUSE Leap 16.0 compatibility + +------------------------------------------------------------------- +Tue Jun 24 05:10:06 UTC 2025 - Alessandro de Oliveira Faria + +- Remove openvino-gcc5-compatibility.patch file + +------------------------------------------------------------------- +Tue Jun 24 02:54:10 UTC 2025 - Alessandro de Oliveira Faria + +Summary of major features and improvements   +- More GenAI coverage and framework integrations to minimize code + changes + * New models supported on CPUs & GPUs: Phi-4, + Mistral-7B-Instruct-v0.3, SD-XL Inpainting 0.1, Stable + Diffusion 3.5 Large Turbo, Phi-4-reasoning, Qwen3, and + Qwen2.5-VL-3B-Instruct. Mistral 7B Instruct v0.3 is also + supported on NPUs. + * Preview: OpenVINO ™ GenAI introduces a text-to-speech + pipeline for the SpeechT5 TTS model, while the new RAG + backend offers developers a simplified API that delivers + reduced memory usage and improved performance. + * Preview: OpenVINO™ GenAI offers a GGUF Reader for seamless + integration of llama.cpp based LLMs, with Python and C++ + pipelines that load GGUF models, build OpenVINO graphs, + and run GPU inference on-the-fly. Validated for popular models: + DeepSeek-R1-Distill-Qwen (1.5B, 7B), Qwen2.5 Instruct + (1.5B, 3B, 7B) & llama-3.2 Instruct (1B, 3B, 8B). +- Broader LLM model support and more model compression + techniques + * Further optimization of LoRA adapters in OpenVINO GenAI + for improved LLM, VLM, and text-to-image model performance + on built-in GPUs. Developers can use LoRA adapters to + quickly customize models for specialized tasks. + * KV cache compression for CPUs is enabled by default for + INT8, providing a reduced memory footprint while maintaining + accuracy compared to FP16. Additionally, it delivers + substantial memory savings for LLMs with INT4 support compared + to INT8. + * Optimizations for Intel® Core™ Ultra Processor Series 2 + built-in GPUs and Intel® Arc™ B Series Graphics with the + Intel® XMX systolic platform to enhance the performance of + VLM models and hybrid quantized image generation models, as + well as improve first-token latency for LLMs through dynamic + quantization. +- More portability and performance to run AI at the edge, in the + cloud, or locally. + * Enhanced Linux* support with the latest GPU driver for + built-in GPUs on Intel® Core™ Ultra Processor Series 2 + (formerly codenamed Arrow Lake H). + * Support for INT4 data-free weights compression for ONNX + models implemented in the Neural Network Compression + Framework (NNCF). + * NPU support for FP16-NF4 precision on Intel® Core™ 200V + Series processors for models with up to 8B parameters is + enabled through symmetrical and channel-wise quantization, + improving accuracy while maintaining performance efficiency. +Support Change and Deprecation Notices +- Discontinued in 2025: + * Runtime components: + + The OpenVINO property of Affinity API is no longer + available. It has been replaced with CPU binding + configurations (ov::hint::enable_cpu_pinning). + + The openvino-nightly PyPI module has been discontinued. + End-users should proceed with the Simple PyPI nightly repo + instead. More information in Release Policy. The + openvino-nightly PyPI module has been discontinued. + End-users should proceed with the Simple PyPI nightly repo + instead. More information in Release Policy. + * Tools: + + The OpenVINO™ Development Tools package (pip install + openvino-dev) is no longer available for OpenVINO releases + in 2025. + + Model Optimizer is no longer available. Consider using the + new conversion methods instead. For more details, see the + model conversion transition guide. + + Intel® Streaming SIMD Extensions (Intel® SSE) are currently + not enabled in the binary package by default. They are + still supported in the source code form. + + Legacy prefixes: l_, w_, and m_ have been removed from + OpenVINO archive names. + * OpenVINO GenAI: + + StreamerBase::put(int64_t token) + + The Bool value for Callback streamer is no longer accepted. + It must now return one of three values of StreamingStatus + enum. + + ChunkStreamerBase is deprecated. Use StreamerBase instead. + * NNCF create_compressed_model() method is now deprecated. + nncf.quantize() method is recommended for + Quantization-Aware Training of PyTorch and TensorFlow models. + * OpenVINO Model Server (OVMS) benchmark client in C++ + using TensorFlow Serving API. +- Deprecated and to be removed in the future: + * Python 3.9 is now deprecated and will be unavailable after + OpenVINO version 2025.4. + * openvino.Type.undefined is now deprecated and will be removed + with version 2026.0. openvino.Type.dynamic should be used + instead. + * APT & YUM Repositories Restructure: Starting with release + 2025.1, users can switch to the new repository structure + for APT and YUM, which no longer uses year-based + subdirectories (like “2025”). The old (legacy) structure + will still be available until 2026, when the change will + be finalized. Detailed instructions are available on the + relevant documentation pages: + + Installation guide - yum + + Installation guide - apt + * OpenCV binaries will be removed from Docker images in 2026. + * Ubuntu 20.04 support will be deprecated in future OpenVINO + releases due to the end of standard support. + * “auto shape” and “auto batch size” (reshaping a model in + runtime) will be removed in the future. OpenVINO’s dynamic + shape models are recommended instead. + * MacOS x86 is no longer recommended for use due to the + discontinuation of validation. Full support will be removed + later in 2025. + * The openvino namespace of the OpenVINO Python API has been + redesigned, removing the nested openvino.runtime module. + The old namespace is now considered deprecated and will be + discontinued in 2026.0. + +------------------------------------------------------------------- +Wed May 21 14:43:02 UTC 2025 - Andreas Schwab + +- Fix file list for riscv64 + +------------------------------------------------------------------- +Mon May 5 07:47:30 UTC 2025 - Dominique Leuenberger + +- Do not force GCC15 on Tumblewed just yet: follow the distro + default compiler, like any other package. + +------------------------------------------------------------------- +Sat May 3 19:19:07 UTC 2025 - Alessandro de Oliveira Faria + +- openvino-gcc5-compatibility.patch to resolve incompatibility + in gcc5 + +------------------------------------------------------------------- +Thu May 1 01:06:52 UTC 2025 - Alessandro de Oliveira Faria + +- Added gcc-14 + +------------------------------------------------------------------- +Mon Apr 14 06:52:03 UTC 2025 - Alessandro de Oliveira Faria + +- Update to 2025.1.0 +- Downgrade from gcc13-c++ to 12 due to incompatibility in tbb + compilation. This was due to C++ libraries (using libstdc++) + resulting in the error: libtbb.so.12: undefined reference to + `__cxa_call_terminate@CXXABI_1.3.15' +- More GenAI coverage and framework integrations to minimize code + changes + * New models supported: Phi-4 Mini, Jina CLIP v1, and Bce + Embedding Base v1. + * OpenVINO™ Model Server now supports VLM models, including + Qwen2-VL, Phi-3.5-Vision, and InternVL2. + * OpenVINO GenAI now includes image-to-image and inpainting + features for transformer-based pipelines, such as Flux.1 and + Stable Diffusion 3 models, enhancing their ability to generate + more realistic content. + * Preview: AI Playground now utilizes the OpenVINO Gen AI backend + to enable highly optimized inferencing performance on AI PCs. + +- Broader LLM model support and more model compression techniques + * Reduced binary size through optimization of the CPU plugin and + removal of the GEMM kernel. + * Optimization of new kernels for the GPU plugin significantly + boosts the performance of Long Short-Term Memory (LSTM) models, + used in many applications, including speech recognition, + language modeling, and time series forecasting. + * Preview: Token Eviction implemented in OpenVINO GenAI to reduce + the memory consumption of KV Cache by eliminating unimportant + tokens. This current Token Eviction implementation is + beneficial for tasks where a long sequence is generated, such + as chatbots and code generation. + * NPU acceleration for text generation is now enabled in + OpenVINO™ Runtime and OpenVINO™ Model Server to support the + power-efficient deployment of VLM models on NPUs for AI PC use + cases with low concurrency. + +- More portability and performance to run AI at the edge, in the + cloud, or locally. + * Additional LLM performance optimizations on Intel® Core™ Ultra + 200H series processors for improved 2nd token latency on + Windows and Linux. + * Enhanced performance and efficient resource utilization with + the implementation of Paged Attention and Continuous Batching + by default in the GPU plugin. + * Preview: The new OpenVINO backend for Executorch will enable + accelerated inference and improved performance on Intel + hardware, including CPUs, GPUs, and NPUs. + +------------------------------------------------------------------- +Tue Mar 4 00:38:30 UTC 2025 - Alessandro de Oliveira Faria + +- Disabled JAX plugin beta. + +------------------------------------------------------------------- +Sun Feb 9 03:36:41 UTC 2025 - Alessandro de Oliveira Faria + +- Update to 2025.0.0 +- More GenAI coverage and framework integrations to minimize code + changes + * New models supported: Qwen 2.5, Deepseek-R1-Distill-Llama-8B, + DeepSeek-R1-Distill-Qwen-7B, and DeepSeek-R1-Distill-Qwen-1.5B, + FLUX.1 Schnell and FLUX.1 Dev + * Whisper Model: Improved performance on CPUs, built-in GPUs, + and discrete GPUs with GenAI API. + * Preview: Introducing NPU support for torch.compile, giving + developers the ability to use the OpenVINO backend to run the + PyTorch API on NPUs. 300+ deep learning models enabled from + the TorchVision, Timm, and TorchBench repositories.. +- Broader Large Language Model (LLM) support and more model + compression techniques. + * Preview: Addition of Prompt Lookup to GenAI API improves 2nd + token latency for LLMs by effectively utilizing predefined + prompts that match the intended use case. + * Preview: The GenAI API now offers image-to-image inpainting + functionality. This feature enables models to generate + realistic content by inpainting specified modifications and + seamlessly integrating them with the original image. + * Asymmetric KV Cache compression is now enabled for INT8 on + CPUs, resulting in lower memory consumption and improved 2nd + token latency, especially when dealing with long prompts that + require significant memory. The option should be explicitly + specified by the user. +- More portability and performance to run AI at the edge, in the + cloud, or locally. + * Support for the latest Intel® Core™ Ultra 200H series + processors (formerly codenamed Arrow Lake-H) + * Integration of the OpenVINO ™ backend with the Triton + Inference Server allows developers to utilize the Triton + server for enhanced model serving performance when deploying + on Intel CPUs. + * Preview: A new OpenVINO ™ backend integration allows + developers to leverage OpenVINO performance optimizations + directly within Keras 3 workflows for faster AI inference on + CPUs, built-in GPUs, discrete GPUs, and NPUs. This feature is + available with the latest Keras 3.8 release. + * The OpenVINO Model Server now supports native Windows Server + deployments, allowing developers to leverage better + performance by eliminating container overhead and simplifying + GPU deployment. +- Support Change and Deprecation Notices + * Now deprecated: + + Legacy prefixes l_, w_, and m_ have been removed from + OpenVINO archive names. + + The runtime namespace for Python API has been marked as + deprecated and designated to be removed for 2026.0. The + new namespace structure has been delivered, and migration + is possible immediately. Details will be communicated + through warnings andvia documentation. + + NNCF create_compressed_model() method is deprecated. + nncf.quantize() method is now recommended for + Quantization-Aware Training of PyTorch and + TensorFlow models. + +------------------------------------------------------------------- +Sun Dec 29 03:41:47 UTC 2024 - Alessandro de Oliveira Faria + +- openvino-onnx-ml-defines.patch and + openvino-remove-npu-compile-tool.patchhas been removed + as it is no longer needed in this version. +- Update to 2024.4.0 +- Summary of major features and improvements   + * OpenVINO 2024.6 release includes updates for enhanced + stability and improved LLM performance. + * Introduced support for Intel® Arc™ B-Series Graphics + (formerly known as Battlemage). + * Implemented optimizations to improve the inference time and + LLM performance on NPUs. + * Improved LLM performance with GenAI API optimizations and + bug fixes. +- Support Change and Deprecation Notices + * Using deprecated features and components is not advised. They + are available to enable a smooth transition to new solutions + and will be discontinued in the future. To keep using + discontinued features, you will have to revert to the last + LTS OpenVINO version supporting them. For more details, refer + to the OpenVINO Legacy Features and Components page. + * Discontinued in 2024.0: + + Runtime components: + - Intel® Gaussian & Neural Accelerator (Intel® GNA).. + Consider using the Neural Processing Unit (NPU) for + low-powered systems like Intel® Core™ Ultra or 14th + generation and beyond. + - OpenVINO C++/C/Python 1.0 APIs (see 2023.3 API transition + guide for reference). + - All ONNX Frontend legacy API (known as ONNX_IMPORTER_API) + - 'PerfomanceMode.UNDEFINED' property as part of the + OpenVINO Python API + + Tools: + - Deployment Manager. See installation and deployment + guides for current distribution options. + - Accuracy Checker. + - Post-Training Optimization Tool (POT). Neural Network + Compression Framework (NNCF) should be used instead. + - A Git patch for NNCF integration with  + huggingface/transformers. The recommended approach is + to use huggingface/optimum-intel for applying NNCF + optimization on top of models from Hugging Face. + - Support for Apache MXNet, Caffe, and Kaldi model formats. + Conversion to ONNX may be used as a solution. + * Deprecated and to be removed in the future: + + The macOS x86_64 debug bins will no longer be provided + with the OpenVINO toolkit, starting with OpenVINO 2024.5. + + Python 3.8 is no longer supported, starting with + OpenVINO 2024.5. + + As MxNet doesn’t support Python version higher than 3.8, + according to the MxNet PyPI project, it is no longer + supported by OpenVINO, either. + + Discrete Keem Bay support is no longer supported, starting + with OpenVINO 2024.5. + + Support for discrete devices (formerly codenamed Raptor + Lake) is no longer available for NPU. + +------------------------------------------------------------------- +Tue Dec 10 15:50:41 UTC 2024 - Giacomo Comes + +- fix build on tumbleweed + * currently openvino does not support protobuf v22 or newer + +------------------------------------------------------------------- +Tue Oct 15 00:56:54 UTC 2024 - Alessandro de Oliveira Faria + +- Temporarily inserted gcc-13 in Tumbleweed/Factory/Slowroll: + Because there is an incompatibility of the source code of the + level-zero library and npu module with gcc-14. I am working + with Intel on tests to return to native gcc. +- Update to 2024.4.0 +- Summary of major features and improvements   + * More Gen AI coverage and framework integrations to minimize + code changes + + Support for GLM-4-9B Chat, MiniCPM-1B, Llama 3 and 3.1, + Phi-3-Mini, Phi-3-Medium and YOLOX-s models. + + Noteworthy notebooks added: Florence-2, NuExtract-tiny + Structure Extraction, Flux.1 Image Generation, PixArt-α: + Photorealistic Text-to-Image Synthesis, and Phi-3-Vision + Visual Language Assistant. + * Broader Large Language Model (LLM) support and more model + compression techniques. + + OpenVINO™ runtime optimized for Intel® Xe Matrix Extensions + (Intel® XMX) systolic arrays on built-in GPUs for efficient + matrix multiplication resulting in significant LLM + performance boost with improved 1st and 2nd token + latency, as well as a smaller memory footprint on + Intel® Core™ Ultra Processors (Series 2). + + Memory sharing enabled for NPUs on Intel® Core™ Ultra + Processors (Series 2) for efficient pipeline integration + without memory copy overhead. + + Addition of the PagedAttention feature for discrete GPUs* + enables a significant boost in throughput for parallel + inferencing when serving LLMs on Intel® Arc™ Graphics + or Intel® Data Center GPU Flex Series. + * More portability and performance to run AI at the edge, + in the cloud, or locally. + + OpenVINO™ Model Server now comes with production-quality + support for OpenAI-compatible API which enables i + significantly higher throughput for parallel inferencing + on Intel® Xeon® processors when serving LLMs to many + concurrent users. + + Improved performance and memory consumption with prefix + caching, KV cache compression, and other optimizations + for serving LLMs using OpenVINO™ Model Server. + + Support for Python 3.12. +- Support Change and Deprecation Notices + * Using deprecated features and components is not advised. + They are available to enable a smooth transition to new + solutions and will be discontinued in the future. + To keep using discontinued features, you will have to + revert to the last LTS OpenVINO version supporting them. + For more details, refer to the OpenVINO Legacy Features + and Components page. + * Discontinued in 2024.0: + + Runtime components: + - Intel® Gaussian & Neural Accelerator (Intel® GNA). + Consider using the Neural Processing Unit (NPU) for + low-powered systems like Intel® Core™ Ultra or + 14th generation and beyond. + - OpenVINO C++/C/Python 1.0 APIs (see 2023.3 API + transition guide for reference). + - All ONNX Frontend legacy API (known as + ONNX_IMPORTER_API) + -'PerfomanceMode.UNDEFINED' property as part of the + OpenVINO Python API + + Tools: + - Deployment Manager. See installation and deployment + guides for current distribution options. + - Accuracy Checker. + - Post-Training Optimization Tool (POT). Neural Network + Compression Framework (NNCF) should be used instead. + - A Git patch for NNCF integration with huggingface/ + transformers. The recommended approach is to use + huggingface/optimum-intel for applying NNCF + optimization on top of models from Hugging Face. + - Support for Apache MXNet, Caffe, and Kaldi model + formats. Conversion to ONNX may be used as a + solution. + * Deprecated and to be removed in the future: + + The macOS x86_64 debug bins will no longer be + provided with the OpenVINO toolkit, starting with + OpenVINO 2024.5. + + Python 3.8 is now considered deprecated, and it will not + be available beyond the 2024.4 OpenVINO version. + + dKMB support is now considered deprecated and will be + fully removed with OpenVINO 2024.5 + + Intel® Streaming SIMD Extensions (Intel® SSE) will be + supported in source code form, but not enabled in the + binary package by default, starting with OpenVINO 2025.0 + + The openvino-nightly PyPI module will soon be discontinued. + End-users should proceed with the Simple PyPI nightly repo + instead. More information in Release Policy. + + The OpenVINO™ Development Tools package (pip install + openvino-dev) will be removed from installation options and + distribution channels beginning with OpenVINO 2025.0. + + Model Optimizer will be discontinued with OpenVINO 2025.0. + Consider using the new conversion methods instead. For more + details, see the model conversion transition guide. + + OpenVINO property Affinity API will be discontinued with + OpenVINO 2025.0. It will be replaced with CPU binding + configurations (ov::hint::enable_cpu_pinning). + + OpenVINO Model Server components: + - “auto shape” and “auto batch size” (reshaping a model in + runtime) will be removed in the future. OpenVINO’s dynamic + shape models are recommended instead. + + A number of notebooks have been deprecated. For an + up-to-date listing of available notebooks, refer to the + OpenVINO™ Notebook index (openvinotoolkit.github.io). + +------------------------------------------------------------------- +Wed Oct 2 20:56:59 UTC 2024 - Giacomo Comes + +- Add Leap15 build +- Remove comment lines in the spec file that cause the insertion + of extra lines during a commit + +------------------------------------------------------------------- +Sat Aug 10 01:41:06 UTC 2024 - Alessandro de Oliveira Faria + +- Remove NPU Compile Tool +* openvino-remove-npu-compile-tool.patch +- Update to 2024.3.0 +- Summary of major features and improvements   + * More Gen AI coverage and framework integrations to minimize + code changes + + OpenVINO pre-optimized models are now available in Hugging + Face making it easier for developers to get started with + these models. + * Broader Large Language Model (LLM) support and more model + compression techniques. + + Significant improvement in LLM performance on Intel + discrete GPUs with the addition of Multi-Head Attention + (MHA) and OneDNN enhancements. + * More portability and performance to run AI at the edge, in the + cloud, or locally. + + Improved CPU performance when serving LLMs with the + inclusion of vLLM and continuous batching in the OpenVINO + Model Server (OVMS). vLLM is an easy-to-use open-source + library that supports efficient LLM inferencing and model + serving. +- Support Change and Deprecation Notices + * Using deprecated features and components is not advised. + They are available to enable a smooth transition to new + solutions and will be discontinued in the future. To keep + using discontinued features, you will have to revert to the + last LTS OpenVINO version supporting them. For more details, + refer to the OpenVINO Legacy Features and Components page. + * Discontinued in 2024.0: + + Runtime components: + - Intel® Gaussian & Neural Accelerator (Intel® GNA)..Consider + using the Neural Processing Unit (NPU) for low-powered + systems like Intel® Core™ Ultra or 14th generation + and beyond. + - OpenVINO C++/C/Python 1.0 APIs (see 2023.3 API transition + guide for reference). + - All ONNX Frontend legacy API (known as ONNX_IMPORTER_API) + - 'PerfomanceMode.UNDEFINED' property as part of the OpenVINO + Python API + + Tools: + - Deployment Manager. See installation and deployment guides + for current distribution options. + - Accuracy Checker. + - Post-Training Optimization Tool (POT). Neural Network + Compression Framework (NNCF) should be used instead. + - A Git patch for NNCF integration with huggingface/ + transformers. The recommended approach is to use + huggingface/optimum-intel for applying NNCF optimization + on top of models from Hugging Face. + - Support for Apache MXNet, Caffe, and Kaldi model formats. + Conversion to ONNX may be used as a solution. + * Deprecated and to be removed in the future: + + The OpenVINO™ Development Tools package (pip install + openvino-dev) will be removed from installation options + and distribution channels beginning with OpenVINO 2025.0. + + Model Optimizer will be discontinued with OpenVINO 2025.0. + Consider using the new conversion methods instead. For + more details, see the model conversion transition guide. + + OpenVINO property Affinity API will be discontinued with + OpenVINO 2025.0. It will be replaced with CPU binding + configurations (ov::hint::enable_cpu_pinning). + + OpenVINO Model Server components: + - “auto shape” and “auto batch size” (reshaping a model + in runtime) will be removed in the future. OpenVINO’s + dynamic shape models are recommended instead. + + A number of notebooks have been deprecated. For an + up-to-date listing of available notebooks, refer to + the OpenVINO™ Notebook index (openvinotoolkit.github.io). + +------------------------------------------------------------------- +Sat Jun 22 12:01:23 UTC 2024 - Andreas Schwab + +- Add riscv-cpu-plugin subpackage + +------------------------------------------------------------------- +Wed Jun 19 21:36:01 UTC 2024 - Alessandro de Oliveira Faria + +- Update to 2024.2.0 +- More Gen AI coverage and framework integrations to minimize code + changes + * Llama 3 optimizations for CPUs, built-in GPUs, and discrete + GPUs for improved performance and efficient memory usage. + * Support for Phi-3-mini, a family of AI models that leverages + the power of small language models for faster, more accurate + and cost-effective text processing. + * Python Custom Operation is now enabled in OpenVINO making it + easier for Python developers to code their custom operations + instead of using C++ custom operations (also supported). + Python Custom Operation empowers users to implement their own + specialized operations into any model. + * Notebooks expansion to ensure better coverage for new models. + Noteworthy notebooks added: DynamiCrafter, YOLOv10, Chatbot + notebook with Phi-3, and QWEN2. +- Broader Large Language Model (LLM) support and more model + compression techniques. + * GPTQ method for 4-bit weight compression added to NNCF for + more efficient inference and improved performance of + compressed LLMs. + * Significant LLM performance improvements and reduced latency + for both built-in GPUs and discrete GPUs. + * Significant improvement in 2nd token latency and memory + footprint of FP16 weight LLMs on AVX2 (13th Gen Intel® Core™ + processors) and AVX512 (3rd Gen Intel® Xeon® Scalable + Processors) based CPU platforms, particularly for small + batch sizes. +- More portability and performance to run AI at the edge, in the + cloud, or locally. + * Model Serving Enhancements: + * Preview: OpenVINO Model Server (OVMS) now supports + OpenAI-compatible API along with Continuous Batching and + PagedAttention, enabling significantly higher throughput + for parallel inferencing, especially on Intel® Xeon® + processors, when serving LLMs to many concurrent users. + * OpenVINO backend for Triton Server now supports built-in + GPUs and discrete GPUs, in addition to dynamic + shapes support. + * Integration of TorchServe through torch.compile OpenVINO + backend for easy model deployment, provisioning to + multiple instances, model versioning, and maintenance. + * Preview: addition of the Generate API, a simplified API + for text generation using large language models with only + a few lines of code. The API is available through the newly + launched OpenVINO GenAI package. + * Support for Intel Atom® Processor X Series. For more details, + see System Requirements. + * Preview: Support for Intel® Xeon® 6 processor. +- Support Change and Deprecation Notices + * Using deprecated features and components is not advised. + They are available to enable a smooth transition to new + solutions and will be discontinued in the future. + To keep using discontinued features, you will have to revert + to the last LTS OpenVINO version supporting them. For more + details, refer to the OpenVINO Legacy Features and + Components page. + * Discontinued in 2024.0: + + Runtime components: + - Intel® Gaussian & Neural Accelerator (Intel® GNA). + Consider using the Neural Processing Unit (NPU) for + low-powered systems like Intel® Core™ Ultra or 14th + generation and beyond. + - OpenVINO C++/C/Python 1.0 APIs (see 2023.3 API + transition guide for reference). + - All ONNX Frontend legacy API (known as ONNX_IMPORTER_API) + - 'PerfomanceMode.UNDEFINED' property as part of the + OpenVINO Python API + + Tools: + - Deployment Manager. See installation and deployment + guides for current distribution options. + - Accuracy Checker. + - Post-Training Optimization Tool (POT). Neural Network + Compression Framework (NNCF) should be used instead. + - A Git patch for NNCF integration with  + huggingface/transformers. The recommended approach  + is to use huggingface/optimum-intel for applying NNCF + optimization on top of models from Hugging Face. + - Support for Apache MXNet, Caffe, and Kaldi model formats. + Conversion to ONNX may be used as a solution. + * Deprecated and to be removed in the future: + + The OpenVINO™ Development Tools package (pip install + openvino-dev) will be removed from installation options + and distribution channels beginning with OpenVINO 2025.0. + + Model Optimizer will be discontinued with OpenVINO 2025.0. + Consider using the new conversion methods instead. For + more details, see the model conversion transition guide. + + OpenVINO property Affinity API will be discontinued with + OpenVINO 2025.0. It will be replaced with CPU binding + configurations (ov::hint::enable_cpu_pinning). + + OpenVINO Model Server components: + + “auto shape” and “auto batch size” (reshaping a model in + runtime) will be removed in the future. OpenVINO’s dynamic + shape models are recommended instead. + + A number of notebooks have been deprecated. For an + up-to-date listing of available notebooks, refer to the + OpenVINO™ Notebook index (openvinotoolkit.github.io). + +------------------------------------------------------------------- +Thu May 9 22:56:53 UTC 2024 - Alessandro de Oliveira Faria + +- Fix sample source path in build script: + * openvino-fix-build-sample-path.patch +- Update to 2024.1.0 +- More Generative AI coverage and framework integrations to + minimize code changes. + * Mixtral and URLNet models optimized for performance + improvements on Intel® Xeon® processors. + * Stable Diffusion 1.5, ChatGLM3-6B, and Qwen-7B models + optimized for improved inference speed on Intel® Core™ + Ultra processors with integrated GPU. + * Support for Falcon-7B-Instruct, a GenAI Large Language Model + (LLM) ready-to-use chat/instruct model with superior + performance metrics. + * New Jupyter Notebooks added: YOLO V9, YOLO V8 + Oriented Bounding Boxes Detection (OOB), Stable Diffusion + in Keras, MobileCLIP, RMBG-v1.4 Background Removal, Magika, + TripoSR, AnimateAnyone, LLaVA-Next, and RAG system with + OpenVINO and LangChain. +- Broader Large Language Model (LLM) support and more model + compression techniques. + * LLM compilation time reduced through additional optimizations + with compressed embedding. Improved 1st token performance of + LLMs on 4th and 5th generations of Intel® Xeon® processors + with Intel® Advanced Matrix Extensions (Intel® AMX). + * Better LLM compression and improved performance with oneDNN, + INT4, and INT8 support for Intel® Arc™ GPUs. + * Significant memory reduction for select smaller GenAI + models on Intel® Core™ Ultra processors with integrated GPU. +- More portability and performance to run AI at the edge, + in the cloud, or locally. + * The preview NPU plugin for Intel® Core™ Ultra processors + is now available in the OpenVINO open-source GitHub + repository, in addition to the main OpenVINO package on PyPI. + * The JavaScript API is now more easily accessible through + the npm repository, enabling JavaScript developers’ seamless + access to the OpenVINO API. + * FP16 inference on ARM processors now enabled for the + Convolutional Neural Network (CNN) by default. +- Support Change and Deprecation Notices + * Using deprecated features and components is not advised. They + are available to enable a smooth transition to new solutions + and will be discontinued in the future. To keep using + Discontinued features, you will have to revert to the last + LTS OpenVINO version supporting them. + * For more details, refer to the OpenVINO Legacy Features + and Components page. + * Discontinued in 2024.0: + + Runtime components: + - Intel® Gaussian & Neural Accelerator (Intel® GNA). + Consider using the Neural Processing Unit (NPU) + for low-powered systems like Intel® Core™ Ultra or + 14th generation and beyond. + - OpenVINO C++/C/Python 1.0 APIs (see 2023.3 API + transition guide for reference). + - All ONNX Frontend legacy API (known as + ONNX_IMPORTER_API) + - 'PerfomanceMode.UNDEFINED' property as part of + the OpenVINO Python API + + Tools: + - Deployment Manager. See installation and deployment + guides for current distribution options. + - Accuracy Checker. + - Post-Training Optimization Tool (POT). Neural Network + Compression Framework (NNCF) should be used instead. + - A Git patch for NNCF integration with  + huggingface/transformers. The recommended approach +  is to use huggingface/optimum-intel for applying + NNCF optimization on top of models from Hugging + Face. + - Support for Apache MXNet, Caffe, and Kaldi model + formats. Conversion to ONNX may be used as + a solution. + * Deprecated and to be removed in the future: + + The OpenVINO™ Development Tools package (pip install + openvino-dev) will be removed from installation options + and distribution channels beginning with OpenVINO 2025.0. + + Model Optimizer will be discontinued with OpenVINO 2025.0. + Consider using the new conversion methods instead. For + more details, see the model conversion transition guide. + + OpenVINO property Affinity API will be discontinued with + OpenVINO 2025.0. It will be replaced with CPU binding + configurations (ov::hint::enable_cpu_pinning). + + OpenVINO Model Server components: + - “auto shape” and “auto batch size” (reshaping a model + in runtime) will be removed in the future. OpenVINO’s + dynamic shape models are recommended instead. + +------------------------------------------------------------------- +Tue Apr 23 18:57:17 UTC 2024 - Atri Bhattacharya + +- License update: play safe and list all third party licenses as + part of the License tag. + +------------------------------------------------------------------- +Tue Apr 23 12:42:32 UTC 2024 - Atri Bhattacharya + +- Switch to _service file as tagged Source tarball does not + include `./thirdparty` submodules. +- Update openvino-fix-install-paths.patch to fix python module + install path. +- Enable python module and split it out into a python subpackage + (for now default python3 only). +- Explicitly build python metadata (dist-info) and install it + (needs simple sed hackery to support "officially" unsupported + platform ppc64le). +- Specify ENABLE_JS=OFF to turn off javascript bindings as + building these requires downloading npm stuff from the network. +- Build with system pybind11. +- Bump _constraints for updated disk space requirements. +- Drop empty %check section, rpmlint was misleading when it + recommended adding this. + +------------------------------------------------------------------- +Fri Apr 19 08:08:02 UTC 2024 - Atri Bhattacharya + +- Numerous specfile cleanups: + * Drop redundant `mv` commands and use `install` where + appropriate. + * Build with system protobuf. + * Fix Summary tags. + * Trim package descriptions. + * Drop forcing CMAKE_BUILD_TYPE=Release, let macro default + RelWithDebInfo be used instead. + * Correct naming of shared library packages. + * Separate out libopenvino_c.so.* into own shared lib package. + * Drop rpmlintrc rule used to hide shlib naming mistakes. + * Rename Source tarball to %{name}-%{version}.EXT pattern. + * Use ldconfig_scriptlet macro for post(un). +- Add openvino-onnx-ml-defines.patch -- Define ONNX_ML at compile + time when using system onnx to allow using 'onnx-ml.pb.h' + instead of 'onnx.pb.h', the latter not being shipped with + openSUSE's onnx-devel package (gh#onnx/onnx#3074). +- Add openvino-fix-install-paths.patch: Change hard-coded install + paths in upstream cmake macro to standard Linux dirs. +- Add openvino-ComputeLibrary-include-string.patch: Include header + for std::string. +- Add external devel packages as Requires for openvino-devel. +- Pass -Wl,-z,noexecstack to %build_ldflags to avoid an exec stack + issue with intel CPU plugin. +- Use ninja for build. +- Adapt _constraits file for correct disk space and memory + requirements. +- Add empty %check section. + +------------------------------------------------------------------- +Mon Apr 15 03:18:33 UTC 2024 - Alessandro de Oliveira Faria + +- Initial package +- Version 2024.0.0 +- Add openvino-rpmlintrc. diff --git a/openvino.obsinfo b/openvino.obsinfo new file mode 100644 index 0000000..84d767e --- /dev/null +++ b/openvino.obsinfo @@ -0,0 +1,4 @@ +name: openvino +version: 2025.3.0 +mtime: 1756212984 +commit: 44526285f241251e9543276572676365fbe542a4 diff --git a/openvino.spec b/openvino.spec new file mode 100644 index 0000000..0837d81 --- /dev/null +++ b/openvino.spec @@ -0,0 +1,441 @@ +# +# spec file for package openvino +# +# Copyright (c) 2025 SUSE LLC +# Copyright (c) 2024 Alessandro de Oliveira Faria (A.K.A. CABELO) or +# +# All modifications and additions to the file contributed by third parties +# remain the property of their copyright owners, unless otherwise agreed +# upon. The license for this file, and modifications and additions to the +# file, is the same license as for the pristine package itself (unless the +# license for the pristine package is not an Open Source License, in which +# case the license is the MIT License). An "Open Source License" is a +# license that conforms to the Open Source Definition (Version 1.9) +# published by the Open Source Initiative. + +# Please submit bugfixes or comments via https://bugs.opensuse.org/ +# + + +%if 0%{?suse_version} < 1600 +%define isLeap %nil +%else +%undefine isLeap +%endif + +# Compilation takes ~1 hr on OBS for a single python, don't try all supported flavours +%if %{defined isLeap} +%define x86_64 x86_64 +%define pythons python311 +%else +%define pythons python3 +%endif +%define __builder ninja +%define so_ver 2530 +%define shlib lib%{name}%{so_ver} +%define shlib_c lib%{name}_c%{so_ver} +%define prj_name OpenVINO + +Name: openvino +Version: 2025.3.0 +Release: 0 +Summary: A toolkit for optimizing and deploying AI inference +# Let's be safe and put all third party licenses here, no matter that we use specific thirdparty libs or not +License: Apache-2.0 AND BSD-2-Clause AND BSD-3-Clause AND HPND AND JSON AND MIT AND OFL-1.1 AND Zlib +URL: https://github.com/openvinotoolkit/openvino +Source0: %{name}-%{version}.tar.zst +Source1: %{name}-rpmlintrc +# PATCH-FEATURE-OPENSUSE openvino-fix-install-paths.patch badshah400@gmail.com -- Fix installation paths hardcoded into upstream defined cmake macros +Patch0: openvino-fix-install-paths.patch +# PATCH-FIX-UPSTREAM openvino-ComputeLibrary-include-string.patch badshah400@gmail.com -- Include header for std::string +Patch1: openvino-ComputeLibrary-include-string.patch +# PATCH-FIX-UPSTREAM openvino-fix-build-sample-path.patch cabelo@opensuse.org -- Fix sample source path in build script +Patch2: openvino-fix-build-sample-path.patch +BuildRequires: ade-devel +BuildRequires: cmake +BuildRequires: fdupes +BuildRequires: ninja +BuildRequires: opencl-cpp-headers +# FIXME: /usr/include/onnx/onnx-ml.pb.h:17:2: error: This file was generated by +# an older version of protoc which is incompatible with your Protocol Buffer +# headers. Please regenerate this file with a newer version of protoc. +#BuildRequires: cmake(ONNX) +BuildRequires: pkgconfig +BuildRequires: %{python_module devel} +BuildRequires: %{python_module pip} +BuildRequires: %{python_module pybind11-devel} +BuildRequires: %{python_module setuptools} +BuildRequires: %{python_module wheel} +BuildRequires: python-rpm-macros +BuildRequires: zstd +BuildRequires: pkgconfig(flatbuffers) +BuildRequires: pkgconfig(libva) +BuildRequires: pkgconfig(nlohmann_json) +BuildRequires: pkgconfig(ocl-icd) +BuildRequires: pkgconfig(protobuf) < 22 +BuildRequires: pkgconfig(pugixml) +%if %{defined isLeap} +BuildRequires: gcc12-c++ +BuildRequires: opencl-headers +BuildRequires: snappy-devel +BuildRequires: tbb-devel +%else +BuildRequires: pkgconfig(OpenCL-Headers) +BuildRequires: pkgconfig(snappy) +BuildRequires: pkgconfig(tbb) +%endif +BuildRequires: pkgconfig(zlib) +%ifarch %{arm64} +BuildRequires: scons +%endif +# No 32-bit support +ExcludeArch: %{ix86} %{arm32} ppc +%define python_subpackage_only 1 +%python_subpackages + +%description +OpenVINO is an open-source toolkit for optimizing and deploying AI inference. + +%package -n %{shlib} +Summary: Shared library for OpenVINO toolkit + +%description -n %{shlib} +OpenVINO is an open-source toolkit for optimizing and deploying AI inference. + +This package provides the shared library for OpenVINO. + +%package -n %{shlib_c} +Summary: Shared C library for OpenVINO toolkit + +%description -n %{shlib_c} +This package provides the C library for OpenVINO. + +%package -n %{name}-devel +Summary: Headers and sources for OpenVINO toolkit +Requires: %{shlib_c} = %{version} +Requires: %{shlib} = %{version} +Requires: lib%{name}_ir_frontend%{so_ver} = %{version} +Requires: lib%{name}_onnx_frontend%{so_ver} = %{version} +Requires: lib%{name}_paddle_frontend%{so_ver} = %{version} +Requires: lib%{name}_pytorch_frontend%{so_ver} = %{version} +Requires: lib%{name}_tensorflow_frontend%{so_ver} = %{version} +Requires: lib%{name}_tensorflow_lite_frontend%{so_ver} = %{version} +Requires: pkgconfig(flatbuffers) +Requires: pkgconfig(libva) +Requires: pkgconfig(nlohmann_json) +Requires: pkgconfig(ocl-icd) +Requires: pkgconfig(protobuf) +Requires: pkgconfig(pugixml) +%if %{defined isLeap} +Requires: opencl-headers +Requires: snappy-devel +Requires: tbb-devel +%else +Requires: pkgconfig(OpenCL-Headers) +Requires: pkgconfig(snappy) +Requires: pkgconfig(tbb) +%endif +Recommends: %{name}-auto-batch-plugin = %{version} +Recommends: %{name}-auto-plugin = %{version} +Recommends: %{name}-hetero-plugin = %{version} +Recommends: %{name}-intel-cpu-plugin = %{version} +%ifarch riscv64 +Recommends: %{name}-riscv-cpu-plugin = %{version} +%endif + +%description -n %{name}-devel +OpenVINO is an open-source toolkit for optimizing and deploying AI inference. + +This package provides the headers and sources for developing applications with +OpenVINO. + +%package -n %{name}-arm-cpu-plugin +Summary: Intel CPU plugin for OpenVINO toolkit + +%description -n %{name}-arm-cpu-plugin +OpenVINO is an open-source toolkit for optimizing and deploying AI inference. + +This package provides the ARM CPU plugin for OpenVINO on %{arm64} archs. + +%package -n %{name}-riscv-cpu-plugin +Summary: RISC-V CPU plugin for OpenVINO toolkit + +%description -n %{name}-riscv-cpu-plugin +OpenVINO is an open-source toolkit for optimizing and deploying AI inference. + +This package provides the RISC-V CPU plugin for OpenVINO on riscv64 archs. + +%package -n %{name}-auto-plugin +Summary: Auto / Multi software plugin for OpenVINO toolkit + +%description -n %{name}-auto-plugin +OpenVINO is an open-source toolkit for optimizing and deploying AI inference. + +This package provides the Auto / Multi software plugin for OpenVINO. + +%package -n %{name}-auto-batch-plugin +Summary: Automatic batch software plugin for OpenVINO toolkit + +%description -n %{name}-auto-batch-plugin +OpenVINO is an open-source toolkit for optimizing and deploying AI inference. + +This package provides the automatic batch software plugin for OpenVINO. + +%package -n %{name}-hetero-plugin +Summary: Hetero frontend for Intel OpenVINO toolkit + +%description -n %{name}-hetero-plugin +OpenVINO is an open-source toolkit for optimizing and deploying AI inference. + +This package provides the hetero frontend for OpenVINO. + +%package -n %{name}-intel-cpu-plugin +Summary: Intel CPU plugin for OpenVINO toolkit + +%description -n %{name}-intel-cpu-plugin +OpenVINO is an open-source toolkit for optimizing and deploying AI inference. + +This package provides the intel CPU plugin for OpenVINO for %{x86_64} archs. + +%package -n %{name}-intel-npu-plugin +Summary: Intel NPU plugin for OpenVINO toolkit + +%description -n %{name}-intel-npu-plugin +OpenVINO is an open-source toolkit for optimizing and deploying AI inference. + +This package provides the intel NPU plugin for OpenVINO for %{x86_64} archs. + +%package -n lib%{name}_ir_frontend%{so_ver} +Summary: Paddle frontend for Intel OpenVINO toolkit + +%description -n lib%{name}_ir_frontend%{so_ver} +OpenVINO is an open-source toolkit for optimizing and deploying AI inference. + +This package provides the ir frontend for OpenVINO. + +%package -n lib%{name}_onnx_frontend%{so_ver} +Summary: Onnx frontend for OpenVINO toolkit + +%description -n lib%{name}_onnx_frontend%{so_ver} +OpenVINO is an open-source toolkit for optimizing and deploying AI inference. + +This package provides the onnx frontend for OpenVINO. + +%package -n lib%{name}_paddle_frontend%{so_ver} +Summary: Paddle frontend for Intel OpenVINO toolkit + +%description -n lib%{name}_paddle_frontend%{so_ver} +OpenVINO is an open-source toolkit for optimizing and deploying AI inference. + +This package provides the paddle frontend for OpenVINO. + +%package -n lib%{name}_pytorch_frontend%{so_ver} +Summary: PyTorch frontend for OpenVINO toolkit + +%description -n lib%{name}_pytorch_frontend%{so_ver} +OpenVINO is an open-source toolkit for optimizing and deploying AI inference. + +This package provides the pytorch frontend for OpenVINO. + +%package -n lib%{name}_tensorflow_frontend%{so_ver} +Summary: TensorFlow frontend for OpenVINO toolkit + +%description -n lib%{name}_tensorflow_frontend%{so_ver} +OpenVINO is an open-source toolkit for optimizing and deploying AI inference. + +This package provides the tensorflow frontend for OpenVINO. + +%package -n lib%{name}_tensorflow_lite_frontend%{so_ver} +Summary: TensorFlow Lite frontend for OpenVINO toolkit + +%description -n lib%{name}_tensorflow_lite_frontend%{so_ver} +OpenVINO is an open-source toolkit for optimizing and deploying AI inference. + +This package provides the tensorflow-lite frontend for OpenVINO. + +%package -n python-openvino +Summary: Python module for openVINO toolkit +Requires: python-numpy < 2 +Requires: python-openvino-telemetry + +%description -n python-openvino +OpenVINO is an open-source toolkit for optimizing and deploying AI inference. + +This package provides a Python module for interfacing with openVINO toolkit. + +%package -n %{name}-sample +Summary: Samples for use with OpenVINO toolkit +BuildArch: noarch + +%description -n %{name}-sample +OpenVINO is an open-source toolkit for optimizing and deploying AI inference. + +This package provides some samples for use with openVINO. + +%prep +%autosetup -p1 + +%build +%if %{defined isLeap} +export CC=gcc-12 CXX=g++-12 +%endif +# Otherwise intel_cpu plugin declares an executable stack +%ifarch %{x86_64} +%define build_ldflags -Wl,-z,noexecstack +%endif +%cmake \ + -DCMAKE_CXX_STANDARD=17 \ + -DBUILD_SHARED_LIBS=ON \ + -DENABLE_OV_ONNX_FRONTEND=ON \ + -DENABLE_OV_PADDLE_FRONTEND=ON \ + -DENABLE_OV_PYTORCH_FRONTEND=ON \ + -DENABLE_OV_IR_FRONTEND=ON \ + -DENABLE_OV_TF_FRONTEND=ON \ + -DENABLE_OV_TF_LITE_FRONTEND=ON \ + -DENABLE_INTEL_GPU=OFF \ + -DENABLE_OV_JAX_FRONTEND=OFF \ + -DENABLE_JS=OFF \ + -DENABLE_PYTHON=ON \ + -DENABLE_WHEEL=OFF \ + -DENABLE_SYSTEM_OPENCL=ON \ + -DENABLE_SYSTEM_PROTOBUF=ON \ + -DENABLE_SYSTEM_PUGIXML=ON \ + -DENABLE_SYSTEM_SNAPPY=ON \ +%if %{defined isLeap} + -DENABLE_TBBBIND_2_5=OFF \ +%endif + -DENABLE_SYSTEM_TBB=ON \ + -DONNX_USE_PROTOBUF_SHARED_LIBS=ON \ + -DProtobuf_USE_STATIC_LIBS=OFF \ + %{nil} +%cmake_build +# Manually generate dist-info dir +export WHEEL_VERSION=%{version} \ + BUILD_TYPE=RelWithDebInfo +%ifarch %{power64} + +# RelWithDebInfo +# Manual hackery for power64 because it not "officially" supported + sed -i "s/{ARCH}/%{_arch}/" ../src/bindings/python/wheel/setup.py +%endif +%python_exec ../setup.py dist_info -o ../ + +%install +%cmake_install + +# Hash-bangs in non-exec python sample scripts +sed -Ei "1{\@/usr/bin/env@d}" \ + %{buildroot}%{_datadir}/%{prj_name}/samples/python/benchmark/bert_benchmark/bert_benchmark.py \ + %{buildroot}%{_datadir}/%{prj_name}/samples/python/benchmark/sync_benchmark/sync_benchmark.py \ + %{buildroot}%{_datadir}/%{prj_name}/samples/python/benchmark/throughput_benchmark/throughput_benchmark.py \ + %{buildroot}%{_datadir}/%{prj_name}/samples/python/classification_sample_async/classification_sample_async.py \ + %{buildroot}%{_datadir}/%{prj_name}/samples/python/hello_classification/hello_classification.py \ + %{buildroot}%{_datadir}/%{prj_name}/samples/python/hello_query_device/hello_query_device.py \ + %{buildroot}%{_datadir}/%{prj_name}/samples/python/hello_reshape_ssd/hello_reshape_ssd.py \ + %{buildroot}%{_datadir}/%{prj_name}/samples/python/model_creation_sample/model_creation_sample.py + +# Unnecessary if we get our package dependencies and lib paths right! +rm -fr %{buildroot}%{_prefix}/install_dependencies \ + %{buildroot}%{_prefix}/setupvars.sh + +%{python_expand rm %{buildroot}%{$python_sitearch}/requirements.txt +chmod -x %{buildroot}%{$python_sitearch}/%{name}/tools/ovc/ovc.py +cp -r %{name}-%{version}.dist-info %{buildroot}%{$python_sitearch}/ +%fdupes %{buildroot}%{$python_sitearch}/%{name}/ +} + +%fdupes %{buildroot}%{_datadir}/ + +# We do not use bundled thirdparty libs +rm -fr %{buildroot}%{_datadir}/licenses/* + +%ldconfig_scriptlets -n %{shlib} +%ldconfig_scriptlets -n %{shlib_c} +%ldconfig_scriptlets -n lib%{name}_ir_frontend%{so_ver} +%ldconfig_scriptlets -n lib%{name}_onnx_frontend%{so_ver} +%ldconfig_scriptlets -n lib%{name}_paddle_frontend%{so_ver} +%ldconfig_scriptlets -n lib%{name}_pytorch_frontend%{so_ver} +%ldconfig_scriptlets -n lib%{name}_tensorflow_lite_frontend%{so_ver} +%ldconfig_scriptlets -n lib%{name}_tensorflow_frontend%{so_ver} + +%files -n %{shlib} +%license LICENSE +%{_libdir}/libopenvino.so.* + +%files -n %{shlib_c} +%license LICENSE +%{_libdir}/libopenvino_c.so.* + +%files -n %{name}-auto-batch-plugin +%dir %{_libdir}/%{prj_name} +%{_libdir}/%{prj_name}/libopenvino_auto_batch_plugin.so + +%files -n %{name}-auto-plugin +%dir %{_libdir}/%{prj_name} +%{_libdir}/%{prj_name}/libopenvino_auto_plugin.so + +%ifarch %{x86_64} +%files -n %{name}-intel-cpu-plugin +%dir %{_libdir}/%{prj_name} +%{_libdir}/%{prj_name}/libopenvino_intel_cpu_plugin.so + +%files -n %{name}-intel-npu-plugin +%dir %{_libdir}/%{prj_name} +%{_libdir}/%{prj_name}/libopenvino_intel_npu_plugin.so +%endif + +%ifarch %{arm64} +%files -n %{name}-arm-cpu-plugin +%dir %{_libdir}/%{prj_name} +%{_libdir}/%{prj_name}/libopenvino_arm_cpu_plugin.so +%endif + +%ifarch riscv64 +%files -n %{name}-riscv-cpu-plugin +%dir %{_libdir}/%{prj_name} +%{_libdir}/%{prj_name}/libopenvino_riscv_cpu_plugin.so +%endif + +%files -n %{name}-hetero-plugin +%dir %{_libdir}/%{prj_name} +%{_libdir}/%{prj_name}/libopenvino_hetero_plugin.so + +%files -n lib%{name}_onnx_frontend%{so_ver} +%{_libdir}/libopenvino_onnx_frontend.so.* + +%files -n lib%{name}_ir_frontend%{so_ver} +%{_libdir}/libopenvino_ir_frontend.so.* + +%files -n lib%{name}_paddle_frontend%{so_ver} +%{_libdir}/libopenvino_paddle_frontend.so.* + +%files -n lib%{name}_pytorch_frontend%{so_ver} +%{_libdir}/libopenvino_pytorch_frontend.so.* + +%files -n lib%{name}_tensorflow_frontend%{so_ver} +%{_libdir}/libopenvino_tensorflow_frontend.so.* + +%files -n lib%{name}_tensorflow_lite_frontend%{so_ver} +%{_libdir}/libopenvino_tensorflow_lite_frontend.so.* + +%files -n %{name}-sample +%license LICENSE +%{_datadir}/%{prj_name}/ +%exclude %{_prefix}/lib/debug/usr/share/OpenVINO/tools/compile_tool/ + +%files -n %{name}-devel +%license LICENSE +%{_includedir}/%{name}/ +%{_libdir}/cmake/%{prj_name}/ +%ifarch riscv64 +%{_libdir}/cmake/xbyak_riscv/ +%endif +%{_libdir}/*.so +%{_libdir}/pkgconfig/openvino.pc + +%files %{python_files openvino} +%license LICENSE +%{python_sitearch}/openvino/ +%{python_sitearch}/openvino*-info/ + +%changelog -- 2.51.1 From 3b365b87989db534b10ebf872d3e41395b9f43f5a3ea1428199c42afc94a06f4 Mon Sep 17 00:00:00 2001 From: Alessandro de Oliveira Faria Date: Wed, 3 Dec 2025 02:10:49 +0000 Subject: [PATCH 2/3] =?UTF-8?q?-=20Update=20to=202025.4.0=20-=20More=20Gen?= =?UTF-8?q?AI=20coverage=20and=20framework=20integrations=20to=20minimize?= =?UTF-8?q?=20code=20=20=20changes=20=20=20*=20New=20models=20supported:?= =?UTF-8?q?=20=20=20=20=20+=20On=20CPUs=20&=20GPUs:=20Qwen3-Embedding-0.6B?= =?UTF-8?q?,=20Qwen3-Reranker-0.6B,=20=20=20=20=20=20=20Mistral-Small-24B-?= =?UTF-8?q?Instruct-2501.=20=20=20=20=20+=20On=20NPUs:=20Gemma-3-4b-it=20a?= =?UTF-8?q?nd=20Qwen2.5-VL-3B-Instruct.=20=20=20*=20Preview:=20Mixture=20o?= =?UTF-8?q?f=20Experts=20(MoE)=20models=20optimized=20for=20CPUs=20=20=20?= =?UTF-8?q?=20=20and=20GPUs,=20validated=20for=20Qwen3-30B-A3B.=20=20=20*?= =?UTF-8?q?=20GenAI=20pipeline=20integrations:=20Qwen3-Embedding-0.6B=20an?= =?UTF-8?q?d=20=20=20=20=20Qwen3-Reranker-0.6B=20for=20enhanced=20retrieva?= =?UTF-8?q?l/ranking,=20and=20=20=20=20=20Qwen2.5VL-7B=20for=20video=20pip?= =?UTF-8?q?eline.=20-=20Broader=20LLM=20model=20support=20and=20more=20mod?= =?UTF-8?q?el=20compression=20=20=20techniques=20=20=20*=20Gold=20support?= =?UTF-8?q?=20for=20Windows=20ML*=20enables=20developers=20to=20deploy=20A?= =?UTF-8?q?I=20=20=20=20=20models=20and=20applications=20effortlessly=20ac?= =?UTF-8?q?ross=20CPUs,=20GPUs,=20and=20=20=20=20=20NPUs=20on=20Intel?= =?UTF-8?q?=C2=AE=20Core=E2=84=A2=20Ultra=20processor-powered=20AI=20PCs.?= =?UTF-8?q?=20=20=20*=20The=20Neural=20Network=20Compression=20Framework?= =?UTF-8?q?=20(NNCF)=20ONNX=20backend=20=20=20=20=20now=20supports=20INT8?= =?UTF-8?q?=20static=20post-training=20quantization=20(PTQ)=20=20=20=20=20?= =?UTF-8?q?and=20INT8/INT4=20weight-only=20compression=20to=20ensure=20acc?= =?UTF-8?q?uracy=20=20=20=20=20parity=20with=20OpenVINO=20IR=20format=20mo?= =?UTF-8?q?dels.=20SmoothQuant=20algorithm=20=20=20=20=20support=20added?= =?UTF-8?q?=20for=20INT8=20quantization.=20=20=20*=20Accelerated=20multi-t?= =?UTF-8?q?oken=20generation=20for=20GenAI,=20leveraging=20=20=20=20=20opt?= =?UTF-8?q?imized=20GPU=20kernels=20to=20deliver=20faster=20inference,=20s?= =?UTF-8?q?marter=20=20=20=20=20KV-cache=20reuse,=20and=20scalable=20LLM?= =?UTF-8?q?=20performance.=20=20=20*=20GPU=20plugin=20updates=20include=20?= =?UTF-8?q?improved=20performance=20with=20prefix=20=20=20=20=20caching=20?= =?UTF-8?q?for=20chat=20history=20scenarios=20and=20enhanced=20LLM=20accur?= =?UTF-8?q?acy=20=20=20=20=20with=20dynamic=20quantization=20support=20for?= =?UTF-8?q?=20INT8.=20-=20More=20portability=20and=20performance=20to=20ru?= =?UTF-8?q?n=20AI=20at=20the=20edge,=20in=20the=20=20=20cloud,=20or=20loca?= =?UTF-8?q?lly.=20=20=20*=20Announcing=20support=20for=20Intel=C2=AE=20Cor?= =?UTF-8?q?e=20Ultra=20Processor=20Series=203.=20=20=20*=20Encrypted=20blo?= =?UTF-8?q?b=20format=20support=20added=20for=20secure=20model=20=20=20=20?= =?UTF-8?q?=20deployment=20with=20OpenVINO=20GenAI.=20Model=20weights=20an?= =?UTF-8?q?d=20artifacts=20=20=20=20=20are=20stored=20and=20transmitted=20?= =?UTF-8?q?in=20an=20encrypted=20format,=20reducing=20=20=20=20=20risks=20?= =?UTF-8?q?of=20IP=20theft=20during=20deployment.=20Developers=20can=20dep?= =?UTF-8?q?loy=20=20=20=20=20with=20minimal=20code=20changes=20using=20Ope?= =?UTF-8?q?nVINO=20GenAI=20pipelines.=20=20=20*=20OpenVINO=E2=84=A2=20Mode?= =?UTF-8?q?l=20Server=20and=20OpenVINO=E2=84=A2=20GenAI=20now=20extend=20?= =?UTF-8?q?=20=20=20=20support=20for=20Agentic=20AI=20scenarios=20with=20n?= =?UTF-8?q?ew=20features=20such=20as=20=20=20=20=20output=20parsing=20and?= =?UTF-8?q?=20improved=20chat=20templates=20for=20reliable=20=20=20=20=20m?= =?UTF-8?q?ulti-turn=20interactions,=20and=20preview=20functionality=20for?= =?UTF-8?q?=20the=20=20=20=20=20Qwen3-30B-A3B=20model.=20OVMS=20also=20int?= =?UTF-8?q?roduces=20a=20preview=20for=20=20=20=20=20audio=20endpoints.=20?= =?UTF-8?q?=20=20*=20NPU=20deployment=20is=20simplified=20with=20batch=20s?= =?UTF-8?q?upport,=20enabling=20=20=20=20=20seamless=20model=20execution?= =?UTF-8?q?=20across=20Intel=C2=AE=20Core=20Ultra=20=20=20=20=20processors?= =?UTF-8?q?=20while=20eliminating=20driver=20dependencies.=20Models=20=20?= =?UTF-8?q?=20=20=20are=20reshaped=20to=20batch=5Fsize=3D1=20before=20comp?= =?UTF-8?q?ilation.=20=20=20*=20The=20improved=20NVIDIA=20Triton=20Server*?= =?UTF-8?q?=20integration=20with=20=20=20=20=20OpenVINO=20backend=20now=20?= =?UTF-8?q?enables=20developers=20to=20utilize=20Intel=20=20=20=20=20GPUs?= =?UTF-8?q?=20or=20NPUs=20for=20deployment.?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit OBS-URL: https://build.opensuse.org/package/show/science:machinelearning/openvino?expand=0&rev=44 --- .gitattributes | 23 + .gitignore | 1 + _constraints | 11 + _service | 16 + openvino-2025.3.0.obscpio | 3 + openvino-2025.4.0.obscpio | 3 + openvino-ComputeLibrary-include-string.patch | 11 + openvino-fix-build-sample-path.patch | 12 + openvino-fix-install-paths.patch | 87 ++ openvino-rpmlintrc | 4 + openvino.changes | 873 +++++++++++++++++++ openvino.obsinfo | 4 + openvino.spec | 441 ++++++++++ 13 files changed, 1489 insertions(+) create mode 100644 .gitattributes create mode 100644 .gitignore create mode 100644 _constraints create mode 100644 _service create mode 100644 openvino-2025.3.0.obscpio create mode 100644 openvino-2025.4.0.obscpio create mode 100644 openvino-ComputeLibrary-include-string.patch create mode 100644 openvino-fix-build-sample-path.patch create mode 100644 openvino-fix-install-paths.patch create mode 100644 openvino-rpmlintrc create mode 100644 openvino.changes create mode 100644 openvino.obsinfo create mode 100644 openvino.spec diff --git a/.gitattributes b/.gitattributes new file mode 100644 index 0000000..9b03811 --- /dev/null +++ b/.gitattributes @@ -0,0 +1,23 @@ +## Default LFS +*.7z filter=lfs diff=lfs merge=lfs -text +*.bsp filter=lfs diff=lfs merge=lfs -text +*.bz2 filter=lfs diff=lfs merge=lfs -text +*.gem filter=lfs diff=lfs merge=lfs -text +*.gz filter=lfs diff=lfs merge=lfs -text +*.jar filter=lfs diff=lfs merge=lfs -text +*.lz filter=lfs diff=lfs merge=lfs -text +*.lzma filter=lfs diff=lfs merge=lfs -text +*.obscpio filter=lfs diff=lfs merge=lfs -text +*.oxt filter=lfs diff=lfs merge=lfs -text +*.pdf filter=lfs diff=lfs merge=lfs -text +*.png filter=lfs diff=lfs merge=lfs -text +*.rpm filter=lfs diff=lfs merge=lfs -text +*.tbz filter=lfs diff=lfs merge=lfs -text +*.tbz2 filter=lfs diff=lfs merge=lfs -text +*.tgz filter=lfs diff=lfs merge=lfs -text +*.ttf filter=lfs diff=lfs merge=lfs -text +*.txz filter=lfs diff=lfs merge=lfs -text +*.whl filter=lfs diff=lfs merge=lfs -text +*.xz filter=lfs diff=lfs merge=lfs -text +*.zip filter=lfs diff=lfs merge=lfs -text +*.zst filter=lfs diff=lfs merge=lfs -text diff --git a/.gitignore b/.gitignore new file mode 100644 index 0000000..57affb6 --- /dev/null +++ b/.gitignore @@ -0,0 +1 @@ +.osc diff --git a/_constraints b/_constraints new file mode 100644 index 0000000..69ac422 --- /dev/null +++ b/_constraints @@ -0,0 +1,11 @@ + + + + + 20 + + + 8 + + + diff --git a/_service b/_service new file mode 100644 index 0000000..8ec83c9 --- /dev/null +++ b/_service @@ -0,0 +1,16 @@ + + + https://github.com/openvinotoolkit/openvino.git + git + 2025.4.0 + 2025.4.0 + enable + openvino + .git + + + + *.tar + zstd + + diff --git a/openvino-2025.3.0.obscpio b/openvino-2025.3.0.obscpio new file mode 100644 index 0000000..dcac4b0 --- /dev/null +++ b/openvino-2025.3.0.obscpio @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:64424d07d5017ed8773c86d9bfa80af8d436d70aa55cb6c4b0a26ca1b2804b1e +size 744031247 diff --git a/openvino-2025.4.0.obscpio b/openvino-2025.4.0.obscpio new file mode 100644 index 0000000..451c150 --- /dev/null +++ b/openvino-2025.4.0.obscpio @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:deda1db3ae8e8acb506d8937ff4709332bfa0380de14393c6f030b88dd2fc5c4 +size 753350671 diff --git a/openvino-ComputeLibrary-include-string.patch b/openvino-ComputeLibrary-include-string.patch new file mode 100644 index 0000000..44006ce --- /dev/null +++ b/openvino-ComputeLibrary-include-string.patch @@ -0,0 +1,11 @@ +diff -uNr openvino-2025.2.0.orig/src/plugins/intel_cpu/thirdparty/ComputeLibrary/arm_compute/core/utils/logging/IPrinter.h openvino-2025.2.0/src/plugins/intel_cpu/thirdparty/ComputeLibrary/arm_compute/core/utils/logging/IPrinter.h +--- openvino-2025.2.0.orig/src/plugins/intel_cpu/thirdparty/ComputeLibrary/arm_compute/core/utils/logging/IPrinter.h 2025-06-22 18:14:56.561471325 -0300 ++++ openvino-2025.2.0/src/plugins/intel_cpu/thirdparty/ComputeLibrary/arm_compute/core/utils/logging/IPrinter.h 2025-06-22 18:15:23.466678704 -0300 +@@ -29,6 +29,7 @@ + */ + + #include "support/Mutex.h" ++#include + + namespace arm_compute + { diff --git a/openvino-fix-build-sample-path.patch b/openvino-fix-build-sample-path.patch new file mode 100644 index 0000000..f0742b1 --- /dev/null +++ b/openvino-fix-build-sample-path.patch @@ -0,0 +1,12 @@ +diff -uNr openvino.orig/samples/cpp/build_samples.sh openvino/samples/cpp/build_samples.sh +--- openvino.orig/samples/cpp/build_samples.sh 2024-04-25 01:04:42.451868881 -0300 ++++ openvino/samples/cpp/build_samples.sh 2024-04-25 01:05:04.678342617 -0300 +@@ -59,7 +59,7 @@ + printf "\nSetting environment variables for building samples...\n" + + if [ -z "$INTEL_OPENVINO_DIR" ]; then +- if [[ "$SAMPLES_SOURCE_DIR" = "/usr/share/openvino"* ]]; then ++ if [[ "$SAMPLES_SOURCE_DIR" = "/usr/share/OpenVINO"* ]]; then + true + elif [ -e "$SAMPLES_SOURCE_DIR/../../setupvars.sh" ]; then + setupvars_path="$SAMPLES_SOURCE_DIR/../../setupvars.sh" diff --git a/openvino-fix-install-paths.patch b/openvino-fix-install-paths.patch new file mode 100644 index 0000000..c18c137 --- /dev/null +++ b/openvino-fix-install-paths.patch @@ -0,0 +1,87 @@ +diff -uNr openvino-2024.6.0.orig/cmake/developer_package/packaging/archive.cmake openvino-2024.6.0/cmake/developer_package/packaging/archive.cmake +--- openvino-2024.6.0.orig/cmake/developer_package/packaging/archive.cmake 2024-12-27 17:04:54.520685198 -0300 ++++ openvino-2024.6.0/cmake/developer_package/packaging/archive.cmake 2024-12-27 17:02:57.644273948 -0300 +@@ -25,14 +25,18 @@ + macro(ov_archive_cpack_set_dirs) + # common "archive" package locations + # TODO: move current variables to OpenVINO specific locations +- set(OV_CPACK_INCLUDEDIR runtime/include) +- set(OV_CPACK_OPENVINO_CMAKEDIR runtime/cmake) +- set(OV_CPACK_DOCDIR docs) +- set(OV_CPACK_LICENSESDIR licenses) +- set(OV_CPACK_SAMPLESDIR samples) +- set(OV_CPACK_WHEELSDIR wheels) +- set(OV_CPACK_DEVREQDIR tools) +- set(OV_CPACK_PYTHONDIR python) ++ set(OV_CPACK_INCLUDEDIR include) ++ set(OV_CPACK_OPENVINO_CMAKEDIR ${CMAKE_INSTALL_LIBDIR}/cmake/${PROJECT_NAME}) ++ set(OV_CPACK_DOCDIR ${CMAKE_INSTALL_DOCDIR}) ++ set(OV_CPACK_LICENSESDIR ${CMAKE_INSTALL_DATAROOTDIR}/licenses/${PROJECT_NAME}) ++ set(OV_CPACK_SAMPLESDIR ${CMAKE_INSTALL_DATAROOTDIR}/${PROJECT_NAME}/samples) ++ if (ENABLE_PYTHON) ++ find_package(Python3 QUIET COMPONENTS Interpreter) ++ file(RELATIVE_PATH OV_PYTHON_MODPATH ${CMAKE_INSTALL_PREFIX} ${Python3_SITEARCH}) ++ set(OV_CPACK_WHEELSDIR tools) ++ set(OV_CPACK_DEVREQDIR tools) ++ set(OV_CPACK_PYTHONDIR ${OV_PYTHON_MODPATH}) ++ endif() + + if(USE_BUILD_TYPE_SUBFOLDER) + set(build_type ${CMAKE_BUILD_TYPE}) +@@ -49,11 +53,12 @@ + set(OV_CPACK_RUNTIMEDIR runtime/lib/${ARCH_FOLDER}/${build_type}) + set(OV_CPACK_ARCHIVEDIR runtime/lib/${ARCH_FOLDER}/${build_type}) + else() +- set(OV_CPACK_LIBRARYDIR runtime/lib/${ARCH_FOLDER}) +- set(OV_CPACK_RUNTIMEDIR runtime/lib/${ARCH_FOLDER}) +- set(OV_CPACK_ARCHIVEDIR runtime/lib/${ARCH_FOLDER}) ++ set(OV_CPACK_LIBRARYDIR ${CMAKE_INSTALL_LIBDIR}) ++ set(OV_CPACK_RUNTIMEDIR ${CMAKE_INSTALL_LIBDIR}) ++ set(OV_CPACK_ARCHIVEDIR ${CMAKE_INSTALL_LIBDIR}) + endif() +- set(OV_CPACK_PLUGINSDIR ${OV_CPACK_RUNTIMEDIR}) ++ set(OV_CPACK_PLUGINSDIR ${OV_CPACK_RUNTIMEDIR}/${PROJECT_NAME}) ++ + endmacro() + + ov_archive_cpack_set_dirs() +diff -uNr openvino-2024.6.0.orig/src/cmake/openvino.cmake openvino-2024.6.0/src/cmake/openvino.cmake +--- openvino-2024.6.0.orig/src/cmake/openvino.cmake 2024-12-27 17:04:55.240687724 -0300 ++++ openvino-2024.6.0/src/cmake/openvino.cmake 2024-12-27 17:03:50.176459053 -0300 +@@ -267,6 +267,7 @@ + + # define relative paths + file(RELATIVE_PATH PKGCONFIG_OpenVINO_PREFIX "/${OV_CPACK_RUNTIMEDIR}/pkgconfig" "/") ++ cmake_path(NORMAL_PATH PKGCONFIG_OpenVINO_PREFIX) + + set(pkgconfig_in "${OpenVINO_SOURCE_DIR}/cmake/templates/openvino.pc.in") + if(CMAKE_VERSION VERSION_GREATER_EQUAL 3.20 AND OV_GENERATOR_MULTI_CONFIG) +diff -uNr openvino-2024.6.0.orig/src/plugins/intel_npu/tools/compile_tool/cmake/standalone.cmake openvino-2024.6.0/src/plugins/intel_npu/tools/compile_tool/cmake/standalone.cmake +--- openvino-2024.6.0.orig/src/plugins/intel_npu/tools/compile_tool/cmake/standalone.cmake 2024-12-27 17:04:56.868693438 -0300 ++++ openvino-2024.6.0/src/plugins/intel_npu/tools/compile_tool/cmake/standalone.cmake 2024-12-28 00:55:18.661614722 -0300 +@@ -43,5 +43,5 @@ + endif() + + install(TARGETS ${TARGET_NAME} +- DESTINATION "tools/${TARGET_NAME}" ++ DESTINATION "${CMAKE_INSTALL_DATAROOTDIR}/${PROJECT_NAME}/tools/${TARGET_NAME}" + COMPONENT npu_tools) +diff -uNr openvino-2024.6.0.orig/src/plugins/intel_npu/tools/compile_tool/CMakeLists.txt openvino-2024.6.0/src/plugins/intel_npu/tools/compile_tool/CMakeLists.txt +--- openvino-2024.6.0.orig/src/plugins/intel_npu/tools/compile_tool/CMakeLists.txt 2024-12-27 17:04:56.868693438 -0300 ++++ openvino-2024.6.0/src/plugins/intel_npu/tools/compile_tool/CMakeLists.txt 2024-12-28 02:18:52.768816190 -0300 +@@ -41,13 +41,13 @@ + # + + install(TARGETS ${TARGET_NAME} +- RUNTIME DESTINATION "tools/${TARGET_NAME}" ++ RUNTIME DESTINATION "${CMAKE_INSTALL_DATAROOTDIR}/${PROJECT_NAME}/tools/${TARGET_NAME}" + COMPONENT ${NPU_INTERNAL_COMPONENT} + ${OV_CPACK_COMP_NPU_INTERNAL_EXCLUDE_ALL}) + + if(EXISTS "${CMAKE_CURRENT_SOURCE_DIR}/README.md") + install(FILES "${CMAKE_CURRENT_SOURCE_DIR}/README.md" +- DESTINATION "tools/${TARGET_NAME}" ++ DESTINATION "${CMAKE_INSTALL_DATAROOTDIR}/${PROJECT_NAME}/tools/${TARGET_NAME}" + COMPONENT ${NPU_INTERNAL_COMPONENT} + ${OV_CPACK_COMP_NPU_INTERNAL_EXCLUDE_ALL}) + endif() diff --git a/openvino-rpmlintrc b/openvino-rpmlintrc new file mode 100644 index 0000000..4ca21df --- /dev/null +++ b/openvino-rpmlintrc @@ -0,0 +1,4 @@ +addFilter("openvino-sample.*: E: devel-file-in-non-devel-package") + +# These files are part of samples, meant for the user to copy and re-use, so env based hashbangs are preferred +addFilter("openvino-sample.*: E: env-script-interpreter") diff --git a/openvino.changes b/openvino.changes new file mode 100644 index 0000000..fe2912f --- /dev/null +++ b/openvino.changes @@ -0,0 +1,873 @@ +------------------------------------------------------------------- +Tue Dec 2 22:43:52 UTC 2025 - Alessandro de Oliveira Faria + +- Update to 2025.4.0 +- More GenAI coverage and framework integrations to minimize code + changes + * New models supported: + + On CPUs & GPUs: Qwen3-Embedding-0.6B, Qwen3-Reranker-0.6B, + Mistral-Small-24B-Instruct-2501. + + On NPUs: Gemma-3-4b-it and Qwen2.5-VL-3B-Instruct. + * Preview: Mixture of Experts (MoE) models optimized for CPUs + and GPUs, validated for Qwen3-30B-A3B. + * GenAI pipeline integrations: Qwen3-Embedding-0.6B and + Qwen3-Reranker-0.6B for enhanced retrieval/ranking, and + Qwen2.5VL-7B for video pipeline. +- Broader LLM model support and more model compression + techniques + * Gold support for Windows ML* enables developers to deploy AI + models and applications effortlessly across CPUs, GPUs, and + NPUs on Intel® Core™ Ultra processor-powered AI PCs. + * The Neural Network Compression Framework (NNCF) ONNX backend + now supports INT8 static post-training quantization (PTQ) + and INT8/INT4 weight-only compression to ensure accuracy + parity with OpenVINO IR format models. SmoothQuant algorithm + support added for INT8 quantization. + * Accelerated multi-token generation for GenAI, leveraging + optimized GPU kernels to deliver faster inference, smarter + KV-cache reuse, and scalable LLM performance. + * GPU plugin updates include improved performance with prefix + caching for chat history scenarios and enhanced LLM accuracy + with dynamic quantization support for INT8. +- More portability and performance to run AI at the edge, in the + cloud, or locally. + * Announcing support for Intel® Core Ultra Processor Series 3. + * Encrypted blob format support added for secure model + deployment with OpenVINO GenAI. Model weights and artifacts + are stored and transmitted in an encrypted format, reducing + risks of IP theft during deployment. Developers can deploy + with minimal code changes using OpenVINO GenAI pipelines. + * OpenVINO™ Model Server and OpenVINO™ GenAI now extend + support for Agentic AI scenarios with new features such as + output parsing and improved chat templates for reliable + multi-turn interactions, and preview functionality for the + Qwen3-30B-A3B model. OVMS also introduces a preview for + audio endpoints. + * NPU deployment is simplified with batch support, enabling + seamless model execution across Intel® Core Ultra + processors while eliminating driver dependencies. Models + are reshaped to batch_size=1 before compilation. + * The improved NVIDIA Triton Server* integration with + OpenVINO backend now enables developers to utilize Intel + GPUs or NPUs for deployment. + +------------------------------------------------------------------- +Sun Sep 7 01:21:19 UTC 2025 - Alessandro de Oliveira Faria + +- Update to 2025.3.0 +- More GenAI coverage and framework integrations to minimize code + changes + * New models supported: Phi-4-mini-reasoning, AFM-4.5B, + Gemma-3-1B-it, Gemma-3-4B-it, and Gemma-3-12B, + * NPU support added for: Qwen3-1.7B, Qwen3-4B, and Qwen3-8B. + * LLMs optimized for NPU now available on OpenVINO Hugging + Face collection. +- Broader LLM model support and more model compression techniques + * The NPU plug-in adds support for longer contexts of up to + 8K tokens, dynamic prompts, and dynamic LoRA for improved + LLM performance. + * The NPU plug-in now supports dynamic batch sizes by reshaping + the model to a batch size of 1 and concurrently managing + multiple inference requests, enhancing performance and + optimizing memory utilization. + * Accuracy improvements for GenAI models on both built-in + and discrete graphics achieved through the implementation + of the key cache compression per channel technique, in + addition to the existing KV cache per-token compression + method. + * OpenVINO™ GenAI introduces TextRerankPipeline for improved + retrieval relevance and RAG pipeline accuracy, plus + Structured Output for enhanced response reliability and + function calling while ensuring adherence to predefined + formats. +- More portability and performance to run AI at the edge, + in the cloud, or locally. + * Announcing support for Intel® Arc™ Pro B-Series + (B50 and B60). + * Preview: Hugging Face models that are GGUF-enabled for + OpenVINO GenAI are now supported by the OpenVINO™ Model + Server for popular LLM model architectures such as + DeepSeek Distill, Qwen2, Qwen2.5, and Llama 3. + This functionality reduces memory footprint and + simplifies integration for GenAI workloads. + * With improved reliability and tool call accuracy, + the OpenVINO™ Model Server boosts support for + agentic AI use cases on AI PCs, while enhancing + performance on Intel CPUs, built-in GPUs, and NPUs. + * int4 data-aware weights compression, now supported in the + Neural Network Compression Framework (NNCF) for ONNX + models, reduces memory footprint while maintaining + accuracy and enables efficient deployment in + resource-constrained environments. + +------------------------------------------------------------------- +Wed Jun 25 01:09:14 UTC 2025 - Alessandro de Oliveira Faria + +- openSUSE Leap 16.0 compatibility + +------------------------------------------------------------------- +Tue Jun 24 05:10:06 UTC 2025 - Alessandro de Oliveira Faria + +- Remove openvino-gcc5-compatibility.patch file + +------------------------------------------------------------------- +Tue Jun 24 02:54:10 UTC 2025 - Alessandro de Oliveira Faria + +Summary of major features and improvements   +- More GenAI coverage and framework integrations to minimize code + changes + * New models supported on CPUs & GPUs: Phi-4, + Mistral-7B-Instruct-v0.3, SD-XL Inpainting 0.1, Stable + Diffusion 3.5 Large Turbo, Phi-4-reasoning, Qwen3, and + Qwen2.5-VL-3B-Instruct. Mistral 7B Instruct v0.3 is also + supported on NPUs. + * Preview: OpenVINO ™ GenAI introduces a text-to-speech + pipeline for the SpeechT5 TTS model, while the new RAG + backend offers developers a simplified API that delivers + reduced memory usage and improved performance. + * Preview: OpenVINO™ GenAI offers a GGUF Reader for seamless + integration of llama.cpp based LLMs, with Python and C++ + pipelines that load GGUF models, build OpenVINO graphs, + and run GPU inference on-the-fly. Validated for popular models: + DeepSeek-R1-Distill-Qwen (1.5B, 7B), Qwen2.5 Instruct + (1.5B, 3B, 7B) & llama-3.2 Instruct (1B, 3B, 8B). +- Broader LLM model support and more model compression + techniques + * Further optimization of LoRA adapters in OpenVINO GenAI + for improved LLM, VLM, and text-to-image model performance + on built-in GPUs. Developers can use LoRA adapters to + quickly customize models for specialized tasks. + * KV cache compression for CPUs is enabled by default for + INT8, providing a reduced memory footprint while maintaining + accuracy compared to FP16. Additionally, it delivers + substantial memory savings for LLMs with INT4 support compared + to INT8. + * Optimizations for Intel® Core™ Ultra Processor Series 2 + built-in GPUs and Intel® Arc™ B Series Graphics with the + Intel® XMX systolic platform to enhance the performance of + VLM models and hybrid quantized image generation models, as + well as improve first-token latency for LLMs through dynamic + quantization. +- More portability and performance to run AI at the edge, in the + cloud, or locally. + * Enhanced Linux* support with the latest GPU driver for + built-in GPUs on Intel® Core™ Ultra Processor Series 2 + (formerly codenamed Arrow Lake H). + * Support for INT4 data-free weights compression for ONNX + models implemented in the Neural Network Compression + Framework (NNCF). + * NPU support for FP16-NF4 precision on Intel® Core™ 200V + Series processors for models with up to 8B parameters is + enabled through symmetrical and channel-wise quantization, + improving accuracy while maintaining performance efficiency. +Support Change and Deprecation Notices +- Discontinued in 2025: + * Runtime components: + + The OpenVINO property of Affinity API is no longer + available. It has been replaced with CPU binding + configurations (ov::hint::enable_cpu_pinning). + + The openvino-nightly PyPI module has been discontinued. + End-users should proceed with the Simple PyPI nightly repo + instead. More information in Release Policy. The + openvino-nightly PyPI module has been discontinued. + End-users should proceed with the Simple PyPI nightly repo + instead. More information in Release Policy. + * Tools: + + The OpenVINO™ Development Tools package (pip install + openvino-dev) is no longer available for OpenVINO releases + in 2025. + + Model Optimizer is no longer available. Consider using the + new conversion methods instead. For more details, see the + model conversion transition guide. + + Intel® Streaming SIMD Extensions (Intel® SSE) are currently + not enabled in the binary package by default. They are + still supported in the source code form. + + Legacy prefixes: l_, w_, and m_ have been removed from + OpenVINO archive names. + * OpenVINO GenAI: + + StreamerBase::put(int64_t token) + + The Bool value for Callback streamer is no longer accepted. + It must now return one of three values of StreamingStatus + enum. + + ChunkStreamerBase is deprecated. Use StreamerBase instead. + * NNCF create_compressed_model() method is now deprecated. + nncf.quantize() method is recommended for + Quantization-Aware Training of PyTorch and TensorFlow models. + * OpenVINO Model Server (OVMS) benchmark client in C++ + using TensorFlow Serving API. +- Deprecated and to be removed in the future: + * Python 3.9 is now deprecated and will be unavailable after + OpenVINO version 2025.4. + * openvino.Type.undefined is now deprecated and will be removed + with version 2026.0. openvino.Type.dynamic should be used + instead. + * APT & YUM Repositories Restructure: Starting with release + 2025.1, users can switch to the new repository structure + for APT and YUM, which no longer uses year-based + subdirectories (like “2025”). The old (legacy) structure + will still be available until 2026, when the change will + be finalized. Detailed instructions are available on the + relevant documentation pages: + + Installation guide - yum + + Installation guide - apt + * OpenCV binaries will be removed from Docker images in 2026. + * Ubuntu 20.04 support will be deprecated in future OpenVINO + releases due to the end of standard support. + * “auto shape” and “auto batch size” (reshaping a model in + runtime) will be removed in the future. OpenVINO’s dynamic + shape models are recommended instead. + * MacOS x86 is no longer recommended for use due to the + discontinuation of validation. Full support will be removed + later in 2025. + * The openvino namespace of the OpenVINO Python API has been + redesigned, removing the nested openvino.runtime module. + The old namespace is now considered deprecated and will be + discontinued in 2026.0. + +------------------------------------------------------------------- +Wed May 21 14:43:02 UTC 2025 - Andreas Schwab + +- Fix file list for riscv64 + +------------------------------------------------------------------- +Mon May 5 07:47:30 UTC 2025 - Dominique Leuenberger + +- Do not force GCC15 on Tumblewed just yet: follow the distro + default compiler, like any other package. + +------------------------------------------------------------------- +Sat May 3 19:19:07 UTC 2025 - Alessandro de Oliveira Faria + +- openvino-gcc5-compatibility.patch to resolve incompatibility + in gcc5 + +------------------------------------------------------------------- +Thu May 1 01:06:52 UTC 2025 - Alessandro de Oliveira Faria + +- Added gcc-14 + +------------------------------------------------------------------- +Mon Apr 14 06:52:03 UTC 2025 - Alessandro de Oliveira Faria + +- Update to 2025.1.0 +- Downgrade from gcc13-c++ to 12 due to incompatibility in tbb + compilation. This was due to C++ libraries (using libstdc++) + resulting in the error: libtbb.so.12: undefined reference to + `__cxa_call_terminate@CXXABI_1.3.15' +- More GenAI coverage and framework integrations to minimize code + changes + * New models supported: Phi-4 Mini, Jina CLIP v1, and Bce + Embedding Base v1. + * OpenVINO™ Model Server now supports VLM models, including + Qwen2-VL, Phi-3.5-Vision, and InternVL2. + * OpenVINO GenAI now includes image-to-image and inpainting + features for transformer-based pipelines, such as Flux.1 and + Stable Diffusion 3 models, enhancing their ability to generate + more realistic content. + * Preview: AI Playground now utilizes the OpenVINO Gen AI backend + to enable highly optimized inferencing performance on AI PCs. + +- Broader LLM model support and more model compression techniques + * Reduced binary size through optimization of the CPU plugin and + removal of the GEMM kernel. + * Optimization of new kernels for the GPU plugin significantly + boosts the performance of Long Short-Term Memory (LSTM) models, + used in many applications, including speech recognition, + language modeling, and time series forecasting. + * Preview: Token Eviction implemented in OpenVINO GenAI to reduce + the memory consumption of KV Cache by eliminating unimportant + tokens. This current Token Eviction implementation is + beneficial for tasks where a long sequence is generated, such + as chatbots and code generation. + * NPU acceleration for text generation is now enabled in + OpenVINO™ Runtime and OpenVINO™ Model Server to support the + power-efficient deployment of VLM models on NPUs for AI PC use + cases with low concurrency. + +- More portability and performance to run AI at the edge, in the + cloud, or locally. + * Additional LLM performance optimizations on Intel® Core™ Ultra + 200H series processors for improved 2nd token latency on + Windows and Linux. + * Enhanced performance and efficient resource utilization with + the implementation of Paged Attention and Continuous Batching + by default in the GPU plugin. + * Preview: The new OpenVINO backend for Executorch will enable + accelerated inference and improved performance on Intel + hardware, including CPUs, GPUs, and NPUs. + +------------------------------------------------------------------- +Tue Mar 4 00:38:30 UTC 2025 - Alessandro de Oliveira Faria + +- Disabled JAX plugin beta. + +------------------------------------------------------------------- +Sun Feb 9 03:36:41 UTC 2025 - Alessandro de Oliveira Faria + +- Update to 2025.0.0 +- More GenAI coverage and framework integrations to minimize code + changes + * New models supported: Qwen 2.5, Deepseek-R1-Distill-Llama-8B, + DeepSeek-R1-Distill-Qwen-7B, and DeepSeek-R1-Distill-Qwen-1.5B, + FLUX.1 Schnell and FLUX.1 Dev + * Whisper Model: Improved performance on CPUs, built-in GPUs, + and discrete GPUs with GenAI API. + * Preview: Introducing NPU support for torch.compile, giving + developers the ability to use the OpenVINO backend to run the + PyTorch API on NPUs. 300+ deep learning models enabled from + the TorchVision, Timm, and TorchBench repositories.. +- Broader Large Language Model (LLM) support and more model + compression techniques. + * Preview: Addition of Prompt Lookup to GenAI API improves 2nd + token latency for LLMs by effectively utilizing predefined + prompts that match the intended use case. + * Preview: The GenAI API now offers image-to-image inpainting + functionality. This feature enables models to generate + realistic content by inpainting specified modifications and + seamlessly integrating them with the original image. + * Asymmetric KV Cache compression is now enabled for INT8 on + CPUs, resulting in lower memory consumption and improved 2nd + token latency, especially when dealing with long prompts that + require significant memory. The option should be explicitly + specified by the user. +- More portability and performance to run AI at the edge, in the + cloud, or locally. + * Support for the latest Intel® Core™ Ultra 200H series + processors (formerly codenamed Arrow Lake-H) + * Integration of the OpenVINO ™ backend with the Triton + Inference Server allows developers to utilize the Triton + server for enhanced model serving performance when deploying + on Intel CPUs. + * Preview: A new OpenVINO ™ backend integration allows + developers to leverage OpenVINO performance optimizations + directly within Keras 3 workflows for faster AI inference on + CPUs, built-in GPUs, discrete GPUs, and NPUs. This feature is + available with the latest Keras 3.8 release. + * The OpenVINO Model Server now supports native Windows Server + deployments, allowing developers to leverage better + performance by eliminating container overhead and simplifying + GPU deployment. +- Support Change and Deprecation Notices + * Now deprecated: + + Legacy prefixes l_, w_, and m_ have been removed from + OpenVINO archive names. + + The runtime namespace for Python API has been marked as + deprecated and designated to be removed for 2026.0. The + new namespace structure has been delivered, and migration + is possible immediately. Details will be communicated + through warnings andvia documentation. + + NNCF create_compressed_model() method is deprecated. + nncf.quantize() method is now recommended for + Quantization-Aware Training of PyTorch and + TensorFlow models. + +------------------------------------------------------------------- +Sun Dec 29 03:41:47 UTC 2024 - Alessandro de Oliveira Faria + +- openvino-onnx-ml-defines.patch and + openvino-remove-npu-compile-tool.patchhas been removed + as it is no longer needed in this version. +- Update to 2024.4.0 +- Summary of major features and improvements   + * OpenVINO 2024.6 release includes updates for enhanced + stability and improved LLM performance. + * Introduced support for Intel® Arc™ B-Series Graphics + (formerly known as Battlemage). + * Implemented optimizations to improve the inference time and + LLM performance on NPUs. + * Improved LLM performance with GenAI API optimizations and + bug fixes. +- Support Change and Deprecation Notices + * Using deprecated features and components is not advised. They + are available to enable a smooth transition to new solutions + and will be discontinued in the future. To keep using + discontinued features, you will have to revert to the last + LTS OpenVINO version supporting them. For more details, refer + to the OpenVINO Legacy Features and Components page. + * Discontinued in 2024.0: + + Runtime components: + - Intel® Gaussian & Neural Accelerator (Intel® GNA).. + Consider using the Neural Processing Unit (NPU) for + low-powered systems like Intel® Core™ Ultra or 14th + generation and beyond. + - OpenVINO C++/C/Python 1.0 APIs (see 2023.3 API transition + guide for reference). + - All ONNX Frontend legacy API (known as ONNX_IMPORTER_API) + - 'PerfomanceMode.UNDEFINED' property as part of the + OpenVINO Python API + + Tools: + - Deployment Manager. See installation and deployment + guides for current distribution options. + - Accuracy Checker. + - Post-Training Optimization Tool (POT). Neural Network + Compression Framework (NNCF) should be used instead. + - A Git patch for NNCF integration with  + huggingface/transformers. The recommended approach is + to use huggingface/optimum-intel for applying NNCF + optimization on top of models from Hugging Face. + - Support for Apache MXNet, Caffe, and Kaldi model formats. + Conversion to ONNX may be used as a solution. + * Deprecated and to be removed in the future: + + The macOS x86_64 debug bins will no longer be provided + with the OpenVINO toolkit, starting with OpenVINO 2024.5. + + Python 3.8 is no longer supported, starting with + OpenVINO 2024.5. + + As MxNet doesn’t support Python version higher than 3.8, + according to the MxNet PyPI project, it is no longer + supported by OpenVINO, either. + + Discrete Keem Bay support is no longer supported, starting + with OpenVINO 2024.5. + + Support for discrete devices (formerly codenamed Raptor + Lake) is no longer available for NPU. + +------------------------------------------------------------------- +Tue Dec 10 15:50:41 UTC 2024 - Giacomo Comes + +- fix build on tumbleweed + * currently openvino does not support protobuf v22 or newer + +------------------------------------------------------------------- +Tue Oct 15 00:56:54 UTC 2024 - Alessandro de Oliveira Faria + +- Temporarily inserted gcc-13 in Tumbleweed/Factory/Slowroll: + Because there is an incompatibility of the source code of the + level-zero library and npu module with gcc-14. I am working + with Intel on tests to return to native gcc. +- Update to 2024.4.0 +- Summary of major features and improvements   + * More Gen AI coverage and framework integrations to minimize + code changes + + Support for GLM-4-9B Chat, MiniCPM-1B, Llama 3 and 3.1, + Phi-3-Mini, Phi-3-Medium and YOLOX-s models. + + Noteworthy notebooks added: Florence-2, NuExtract-tiny + Structure Extraction, Flux.1 Image Generation, PixArt-α: + Photorealistic Text-to-Image Synthesis, and Phi-3-Vision + Visual Language Assistant. + * Broader Large Language Model (LLM) support and more model + compression techniques. + + OpenVINO™ runtime optimized for Intel® Xe Matrix Extensions + (Intel® XMX) systolic arrays on built-in GPUs for efficient + matrix multiplication resulting in significant LLM + performance boost with improved 1st and 2nd token + latency, as well as a smaller memory footprint on + Intel® Core™ Ultra Processors (Series 2). + + Memory sharing enabled for NPUs on Intel® Core™ Ultra + Processors (Series 2) for efficient pipeline integration + without memory copy overhead. + + Addition of the PagedAttention feature for discrete GPUs* + enables a significant boost in throughput for parallel + inferencing when serving LLMs on Intel® Arc™ Graphics + or Intel® Data Center GPU Flex Series. + * More portability and performance to run AI at the edge, + in the cloud, or locally. + + OpenVINO™ Model Server now comes with production-quality + support for OpenAI-compatible API which enables i + significantly higher throughput for parallel inferencing + on Intel® Xeon® processors when serving LLMs to many + concurrent users. + + Improved performance and memory consumption with prefix + caching, KV cache compression, and other optimizations + for serving LLMs using OpenVINO™ Model Server. + + Support for Python 3.12. +- Support Change and Deprecation Notices + * Using deprecated features and components is not advised. + They are available to enable a smooth transition to new + solutions and will be discontinued in the future. + To keep using discontinued features, you will have to + revert to the last LTS OpenVINO version supporting them. + For more details, refer to the OpenVINO Legacy Features + and Components page. + * Discontinued in 2024.0: + + Runtime components: + - Intel® Gaussian & Neural Accelerator (Intel® GNA). + Consider using the Neural Processing Unit (NPU) for + low-powered systems like Intel® Core™ Ultra or + 14th generation and beyond. + - OpenVINO C++/C/Python 1.0 APIs (see 2023.3 API + transition guide for reference). + - All ONNX Frontend legacy API (known as + ONNX_IMPORTER_API) + -'PerfomanceMode.UNDEFINED' property as part of the + OpenVINO Python API + + Tools: + - Deployment Manager. See installation and deployment + guides for current distribution options. + - Accuracy Checker. + - Post-Training Optimization Tool (POT). Neural Network + Compression Framework (NNCF) should be used instead. + - A Git patch for NNCF integration with huggingface/ + transformers. The recommended approach is to use + huggingface/optimum-intel for applying NNCF + optimization on top of models from Hugging Face. + - Support for Apache MXNet, Caffe, and Kaldi model + formats. Conversion to ONNX may be used as a + solution. + * Deprecated and to be removed in the future: + + The macOS x86_64 debug bins will no longer be + provided with the OpenVINO toolkit, starting with + OpenVINO 2024.5. + + Python 3.8 is now considered deprecated, and it will not + be available beyond the 2024.4 OpenVINO version. + + dKMB support is now considered deprecated and will be + fully removed with OpenVINO 2024.5 + + Intel® Streaming SIMD Extensions (Intel® SSE) will be + supported in source code form, but not enabled in the + binary package by default, starting with OpenVINO 2025.0 + + The openvino-nightly PyPI module will soon be discontinued. + End-users should proceed with the Simple PyPI nightly repo + instead. More information in Release Policy. + + The OpenVINO™ Development Tools package (pip install + openvino-dev) will be removed from installation options and + distribution channels beginning with OpenVINO 2025.0. + + Model Optimizer will be discontinued with OpenVINO 2025.0. + Consider using the new conversion methods instead. For more + details, see the model conversion transition guide. + + OpenVINO property Affinity API will be discontinued with + OpenVINO 2025.0. It will be replaced with CPU binding + configurations (ov::hint::enable_cpu_pinning). + + OpenVINO Model Server components: + - “auto shape” and “auto batch size” (reshaping a model in + runtime) will be removed in the future. OpenVINO’s dynamic + shape models are recommended instead. + + A number of notebooks have been deprecated. For an + up-to-date listing of available notebooks, refer to the + OpenVINO™ Notebook index (openvinotoolkit.github.io). + +------------------------------------------------------------------- +Wed Oct 2 20:56:59 UTC 2024 - Giacomo Comes + +- Add Leap15 build +- Remove comment lines in the spec file that cause the insertion + of extra lines during a commit + +------------------------------------------------------------------- +Sat Aug 10 01:41:06 UTC 2024 - Alessandro de Oliveira Faria + +- Remove NPU Compile Tool +* openvino-remove-npu-compile-tool.patch +- Update to 2024.3.0 +- Summary of major features and improvements   + * More Gen AI coverage and framework integrations to minimize + code changes + + OpenVINO pre-optimized models are now available in Hugging + Face making it easier for developers to get started with + these models. + * Broader Large Language Model (LLM) support and more model + compression techniques. + + Significant improvement in LLM performance on Intel + discrete GPUs with the addition of Multi-Head Attention + (MHA) and OneDNN enhancements. + * More portability and performance to run AI at the edge, in the + cloud, or locally. + + Improved CPU performance when serving LLMs with the + inclusion of vLLM and continuous batching in the OpenVINO + Model Server (OVMS). vLLM is an easy-to-use open-source + library that supports efficient LLM inferencing and model + serving. +- Support Change and Deprecation Notices + * Using deprecated features and components is not advised. + They are available to enable a smooth transition to new + solutions and will be discontinued in the future. To keep + using discontinued features, you will have to revert to the + last LTS OpenVINO version supporting them. For more details, + refer to the OpenVINO Legacy Features and Components page. + * Discontinued in 2024.0: + + Runtime components: + - Intel® Gaussian & Neural Accelerator (Intel® GNA)..Consider + using the Neural Processing Unit (NPU) for low-powered + systems like Intel® Core™ Ultra or 14th generation + and beyond. + - OpenVINO C++/C/Python 1.0 APIs (see 2023.3 API transition + guide for reference). + - All ONNX Frontend legacy API (known as ONNX_IMPORTER_API) + - 'PerfomanceMode.UNDEFINED' property as part of the OpenVINO + Python API + + Tools: + - Deployment Manager. See installation and deployment guides + for current distribution options. + - Accuracy Checker. + - Post-Training Optimization Tool (POT). Neural Network + Compression Framework (NNCF) should be used instead. + - A Git patch for NNCF integration with huggingface/ + transformers. The recommended approach is to use + huggingface/optimum-intel for applying NNCF optimization + on top of models from Hugging Face. + - Support for Apache MXNet, Caffe, and Kaldi model formats. + Conversion to ONNX may be used as a solution. + * Deprecated and to be removed in the future: + + The OpenVINO™ Development Tools package (pip install + openvino-dev) will be removed from installation options + and distribution channels beginning with OpenVINO 2025.0. + + Model Optimizer will be discontinued with OpenVINO 2025.0. + Consider using the new conversion methods instead. For + more details, see the model conversion transition guide. + + OpenVINO property Affinity API will be discontinued with + OpenVINO 2025.0. It will be replaced with CPU binding + configurations (ov::hint::enable_cpu_pinning). + + OpenVINO Model Server components: + - “auto shape” and “auto batch size” (reshaping a model + in runtime) will be removed in the future. OpenVINO’s + dynamic shape models are recommended instead. + + A number of notebooks have been deprecated. For an + up-to-date listing of available notebooks, refer to + the OpenVINO™ Notebook index (openvinotoolkit.github.io). + +------------------------------------------------------------------- +Sat Jun 22 12:01:23 UTC 2024 - Andreas Schwab + +- Add riscv-cpu-plugin subpackage + +------------------------------------------------------------------- +Wed Jun 19 21:36:01 UTC 2024 - Alessandro de Oliveira Faria + +- Update to 2024.2.0 +- More Gen AI coverage and framework integrations to minimize code + changes + * Llama 3 optimizations for CPUs, built-in GPUs, and discrete + GPUs for improved performance and efficient memory usage. + * Support for Phi-3-mini, a family of AI models that leverages + the power of small language models for faster, more accurate + and cost-effective text processing. + * Python Custom Operation is now enabled in OpenVINO making it + easier for Python developers to code their custom operations + instead of using C++ custom operations (also supported). + Python Custom Operation empowers users to implement their own + specialized operations into any model. + * Notebooks expansion to ensure better coverage for new models. + Noteworthy notebooks added: DynamiCrafter, YOLOv10, Chatbot + notebook with Phi-3, and QWEN2. +- Broader Large Language Model (LLM) support and more model + compression techniques. + * GPTQ method for 4-bit weight compression added to NNCF for + more efficient inference and improved performance of + compressed LLMs. + * Significant LLM performance improvements and reduced latency + for both built-in GPUs and discrete GPUs. + * Significant improvement in 2nd token latency and memory + footprint of FP16 weight LLMs on AVX2 (13th Gen Intel® Core™ + processors) and AVX512 (3rd Gen Intel® Xeon® Scalable + Processors) based CPU platforms, particularly for small + batch sizes. +- More portability and performance to run AI at the edge, in the + cloud, or locally. + * Model Serving Enhancements: + * Preview: OpenVINO Model Server (OVMS) now supports + OpenAI-compatible API along with Continuous Batching and + PagedAttention, enabling significantly higher throughput + for parallel inferencing, especially on Intel® Xeon® + processors, when serving LLMs to many concurrent users. + * OpenVINO backend for Triton Server now supports built-in + GPUs and discrete GPUs, in addition to dynamic + shapes support. + * Integration of TorchServe through torch.compile OpenVINO + backend for easy model deployment, provisioning to + multiple instances, model versioning, and maintenance. + * Preview: addition of the Generate API, a simplified API + for text generation using large language models with only + a few lines of code. The API is available through the newly + launched OpenVINO GenAI package. + * Support for Intel Atom® Processor X Series. For more details, + see System Requirements. + * Preview: Support for Intel® Xeon® 6 processor. +- Support Change and Deprecation Notices + * Using deprecated features and components is not advised. + They are available to enable a smooth transition to new + solutions and will be discontinued in the future. + To keep using discontinued features, you will have to revert + to the last LTS OpenVINO version supporting them. For more + details, refer to the OpenVINO Legacy Features and + Components page. + * Discontinued in 2024.0: + + Runtime components: + - Intel® Gaussian & Neural Accelerator (Intel® GNA). + Consider using the Neural Processing Unit (NPU) for + low-powered systems like Intel® Core™ Ultra or 14th + generation and beyond. + - OpenVINO C++/C/Python 1.0 APIs (see 2023.3 API + transition guide for reference). + - All ONNX Frontend legacy API (known as ONNX_IMPORTER_API) + - 'PerfomanceMode.UNDEFINED' property as part of the + OpenVINO Python API + + Tools: + - Deployment Manager. See installation and deployment + guides for current distribution options. + - Accuracy Checker. + - Post-Training Optimization Tool (POT). Neural Network + Compression Framework (NNCF) should be used instead. + - A Git patch for NNCF integration with  + huggingface/transformers. The recommended approach  + is to use huggingface/optimum-intel for applying NNCF + optimization on top of models from Hugging Face. + - Support for Apache MXNet, Caffe, and Kaldi model formats. + Conversion to ONNX may be used as a solution. + * Deprecated and to be removed in the future: + + The OpenVINO™ Development Tools package (pip install + openvino-dev) will be removed from installation options + and distribution channels beginning with OpenVINO 2025.0. + + Model Optimizer will be discontinued with OpenVINO 2025.0. + Consider using the new conversion methods instead. For + more details, see the model conversion transition guide. + + OpenVINO property Affinity API will be discontinued with + OpenVINO 2025.0. It will be replaced with CPU binding + configurations (ov::hint::enable_cpu_pinning). + + OpenVINO Model Server components: + + “auto shape” and “auto batch size” (reshaping a model in + runtime) will be removed in the future. OpenVINO’s dynamic + shape models are recommended instead. + + A number of notebooks have been deprecated. For an + up-to-date listing of available notebooks, refer to the + OpenVINO™ Notebook index (openvinotoolkit.github.io). + +------------------------------------------------------------------- +Thu May 9 22:56:53 UTC 2024 - Alessandro de Oliveira Faria + +- Fix sample source path in build script: + * openvino-fix-build-sample-path.patch +- Update to 2024.1.0 +- More Generative AI coverage and framework integrations to + minimize code changes. + * Mixtral and URLNet models optimized for performance + improvements on Intel® Xeon® processors. + * Stable Diffusion 1.5, ChatGLM3-6B, and Qwen-7B models + optimized for improved inference speed on Intel® Core™ + Ultra processors with integrated GPU. + * Support for Falcon-7B-Instruct, a GenAI Large Language Model + (LLM) ready-to-use chat/instruct model with superior + performance metrics. + * New Jupyter Notebooks added: YOLO V9, YOLO V8 + Oriented Bounding Boxes Detection (OOB), Stable Diffusion + in Keras, MobileCLIP, RMBG-v1.4 Background Removal, Magika, + TripoSR, AnimateAnyone, LLaVA-Next, and RAG system with + OpenVINO and LangChain. +- Broader Large Language Model (LLM) support and more model + compression techniques. + * LLM compilation time reduced through additional optimizations + with compressed embedding. Improved 1st token performance of + LLMs on 4th and 5th generations of Intel® Xeon® processors + with Intel® Advanced Matrix Extensions (Intel® AMX). + * Better LLM compression and improved performance with oneDNN, + INT4, and INT8 support for Intel® Arc™ GPUs. + * Significant memory reduction for select smaller GenAI + models on Intel® Core™ Ultra processors with integrated GPU. +- More portability and performance to run AI at the edge, + in the cloud, or locally. + * The preview NPU plugin for Intel® Core™ Ultra processors + is now available in the OpenVINO open-source GitHub + repository, in addition to the main OpenVINO package on PyPI. + * The JavaScript API is now more easily accessible through + the npm repository, enabling JavaScript developers’ seamless + access to the OpenVINO API. + * FP16 inference on ARM processors now enabled for the + Convolutional Neural Network (CNN) by default. +- Support Change and Deprecation Notices + * Using deprecated features and components is not advised. They + are available to enable a smooth transition to new solutions + and will be discontinued in the future. To keep using + Discontinued features, you will have to revert to the last + LTS OpenVINO version supporting them. + * For more details, refer to the OpenVINO Legacy Features + and Components page. + * Discontinued in 2024.0: + + Runtime components: + - Intel® Gaussian & Neural Accelerator (Intel® GNA). + Consider using the Neural Processing Unit (NPU) + for low-powered systems like Intel® Core™ Ultra or + 14th generation and beyond. + - OpenVINO C++/C/Python 1.0 APIs (see 2023.3 API + transition guide for reference). + - All ONNX Frontend legacy API (known as + ONNX_IMPORTER_API) + - 'PerfomanceMode.UNDEFINED' property as part of + the OpenVINO Python API + + Tools: + - Deployment Manager. See installation and deployment + guides for current distribution options. + - Accuracy Checker. + - Post-Training Optimization Tool (POT). Neural Network + Compression Framework (NNCF) should be used instead. + - A Git patch for NNCF integration with  + huggingface/transformers. The recommended approach +  is to use huggingface/optimum-intel for applying + NNCF optimization on top of models from Hugging + Face. + - Support for Apache MXNet, Caffe, and Kaldi model + formats. Conversion to ONNX may be used as + a solution. + * Deprecated and to be removed in the future: + + The OpenVINO™ Development Tools package (pip install + openvino-dev) will be removed from installation options + and distribution channels beginning with OpenVINO 2025.0. + + Model Optimizer will be discontinued with OpenVINO 2025.0. + Consider using the new conversion methods instead. For + more details, see the model conversion transition guide. + + OpenVINO property Affinity API will be discontinued with + OpenVINO 2025.0. It will be replaced with CPU binding + configurations (ov::hint::enable_cpu_pinning). + + OpenVINO Model Server components: + - “auto shape” and “auto batch size” (reshaping a model + in runtime) will be removed in the future. OpenVINO’s + dynamic shape models are recommended instead. + +------------------------------------------------------------------- +Tue Apr 23 18:57:17 UTC 2024 - Atri Bhattacharya + +- License update: play safe and list all third party licenses as + part of the License tag. + +------------------------------------------------------------------- +Tue Apr 23 12:42:32 UTC 2024 - Atri Bhattacharya + +- Switch to _service file as tagged Source tarball does not + include `./thirdparty` submodules. +- Update openvino-fix-install-paths.patch to fix python module + install path. +- Enable python module and split it out into a python subpackage + (for now default python3 only). +- Explicitly build python metadata (dist-info) and install it + (needs simple sed hackery to support "officially" unsupported + platform ppc64le). +- Specify ENABLE_JS=OFF to turn off javascript bindings as + building these requires downloading npm stuff from the network. +- Build with system pybind11. +- Bump _constraints for updated disk space requirements. +- Drop empty %check section, rpmlint was misleading when it + recommended adding this. + +------------------------------------------------------------------- +Fri Apr 19 08:08:02 UTC 2024 - Atri Bhattacharya + +- Numerous specfile cleanups: + * Drop redundant `mv` commands and use `install` where + appropriate. + * Build with system protobuf. + * Fix Summary tags. + * Trim package descriptions. + * Drop forcing CMAKE_BUILD_TYPE=Release, let macro default + RelWithDebInfo be used instead. + * Correct naming of shared library packages. + * Separate out libopenvino_c.so.* into own shared lib package. + * Drop rpmlintrc rule used to hide shlib naming mistakes. + * Rename Source tarball to %{name}-%{version}.EXT pattern. + * Use ldconfig_scriptlet macro for post(un). +- Add openvino-onnx-ml-defines.patch -- Define ONNX_ML at compile + time when using system onnx to allow using 'onnx-ml.pb.h' + instead of 'onnx.pb.h', the latter not being shipped with + openSUSE's onnx-devel package (gh#onnx/onnx#3074). +- Add openvino-fix-install-paths.patch: Change hard-coded install + paths in upstream cmake macro to standard Linux dirs. +- Add openvino-ComputeLibrary-include-string.patch: Include header + for std::string. +- Add external devel packages as Requires for openvino-devel. +- Pass -Wl,-z,noexecstack to %build_ldflags to avoid an exec stack + issue with intel CPU plugin. +- Use ninja for build. +- Adapt _constraits file for correct disk space and memory + requirements. +- Add empty %check section. + +------------------------------------------------------------------- +Mon Apr 15 03:18:33 UTC 2024 - Alessandro de Oliveira Faria + +- Initial package +- Version 2024.0.0 +- Add openvino-rpmlintrc. diff --git a/openvino.obsinfo b/openvino.obsinfo new file mode 100644 index 0000000..83ca866 --- /dev/null +++ b/openvino.obsinfo @@ -0,0 +1,4 @@ +name: openvino +version: 2025.4.0 +mtime: 1763052589 +commit: 7a975177ff432c687e5619e8fb22e4bf265e48b7 diff --git a/openvino.spec b/openvino.spec new file mode 100644 index 0000000..49f55ca --- /dev/null +++ b/openvino.spec @@ -0,0 +1,441 @@ +# +# spec file for package openvino +# +# Copyright (c) 2025 SUSE LLC +# Copyright (c) 2024 Alessandro de Oliveira Faria (A.K.A. CABELO) or +# +# All modifications and additions to the file contributed by third parties +# remain the property of their copyright owners, unless otherwise agreed +# upon. The license for this file, and modifications and additions to the +# file, is the same license as for the pristine package itself (unless the +# license for the pristine package is not an Open Source License, in which +# case the license is the MIT License). An "Open Source License" is a +# license that conforms to the Open Source Definition (Version 1.9) +# published by the Open Source Initiative. + +# Please submit bugfixes or comments via https://bugs.opensuse.org/ +# + + +%if 0%{?suse_version} < 1600 +%define isLeap %nil +%else +%undefine isLeap +%endif + +# Compilation takes ~1 hr on OBS for a single python, don't try all supported flavours +%if %{defined isLeap} +%define x86_64 x86_64 +%define pythons python311 +%else +%define pythons python3 +%endif +%define __builder ninja +%define so_ver 2540 +%define shlib lib%{name}%{so_ver} +%define shlib_c lib%{name}_c%{so_ver} +%define prj_name OpenVINO + +Name: openvino +Version: 2025.4.0 +Release: 0 +Summary: A toolkit for optimizing and deploying AI inference +# Let's be safe and put all third party licenses here, no matter that we use specific thirdparty libs or not +License: Apache-2.0 AND BSD-2-Clause AND BSD-3-Clause AND HPND AND JSON AND MIT AND OFL-1.1 AND Zlib +URL: https://github.com/openvinotoolkit/openvino +Source0: %{name}-%{version}.tar.zst +Source1: %{name}-rpmlintrc +# PATCH-FEATURE-OPENSUSE openvino-fix-install-paths.patch badshah400@gmail.com -- Fix installation paths hardcoded into upstream defined cmake macros +Patch0: openvino-fix-install-paths.patch +# PATCH-FIX-UPSTREAM openvino-ComputeLibrary-include-string.patch badshah400@gmail.com -- Include header for std::string +Patch1: openvino-ComputeLibrary-include-string.patch +# PATCH-FIX-UPSTREAM openvino-fix-build-sample-path.patch cabelo@opensuse.org -- Fix sample source path in build script +Patch2: openvino-fix-build-sample-path.patch +BuildRequires: ade-devel +BuildRequires: cmake +BuildRequires: fdupes +BuildRequires: ninja +BuildRequires: opencl-cpp-headers +# FIXME: /usr/include/onnx/onnx-ml.pb.h:17:2: error: This file was generated by +# an older version of protoc which is incompatible with your Protocol Buffer +# headers. Please regenerate this file with a newer version of protoc. +#BuildRequires: cmake(ONNX) +BuildRequires: pkgconfig +BuildRequires: %{python_module devel} +BuildRequires: %{python_module pip} +BuildRequires: %{python_module pybind11-devel} +BuildRequires: %{python_module setuptools} +BuildRequires: %{python_module wheel} +BuildRequires: python-rpm-macros +BuildRequires: zstd +BuildRequires: pkgconfig(flatbuffers) +BuildRequires: pkgconfig(libva) +BuildRequires: pkgconfig(nlohmann_json) +BuildRequires: pkgconfig(ocl-icd) +BuildRequires: pkgconfig(protobuf) < 22 +BuildRequires: pkgconfig(pugixml) +%if %{defined isLeap} +BuildRequires: gcc12-c++ +BuildRequires: opencl-headers +BuildRequires: snappy-devel +BuildRequires: tbb-devel +%else +BuildRequires: pkgconfig(OpenCL-Headers) +BuildRequires: pkgconfig(snappy) +BuildRequires: pkgconfig(tbb) +%endif +BuildRequires: pkgconfig(zlib) +%ifarch %{arm64} +BuildRequires: scons +%endif +# No 32-bit support +ExcludeArch: %{ix86} %{arm32} ppc +%define python_subpackage_only 1 +%python_subpackages + +%description +OpenVINO is an open-source toolkit for optimizing and deploying AI inference. + +%package -n %{shlib} +Summary: Shared library for OpenVINO toolkit + +%description -n %{shlib} +OpenVINO is an open-source toolkit for optimizing and deploying AI inference. + +This package provides the shared library for OpenVINO. + +%package -n %{shlib_c} +Summary: Shared C library for OpenVINO toolkit + +%description -n %{shlib_c} +This package provides the C library for OpenVINO. + +%package -n %{name}-devel +Summary: Headers and sources for OpenVINO toolkit +Requires: %{shlib_c} = %{version} +Requires: %{shlib} = %{version} +Requires: lib%{name}_ir_frontend%{so_ver} = %{version} +Requires: lib%{name}_onnx_frontend%{so_ver} = %{version} +Requires: lib%{name}_paddle_frontend%{so_ver} = %{version} +Requires: lib%{name}_pytorch_frontend%{so_ver} = %{version} +Requires: lib%{name}_tensorflow_frontend%{so_ver} = %{version} +Requires: lib%{name}_tensorflow_lite_frontend%{so_ver} = %{version} +Requires: pkgconfig(flatbuffers) +Requires: pkgconfig(libva) +Requires: pkgconfig(nlohmann_json) +Requires: pkgconfig(ocl-icd) +Requires: pkgconfig(protobuf) +Requires: pkgconfig(pugixml) +%if %{defined isLeap} +Requires: opencl-headers +Requires: snappy-devel +Requires: tbb-devel +%else +Requires: pkgconfig(OpenCL-Headers) +Requires: pkgconfig(snappy) +Requires: pkgconfig(tbb) +%endif +Recommends: %{name}-auto-batch-plugin = %{version} +Recommends: %{name}-auto-plugin = %{version} +Recommends: %{name}-hetero-plugin = %{version} +Recommends: %{name}-intel-cpu-plugin = %{version} +%ifarch riscv64 +Recommends: %{name}-riscv-cpu-plugin = %{version} +%endif + +%description -n %{name}-devel +OpenVINO is an open-source toolkit for optimizing and deploying AI inference. + +This package provides the headers and sources for developing applications with +OpenVINO. + +%package -n %{name}-arm-cpu-plugin +Summary: Intel CPU plugin for OpenVINO toolkit + +%description -n %{name}-arm-cpu-plugin +OpenVINO is an open-source toolkit for optimizing and deploying AI inference. + +This package provides the ARM CPU plugin for OpenVINO on %{arm64} archs. + +%package -n %{name}-riscv-cpu-plugin +Summary: RISC-V CPU plugin for OpenVINO toolkit + +%description -n %{name}-riscv-cpu-plugin +OpenVINO is an open-source toolkit for optimizing and deploying AI inference. + +This package provides the RISC-V CPU plugin for OpenVINO on riscv64 archs. + +%package -n %{name}-auto-plugin +Summary: Auto / Multi software plugin for OpenVINO toolkit + +%description -n %{name}-auto-plugin +OpenVINO is an open-source toolkit for optimizing and deploying AI inference. + +This package provides the Auto / Multi software plugin for OpenVINO. + +%package -n %{name}-auto-batch-plugin +Summary: Automatic batch software plugin for OpenVINO toolkit + +%description -n %{name}-auto-batch-plugin +OpenVINO is an open-source toolkit for optimizing and deploying AI inference. + +This package provides the automatic batch software plugin for OpenVINO. + +%package -n %{name}-hetero-plugin +Summary: Hetero frontend for Intel OpenVINO toolkit + +%description -n %{name}-hetero-plugin +OpenVINO is an open-source toolkit for optimizing and deploying AI inference. + +This package provides the hetero frontend for OpenVINO. + +%package -n %{name}-intel-cpu-plugin +Summary: Intel CPU plugin for OpenVINO toolkit + +%description -n %{name}-intel-cpu-plugin +OpenVINO is an open-source toolkit for optimizing and deploying AI inference. + +This package provides the intel CPU plugin for OpenVINO for %{x86_64} archs. + +%package -n %{name}-intel-npu-plugin +Summary: Intel NPU plugin for OpenVINO toolkit + +%description -n %{name}-intel-npu-plugin +OpenVINO is an open-source toolkit for optimizing and deploying AI inference. + +This package provides the intel NPU plugin for OpenVINO for %{x86_64} archs. + +%package -n lib%{name}_ir_frontend%{so_ver} +Summary: Paddle frontend for Intel OpenVINO toolkit + +%description -n lib%{name}_ir_frontend%{so_ver} +OpenVINO is an open-source toolkit for optimizing and deploying AI inference. + +This package provides the ir frontend for OpenVINO. + +%package -n lib%{name}_onnx_frontend%{so_ver} +Summary: Onnx frontend for OpenVINO toolkit + +%description -n lib%{name}_onnx_frontend%{so_ver} +OpenVINO is an open-source toolkit for optimizing and deploying AI inference. + +This package provides the onnx frontend for OpenVINO. + +%package -n lib%{name}_paddle_frontend%{so_ver} +Summary: Paddle frontend for Intel OpenVINO toolkit + +%description -n lib%{name}_paddle_frontend%{so_ver} +OpenVINO is an open-source toolkit for optimizing and deploying AI inference. + +This package provides the paddle frontend for OpenVINO. + +%package -n lib%{name}_pytorch_frontend%{so_ver} +Summary: PyTorch frontend for OpenVINO toolkit + +%description -n lib%{name}_pytorch_frontend%{so_ver} +OpenVINO is an open-source toolkit for optimizing and deploying AI inference. + +This package provides the pytorch frontend for OpenVINO. + +%package -n lib%{name}_tensorflow_frontend%{so_ver} +Summary: TensorFlow frontend for OpenVINO toolkit + +%description -n lib%{name}_tensorflow_frontend%{so_ver} +OpenVINO is an open-source toolkit for optimizing and deploying AI inference. + +This package provides the tensorflow frontend for OpenVINO. + +%package -n lib%{name}_tensorflow_lite_frontend%{so_ver} +Summary: TensorFlow Lite frontend for OpenVINO toolkit + +%description -n lib%{name}_tensorflow_lite_frontend%{so_ver} +OpenVINO is an open-source toolkit for optimizing and deploying AI inference. + +This package provides the tensorflow-lite frontend for OpenVINO. + +%package -n python-openvino +Summary: Python module for openVINO toolkit +Requires: python-numpy < 2 +Requires: python-openvino-telemetry + +%description -n python-openvino +OpenVINO is an open-source toolkit for optimizing and deploying AI inference. + +This package provides a Python module for interfacing with openVINO toolkit. + +%package -n %{name}-sample +Summary: Samples for use with OpenVINO toolkit +BuildArch: noarch + +%description -n %{name}-sample +OpenVINO is an open-source toolkit for optimizing and deploying AI inference. + +This package provides some samples for use with openVINO. + +%prep +%autosetup -p1 + +%build +%if %{defined isLeap} +export CC=gcc-12 CXX=g++-12 +%endif +# Otherwise intel_cpu plugin declares an executable stack +%ifarch %{x86_64} +%define build_ldflags -Wl,-z,noexecstack +%endif +%cmake \ + -DCMAKE_CXX_STANDARD=17 \ + -DBUILD_SHARED_LIBS=ON \ + -DENABLE_OV_ONNX_FRONTEND=ON \ + -DENABLE_OV_PADDLE_FRONTEND=ON \ + -DENABLE_OV_PYTORCH_FRONTEND=ON \ + -DENABLE_OV_IR_FRONTEND=ON \ + -DENABLE_OV_TF_FRONTEND=ON \ + -DENABLE_OV_TF_LITE_FRONTEND=ON \ + -DENABLE_INTEL_GPU=OFF \ + -DENABLE_OV_JAX_FRONTEND=OFF \ + -DENABLE_JS=OFF \ + -DENABLE_PYTHON=ON \ + -DENABLE_WHEEL=OFF \ + -DENABLE_SYSTEM_OPENCL=ON \ + -DENABLE_SYSTEM_PROTOBUF=ON \ + -DENABLE_SYSTEM_PUGIXML=ON \ + -DENABLE_SYSTEM_SNAPPY=ON \ +%if %{defined isLeap} + -DENABLE_TBBBIND_2_5=OFF \ +%endif + -DENABLE_SYSTEM_TBB=ON \ + -DONNX_USE_PROTOBUF_SHARED_LIBS=ON \ + -DProtobuf_USE_STATIC_LIBS=OFF \ + %{nil} +%cmake_build +# Manually generate dist-info dir +export WHEEL_VERSION=%{version} \ + BUILD_TYPE=RelWithDebInfo +%ifarch %{power64} + +# RelWithDebInfo +# Manual hackery for power64 because it not "officially" supported + sed -i "s/{ARCH}/%{_arch}/" ../src/bindings/python/wheel/setup.py +%endif +%python_exec ../setup.py dist_info -o ../ + +%install +%cmake_install + +# Hash-bangs in non-exec python sample scripts +sed -Ei "1{\@/usr/bin/env@d}" \ + %{buildroot}%{_datadir}/%{prj_name}/samples/python/benchmark/bert_benchmark/bert_benchmark.py \ + %{buildroot}%{_datadir}/%{prj_name}/samples/python/benchmark/sync_benchmark/sync_benchmark.py \ + %{buildroot}%{_datadir}/%{prj_name}/samples/python/benchmark/throughput_benchmark/throughput_benchmark.py \ + %{buildroot}%{_datadir}/%{prj_name}/samples/python/classification_sample_async/classification_sample_async.py \ + %{buildroot}%{_datadir}/%{prj_name}/samples/python/hello_classification/hello_classification.py \ + %{buildroot}%{_datadir}/%{prj_name}/samples/python/hello_query_device/hello_query_device.py \ + %{buildroot}%{_datadir}/%{prj_name}/samples/python/hello_reshape_ssd/hello_reshape_ssd.py \ + %{buildroot}%{_datadir}/%{prj_name}/samples/python/model_creation_sample/model_creation_sample.py + +# Unnecessary if we get our package dependencies and lib paths right! +rm -fr %{buildroot}%{_prefix}/install_dependencies \ + %{buildroot}%{_prefix}/setupvars.sh + +%{python_expand rm %{buildroot}%{$python_sitearch}/requirements.txt +chmod -x %{buildroot}%{$python_sitearch}/%{name}/tools/ovc/ovc.py +cp -r %{name}-%{version}.dist-info %{buildroot}%{$python_sitearch}/ +%fdupes %{buildroot}%{$python_sitearch}/%{name}/ +} + +%fdupes %{buildroot}%{_datadir}/ + +# We do not use bundled thirdparty libs +rm -fr %{buildroot}%{_datadir}/licenses/* + +%ldconfig_scriptlets -n %{shlib} +%ldconfig_scriptlets -n %{shlib_c} +%ldconfig_scriptlets -n lib%{name}_ir_frontend%{so_ver} +%ldconfig_scriptlets -n lib%{name}_onnx_frontend%{so_ver} +%ldconfig_scriptlets -n lib%{name}_paddle_frontend%{so_ver} +%ldconfig_scriptlets -n lib%{name}_pytorch_frontend%{so_ver} +%ldconfig_scriptlets -n lib%{name}_tensorflow_lite_frontend%{so_ver} +%ldconfig_scriptlets -n lib%{name}_tensorflow_frontend%{so_ver} + +%files -n %{shlib} +%license LICENSE +%{_libdir}/libopenvino.so.* + +%files -n %{shlib_c} +%license LICENSE +%{_libdir}/libopenvino_c.so.* + +%files -n %{name}-auto-batch-plugin +%dir %{_libdir}/%{prj_name} +%{_libdir}/%{prj_name}/libopenvino_auto_batch_plugin.so + +%files -n %{name}-auto-plugin +%dir %{_libdir}/%{prj_name} +%{_libdir}/%{prj_name}/libopenvino_auto_plugin.so + +%ifarch %{x86_64} +%files -n %{name}-intel-cpu-plugin +%dir %{_libdir}/%{prj_name} +%{_libdir}/%{prj_name}/libopenvino_intel_cpu_plugin.so + +%files -n %{name}-intel-npu-plugin +%dir %{_libdir}/%{prj_name} +%{_libdir}/%{prj_name}/libopenvino_intel_npu_plugin.so +%endif + +%ifarch %{arm64} +%files -n %{name}-arm-cpu-plugin +%dir %{_libdir}/%{prj_name} +%{_libdir}/%{prj_name}/libopenvino_arm_cpu_plugin.so +%endif + +%ifarch riscv64 +%files -n %{name}-riscv-cpu-plugin +%dir %{_libdir}/%{prj_name} +%{_libdir}/%{prj_name}/libopenvino_riscv_cpu_plugin.so +%endif + +%files -n %{name}-hetero-plugin +%dir %{_libdir}/%{prj_name} +%{_libdir}/%{prj_name}/libopenvino_hetero_plugin.so + +%files -n lib%{name}_onnx_frontend%{so_ver} +%{_libdir}/libopenvino_onnx_frontend.so.* + +%files -n lib%{name}_ir_frontend%{so_ver} +%{_libdir}/libopenvino_ir_frontend.so.* + +%files -n lib%{name}_paddle_frontend%{so_ver} +%{_libdir}/libopenvino_paddle_frontend.so.* + +%files -n lib%{name}_pytorch_frontend%{so_ver} +%{_libdir}/libopenvino_pytorch_frontend.so.* + +%files -n lib%{name}_tensorflow_frontend%{so_ver} +%{_libdir}/libopenvino_tensorflow_frontend.so.* + +%files -n lib%{name}_tensorflow_lite_frontend%{so_ver} +%{_libdir}/libopenvino_tensorflow_lite_frontend.so.* + +%files -n %{name}-sample +%license LICENSE +%{_datadir}/%{prj_name}/ +%exclude %{_prefix}/lib/debug/usr/share/OpenVINO/tools/compile_tool/ + +%files -n %{name}-devel +%license LICENSE +%{_includedir}/%{name}/ +%{_libdir}/cmake/%{prj_name}/ +%ifarch riscv64 +%{_libdir}/cmake/xbyak_riscv/ +%endif +%{_libdir}/*.so +%{_libdir}/pkgconfig/openvino.pc + +%files %{python_files openvino} +%license LICENSE +%{python_sitearch}/openvino/ +%{python_sitearch}/openvino*-info/ + +%changelog -- 2.51.1 From 7b3f58176b8ae71a664922eed40b69279c5f5a92e804cde39c18ad451069b949 Mon Sep 17 00:00:00 2001 From: Alessandro de Oliveira Faria Date: Wed, 3 Dec 2025 04:09:16 +0000 Subject: [PATCH 3/3] =?UTF-8?q?-=20Update=20to=202025.4.0=20-=20More=20Gen?= =?UTF-8?q?AI=20coverage=20and=20framework=20integrations=20to=20minimize?= =?UTF-8?q?=20code=20=20=20changes=20=20=20*=20New=20models=20supported:?= =?UTF-8?q?=20=20=20=20=20+=20On=20CPUs=20&=20GPUs:=20Qwen3-Embedding-0.6B?= =?UTF-8?q?,=20Qwen3-Reranker-0.6B,=20=20=20=20=20=20=20Mistral-Small-24B-?= =?UTF-8?q?Instruct-2501.=20=20=20=20=20+=20On=20NPUs:=20Gemma-3-4b-it=20a?= =?UTF-8?q?nd=20Qwen2.5-VL-3B-Instruct.=20=20=20*=20Preview:=20Mixture=20o?= =?UTF-8?q?f=20Experts=20(MoE)=20models=20optimized=20for=20CPUs=20=20=20?= =?UTF-8?q?=20=20and=20GPUs,=20validated=20for=20Qwen3-30B-A3B.=20=20=20*?= =?UTF-8?q?=20GenAI=20pipeline=20integrations:=20Qwen3-Embedding-0.6B=20an?= =?UTF-8?q?d=20=20=20=20=20Qwen3-Reranker-0.6B=20for=20enhanced=20retrieva?= =?UTF-8?q?l/ranking,=20and=20=20=20=20=20Qwen2.5VL-7B=20for=20video=20pip?= =?UTF-8?q?eline.=20-=20Broader=20LLM=20model=20support=20and=20more=20mod?= =?UTF-8?q?el=20compression=20=20=20techniques=20=20=20*=20The=20Neural=20?= =?UTF-8?q?Network=20Compression=20Framework=20(NNCF)=20ONNX=20backend=20?= =?UTF-8?q?=20=20=20=20now=20supports=20INT8=20static=20post-training=20qu?= =?UTF-8?q?antization=20(PTQ)=20=20=20=20=20and=20INT8/INT4=20weight-only?= =?UTF-8?q?=20compression=20to=20ensure=20accuracy=20=20=20=20=20parity=20?= =?UTF-8?q?with=20OpenVINO=20IR=20format=20models.=20SmoothQuant=20algorit?= =?UTF-8?q?hm=20=20=20=20=20support=20added=20for=20INT8=20quantization.?= =?UTF-8?q?=20=20=20*=20Accelerated=20multi-token=20generation=20for=20Gen?= =?UTF-8?q?AI,=20leveraging=20=20=20=20=20optimized=20GPU=20kernels=20to?= =?UTF-8?q?=20deliver=20faster=20inference,=20smarter=20=20=20=20=20KV-cac?= =?UTF-8?q?he=20reuse,=20and=20scalable=20LLM=20performance.=20=20=20*=20G?= =?UTF-8?q?PU=20plugin=20updates=20include=20improved=20performance=20with?= =?UTF-8?q?=20prefix=20=20=20=20=20caching=20for=20chat=20history=20scenar?= =?UTF-8?q?ios=20and=20enhanced=20LLM=20accuracy=20=20=20=20=20with=20dyna?= =?UTF-8?q?mic=20quantization=20support=20for=20INT8.=20-=20More=20portabi?= =?UTF-8?q?lity=20and=20performance=20to=20run=20AI=20at=20the=20edge,=20i?= =?UTF-8?q?n=20the=20=20=20cloud,=20or=20locally.=20=20=20*=20Announcing?= =?UTF-8?q?=20support=20for=20Intel=C2=AE=20Core=20Ultra=20Processor=20Ser?= =?UTF-8?q?ies=203.=20=20=20*=20Encrypted=20blob=20format=20support=20adde?= =?UTF-8?q?d=20for=20secure=20model=20=20=20=20=20deployment=20with=20Open?= =?UTF-8?q?VINO=20GenAI.=20Model=20weights=20and=20artifacts=20=20=20=20?= =?UTF-8?q?=20are=20stored=20and=20transmitted=20in=20an=20encrypted=20for?= =?UTF-8?q?mat,=20reducing=20=20=20=20=20risks=20of=20IP=20theft=20during?= =?UTF-8?q?=20deployment.=20Developers=20can=20deploy=20=20=20=20=20with?= =?UTF-8?q?=20minimal=20code=20changes=20using=20OpenVINO=20GenAI=20pipeli?= =?UTF-8?q?nes.=20=20=20*=20OpenVINO=E2=84=A2=20Model=20Server=20and=20Ope?= =?UTF-8?q?nVINO=E2=84=A2=20GenAI=20now=20extend=20=20=20=20=20support=20f?= =?UTF-8?q?or=20Agentic=20AI=20scenarios=20with=20new=20features=20such=20?= =?UTF-8?q?as=20=20=20=20=20output=20parsing=20and=20improved=20chat=20tem?= =?UTF-8?q?plates=20for=20reliable=20=20=20=20=20multi-turn=20interactions?= =?UTF-8?q?,=20and=20preview=20functionality=20for=20the=20=20=20=20=20Qwe?= =?UTF-8?q?n3-30B-A3B=20model.=20OVMS=20also=20introduces=20a=20preview=20?= =?UTF-8?q?for=20=20=20=20=20audio=20endpoints.=20=20=20*=20NPU=20deployme?= =?UTF-8?q?nt=20is=20simplified=20with=20batch=20support,=20enabling=20=20?= =?UTF-8?q?=20=20=20seamless=20model=20execution=20across=20Intel=C2=AE=20?= =?UTF-8?q?Core=20Ultra=20=20=20=20=20processors=20while=20eliminating=20d?= =?UTF-8?q?river=20dependencies.=20Models=20=20=20=20=20are=20reshaped=20t?= =?UTF-8?q?o=20batch=5Fsize=3D1=20before=20compilation.=20=20=20*=20The=20?= =?UTF-8?q?improved=20NVIDIA=20Triton=20Server*=20integration=20with=20=20?= =?UTF-8?q?=20=20=20OpenVINO=20backend=20now=20enables=20developers=20to?= =?UTF-8?q?=20utilize=20Intel=20=20=20=20=20GPUs=20or=20NPUs=20for=20deplo?= =?UTF-8?q?yment.=20-=20Update=20to=202025.3.0=20-=20More=20GenAI=20covera?= =?UTF-8?q?ge=20and=20framework=20integrations=20to=20minimize=20code=20?= =?UTF-8?q?=20=20changes=20=20=20*=20New=20models=20supported:=20Phi-4-min?= =?UTF-8?q?i-reasoning,=20AFM-4.5B,=20=20=20=20=20Gemma-3-1B-it,=20Gemma-3?= =?UTF-8?q?-4B-it,=20and=20Gemma-3-12B,=20=20=20*=20NPU=20support=20added?= =?UTF-8?q?=20for:=20Qwen3-1.7B,=20Qwen3-4B,=20and=20Qwen3-8B.=20=20=20*?= =?UTF-8?q?=20LLMs=20optimized=20for=20NPU=20now=20available=20on=20OpenVI?= =?UTF-8?q?NO=20Hugging=20=20=20=20=20Face=20collection.=20-=20Broader=20L?= =?UTF-8?q?LM=20model=20support=20and=20more=20model=20compression=20techn?= =?UTF-8?q?iques=20=20=20*=20The=20NPU=20plug-in=20adds=20support=20for=20?= =?UTF-8?q?longer=20contexts=20of=20up=20to=20=20=20=20=208K=20tokens,=20d?= =?UTF-8?q?ynamic=20prompts,=20and=20dynamic=20LoRA=20for=20improved=20=20?= =?UTF-8?q?=20=20=20LLM=20performance.=20=20=20*=20The=20NPU=20plug-in=20n?= =?UTF-8?q?ow=20supports=20dynamic=20batch=20sizes=20by=20reshaping=20=20?= =?UTF-8?q?=20=20=20the=20model=20to=20a=20batch=20size=20of=201=20and=20c?= =?UTF-8?q?oncurrently=20managing=20=20=20=20=20multiple=20inference=20req?= =?UTF-8?q?uests,=20enhancing=20performance=20and=20=20=20=20=20optimizing?= =?UTF-8?q?=20memory=20utilization.=20=20=20*=20Accuracy=20improvements=20?= =?UTF-8?q?for=20GenAI=20models=20on=20both=20built-in=20=20=20=20=20and?= =?UTF-8?q?=20discrete=20graphics=20achieved=20through=20the=20implementat?= =?UTF-8?q?ion=20=20=20=20=20of=20the=20key=20cache=20compression=20per=20?= =?UTF-8?q?channel=20technique,=20in=20=20=20=20=20addition=20to=20the=20e?= =?UTF-8?q?xisting=20KV=20cache=20per-token=20compression=20=20=20=20=20me?= =?UTF-8?q?thod.=20=20=20*=20OpenVINO=E2=84=A2=20GenAI=20introduces=20Text?= =?UTF-8?q?RerankPipeline=20for=20improved=20=20=20=20=20retrieval=20relev?= =?UTF-8?q?ance=20and=20RAG=20pipeline=20accuracy,=20plus=20=20=20=20=20St?= =?UTF-8?q?ructured=20Output=20for=20enhanced=20response=20reliability=20a?= =?UTF-8?q?nd=20=20=20=20=20function=20calling=20while=20ensuring=20adhere?= =?UTF-8?q?nce=20to=20predefined=20=20=20=20=20formats.=20-=20More=20porta?= =?UTF-8?q?bility=20and=20performance=20to=20run=20AI=20at=20the=20edge,?= =?UTF-8?q?=20=20=20in=20the=20cloud,=20or=20locally.=20=20=20*=20Announci?= =?UTF-8?q?ng=20support=20for=20Intel=C2=AE=20Arc=E2=84=A2=20Pro=20B-Serie?= =?UTF-8?q?s=20=20=20=20=20(B50=20and=20B60).=20=20=20*=20Preview:=20Huggi?= =?UTF-8?q?ng=20Face=20models=20that=20are=20GGUF-enabled=20for=20=20=20?= =?UTF-8?q?=20=20OpenVINO=20GenAI=20are=20now=20supported=20by=20the=20Ope?= =?UTF-8?q?nVINO=E2=84=A2=20Model=20=20=20=20=20Server=20for=20popular=20L?= =?UTF-8?q?LM=20model=20architectures=20such=20as=20=20=20=20=20DeepSeek?= =?UTF-8?q?=20Distill,=20Qwen2,=20Qwen2.5,=20and=20Llama=203.=20=20=20=20?= =?UTF-8?q?=20This=20functionality=20reduces=20memory=20footprint=20and=20?= =?UTF-8?q?=20=20=20=20simplifies=20integration=20for=20GenAI=20workloads.?= =?UTF-8?q?=20=20=20*=20With=20improved=20reliability=20and=20tool=20call?= =?UTF-8?q?=20accuracy,=20=20=20=20=20the=20OpenVINO=E2=84=A2=20Model=20Se?= =?UTF-8?q?rver=20boosts=20support=20for=20=20=20=20=20agentic=20AI=20use?= =?UTF-8?q?=20cases=20on=20AI=20PCs,=20while=20enhancing=20=20=20=20=20per?= =?UTF-8?q?formance=20on=20Intel=20CPUs,=20built-in=20GPUs,=20and=20NPUs.?= =?UTF-8?q?=20=20=20*=20int4=20data-aware=20weights=20compression,=20now?= =?UTF-8?q?=20supported=20in=20the=20=20=20=20=20Neural=20Network=20Compre?= =?UTF-8?q?ssion=20Framework=20(NNCF)=20for=20ONNX=20=20=20=20=20models,?= =?UTF-8?q?=20reduces=20memory=20footprint=20while=20maintaining=20=20=20?= =?UTF-8?q?=20=20accuracy=20and=20enables=20efficient=20deployment=20in=20?= =?UTF-8?q?=20=20=20=20resource-constrained=20environments.=20-=20openSUSE?= =?UTF-8?q?=20Leap=2016.0=20compatibility=20-=20Remove=20openvino-gcc5-com?= =?UTF-8?q?patibility.patch=20file=20Summary=20of=20major=20features=20and?= =?UTF-8?q?=20improvements=E2=80=AF=E2=80=AF=20-=20More=20GenAI=20coverage?= =?UTF-8?q?=20and=20framework=20integrations=20to=20minimize=20code=20=20?= =?UTF-8?q?=20changes=20=20=20*=20New=20models=20supported=20on=20CPUs=20&?= =?UTF-8?q?=20GPUs:=20Phi-4,=20=20=20=20=20Mistral-7B-Instruct-v0.3,=20SD-?= =?UTF-8?q?XL=20Inpainting=200.1,=20Stable=20=20=20=20=20Diffusion=203.5?= =?UTF-8?q?=20Large=20Turbo,=20Phi-4-reasoning,=20Qwen3,=20and=20=20=20=20?= =?UTF-8?q?=20Qwen2.5-VL-3B-Instruct.=20Mistral=207B=20Instruct=20v0.3=20i?= =?UTF-8?q?s=20also=20=20=20=20=20supported=20on=20NPUs.=20=20=20*=20Previ?= =?UTF-8?q?ew:=20OpenVINO=20=E2=84=A2=20GenAI=20introduces=20a=20text-to-s?= =?UTF-8?q?peech=20=20=20=20=20pipeline=20for=20the=20SpeechT5=20TTS=20mod?= =?UTF-8?q?el,=20while=20the=20new=20RAG=20=20=20=20=20backend=20offers=20?= =?UTF-8?q?developers=20a=20simplified=20API=20that=20delivers=20=20=20=20?= =?UTF-8?q?=20reduced=20memory=20usage=20and=20improved=20performance.=20?= =?UTF-8?q?=20=20*=20Preview:=20OpenVINO=E2=84=A2=20GenAI=20offers=20a=20G?= =?UTF-8?q?GUF=20Reader=20for=20seamless=20=20=20=20=20integration=20of=20?= =?UTF-8?q?llama.cpp=20based=20LLMs,=20with=20Python=20and=20C++=20=20=20?= =?UTF-8?q?=20=20pipelines=20that=20load=20GGUF=20models,=20build=20OpenVI?= =?UTF-8?q?NO=20graphs,=20=20=20=20=20and=20run=20GPU=20inference=20on-the?= =?UTF-8?q?-fly.=20Validated=20for=20popular=20models:=20=20=20=20=20DeepS?= =?UTF-8?q?eek-R1-Distill-Qwen=20(1.5B,=207B),=20Qwen2.5=20Instruct=20=20?= =?UTF-8?q?=20=20=20(1.5B,=203B,=207B)=20&=20llama-3.2=20Instruct=20(1B,?= =?UTF-8?q?=203B,=208B).=20-=20Broader=20LLM=20model=20support=20and=20mor?= =?UTF-8?q?e=20model=20compression=20=20=20techniques=20=20=20*=20Further?= =?UTF-8?q?=20optimization=20of=20LoRA=20adapters=20in=20OpenVINO=20GenAI?= =?UTF-8?q?=20=20=20=20=20for=20improved=20LLM,=20VLM,=20and=20text-to-ima?= =?UTF-8?q?ge=20model=20performance=20=20=20=20=20on=20built-in=20GPUs.=20?= =?UTF-8?q?Developers=20can=20use=20LoRA=20adapters=20to=20=20=20=20=20qui?= =?UTF-8?q?ckly=20customize=20models=20for=20specialized=20tasks.=20=20=20?= =?UTF-8?q?*=20KV=20cache=20compression=20for=20CPUs=20is=20enabled=20by?= =?UTF-8?q?=20default=20for=20=20=20=20=20INT8,=20providing=20a=20reduced?= =?UTF-8?q?=20memory=20footprint=20while=20maintaining=20=20=20=20=20accur?= =?UTF-8?q?acy=20compared=20to=20FP16.=20Additionally,=20it=20delivers=20?= =?UTF-8?q?=20=20=20=20substantial=20memory=20savings=20for=20LLMs=20with?= =?UTF-8?q?=20INT4=20support=20compared=20=20=20=20=20to=20INT8.=20=20=20*?= =?UTF-8?q?=20Optimizations=20for=20Intel=C2=AE=20Core=E2=84=A2=20Ultra=20?= =?UTF-8?q?Processor=20Series=202=20=20=20=20=20built-in=20GPUs=20and=20In?= =?UTF-8?q?tel=C2=AE=20Arc=E2=84=A2=20B=20Series=20Graphics=20with=20the?= =?UTF-8?q?=20=20=20=20=20Intel=C2=AE=20XMX=20systolic=20platform=20to=20e?= =?UTF-8?q?nhance=20the=20performance=20of=20=20=20=20=20VLM=20models=20an?= =?UTF-8?q?d=20hybrid=20quantized=20image=20generation=20models,=20as=20?= =?UTF-8?q?=20=20=20=20well=20as=20improve=20first-token=20latency=20for?= =?UTF-8?q?=20LLMs=20through=20dynamic=20=20=20=20=20quantization.=20-=20M?= =?UTF-8?q?ore=20portability=20and=20performance=20to=20run=20AI=20at=20th?= =?UTF-8?q?e=20edge,=20in=20the=20=20=20cloud,=20or=20locally.=20=20=20*?= =?UTF-8?q?=20Enhanced=20Linux*=20support=20with=20the=20latest=20GPU=20dr?= =?UTF-8?q?iver=20for=20=20=20=20=20built-in=20GPUs=20on=20Intel=C2=AE=20C?= =?UTF-8?q?ore=E2=84=A2=20Ultra=20Processor=20Series=202=20=20=20=20=20(fo?= =?UTF-8?q?rmerly=20codenamed=20Arrow=20Lake=20H).=20=20=20*=20Support=20f?= =?UTF-8?q?or=20INT4=20data-free=20weights=20compression=20for=20ONNX=20?= =?UTF-8?q?=20=20=20=20models=20implemented=20in=20the=20Neural=20Network?= =?UTF-8?q?=20Compression=20=20=20=20=20Framework=20(NNCF).=20=20=20*=20NP?= =?UTF-8?q?U=20support=20for=20FP16-NF4=20precision=20on=20Intel=C2=AE=20C?= =?UTF-8?q?ore=E2=84=A2=20200V=20=20=20=20=20Series=20processors=20for=20m?= =?UTF-8?q?odels=20with=20up=20to=208B=20parameters=20is=20=20=20=20=20ena?= =?UTF-8?q?bled=20through=20symmetrical=20and=20channel-wise=20quantizatio?= =?UTF-8?q?n,=20=20=20=20=20improving=20accuracy=20while=20maintaining=20p?= =?UTF-8?q?erformance=20efficiency.=20Support=20Change=20and=20Deprecation?= =?UTF-8?q?=20Notices=20-=20Discontinued=20in=202025:=20=20=20*=20Runtime?= =?UTF-8?q?=20components:=20=20=20=20=20+=20The=20OpenVINO=20property=20of?= =?UTF-8?q?=20Affinity=20API=20is=20no=20longer=20=20=20=20=20=20=20availa?= =?UTF-8?q?ble.=20It=20has=20been=20replaced=20with=20CPU=20binding=20=20?= =?UTF-8?q?=20=20=20=20=20configurations=20(ov::hint::enable=5Fcpu=5Fpinni?= =?UTF-8?q?ng).=20=20=20=20=20+=20The=20openvino-nightly=20PyPI=20module?= =?UTF-8?q?=20has=20been=20discontinued.=20=20=20=20=20=20=20End-users=20s?= =?UTF-8?q?hould=20proceed=20with=20the=20Simple=20PyPI=20nightly=20repo?= =?UTF-8?q?=20=20=20=20=20=20=20instead.=20More=20information=20in=20Relea?= =?UTF-8?q?se=20Policy.=20The=20=20=20=20=20=20=20openvino-nightly=20PyPI?= =?UTF-8?q?=20module=20has=20been=20discontinued.=20=20=20=20=20=20=20End-?= =?UTF-8?q?users=20should=20proceed=20with=20the=20Simple=20PyPI=20nightly?= =?UTF-8?q?=20repo=20=20=20=20=20=20=20instead.=20More=20information=20in?= =?UTF-8?q?=20Release=20Policy.=20=20=20*=20Tools:=20=20=20=20=20+=20The?= =?UTF-8?q?=20OpenVINO=E2=84=A2=20Development=20Tools=20package=20(pip=20i?= =?UTF-8?q?nstall=20=20=20=20=20=20=20openvino-dev)=20is=20no=20longer=20a?= =?UTF-8?q?vailable=20for=20OpenVINO=20releases=20=20=20=20=20=20=20in=202?= =?UTF-8?q?025.=20=20=20=20=20+=20Model=20Optimizer=20is=20no=20longer=20a?= =?UTF-8?q?vailable.=20Consider=20using=20the=20=20=20=20=20=20=20new=20co?= =?UTF-8?q?nversion=20methods=20instead.=20For=20more=20details,=20see=20t?= =?UTF-8?q?he=20=20=20=20=20=20=20model=20conversion=20transition=20guide.?= =?UTF-8?q?=20=20=20=20=20+=20Intel=C2=AE=20Streaming=20SIMD=20Extensions?= =?UTF-8?q?=20(Intel=C2=AE=20SSE)=20are=20currently=20=20=20=20=20=20=20no?= =?UTF-8?q?t=20enabled=20in=20the=20binary=20package=20by=20default.=20The?= =?UTF-8?q?y=20are=20=20=20=20=20=20=20still=20supported=20in=20the=20sour?= =?UTF-8?q?ce=20code=20form.=20=20=20=20=20+=20Legacy=20prefixes:=20l=5F,?= =?UTF-8?q?=20w=5F,=20and=20m=5F=20have=20been=20removed=20from=20=20=20?= =?UTF-8?q?=20=20=20=20OpenVINO=20archive=20names.=20=20=20*=20OpenVINO=20?= =?UTF-8?q?GenAI:=20=20=20=20=20+=20StreamerBase::put(int64=5Ft=20token)?= =?UTF-8?q?=20=20=20=20=20+=20The=20Bool=20value=20for=20Callback=20stream?= =?UTF-8?q?er=20is=20no=20longer=20accepted.=20=20=20=20=20=20=20It=20must?= =?UTF-8?q?=20now=20return=20one=20of=20three=20values=20of=20StreamingSta?= =?UTF-8?q?tus=20=20=20=20=20=20=20enum.=20=20=20=20=20+=20ChunkStreamerBa?= =?UTF-8?q?se=20is=20deprecated.=20Use=20StreamerBase=20instead.=20=20=20*?= =?UTF-8?q?=20NNCF=20create=5Fcompressed=5Fmodel()=20method=20is=20now=20d?= =?UTF-8?q?eprecated.=20=20=20=20=20nncf.quantize()=20method=20is=20recomm?= =?UTF-8?q?ended=20for=20=20=20=20=20Quantization-Aware=20Training=20of=20?= =?UTF-8?q?PyTorch=20and=20TensorFlow=20models.=20=20=20*=20OpenVINO=20Mod?= =?UTF-8?q?el=20Server=20(OVMS)=20benchmark=20client=20in=20C++=20=20=20?= =?UTF-8?q?=20=20using=20TensorFlow=20Serving=20API.=20-=20Deprecated=20an?= =?UTF-8?q?d=20to=20be=20removed=20in=20the=20future:=20=20=20*=20Python?= =?UTF-8?q?=203.9=20is=20now=20deprecated=20and=20will=20be=20unavailable?= =?UTF-8?q?=20after=20=20=20=20=20OpenVINO=20version=202025.4.=20=20=20*?= =?UTF-8?q?=20openvino.Type.undefined=20is=20now=20deprecated=20and=20will?= =?UTF-8?q?=20be=20removed=20=20=20=20=20with=20version=202026.0.=20openvi?= =?UTF-8?q?no.Type.dynamic=20should=20be=20used=20=20=20=20=20instead.=20?= =?UTF-8?q?=20=20*=20APT=20&=20YUM=20Repositories=20Restructure:=20Startin?= =?UTF-8?q?g=20with=20release=20=20=20=20=202025.1,=20users=20can=20switch?= =?UTF-8?q?=20to=20the=20new=20repository=20structure=20=20=20=20=20for=20?= =?UTF-8?q?APT=20and=20YUM,=20which=20no=20longer=20uses=20year-based=20?= =?UTF-8?q?=20=20=20=20subdirectories=20(like=20=E2=80=9C2025=E2=80=9D).?= =?UTF-8?q?=20The=20old=20(legacy)=20structure=20=20=20=20=20will=20still?= =?UTF-8?q?=20be=20available=20until=202026,=20when=20the=20change=20will?= =?UTF-8?q?=20=20=20=20=20be=20finalized.=20Detailed=20instructions=20are?= =?UTF-8?q?=20available=20on=20the=20=20=20=20=20relevant=20documentation?= =?UTF-8?q?=20pages:=20=20=20=20=20+=20Installation=20guide=20-=20yum=20?= =?UTF-8?q?=20=20=20=20+=20Installation=20guide=20-=20apt=20=20=20*=20Open?= =?UTF-8?q?CV=20binaries=20will=20be=20removed=20from=20Docker=20images=20?= =?UTF-8?q?in=202026.=20=20=20*=20Ubuntu=2020.04=20support=20will=20be=20d?= =?UTF-8?q?eprecated=20in=20future=20OpenVINO=20=20=20=20=20releases=20due?= =?UTF-8?q?=20to=20the=20end=20of=20standard=20support.=20=20=20*=20?= =?UTF-8?q?=E2=80=9Cauto=20shape=E2=80=9D=20and=20=E2=80=9Cauto=20batch=20?= =?UTF-8?q?size=E2=80=9D=20(reshaping=20a=20model=20in=20=20=20=20=20runti?= =?UTF-8?q?me)=20will=20be=20removed=20in=20the=20future.=20OpenVINO?= =?UTF-8?q?=E2=80=99s=20dynamic=20=20=20=20=20shape=20models=20are=20recom?= =?UTF-8?q?mended=20instead.=20=20=20*=20MacOS=20x86=20is=20no=20longer=20?= =?UTF-8?q?recommended=20for=20use=20due=20to=20the=20=20=20=20=20disconti?= =?UTF-8?q?nuation=20of=20validation.=20Full=20support=20will=20be=20remov?= =?UTF-8?q?ed=20=20=20=20=20later=20in=202025.=20=20=20*=20The=20openvino?= =?UTF-8?q?=20namespace=20of=20the=20OpenVINO=20Python=20API=20has=20been?= =?UTF-8?q?=20=20=20=20=20redesigned,=20removing=20the=20nested=20openvino?= =?UTF-8?q?.runtime=20module.=20=20=20=20=20The=20old=20namespace=20is=20n?= =?UTF-8?q?ow=20considered=20deprecated=20and=20will=20be=20=20=20=20=20di?= =?UTF-8?q?scontinued=20in=202026.0.=20-=20Fix=20file=20list=20for=20riscv?= =?UTF-8?q?64=20-=20Do=20not=20force=20GCC15=20on=20Tumblewed=20just=20yet?= =?UTF-8?q?:=20follow=20the=20distro=20=20=20default=20compiler,=20like=20?= =?UTF-8?q?any=20other=20package.=20-=20openvino-gcc5-compatibility.patch?= =?UTF-8?q?=20to=20resolve=20incompatibility=20=20=20in=20gcc5=20-=20Added?= =?UTF-8?q?=20gcc-14=20-=20Update=20to=202025.1.0=20-=20Downgrade=20from?= =?UTF-8?q?=20gcc13-c++=20to=2012=20due=20to=20incompatibility=20in=20tbb?= =?UTF-8?q?=20=20=20compilation.=20=20This=20was=20due=20to=20C++=20librar?= =?UTF-8?q?ies=20(using=20libstdc++)=20=20=20resulting=20in=20the=20error:?= =?UTF-8?q?=20libtbb.so.12:=20undefined=20reference=20to=20=20=20`=5F=5Fcx?= =?UTF-8?q?a=5Fcall=5Fterminate@CXXABI=5F1.3.15'=20-=20More=20GenAI=20cove?= =?UTF-8?q?rage=20and=20framework=20integrations=20to=20minimize=20code=20?= =?UTF-8?q?=20=20changes=20=20=20*=20New=20models=20supported:=20Phi-4=20M?= =?UTF-8?q?ini,=20Jina=20CLIP=20v1,=20and=20Bce=20=20=20=20=20Embedding=20?= =?UTF-8?q?Base=20v1.=20=20=20*=20OpenVINO=E2=84=A2=20Model=20Server=20now?= =?UTF-8?q?=20supports=20VLM=20models,=20including=20=20=20=20=20Qwen2-VL,?= =?UTF-8?q?=20Phi-3.5-Vision,=20and=20InternVL2.=20=20=20*=20OpenVINO=20Ge?= =?UTF-8?q?nAI=20now=20includes=20image-to-image=20and=20inpainting=20=20?= =?UTF-8?q?=20=20=20features=20for=20transformer-based=20pipelines,=20such?= =?UTF-8?q?=20as=20Flux.1=20and=20=20=20=20=20Stable=20Diffusion=203=20mod?= =?UTF-8?q?els,=20enhancing=20their=20ability=20to=20generate=20=20=20=20?= =?UTF-8?q?=20more=20realistic=20content.=20=20=20*=20Preview:=20AI=20Play?= =?UTF-8?q?ground=20now=20utilizes=20the=20OpenVINO=20Gen=20AI=20backend?= =?UTF-8?q?=20=20=20=20=20to=20enable=20highly=20optimized=20inferencing?= =?UTF-8?q?=20performance=20on=20AI=20PCs.=20-=20Broader=20LLM=20model=20s?= =?UTF-8?q?upport=20and=20more=20model=20compression=20techniques=20=20=20?= =?UTF-8?q?*=20Reduced=20binary=20size=20through=20optimization=20of=20the?= =?UTF-8?q?=20CPU=20plugin=20and=20=20=20=20=20removal=20of=20the=20GEMM?= =?UTF-8?q?=20kernel.=20=20=20*=20Optimization=20of=20new=20kernels=20for?= =?UTF-8?q?=20the=20GPU=20plugin=20significantly=20=20=20=20=20boosts=20th?= =?UTF-8?q?e=20performance=20of=20Long=20Short-Term=20Memory=20(LSTM)=20mo?= =?UTF-8?q?dels,=20=20=20=20=20used=20in=20many=20applications,=20includin?= =?UTF-8?q?g=20speech=20recognition,=20=20=20=20=20language=20modeling,=20?= =?UTF-8?q?and=20time=20series=20forecasting.=20=20=20*=20Preview:=20Token?= =?UTF-8?q?=20Eviction=20implemented=20in=20OpenVINO=20GenAI=20to=20reduce?= =?UTF-8?q?=20=20=20=20=20the=20memory=20consumption=20of=20KV=20Cache=20b?= =?UTF-8?q?y=20eliminating=20unimportant=20=20=20=20=20tokens.=20This=20cu?= =?UTF-8?q?rrent=20Token=20Eviction=20implementation=20is=20=20=20=20=20be?= =?UTF-8?q?neficial=20for=20tasks=20where=20a=20long=20sequence=20is=20gen?= =?UTF-8?q?erated,=20such=20=20=20=20=20as=20chatbots=20and=20code=20gener?= =?UTF-8?q?ation.=20=20=20*=20NPU=20acceleration=20for=20text=20generation?= =?UTF-8?q?=20is=20now=20enabled=20in=20=20=20=20=20OpenVINO=E2=84=A2=20Ru?= =?UTF-8?q?ntime=20and=20OpenVINO=E2=84=A2=20Model=20Server=20to=20support?= =?UTF-8?q?=20the=20=20=20=20=20power-efficient=20deployment=20of=20VLM=20?= =?UTF-8?q?models=20on=20NPUs=20for=20AI=20PC=20use=20=20=20=20=20cases=20?= =?UTF-8?q?with=20low=20concurrency.=20-=20More=20portability=20and=20perf?= =?UTF-8?q?ormance=20to=20run=20AI=20at=20the=20edge,=20in=20the=20=20=20c?= =?UTF-8?q?loud,=20or=20locally.=20=20=20*=20Additional=20LLM=20performanc?= =?UTF-8?q?e=20optimizations=20on=20Intel=C2=AE=20Core=E2=84=A2=20Ultra=20?= =?UTF-8?q?=20=20=20=20200H=20series=20processors=20for=20improved=202nd?= =?UTF-8?q?=20token=20latency=20on=20=20=20=20=20Linux.=20=20=20*=20Enhanc?= =?UTF-8?q?ed=20performance=20and=20efficient=20resource=20utilization=20w?= =?UTF-8?q?ith=20=20=20=20=20the=20implementation=20of=20Paged=20Attention?= =?UTF-8?q?=20and=20Continuous=20Batching=20=20=20=20=20by=20default=20in?= =?UTF-8?q?=20the=20GPU=20plugin.=20=20=20*=20Preview:=20The=20new=20OpenV?= =?UTF-8?q?INO=20backend=20for=20Executorch=20will=20enable=20=20=20=20=20?= =?UTF-8?q?accelerated=20inference=20and=20improved=20performance=20on=20I?= =?UTF-8?q?ntel=20=20=20=20=20hardware,=20including=20CPUs,=20GPUs,=20and?= =?UTF-8?q?=20NPUs.=20-=20Disabled=20JAX=20plugin=20beta.=20-=20Update=20t?= =?UTF-8?q?o=202025.0.0=20-=20More=20GenAI=20coverage=20and=20framework=20?= =?UTF-8?q?integrations=20to=20minimize=20code=20=20=20changes=20=20=20*?= =?UTF-8?q?=20New=20models=20supported:=20Qwen=202.5,=20Deepseek-R1-Distil?= =?UTF-8?q?l-Llama-8B,=20=20=20=20=20DeepSeek-R1-Distill-Qwen-7B,=20and=20?= =?UTF-8?q?DeepSeek-R1-Distill-Qwen-1.5B,=20=20=20=20=20FLUX.1=20Schnell?= =?UTF-8?q?=20and=20FLUX.1=20Dev=20=20=20*=20Whisper=20Model:=20Improved?= =?UTF-8?q?=20performance=20on=20CPUs,=20built-in=20GPUs,=20=20=20=20=20an?= =?UTF-8?q?d=20discrete=20GPUs=20with=20GenAI=20API.=20=20=20*=20Preview:?= =?UTF-8?q?=20Introducing=20NPU=20support=20for=20torch.compile,=20giving?= =?UTF-8?q?=20=20=20=20=20developers=20the=20ability=20to=20use=20the=20Op?= =?UTF-8?q?enVINO=20backend=20to=20run=20the=20=20=20=20=20PyTorch=20API?= =?UTF-8?q?=20on=20NPUs.=20300+=20deep=20learning=20models=20enabled=20fro?= =?UTF-8?q?m=20=20=20=20=20the=20TorchVision,=20Timm,=20and=20TorchBench?= =?UTF-8?q?=20repositories..=20-=20Broader=20Large=20Language=20Model=20(L?= =?UTF-8?q?LM)=20support=20and=20more=20model=20=20=20compression=20techni?= =?UTF-8?q?ques.=20=20=20*=20Preview:=20Addition=20of=20Prompt=20Lookup=20?= =?UTF-8?q?to=20GenAI=20API=20improves=202nd=20=20=20=20=20token=20latency?= =?UTF-8?q?=20for=20LLMs=20by=20effectively=20utilizing=20predefined=20=20?= =?UTF-8?q?=20=20=20prompts=20that=20match=20the=20intended=20use=20case.?= =?UTF-8?q?=20=20=20*=20Preview:=20The=20GenAI=20API=20now=20offers=20imag?= =?UTF-8?q?e-to-image=20inpainting=20=20=20=20=20functionality.=20This=20f?= =?UTF-8?q?eature=20enables=20models=20to=20generate=20=20=20=20=20realist?= =?UTF-8?q?ic=20content=20by=20inpainting=20specified=20modifications=20an?= =?UTF-8?q?d=20=20=20=20=20seamlessly=20integrating=20them=20with=20the=20?= =?UTF-8?q?original=20image.=20=20=20*=20Asymmetric=20KV=20Cache=20compres?= =?UTF-8?q?sion=20is=20now=20enabled=20for=20INT8=20on=20=20=20=20=20CPUs,?= =?UTF-8?q?=20resulting=20in=20lower=20memory=20consumption=20and=20improv?= =?UTF-8?q?ed=202nd=20=20=20=20=20token=20latency,=20especially=20when=20d?= =?UTF-8?q?ealing=20with=20long=20prompts=20that=20=20=20=20=20require=20s?= =?UTF-8?q?ignificant=20memory.=20The=20option=20should=20be=20explicitly?= =?UTF-8?q?=20=20=20=20=20specified=20by=20the=20user.=20-=20More=20portab?= =?UTF-8?q?ility=20and=20performance=20to=20run=20AI=20at=20the=20edge,=20?= =?UTF-8?q?in=20the=20=20=20cloud,=20or=20locally.=20=20=20*=20Support=20f?= =?UTF-8?q?or=20the=20latest=20Intel=C2=AE=20Core=E2=84=A2=20Ultra=20200H?= =?UTF-8?q?=20series=20=20=20=20=20processors=20(formerly=20codenamed=20Ar?= =?UTF-8?q?row=20Lake-H)=20=20=20*=20Integration=20of=20the=20OpenVINO=20?= =?UTF-8?q?=E2=84=A2=20backend=20with=20the=20Triton=20=20=20=20=20Inferen?= =?UTF-8?q?ce=20Server=20allows=20developers=20to=20utilize=20the=20Triton?= =?UTF-8?q?=20=20=20=20=20server=20for=20enhanced=20model=20serving=20perf?= =?UTF-8?q?ormance=20when=20deploying=20=20=20=20=20on=20Intel=20CPUs.=20?= =?UTF-8?q?=20=20*=20Preview:=20A=20new=20OpenVINO=20=E2=84=A2=20backend?= =?UTF-8?q?=20integration=20allows=20=20=20=20=20developers=20to=20leverag?= =?UTF-8?q?e=20OpenVINO=20performance=20optimizations=20=20=20=20=20direct?= =?UTF-8?q?ly=20within=20Keras=203=20workflows=20for=20faster=20AI=20infer?= =?UTF-8?q?ence=20on=20=20=20=20=20CPUs,=20built-in=20GPUs,=20discrete=20G?= =?UTF-8?q?PUs,=20and=20NPUs.=20This=20feature=20is=20=20=20=20=20availabl?= =?UTF-8?q?e=20with=20the=20latest=20Keras=203.8=20release.=20-=20Support?= =?UTF-8?q?=20Change=20and=20Deprecation=20Notices=20=20=20*=20Now=20depre?= =?UTF-8?q?cated:=20=20=20=20=20+=20Legacy=20prefixes=20l=5F,=20w=5F,=20an?= =?UTF-8?q?d=20m=5F=20have=20been=20removed=20from=20=20=20=20=20=20=20Ope?= =?UTF-8?q?nVINO=20archive=20names.=20=20=20=20=20+=20The=20runtime=20name?= =?UTF-8?q?space=20for=20Python=20API=20has=20been=20marked=20as=20=20=20?= =?UTF-8?q?=20=20=20=20deprecated=20and=20designated=20to=20be=20removed?= =?UTF-8?q?=20for=202026.0.=20The=20=20=20=20=20=20=20new=20namespace=20st?= =?UTF-8?q?ructure=20has=20been=20delivered,=20and=20migration=20=20=20=20?= =?UTF-8?q?=20=20=20is=20possible=20=20immediately.=20Details=20will=20be?= =?UTF-8?q?=20communicated=20=20=20=20=20=20=20through=20warnings=20andvia?= =?UTF-8?q?=20documentation.=20=20=20=20=20+=20NNCF=20create=5Fcompressed?= =?UTF-8?q?=5Fmodel()=20method=20is=20deprecated.=20=20=20=20=20=20=20nncf?= =?UTF-8?q?.quantize()=20method=20is=20now=20recommended=20for=20=20=20=20?= =?UTF-8?q?=20=20=20Quantization-Aware=20Training=20of=20PyTorch=20and=20?= =?UTF-8?q?=20=20=20=20=20=20TensorFlow=20models.=20-=20openvino-onnx-ml-d?= =?UTF-8?q?efines.patch=20and=20=20=20openvino-remove-npu-compile-tool.pat?= =?UTF-8?q?chhas=20been=20removed=20=20=20as=20it=20is=20no=20longer=20nee?= =?UTF-8?q?ded=20in=20this=20version.=20-=20Update=20to=202024.4.0=20-=20S?= =?UTF-8?q?ummary=20of=20major=20features=20and=20improvements=E2=80=AF?= =?UTF-8?q?=E2=80=AF=20=20=20*=20OpenVINO=202024.6=20release=20includes=20?= =?UTF-8?q?updates=20for=20enhanced=20=20=20=20=20stability=20and=20improv?= =?UTF-8?q?ed=20LLM=20performance.=20=20=20*=20Introduced=20support=20for?= =?UTF-8?q?=20Intel=C2=AE=20Arc=E2=84=A2=20B-Series=20Graphics=20=20=20=20?= =?UTF-8?q?=20(formerly=20known=20as=20Battlemage).=20=20=20*=20Implemente?= =?UTF-8?q?d=20optimizations=20to=20improve=20the=20inference=20time=20and?= =?UTF-8?q?=20=20=20=20=20LLM=20performance=20on=20NPUs.=20=20=20*=20Impro?= =?UTF-8?q?ved=20LLM=20performance=20with=20GenAI=20API=20optimizations=20?= =?UTF-8?q?and=20=20=20=20=20bug=20fixes.=20-=20Support=20Change=20and=20D?= =?UTF-8?q?eprecation=20Notices=20=20=20*=20Using=20deprecated=20features?= =?UTF-8?q?=20and=20components=20is=20not=20advised.=20They=20=20=20=20=20?= =?UTF-8?q?are=20available=20to=20enable=20a=20smooth=20transition=20to=20?= =?UTF-8?q?new=20solutions=20=20=20=20=20and=20will=20be=20discontinued=20?= =?UTF-8?q?in=20the=20future.=20To=20keep=20using=20=20=20=20=20discontinu?= =?UTF-8?q?ed=20features,=20you=20will=20have=20to=20revert=20to=20the=20l?= =?UTF-8?q?ast=20=20=20=20=20LTS=20OpenVINO=20version=20supporting=20them.?= =?UTF-8?q?=20For=20more=20details,=20refer=20=20=20=20=20to=20the=20OpenV?= =?UTF-8?q?INO=20Legacy=20Features=20and=20Components=20page.=20=20=20*=20?= =?UTF-8?q?Discontinued=20in=202024.0:=20=20=20=20=20+=20Runtime=20compone?= =?UTF-8?q?nts:=20=20=20=20=20=20=20-=20Intel=C2=AE=20Gaussian=20&=20Neura?= =?UTF-8?q?l=20Accelerator=20(Intel=C2=AE=E2=80=AFGNA)..=20=20=20=20=20=20?= =?UTF-8?q?=20=20=20Consider=20using=20the=20Neural=20Processing=20Unit=20?= =?UTF-8?q?(NPU)=20for=20=20=20=20=20=20=20=20=20low-powered=20systems=20l?= =?UTF-8?q?ike=20Intel=C2=AE=20Core=E2=84=A2=20Ultra=20or=2014th=20=20=20?= =?UTF-8?q?=20=20=20=20=20=20generation=20and=20beyond.=20=20=20=20=20=20?= =?UTF-8?q?=20-=20OpenVINO=20C++/C/Python=201.0=20APIs=20(see=202023.3=20A?= =?UTF-8?q?PI=20transition=20=20=20=20=20=20=20=20=20guide=20for=20referen?= =?UTF-8?q?ce).=20=20=20=20=20=20=20-=20All=20ONNX=20Frontend=20legacy=20A?= =?UTF-8?q?PI=20(known=20as=20ONNX=5FIMPORTER=5FAPI)=20=20=20=20=20=20=20-?= =?UTF-8?q?=20'PerfomanceMode.UNDEFINED'=20property=20as=20part=20of=20the?= =?UTF-8?q?=20=20=20=20=20=20=20=20=20OpenVINO=20Python=20API=20=20=20=20?= =?UTF-8?q?=20+=20Tools:=20=20=20=20=20=20=20-=20Deployment=20Manager.=20S?= =?UTF-8?q?ee=20installation=20and=20deployment=20=20=20=20=20=20=20=20=20?= =?UTF-8?q?guides=20for=20current=20distribution=20options.=20=20=20=20=20?= =?UTF-8?q?=20=20-=20Accuracy=20Checker.=20=20=20=20=20=20=20-=20Post-Trai?= =?UTF-8?q?ning=20Optimization=20Tool=20(POT).=E2=80=AFNeural=20Network=20?= =?UTF-8?q?=20=20=20=20=20=20=20=20Compression=20Framework=20(NNCF)=20shou?= =?UTF-8?q?ld=20be=20used=20instead.=20=20=20=20=20=20=20-=20A=20Git=20pat?= =?UTF-8?q?ch=E2=80=AFfor=20NNCF=20integration=20with=E2=80=AF=20=20=20=20?= =?UTF-8?q?=20=20=20=20=20huggingface/transformers.=20The=20recommended=20?= =?UTF-8?q?approach=E2=80=AFis=20=20=20=20=20=20=20=20=20to=20use=E2=80=AF?= =?UTF-8?q?huggingface/optimum-intel=E2=80=AFfor=20applying=20NNCF=20=20?= =?UTF-8?q?=20=20=20=20=20=20=20optimization=20on=20top=20of=20models=20fr?= =?UTF-8?q?om=20Hugging=20Face.=20=20=20=20=20=20=20-=20Support=20for=20Ap?= =?UTF-8?q?ache=20MXNet,=20Caffe,=20and=20Kaldi=20model=20formats.=20=20?= =?UTF-8?q?=20=20=20=20=20=20=20Conversion=20to=20ONNX=20may=20be=20used?= =?UTF-8?q?=20as=20a=20solution.=20=20=20*=20Deprecated=20and=20to=20be=20?= =?UTF-8?q?removed=20in=20the=20future:=20=20=20=20=20+=20The=20macOS=20x8?= =?UTF-8?q?6=5F64=20debug=20bins=20will=20no=20longer=20be=20provided=20?= =?UTF-8?q?=20=20=20=20=20=20with=20the=20OpenVINO=20toolkit,=20starting?= =?UTF-8?q?=20with=20OpenVINO=202024.5.=20=20=20=20=20+=20Python=203.8=20i?= =?UTF-8?q?s=20no=20longer=20supported,=20starting=20with=20=20=20=20=20?= =?UTF-8?q?=20=20OpenVINO=202024.5.=20=20=20=20=20+=20As=20MxNet=20doesn?= =?UTF-8?q?=E2=80=99t=20support=20Python=20version=20higher=20than=203.8,?= =?UTF-8?q?=20=20=20=20=20=20=20according=20to=20the=20MxNet=20PyPI=20proj?= =?UTF-8?q?ect,=20it=20is=20no=20longer=20=20=20=20=20=20=20supported=20by?= =?UTF-8?q?=20OpenVINO,=20either.=20=20=20=20=20+=20Discrete=20Keem=20Bay?= =?UTF-8?q?=20support=20is=20no=20longer=20supported,=20starting=20=20=20?= =?UTF-8?q?=20=20=20=20with=20OpenVINO=202024.5.=20=20=20=20=20+=20Support?= =?UTF-8?q?=20for=20discrete=20devices=20(formerly=20codenamed=20Raptor=20?= =?UTF-8?q?=20=20=20=20=20=20Lake)=20is=20no=20longer=20available=20for=20?= =?UTF-8?q?NPU.=20-=20fix=20build=20on=20tumbleweed=20=20=20*=20currently?= =?UTF-8?q?=20openvino=20does=20not=20support=20protobuf=20v22=20or=20newe?= =?UTF-8?q?r=20-=20Temporarily=20inserted=20gcc-13=20in=20Tumbleweed/Facto?= =?UTF-8?q?ry/Slowroll:=20=20=20Because=20there=20is=20an=20incompatibilit?= =?UTF-8?q?y=20of=20the=20source=20code=20of=20the=20=20=20level-zero=20li?= =?UTF-8?q?brary=20and=20npu=20module=20with=20gcc-14.=20I=20am=20working?= =?UTF-8?q?=20=20=20with=20Intel=20on=20tests=20to=20return=20to=20native?= =?UTF-8?q?=20gcc.=20-=20Update=20to=202024.4.0=20-=20Summary=20of=20major?= =?UTF-8?q?=20features=20and=20improvements=E2=80=AF=E2=80=AF=20=20=20*=20?= =?UTF-8?q?More=20Gen=20AI=20coverage=20and=20framework=20integrations=20t?= =?UTF-8?q?o=20minimize=20=20=20=20=20code=20changes=20=20=20=20=20+=20Sup?= =?UTF-8?q?port=20for=20GLM-4-9B=20Chat,=20MiniCPM-1B,=20Llama=203=20and?= =?UTF-8?q?=203.1,=20=20=20=20=20=20=20Phi-3-Mini,=20Phi-3-Medium=20and=20?= =?UTF-8?q?YOLOX-s=20models.=20=20=20=20=20+=20Noteworthy=20notebooks=20ad?= =?UTF-8?q?ded:=20Florence-2,=20NuExtract-tiny=20=20=20=20=20=20=20Structu?= =?UTF-8?q?re=20Extraction,=20Flux.1=20Image=20Generation,=20PixArt-=CE=B1?= =?UTF-8?q?:=20=20=20=20=20=20=20Photorealistic=20Text-to-Image=20Synthesi?= =?UTF-8?q?s,=20and=20Phi-3-Vision=20=20=20=20=20=20=20Visual=20Language?= =?UTF-8?q?=20Assistant.=20=20=20*=20Broader=20Large=20Language=20Model=20?= =?UTF-8?q?(LLM)=20support=20and=20more=20model=20=20=20=20=20compression?= =?UTF-8?q?=20techniques.=20=20=20=20=20+=20OpenVINO=E2=84=A2=20runtime=20?= =?UTF-8?q?optimized=20for=20Intel=C2=AE=20Xe=20Matrix=20Extensions=20=20?= =?UTF-8?q?=20=20=20=20=20(Intel=C2=AE=20XMX)=20systolic=20arrays=20on=20b?= =?UTF-8?q?uilt-in=20GPUs=20for=20efficient=20=20=20=20=20=20=20matrix=20m?= =?UTF-8?q?ultiplication=20resulting=20in=20significant=20LLM=20=20=20=20?= =?UTF-8?q?=20=20=20performance=20boost=20with=20improved=201st=20and=202n?= =?UTF-8?q?d=20token=20=20=20=20=20=20=20latency,=20as=20well=20as=20a=20s?= =?UTF-8?q?maller=20memory=20footprint=20on=20=20=20=20=20=20=20Intel?= =?UTF-8?q?=C2=AE=20Core=E2=84=A2=20Ultra=20Processors=20(Series=202).=20?= =?UTF-8?q?=20=20=20=20+=20Memory=20sharing=20enabled=20for=20NPUs=20on=20?= =?UTF-8?q?Intel=C2=AE=20Core=E2=84=A2=20Ultra=20=20=20=20=20=20=20Process?= =?UTF-8?q?ors=20(Series=202)=20for=20efficient=20pipeline=20integration?= =?UTF-8?q?=20=20=20=20=20=20=20without=20memory=20copy=20overhead.=20=20?= =?UTF-8?q?=20=20=20+=20Addition=20of=20the=20PagedAttention=20feature=20f?= =?UTF-8?q?or=20discrete=20GPUs*=20=20=20=20=20=20=20enables=20a=20signifi?= =?UTF-8?q?cant=20boost=20in=20throughput=20for=20parallel=20=20=20=20=20?= =?UTF-8?q?=20=20inferencing=20when=20serving=20LLMs=20on=20Intel=C2=AE=20?= =?UTF-8?q?Arc=E2=84=A2=20Graphics=20=20=20=20=20=20=20or=20Intel=C2=AE=20?= =?UTF-8?q?Data=20Center=20GPU=20Flex=20Series.=20=20=20*=20More=20portabi?= =?UTF-8?q?lity=20and=20performance=20to=20run=20AI=20at=20the=20edge,=20?= =?UTF-8?q?=20=20=20=20in=20the=20cloud,=20or=20locally.=20=20=20=20=20+?= =?UTF-8?q?=20OpenVINO=E2=84=A2=20Model=20Server=20now=20comes=20with=20pr?= =?UTF-8?q?oduction-quality=20=20=20=20=20=20=20support=20for=20OpenAI-com?= =?UTF-8?q?patible=20API=20which=20enables=20i=20=20=20=20=20=20=20signifi?= =?UTF-8?q?cantly=20higher=20throughput=20for=20parallel=20inferencing=20?= =?UTF-8?q?=20=20=20=20=20=20on=20Intel=C2=AE=20Xeon=C2=AE=20processors=20?= =?UTF-8?q?when=20serving=20LLMs=20to=20many=20=20=20=20=20=20=20concurren?= =?UTF-8?q?t=20users.=20=20=20=20=20+=20Improved=20performance=20and=20mem?= =?UTF-8?q?ory=20consumption=20with=20prefix=20=20=20=20=20=20=20caching,?= =?UTF-8?q?=20KV=20cache=20compression,=20and=20other=20optimizations=20?= =?UTF-8?q?=20=20=20=20=20=20for=20serving=20LLMs=20using=20OpenVINO?= =?UTF-8?q?=E2=84=A2=20Model=20Server.=20=20=20=20=20+=20Support=20for=20P?= =?UTF-8?q?ython=203.12.=20-=20Support=20Change=20and=20Deprecation=20Noti?= =?UTF-8?q?ces=20=20=20*=20Using=20deprecated=20features=20and=20component?= =?UTF-8?q?s=20is=20not=20advised.=20=20=20=20=20They=20are=20available=20?= =?UTF-8?q?to=20enable=20a=20smooth=20transition=20to=20new=20=20=20=20=20?= =?UTF-8?q?solutions=20and=20will=20be=20discontinued=20in=20the=20future.?= =?UTF-8?q?=20=20=20=20=20To=20keep=20using=20discontinued=20features,=20y?= =?UTF-8?q?ou=20will=20have=20to=20=20=20=20=20revert=20to=20the=20last=20?= =?UTF-8?q?LTS=20OpenVINO=20version=20supporting=20them.=20=20=20=20=20For?= =?UTF-8?q?=20more=20details,=20refer=20to=20the=20OpenVINO=20Legacy=20Fea?= =?UTF-8?q?tures=20=20=20=20=20and=20Components=20page.=20=20=20*=20Discon?= =?UTF-8?q?tinued=20in=202024.0:=20=20=20=20=20+=20Runtime=20components:?= =?UTF-8?q?=20=20=20=20=20=20=20-=20Intel=C2=AE=20Gaussian=20&=20Neural=20?= =?UTF-8?q?Accelerator=20(Intel=C2=AE=E2=80=AFGNA).=20=20=20=20=20=20=20?= =?UTF-8?q?=20=20Consider=20using=20the=20Neural=20Processing=20Unit=20(NP?= =?UTF-8?q?U)=20for=20=20=20=20=20=20=20=20=20low-powered=20systems=20like?= =?UTF-8?q?=20Intel=C2=AE=20Core=E2=84=A2=20Ultra=20or=20=20=20=20=20=20?= =?UTF-8?q?=20=20=2014th=20generation=20and=20beyond.=20=20=20=20=20=20=20?= =?UTF-8?q?-=20OpenVINO=20C++/C/Python=201.0=20APIs=20(see=202023.3=20API?= =?UTF-8?q?=20=20=20=20=20=20=20=20=20transition=20guide=20for=20reference?= =?UTF-8?q?).=20=20=20=20=20=20=20-=20All=20ONNX=20Frontend=20legacy=20API?= =?UTF-8?q?=20(known=20as=20=20=20=20=20=20=20=20=20ONNX=5FIMPORTER=5FAPI)?= =?UTF-8?q?=20=20=20=20=20=20=20-'PerfomanceMode.UNDEFINED'=20property=20a?= =?UTF-8?q?s=20part=20of=20the=20=20=20=20=20=20=20=20=20OpenVINO=20Python?= =?UTF-8?q?=20API=20=20=20=20=20+=20Tools:=20=20=20=20=20=20=20=20-=20Depl?= =?UTF-8?q?oyment=20Manager.=20See=20installation=20and=20deployment=20=20?= =?UTF-8?q?=20=20=20=20=20=20=20=20guides=20for=20current=20distribution?= =?UTF-8?q?=20options.=20=20=20=20=20=20=20=20-=20Accuracy=20Checker.=20?= =?UTF-8?q?=20=20=20=20=20=20=20-=20Post-Training=20Optimization=20Tool=20?= =?UTF-8?q?(POT).=E2=80=AFNeural=20Network=20=20=20=20=20=20=20=20=20=20Co?= =?UTF-8?q?mpression=20Framework=20(NNCF)=20should=20be=20used=20instead.?= =?UTF-8?q?=20=20=20=20=20=20=20=20-=20A=20Git=20patch=E2=80=AFfor=20NNCF?= =?UTF-8?q?=20integration=20with=E2=80=AFhuggingface/=20=20=20=20=20=20=20?= =?UTF-8?q?=20=20=20transformers.=20The=20recommended=20approach=E2=80=AFi?= =?UTF-8?q?s=20to=20use=20=20=20=20=20=20=20=20=20=20huggingface/optimum-i?= =?UTF-8?q?ntel=E2=80=AFfor=20applying=20NNCF=20=20=20=20=20=20=20=20=20?= =?UTF-8?q?=20optimization=20on=20top=20of=20models=20from=20Hugging=20Fac?= =?UTF-8?q?e.=20=20=20=20=20=20=20=20-=20Support=20for=20Apache=20MXNet,?= =?UTF-8?q?=20Caffe,=20and=20Kaldi=20model=20=20=20=20=20=20=20=20=20=20fo?= =?UTF-8?q?rmats.=20Conversion=20to=20ONNX=20may=20be=20used=20as=20a=20?= =?UTF-8?q?=20=20=20=20=20=20=20=20=20solution.=20=20=20*=20Deprecated=20a?= =?UTF-8?q?nd=20to=20be=20removed=20in=20the=20future:=20=20=20=20=20+=20T?= =?UTF-8?q?he=20macOS=20x86=5F64=20debug=20bins=20will=20no=20longer=20be?= =?UTF-8?q?=20=20=20=20=20=20=20provided=20with=20the=20OpenVINO=20toolkit?= =?UTF-8?q?,=20starting=20with=20=20=20=20=20=20=20OpenVINO=202024.5.=20?= =?UTF-8?q?=20=20=20=20+=20Python=203.8=20is=20now=20considered=20deprecat?= =?UTF-8?q?ed,=20and=20it=20will=20not=20=20=20=20=20=20=20be=20available?= =?UTF-8?q?=20beyond=20the=202024.4=20OpenVINO=20version.=20=20=20=20=20+?= =?UTF-8?q?=20dKMB=20support=20is=20now=20considered=20deprecated=20and=20?= =?UTF-8?q?will=20be=20=20=20=20=20=20=20fully=20removed=20with=20OpenVINO?= =?UTF-8?q?=202024.5=20=20=20=20=20+=20Intel=C2=AE=20Streaming=20SIMD=20Ex?= =?UTF-8?q?tensions=20(Intel=C2=AE=20SSE)=20will=20be=20=20=20=20=20=20=20?= =?UTF-8?q?supported=20in=20source=20code=20form,=20but=20not=20enabled=20?= =?UTF-8?q?in=20the=20=20=20=20=20=20=20binary=20package=20by=20default,?= =?UTF-8?q?=20starting=20with=20OpenVINO=202025.0=20=20=20=20=20+=20The=20?= =?UTF-8?q?openvino-nightly=20PyPI=20module=20will=20soon=20be=20discontin?= =?UTF-8?q?ued.=20=20=20=20=20=20=20End-users=20should=20proceed=20with=20?= =?UTF-8?q?the=20Simple=20PyPI=20nightly=20repo=20=20=20=20=20=20=20instea?= =?UTF-8?q?d.=20More=20information=20in=20Release=20Policy.=20=20=20=20=20?= =?UTF-8?q?+=20The=20OpenVINO=E2=84=A2=20Development=20Tools=20package=20(?= =?UTF-8?q?pip=20install=20=20=20=20=20=20=20openvino-dev)=20will=20be=20r?= =?UTF-8?q?emoved=20from=20installation=20options=20and=20=20=20=20=20=20?= =?UTF-8?q?=20distribution=20channels=20beginning=20with=20OpenVINO=202025?= =?UTF-8?q?.0.=20=20=20=20=20+=20Model=20Optimizer=20will=20be=20discontin?= =?UTF-8?q?ued=20with=20OpenVINO=202025.0.=20=20=20=20=20=20=20Consider=20?= =?UTF-8?q?using=20the=20new=20conversion=20methods=20instead.=20For=20mor?= =?UTF-8?q?e=20=20=20=20=20=20=20details,=20see=20the=20model=20conversion?= =?UTF-8?q?=20transition=20guide.=20=20=20=20=20+=20OpenVINO=20property=20?= =?UTF-8?q?Affinity=20API=20will=20be=20discontinued=20with=20=20=20=20=20?= =?UTF-8?q?=20=20OpenVINO=202025.0.=20It=20will=20be=20replaced=20with=20C?= =?UTF-8?q?PU=20binding=20=20=20=20=20=20=20configurations=20(ov::hint::en?= =?UTF-8?q?able=5Fcpu=5Fpinning).=20=20=20=20=20+=20OpenVINO=20Model=20Ser?= =?UTF-8?q?ver=20components:=20=20=20=20=20=20=20-=20=E2=80=9Cauto=20shape?= =?UTF-8?q?=E2=80=9D=20and=20=E2=80=9Cauto=20batch=20size=E2=80=9D=20(resh?= =?UTF-8?q?aping=20a=20model=20in=20=20=20=20=20=20=20runtime)=20will=20be?= =?UTF-8?q?=20removed=20in=20the=20future.=20OpenVINO=E2=80=99s=20dynamic?= =?UTF-8?q?=20=20=20=20=20=20=20shape=20models=20are=20recommended=20inste?= =?UTF-8?q?ad.=20=20=20=20=20+=20A=20number=20of=20notebooks=20have=20been?= =?UTF-8?q?=20deprecated.=20For=20an=20=20=20=20=20=20=20up-to-date=20list?= =?UTF-8?q?ing=20of=20available=20notebooks,=20refer=20to=20the=20=20=20?= =?UTF-8?q?=20=20=20=20OpenVINO=E2=84=A2=20Notebook=20index=20(openvinotoo?= =?UTF-8?q?lkit.github.io).=20-=20Add=20Leap15=20build=20-=20Remove=20comm?= =?UTF-8?q?ent=20lines=20in=20the=20spec=20file=20that=20cause=20the=20ins?= =?UTF-8?q?ertion=20=20=20of=20extra=20lines=20during=20a=20commit=20-=20R?= =?UTF-8?q?emove=20NPU=20Compile=20Tool=20*=20openvino-remove-npu-compile-?= =?UTF-8?q?tool.patch=20-=20Update=20to=202024.3.0=20-=20Summary=20of=20ma?= =?UTF-8?q?jor=20features=20and=20improvements=E2=80=AF=E2=80=AF=20=20=20*?= =?UTF-8?q?=20More=20Gen=20AI=20coverage=20and=20framework=20integrations?= =?UTF-8?q?=20to=20minimize=20=20=20=20=20code=20changes=20=20=20=20=20+?= =?UTF-8?q?=20OpenVINO=20pre-optimized=20models=20are=20now=20available=20?= =?UTF-8?q?in=20Hugging=20=20=20=20=20=20=20Face=20making=20it=20easier=20?= =?UTF-8?q?for=20developers=20to=20get=20started=20with=20=20=20=20=20=20?= =?UTF-8?q?=20these=20models.=20=20=20*=20Broader=20Large=20Language=20Mod?= =?UTF-8?q?el=20(LLM)=20support=20and=20more=20model=20=20=20=20=20compres?= =?UTF-8?q?sion=20techniques.=20=20=20=20=20+=20Significant=20improvement?= =?UTF-8?q?=20in=20LLM=20performance=20on=20Intel=20=20=20=20=20=20=20disc?= =?UTF-8?q?rete=20GPUs=20with=20the=20addition=20of=20Multi-Head=20Attenti?= =?UTF-8?q?on=20=20=20=20=20=20=20(MHA)=20and=20OneDNN=20enhancements.=20?= =?UTF-8?q?=20=20*=20More=20portability=20and=20performance=20to=20run=20A?= =?UTF-8?q?I=20at=20the=20edge,=20in=20the=20=20=20=20=20cloud,=20or=20loc?= =?UTF-8?q?ally.=20=20=20=20=20+=20Improved=20CPU=20performance=20when=20s?= =?UTF-8?q?erving=20LLMs=20with=20the=20=20=20=20=20=20=20inclusion=20of?= =?UTF-8?q?=20vLLM=20and=20continuous=20batching=20in=20the=20OpenVINO=20?= =?UTF-8?q?=20=20=20=20=20=20Model=20Server=20(OVMS).=20vLLM=20is=20an=20e?= =?UTF-8?q?asy-to-use=20open-source=20=20=20=20=20=20=20library=20that=20s?= =?UTF-8?q?upports=20efficient=20LLM=20inferencing=20and=20model=20=20=20?= =?UTF-8?q?=20=20=20=20serving.=20-=20Support=20Change=20and=20Deprecation?= =?UTF-8?q?=20Notices=20=20=20*=20Using=20deprecated=20features=20and=20co?= =?UTF-8?q?mponents=20is=20not=20advised.=20=20=20=20=20They=20are=20avail?= =?UTF-8?q?able=20to=20enable=20a=20smooth=20transition=20to=20new=20=20?= =?UTF-8?q?=20=20=20solutions=20and=20will=20be=20discontinued=20in=20the?= =?UTF-8?q?=20future.=20To=20keep=20=20=20=20=20using=20discontinued=20fea?= =?UTF-8?q?tures,=20you=20will=20have=20to=20revert=20to=20the=20=20=20=20?= =?UTF-8?q?=20last=20LTS=20OpenVINO=20version=20supporting=20them.=20For?= =?UTF-8?q?=20more=20details,=20=20=20=20=20refer=20to=20the=20OpenVINO=20?= =?UTF-8?q?Legacy=20Features=20and=20Components=20page.=20=20=20*=20Discon?= =?UTF-8?q?tinued=20in=202024.0:=20=20=20=20=20+=20Runtime=20components:?= =?UTF-8?q?=20=20=20=20=20=20=20-=20Intel=C2=AE=20Gaussian=20&=20Neural=20?= =?UTF-8?q?Accelerator=20(Intel=C2=AE=E2=80=AFGNA)..Consider=20=20=20=20?= =?UTF-8?q?=20=20=20=20=20using=20the=20Neural=20Processing=20Unit=20(NPU)?= =?UTF-8?q?=20for=20low-powered=20=20=20=20=20=20=20=20=20systems=20like?= =?UTF-8?q?=20Intel=C2=AE=20Core=E2=84=A2=20Ultra=20or=2014th=20generation?= =?UTF-8?q?=20=20=20=20=20=20=20=20=20and=20beyond.=20=20=20=20=20=20=20-?= =?UTF-8?q?=20OpenVINO=20C++/C/Python=201.0=20APIs=20(see=202023.3=20API?= =?UTF-8?q?=20transition=20=20=20=20=20=20=20=20=20guide=20for=20reference?= =?UTF-8?q?).=20=20=20=20=20=20=20-=20All=20ONNX=20Frontend=20legacy=20API?= =?UTF-8?q?=20(known=20as=20ONNX=5FIMPORTER=5FAPI)=20=20=20=20=20=20=20-?= =?UTF-8?q?=20'PerfomanceMode.UNDEFINED'=20property=20as=20part=20of=20the?= =?UTF-8?q?=20OpenVINO=20=20=20=20=20=20=20=20=20=20Python=20API=20=20=20?= =?UTF-8?q?=20=20+=20Tools:=20=20=20=20=20=20=20-=20Deployment=20Manager.?= =?UTF-8?q?=20See=20installation=20and=20deployment=20guides=20=20=20=20?= =?UTF-8?q?=20=20=20=20=20for=20current=20distribution=20options.=20=20=20?= =?UTF-8?q?=20=20=20=20-=20Accuracy=20Checker.=20=20=20=20=20=20=20-=20Pos?= =?UTF-8?q?t-Training=20Optimization=20Tool=20(POT).=E2=80=AFNeural=20Netw?= =?UTF-8?q?ork=20=20=20=20=20=20=20=20=20Compression=20Framework=20(NNCF)?= =?UTF-8?q?=20should=20be=20used=20instead.=20=20=20=20=20=20=20-=20A=20Gi?= =?UTF-8?q?t=20patch=E2=80=AFfor=20NNCF=20integration=20with=E2=80=AFhuggi?= =?UTF-8?q?ngface/=20=20=20=20=20=20=20=20=20transformers.=20The=20recomme?= =?UTF-8?q?nded=20approach=E2=80=AFis=20to=20use=20=20=20=20=20=20=20=20?= =?UTF-8?q?=20huggingface/optimum-intel=E2=80=AFfor=20applying=20NNCF=20op?= =?UTF-8?q?timization=20=20=20=20=20=20=20=20=20on=20top=20of=20models=20f?= =?UTF-8?q?rom=20Hugging=20Face.=20=20=20=20=20=20=20-=20Support=20for=20A?= =?UTF-8?q?pache=20MXNet,=20Caffe,=20and=20Kaldi=20model=20formats.=20=20?= =?UTF-8?q?=20=20=20=20=20=20=20Conversion=20to=20ONNX=20may=20be=20used?= =?UTF-8?q?=20as=20a=20solution.=20=20=20*=20Deprecated=20and=20to=20be=20?= =?UTF-8?q?removed=20in=20the=20future:=20=20=20=20=20+=20The=20OpenVINO?= =?UTF-8?q?=E2=84=A2=20Development=20Tools=20package=20(pip=20install=20?= =?UTF-8?q?=20=20=20=20=20=20openvino-dev)=20will=20be=20removed=20from=20?= =?UTF-8?q?installation=20options=20=20=20=20=20=20=20and=20distribution?= =?UTF-8?q?=20channels=20beginning=20with=20OpenVINO=202025.0.=20=20=20=20?= =?UTF-8?q?=20+=20Model=20Optimizer=20will=20be=20discontinued=20with=20Op?= =?UTF-8?q?enVINO=202025.0.=20=20=20=20=20=20=20Consider=20using=20the=20n?= =?UTF-8?q?ew=20conversion=20methods=20instead.=20For=20=20=20=20=20=20=20?= =?UTF-8?q?more=20details,=20see=20the=20model=20conversion=20transition?= =?UTF-8?q?=20guide.=20=20=20=20=20+=20OpenVINO=20property=20Affinity=20AP?= =?UTF-8?q?I=20will=20be=20discontinued=20with=20=20=20=20=20=20=20OpenVIN?= =?UTF-8?q?O=202025.0.=20It=20will=20be=20replaced=20with=20CPU=20binding?= =?UTF-8?q?=20=20=20=20=20=20=20configurations=20(ov::hint::enable=5Fcpu?= =?UTF-8?q?=5Fpinning).=20=20=20=20=20+=20OpenVINO=20Model=20Server=20comp?= =?UTF-8?q?onents:=20=20=20=20=20=20=20-=20=E2=80=9Cauto=20shape=E2=80=9D?= =?UTF-8?q?=20and=20=E2=80=9Cauto=20batch=20size=E2=80=9D=20(reshaping=20a?= =?UTF-8?q?=20model=20=20=20=20=20=20=20=20=20in=20runtime)=20will=20be=20?= =?UTF-8?q?removed=20in=20the=20future.=20OpenVINO=E2=80=99s=20=20=20=20?= =?UTF-8?q?=20=20=20=20=20dynamic=20shape=20models=20are=20recommended=20i?= =?UTF-8?q?nstead.=20=20=20=20=20+=20A=20number=20of=20notebooks=20have=20?= =?UTF-8?q?been=20deprecated.=20For=20an=20=20=20=20=20=20=20up-to-date=20?= =?UTF-8?q?listing=20of=20available=20notebooks,=20refer=20to=20=20=20=20?= =?UTF-8?q?=20=20=20the=20OpenVINO=E2=84=A2=20Notebook=20index=20(openvino?= =?UTF-8?q?toolkit.github.io).=20-=20Add=20riscv-cpu-plugin=20subpackage?= =?UTF-8?q?=20-=20Update=20to=202024.2.0=20-=20More=20Gen=20AI=20coverage?= =?UTF-8?q?=20and=20framework=20integrations=20to=20minimize=20code=20=20?= =?UTF-8?q?=20changes=20=20=20*=20Llama=203=20optimizations=20for=20CPUs,?= =?UTF-8?q?=20built-in=20GPUs,=20and=20discrete=20=20=20=20=20GPUs=20for?= =?UTF-8?q?=20improved=20performance=20and=20efficient=20memory=20usage.?= =?UTF-8?q?=20=20=20*=20Support=20for=20Phi-3-mini,=20a=20family=20of=20AI?= =?UTF-8?q?=20models=20that=20leverages=20=20=20=20=20the=20power=20of=20s?= =?UTF-8?q?mall=20language=20models=20for=20faster,=20more=20accurate=20?= =?UTF-8?q?=20=20=20=20and=20cost-effective=20text=20processing.=20=20=20*?= =?UTF-8?q?=20Python=20Custom=20Operation=20is=20now=20enabled=20in=20Open?= =?UTF-8?q?VINO=20making=20it=20=20=20=20=20easier=20for=20Python=20develo?= =?UTF-8?q?pers=20to=20code=20their=20custom=20operations=20=20=20=20=20in?= =?UTF-8?q?stead=20of=20using=20C++=20custom=20operations=20(also=20suppor?= =?UTF-8?q?ted).=20=20=20=20=20Python=20Custom=20Operation=20empowers=20us?= =?UTF-8?q?ers=20to=20implement=20their=20own=20=20=20=20=20specialized=20?= =?UTF-8?q?operations=20into=20any=20model.=20=20=20*=20Notebooks=20expans?= =?UTF-8?q?ion=20to=20ensure=20better=20coverage=20for=20new=20models.=20?= =?UTF-8?q?=20=20=20=20Noteworthy=20notebooks=20added:=20DynamiCrafter,=20?= =?UTF-8?q?YOLOv10,=20Chatbot=20=20=20=20=20notebook=20with=20Phi-3,=20and?= =?UTF-8?q?=20QWEN2.=20-=20Broader=20Large=20Language=20Model=20(LLM)=20su?= =?UTF-8?q?pport=20and=20more=20model=20=20=20compression=20techniques.=20?= =?UTF-8?q?=20=20*=20GPTQ=20method=20for=204-bit=20weight=20compression=20?= =?UTF-8?q?added=20to=20NNCF=20for=20=20=20=20=20more=20efficient=20infere?= =?UTF-8?q?nce=20and=20improved=20performance=20of=20=20=20=20=20compresse?= =?UTF-8?q?d=20LLMs.=20=20=20*=20Significant=20LLM=20performance=20improve?= =?UTF-8?q?ments=20and=20reduced=20latency=20=20=20=20=20for=20both=20buil?= =?UTF-8?q?t-in=20GPUs=20and=20discrete=20GPUs.=20=20=20*=20Significant=20?= =?UTF-8?q?improvement=20in=202nd=20token=20latency=20and=20memory=20=20?= =?UTF-8?q?=20=20=20footprint=20of=20FP16=20weight=20LLMs=20on=20AVX2=20(1?= =?UTF-8?q?3th=20Gen=20Intel=C2=AE=20Core=E2=84=A2=20=20=20=20=20processor?= =?UTF-8?q?s)=20and=20AVX512=20(3rd=20Gen=20Intel=C2=AE=20Xeon=C2=AE=20Sca?= =?UTF-8?q?lable=20=20=20=20=20Processors)=20based=20CPU=20platforms,=20pa?= =?UTF-8?q?rticularly=20for=20small=20=20=20=20=20batch=20sizes.=20-=20Mor?= =?UTF-8?q?e=20portability=20and=20performance=20to=20run=20AI=20at=20the?= =?UTF-8?q?=20edge,=20in=20the=20=20=20cloud,=20or=20locally.=20=20=20*=20?= =?UTF-8?q?Model=20Serving=20Enhancements:=20=20=20*=20Preview:=20OpenVINO?= =?UTF-8?q?=20Model=20Server=20(OVMS)=20now=20supports=20=20=20=20=20OpenA?= =?UTF-8?q?I-compatible=20API=20along=20with=20Continuous=20Batching=20and?= =?UTF-8?q?=20=20=20=20=20PagedAttention,=20enabling=20significantly=20hig?= =?UTF-8?q?her=20throughput=20=20=20=20=20for=20parallel=20inferencing,=20?= =?UTF-8?q?especially=20on=20Intel=C2=AE=20Xeon=C2=AE=20=20=20=20=20proces?= =?UTF-8?q?sors,=20when=20serving=20LLMs=20to=20many=20concurrent=20users.?= =?UTF-8?q?=20=20=20*=20OpenVINO=20backend=20for=20Triton=20Server=20now?= =?UTF-8?q?=20supports=20built-in=20=20=20=20=20GPUs=20and=20discrete=20GP?= =?UTF-8?q?Us,=20in=20addition=20to=20dynamic=20=20=20=20=20shapes=20suppo?= =?UTF-8?q?rt.=20=20=20*=20Integration=20of=20TorchServe=20through=20torch?= =?UTF-8?q?.compile=20OpenVINO=20=20=20=20=20backend=20for=20easy=20model?= =?UTF-8?q?=20deployment,=20provisioning=20to=20=20=20=20=20multiple=20ins?= =?UTF-8?q?tances,=20model=20versioning,=20and=20maintenance.=20=20=20*=20?= =?UTF-8?q?Preview:=20addition=20of=20the=20Generate=20API,=20a=20simplifi?= =?UTF-8?q?ed=20API=20=20=20=20=20for=20text=20generation=20using=20large?= =?UTF-8?q?=20language=20models=20with=20only=20=20=20=20=20a=20few=20line?= =?UTF-8?q?s=20of=20code.=20The=20API=20is=20available=20through=20the=20n?= =?UTF-8?q?ewly=20=20=20=20=20launched=20OpenVINO=20GenAI=20package.=20=20?= =?UTF-8?q?=20*=20Support=20for=20Intel=20Atom=C2=AE=20Processor=20X=20Ser?= =?UTF-8?q?ies.=20For=20more=20details,=20=20=20=20=20see=20System=20Requi?= =?UTF-8?q?rements.=20=20=20*=20Preview:=20Support=20for=20Intel=C2=AE=20X?= =?UTF-8?q?eon=C2=AE=206=20processor.=20-=20Support=20Change=20and=20Depre?= =?UTF-8?q?cation=20Notices=20=20=20*=20Using=20deprecated=20features=20an?= =?UTF-8?q?d=20components=20is=20not=20advised.=20=20=20=20=20They=20are?= =?UTF-8?q?=20available=20to=20enable=20a=20smooth=20transition=20to=20new?= =?UTF-8?q?=20=20=20=20=20solutions=20and=20will=20be=20discontinued=20in?= =?UTF-8?q?=20the=20future.=20=20=20=20=20To=20keep=20using=20discontinued?= =?UTF-8?q?=20features,=20you=20will=20have=20to=20revert=20=20=20=20=20to?= =?UTF-8?q?=20the=20last=20LTS=20OpenVINO=20version=20supporting=20them.?= =?UTF-8?q?=20For=20more=20=20=20=20=20details,=20refer=20to=20the=20OpenV?= =?UTF-8?q?INO=20Legacy=20Features=20and=20=20=20=20=20Components=20page.?= =?UTF-8?q?=20=20=20*=20Discontinued=20in=202024.0:=20=20=20=20=20+=20Runt?= =?UTF-8?q?ime=20components:=20=20=20=20=20=20=20-=20Intel=C2=AE=20Gaussia?= =?UTF-8?q?n=20&=20Neural=20Accelerator=20(Intel=C2=AE=E2=80=AFGNA).=20=20?= =?UTF-8?q?=20=20=20=20=20=20=20Consider=20using=20the=20Neural=20Processi?= =?UTF-8?q?ng=20Unit=20(NPU)=20for=20=20=20=20=20=20=20=20=20low-powered?= =?UTF-8?q?=20systems=20like=20Intel=C2=AE=20Core=E2=84=A2=20Ultra=20or=20?= =?UTF-8?q?14th=20=20=20=20=20=20=20=20=20generation=20and=20beyond.=20=20?= =?UTF-8?q?=20=20=20=20=20-=20OpenVINO=20C++/C/Python=201.0=20APIs=20(see?= =?UTF-8?q?=202023.3=20API=20=20=20=20=20=20=20=20=20transition=20guide=20?= =?UTF-8?q?for=20reference).=20=20=20=20=20=20=20-=20All=20ONNX=20Frontend?= =?UTF-8?q?=20legacy=20API=20(known=20as=20ONNX=5FIMPORTER=5FAPI)=20=20=20?= =?UTF-8?q?=20=20=20=20-=20'PerfomanceMode.UNDEFINED'=20property=20as=20pa?= =?UTF-8?q?rt=20of=20the=20=20=20=20=20=20=20=20=20OpenVINO=20Python=20API?= =?UTF-8?q?=20=20=20=20=20+=20Tools:=20=20=20=20=20=20=20-=20Deployment=20?= =?UTF-8?q?Manager.=20See=20installation=20and=20deployment=20=20=20=20=20?= =?UTF-8?q?=20=20=20=20guides=20for=20current=20distribution=20options.=20?= =?UTF-8?q?=20=20=20=20=20=20-=20Accuracy=20Checker.=20=20=20=20=20=20=20-?= =?UTF-8?q?=20Post-Training=20Optimization=20Tool=20(POT).=E2=80=AFNeural?= =?UTF-8?q?=20Network=20=20=20=20=20=20=20=20=20Compression=20Framework=20?= =?UTF-8?q?(NNCF)=20should=20be=20used=20instead.=20=20=20=20=20=20=20-=20?= =?UTF-8?q?A=20Git=20patch=E2=80=AFfor=20NNCF=20integration=20with?= =?UTF-8?q?=E2=80=AF=20=20=20=20=20=20=20=20=20huggingface/transformers.?= =?UTF-8?q?=20The=20recommended=20approach=E2=80=AF=20=20=20=20=20=20=20?= =?UTF-8?q?=20=20is=20to=20use=E2=80=AFhuggingface/optimum-intel=E2=80=AFf?= =?UTF-8?q?or=20applying=20NNCF=20=20=20=20=20=20=20=20=20optimization=20o?= =?UTF-8?q?n=20top=20of=20models=20from=20Hugging=20Face.=20=20=20=20=20?= =?UTF-8?q?=20=20-=20Support=20for=20Apache=20MXNet,=20Caffe,=20and=20Kald?= =?UTF-8?q?i=20model=20formats.=20=20=20=20=20=20=20=20=20Conversion=20to?= =?UTF-8?q?=20ONNX=20may=20be=20used=20as=20a=20solution.=20=20=20*=20Depr?= =?UTF-8?q?ecated=20and=20to=20be=20removed=20in=20the=20future:=20=20=20?= =?UTF-8?q?=20=20+=20The=20OpenVINO=E2=84=A2=20Development=20Tools=20packa?= =?UTF-8?q?ge=20(pip=20install=20=20=20=20=20=20=20openvino-dev)=20will=20?= =?UTF-8?q?be=20removed=20from=20installation=20options=20=20=20=20=20=20?= =?UTF-8?q?=20and=20distribution=20channels=20beginning=20with=20OpenVINO?= =?UTF-8?q?=202025.0.=20=20=20=20=20+=20Model=20Optimizer=20will=20be=20di?= =?UTF-8?q?scontinued=20with=20OpenVINO=202025.0.=20=20=20=20=20=20=20Cons?= =?UTF-8?q?ider=20using=20the=20new=20conversion=20methods=20instead.=20Fo?= =?UTF-8?q?r=20=20=20=20=20=20=20more=20details,=20see=20the=20model=20con?= =?UTF-8?q?version=20transition=20guide.=20=20=20=20=20+=20OpenVINO=20prop?= =?UTF-8?q?erty=20Affinity=20API=20will=20be=20discontinued=20with=20=20?= =?UTF-8?q?=20=20=20=20=20OpenVINO=202025.0.=20It=20will=20be=20replaced?= =?UTF-8?q?=20with=20CPU=20binding=20=20=20=20=20=20=20configurations=20(o?= =?UTF-8?q?v::hint::enable=5Fcpu=5Fpinning).=20=20=20=20=20+=20OpenVINO=20?= =?UTF-8?q?Model=20Server=20components:=20=20=20=20=20+=20=E2=80=9Cauto=20?= =?UTF-8?q?shape=E2=80=9D=20and=20=E2=80=9Cauto=20batch=20size=E2=80=9D=20?= =?UTF-8?q?(reshaping=20a=20model=20in=20=20=20=20=20=20=20runtime)=20will?= =?UTF-8?q?=20be=20removed=20in=20the=20future.=20OpenVINO=E2=80=99s=20dyn?= =?UTF-8?q?amic=20=20=20=20=20=20=20shape=20models=20are=20recommended=20i?= =?UTF-8?q?nstead.=20=20=20=20=20+=20A=20number=20of=20notebooks=20have=20?= =?UTF-8?q?been=20deprecated.=20For=20an=20=20=20=20=20=20=20up-to-date=20?= =?UTF-8?q?listing=20of=20available=20notebooks,=20refer=20to=20the=20=20?= =?UTF-8?q?=20=20=20=20=20OpenVINO=E2=84=A2=20Notebook=20index=20(openvino?= =?UTF-8?q?toolkit.github.io).=20-=20Fix=20sample=20source=20path=20in=20b?= =?UTF-8?q?uild=20script:=20=20=20*=20openvino-fix-build-sample-path.patch?= =?UTF-8?q?=20-=20Update=20to=202024.1.0=20-=20More=20Generative=20AI=20co?= =?UTF-8?q?verage=20and=20framework=20integrations=20to=20=20=20minimize?= =?UTF-8?q?=20code=20changes.=20=20=20*=20Mixtral=20and=20URLNet=20models?= =?UTF-8?q?=20optimized=20for=20performance=20=20=20=20=20improvements=20o?= =?UTF-8?q?n=20Intel=C2=AE=20Xeon=C2=AE=20processors.=20=20=20*=20Stable?= =?UTF-8?q?=20Diffusion=201.5,=20ChatGLM3-6B,=20and=20Qwen-7B=20models=20?= =?UTF-8?q?=20=20=20=20optimized=20for=20improved=20inference=20speed=20on?= =?UTF-8?q?=20Intel=C2=AE=20Core=E2=84=A2=20=20=20=20=20Ultra=20processors?= =?UTF-8?q?=20with=20integrated=20GPU.=20=20=20*=20Support=20for=20Falcon-?= =?UTF-8?q?7B-Instruct,=20a=20GenAI=20Large=20Language=20Model=20=20=20=20?= =?UTF-8?q?=20(LLM)=20ready-to-use=20chat/instruct=20model=20with=20superi?= =?UTF-8?q?or=20=20=20=20=20performance=20metrics.=20=20=20*=20New=20Jupyt?= =?UTF-8?q?er=20Notebooks=20added:=20YOLO=20V9,=20YOLO=20V8=20=20=20=20=20?= =?UTF-8?q?Oriented=20Bounding=20Boxes=20Detection=20(OOB),=20Stable=20Dif?= =?UTF-8?q?fusion=20=20=20=20=20in=20Keras,=20MobileCLIP,=20RMBG-v1.4=20Ba?= =?UTF-8?q?ckground=20Removal,=20Magika,=20=20=20=20=20TripoSR,=20AnimateA?= =?UTF-8?q?nyone,=20LLaVA-Next,=20and=20RAG=20system=20with=20=20=20=20=20?= =?UTF-8?q?OpenVINO=20and=20LangChain.=20-=20Broader=20Large=20Language=20?= =?UTF-8?q?Model=20(LLM)=20support=20and=20more=20model=20=20=20compressio?= =?UTF-8?q?n=20techniques.=20=20=20*=20LLM=20compilation=20time=20reduced?= =?UTF-8?q?=20through=20additional=20optimizations=20=20=20=20=20with=20co?= =?UTF-8?q?mpressed=20embedding.=20Improved=201st=20token=20performance=20?= =?UTF-8?q?of=20=20=20=20=20LLMs=20on=204th=20and=205th=20generations=20of?= =?UTF-8?q?=20Intel=C2=AE=20Xeon=C2=AE=20processors=20=20=20=20=20with=20I?= =?UTF-8?q?ntel=C2=AE=20Advanced=20Matrix=20Extensions=20(Intel=C2=AE=20AM?= =?UTF-8?q?X).=20=20=20*=20Better=20LLM=20compression=20and=20improved=20p?= =?UTF-8?q?erformance=20with=20oneDNN,=20=20=20=20=20INT4,=20and=20INT8=20?= =?UTF-8?q?support=20for=20Intel=C2=AE=20Arc=E2=84=A2=20GPUs.=20=20=20*=20?= =?UTF-8?q?Significant=20memory=20reduction=20for=20select=20smaller=20Gen?= =?UTF-8?q?AI=20=20=20=20=20models=20on=20Intel=C2=AE=20Core=E2=84=A2=20Ul?= =?UTF-8?q?tra=20processors=20with=20integrated=20GPU.=20-=20More=20portab?= =?UTF-8?q?ility=20and=20performance=20to=20run=20AI=20at=20the=20edge,=20?= =?UTF-8?q?=20=20in=20the=20cloud,=20or=20locally.=20=20=20*=20The=20previ?= =?UTF-8?q?ew=20NPU=20plugin=20for=20Intel=C2=AE=20Core=E2=84=A2=20Ultra?= =?UTF-8?q?=20processors=20=20=20=20=20is=20now=20available=20in=20the=20O?= =?UTF-8?q?penVINO=20open-source=20GitHub=20=20=20=20=20repository,=20in?= =?UTF-8?q?=20addition=20to=20the=20main=20OpenVINO=20package=20on=20PyPI.?= =?UTF-8?q?=20=20=20*=20The=20JavaScript=20API=20is=20now=20more=20easily?= =?UTF-8?q?=20accessible=20through=20=20=20=20=20the=20npm=20repository,?= =?UTF-8?q?=20enabling=20JavaScript=20developers=E2=80=99=20seamless=20=20?= =?UTF-8?q?=20=20=20access=20to=20the=20OpenVINO=20API.=20=20=20*=20FP16?= =?UTF-8?q?=20inference=20on=20ARM=20processors=20now=20enabled=20for=20th?= =?UTF-8?q?e=20=20=20=20=20Convolutional=20Neural=20Network=20(CNN)=20by?= =?UTF-8?q?=20default.=20-=20Support=20Change=20and=20Deprecation=20Notice?= =?UTF-8?q?s=20=20=20*=20Using=20deprecated=20features=20and=20components?= =?UTF-8?q?=20is=20not=20advised.=20They=20=20=20=20=20are=20available=20t?= =?UTF-8?q?o=20enable=20a=20smooth=20transition=20to=20new=20solutions=20?= =?UTF-8?q?=20=20=20=20and=20will=20be=20discontinued=20in=20the=20future.?= =?UTF-8?q?=20To=20keep=20using=20=20=20=20=20Discontinued=20features,=20y?= =?UTF-8?q?ou=20will=20have=20to=20revert=20to=20the=20last=20=20=20=20=20?= =?UTF-8?q?LTS=20OpenVINO=20version=20supporting=20them.=20=20=20*=20For?= =?UTF-8?q?=20more=20details,=20refer=20to=20the=E2=80=AFOpenVINO=20Legacy?= =?UTF-8?q?=20Features=20=20=20=20=20and=20Components=20page.=20=20=20*=20?= =?UTF-8?q?Discontinued=20in=202024.0:=20=20=20=20=20+=20Runtime=20compone?= =?UTF-8?q?nts:=20=20=20=20=20=20=20-=20Intel=C2=AE=20Gaussian=20&=20Neura?= =?UTF-8?q?l=20Accelerator=20(Intel=C2=AE=E2=80=AFGNA).=20=20=20=20=20=20?= =?UTF-8?q?=20=20=20Consider=20using=20the=20Neural=20Processing=20Unit=20?= =?UTF-8?q?(NPU)=20=20=20=20=20=20=20=20=20for=20low-powered=20systems=20l?= =?UTF-8?q?ike=20Intel=C2=AE=20Core=E2=84=A2=20Ultra=20or=20=20=20=20=20?= =?UTF-8?q?=20=20=20=2014th=E2=80=AFgeneration=20and=20beyond.=20=20=20=20?= =?UTF-8?q?=20=20=20-=20OpenVINO=20C++/C/Python=201.0=20APIs=20(see=202023?= =?UTF-8?q?.3=20API=20=20=20=20=20=20=20=20=20transition=20guide=20for=20r?= =?UTF-8?q?eference).=20=20=20=20=20=20=20-=20All=20ONNX=20Frontend=20lega?= =?UTF-8?q?cy=20API=20(known=20as=20=20=20=20=20=20=20=20=20ONNX=5FIMPORTE?= =?UTF-8?q?R=5FAPI)=20=20=20=20=20=20=20-=20'PerfomanceMode.UNDEFINED'=20p?= =?UTF-8?q?roperty=20as=20part=20of=20=20=20=20=20=20=20=20=20=20the=20Ope?= =?UTF-8?q?nVINO=20Python=20API=20=20=20=20=20+=20Tools:=20=20=20=20=20=20?= =?UTF-8?q?=20-=20Deployment=20Manager.=20See=20installation=20and=20deplo?= =?UTF-8?q?yment=20=20=20=20=20=20=20=20=20guides=20for=20current=20distri?= =?UTF-8?q?bution=20options.=20=20=20=20=20=20=20-=20Accuracy=20Checker.?= =?UTF-8?q?=20=20=20=20=20=20=20-=20Post-Training=20Optimization=20Tool=20?= =?UTF-8?q?(POT).=E2=80=AFNeural=20Network=20=20=20=20=20=20=20=20=20Compr?= =?UTF-8?q?ession=20Framework=20(NNCF)=20should=20be=20used=20instead.=20?= =?UTF-8?q?=20=20=20=20=20=20-=20A=20Git=20patch=E2=80=AFfor=20NNCF=20inte?= =?UTF-8?q?gration=20with=E2=80=AF=20=20=20=20=20=20=20=20=20huggingface/t?= =?UTF-8?q?ransformers.=20The=20recommended=20approach=20=20=20=20=20=20?= =?UTF-8?q?=20=20=E2=80=AFis=20to=20use=E2=80=AFhuggingface/optimum-intel?= =?UTF-8?q?=E2=80=AFfor=20applying=20=20=20=20=20=20=20=20=20NNCF=20optimi?= =?UTF-8?q?zation=20on=20top=20of=20models=20from=20Hugging=20=20=20=20=20?= =?UTF-8?q?=20=20=20=20Face.=20=20=20=20=20=20=20-=20Support=20for=20Apach?= =?UTF-8?q?e=20MXNet,=20Caffe,=20and=20Kaldi=20model=20=20=20=20=20=20=20?= =?UTF-8?q?=20=20formats.=20Conversion=20to=20ONNX=20may=20be=20used=20as?= =?UTF-8?q?=20=20=20=20=20=20=20=20=20a=20solution.=20=20=20*=20Deprecated?= =?UTF-8?q?=20and=20to=20be=20removed=20in=20the=20future:=20=20=20=20=20+?= =?UTF-8?q?=20The=20OpenVINO=E2=84=A2=20Development=20Tools=20package=20(p?= =?UTF-8?q?ip=20install=20=20=20=20=20=20=20openvino-dev)=20will=20be=20re?= =?UTF-8?q?moved=20from=20installation=20options=20=20=20=20=20=20=20and?= =?UTF-8?q?=20distribution=20channels=20beginning=20with=20OpenVINO=202025?= =?UTF-8?q?.0.=20=20=20=20=20+=20Model=20Optimizer=20will=20be=20discontin?= =?UTF-8?q?ued=20with=20OpenVINO=202025.0.=20=20=20=20=20=20=20Consider=20?= =?UTF-8?q?using=20the=20new=20conversion=20methods=20instead.=20For=20=20?= =?UTF-8?q?=20=20=20=20=20more=20details,=20see=20the=20model=20conversion?= =?UTF-8?q?=20transition=20guide.=20=20=20=20=20+=20OpenVINO=20property=20?= =?UTF-8?q?Affinity=20API=20will=20be=20discontinued=20with=20=20=20=20=20?= =?UTF-8?q?=20=20OpenVINO=202025.0.=20It=20will=20be=20replaced=20with=20C?= =?UTF-8?q?PU=20binding=20=20=20=20=20=20=20configurations=20(ov::hint::en?= =?UTF-8?q?able=5Fcpu=5Fpinning).=20=20=20=20=20+=20OpenVINO=20Model=20Ser?= =?UTF-8?q?ver=20components:=20=20=20=20=20=20=20-=20=E2=80=9Cauto=20shape?= =?UTF-8?q?=E2=80=9D=20and=20=E2=80=9Cauto=20batch=20size=E2=80=9D=20(resh?= =?UTF-8?q?aping=20a=20model=20=20=20=20=20=20=20=20=20in=20runtime)=20wil?= =?UTF-8?q?l=20be=20removed=20in=20the=20future.=20OpenVINO=E2=80=99s=20?= =?UTF-8?q?=20=20=20=20=20=20=20=20dynamic=20shape=20models=20are=20recomm?= =?UTF-8?q?ended=20instead.=20-=20License=20update:=20play=20safe=20and=20?= =?UTF-8?q?list=20all=20third=20party=20licenses=20as=20=20=20part=20of=20?= =?UTF-8?q?the=20License=20tag.=20-=20Switch=20to=20=5Fservice=20file=20as?= =?UTF-8?q?=20tagged=20Source=20tarball=20does=20not=20=20=20include=20`./?= =?UTF-8?q?thirdparty`=20submodules.=20-=20Update=20openvino-fix-install-p?= =?UTF-8?q?aths.patch=20to=20fix=20python=20module=20=20=20install=20path.?= =?UTF-8?q?=20-=20Enable=20python=20module=20and=20split=20it=20out=20into?= =?UTF-8?q?=20a=20python=20subpackage=20=20=20(for=20now=20default=20pytho?= =?UTF-8?q?n3=20only).=20-=20Explicitly=20build=20python=20metadata=20(dis?= =?UTF-8?q?t-info)=20and=20install=20it=20=20=20(needs=20simple=20sed=20ha?= =?UTF-8?q?ckery=20to=20support=20"officially"=20unsupported=20=20=20platf?= =?UTF-8?q?orm=20ppc64le).=20-=20Specify=20ENABLE=5FJS=3DOFF=20to=20turn?= =?UTF-8?q?=20off=20javascript=20bindings=20as=20=20=20building=20these=20?= =?UTF-8?q?requires=20downloading=20npm=20stuff=20from=20the=20network.=20?= =?UTF-8?q?-=20Build=20with=20system=20pybind11.=20-=20Bump=20=5Fconstrain?= =?UTF-8?q?ts=20for=20updated=20disk=20space=20requirements.=20-=20Drop=20?= =?UTF-8?q?empty=20%check=20section,=20rpmlint=20was=20misleading=20when?= =?UTF-8?q?=20it=20=20=20recommended=20adding=20this.=20-=20Numerous=20spe?= =?UTF-8?q?cfile=20cleanups:=20=20=20*=20Drop=20redundant=20`mv`=20command?= =?UTF-8?q?s=20and=20use=20`install`=20where=20=20=20=20=20appropriate.=20?= =?UTF-8?q?=20=20*=20Build=20with=20system=20protobuf.=20=20=20*=20Fix=20S?= =?UTF-8?q?ummary=20tags.=20=20=20*=20Trim=20package=20descriptions.=20=20?= =?UTF-8?q?=20*=20Drop=20forcing=20CMAKE=5FBUILD=5FTYPE=3DRelease,=20let?= =?UTF-8?q?=20macro=20default=20=20=20=20=20RelWithDebInfo=20be=20used=20i?= =?UTF-8?q?nstead.=20=20=20*=20Correct=20naming=20of=20shared=20library=20?= =?UTF-8?q?packages.=20=20=20*=20Separate=20out=20libopenvino=5Fc.so.*=20i?= =?UTF-8?q?nto=20own=20shared=20lib=20package.=20=20=20*=20Drop=20rpmlintr?= =?UTF-8?q?c=20rule=20used=20to=20hide=20shlib=20naming=20mistakes.=20=20?= =?UTF-8?q?=20*=20Rename=20Source=20tarball=20to=20%{name}-%{version}.EXT?= =?UTF-8?q?=20pattern.=20=20=20*=20Use=20ldconfig=5Fscriptlet=20macro=20fo?= =?UTF-8?q?r=20post(un).=20-=20Add=20openvino-onnx-ml-defines.patch=20--?= =?UTF-8?q?=20Define=20ONNX=5FML=20at=20compile=20=20=20time=20when=20usin?= =?UTF-8?q?g=20system=20onnx=20to=20allow=20using=20'onnx-ml.pb.h'=20=20?= =?UTF-8?q?=20instead=20of=20'onnx.pb.h',=20the=20latter=20not=20being=20s?= =?UTF-8?q?hipped=20with=20=20=20openSUSE's=20onnx-devel=20package=20(gh#o?= =?UTF-8?q?nnx/onnx#3074).=20-=20Add=20openvino-fix-install-paths.patch:?= =?UTF-8?q?=20Change=20hard-coded=20install=20=20=20paths=20in=20upstream?= =?UTF-8?q?=20cmake=20macro=20to=20standard=20Linux=20dirs.=20-=20Add=20op?= =?UTF-8?q?envino-ComputeLibrary-include-string.patch:=20Include=20header?= =?UTF-8?q?=20=20=20for=20std::string.=20-=20Add=20external=20devel=20pack?= =?UTF-8?q?ages=20as=20Requires=20for=20openvino-devel.=20-=20Pass=20-Wl,-?= =?UTF-8?q?z,noexecstack=20to=20%build=5Fldflags=20to=20avoid=20an=20exec?= =?UTF-8?q?=20stack=20=20=20issue=20with=20intel=20CPU=20plugin.=20-=20Use?= =?UTF-8?q?=20ninja=20for=20build.=20-=20Adapt=20=5Fconstraits=20file=20fo?= =?UTF-8?q?r=20correct=20disk=20space=20and=20memory=20=20=20requirements.?= =?UTF-8?q?=20-=20Add=20empty=20%check=20section.=20-=20Initial=20package?= =?UTF-8?q?=20-=20Version=202024.0.0=20-=20Add=20openvino-rpmlintrc.?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit OBS-URL: https://build.opensuse.org/package/show/science:machinelearning/openvino?expand=0&rev=45 --- openvino.changes | 9 +-------- 1 file changed, 1 insertion(+), 8 deletions(-) diff --git a/openvino.changes b/openvino.changes index fe2912f..8c8b5bc 100644 --- a/openvino.changes +++ b/openvino.changes @@ -15,9 +15,6 @@ Tue Dec 2 22:43:52 UTC 2025 - Alessandro de Oliveira Faria