[CI] [GHA] Migrate mac workflow to mac-14 arm64

.github/workflows/mac.yml

@@ -1,4 +1,4 @@
-name: macOS (13, Python 3.10)
+name: macOS (14, Python 3.10)
 on:
   workflow_dispatch:
   pull_request:
@@ -19,7 +19,7 @@ env:
   MACOSX_DEPLOYMENT_TARGET: '11.0'
   PYTHON_VERSION: '3.10'
   TARGET_BRANCH: ${{ github.base_ref || github.event.merge_group.base_ref || github.ref }}
-  BASE_PRODUCT_TYPE: public_darwin
+  BASE_PRODUCT_TYPE: public_macos_arm64
   CCACHE_MAXSIZE: 500Mi
   HF_HOME: ~/.cache/hf
   OV_CACHE: ~/.cache/ov_cache/194c936
@@ -81,7 +81,8 @@ jobs:
     - uses: openvinotoolkit/openvino/.github/actions/openvino_provider@master
       id: openvino_download
       with:
-        platform: macos_12_6
+        platform: macos_14_7
+        arch: 'arm64'
         commit_packages_to_provide: wheels,developer_package.tar.gz,openvino_node_npm_package.tar.gz
         revision: latest_available_commit
 
@@ -96,7 +97,7 @@ jobs:
     defaults:
       run:
         shell: bash
-    runs-on: macos-13
+    runs-on: macos-14
     env:
       CMAKE_GENERATOR: Ninja
       CMAKE_CXX_COMPILER_LAUNCHER: ccache
@@ -148,7 +149,7 @@ jobs:
         with:
           repos: ${{ env.SRC_DIR }}
           product_type: ${{ env.BASE_PRODUCT_TYPE }}_${{ matrix.build-type }}
-          target_arch: 'x86_64'
+          target_arch: 'arm64'
           build_type: ${{ matrix.build-type }}
           save_to: ${{ env.MANIFEST_PATH }}
 
@@ -161,7 +162,7 @@ jobs:
           cmake -DOpenVINODeveloperPackage_DIR=${{ env.OV_INSTALL_DIR }}/developer_package/cmake \
                 -DENABLE_PYTHON=ON \
                 -DCMAKE_BUILD_TYPE=${{ matrix.build-type }} \
-                -S ${{ env.SRC_DIR}} \
+                -S ${{ env.SRC_DIR }} \
                 -B ${{ env.BUILD_DIR }}
           cmake --build ${{ env.BUILD_DIR}} --config ${{ matrix.build-type }} --parallel $(nproc) --verbose
           cmake --install ${{ env.BUILD_DIR }} --config ${{ matrix.build-type }} --prefix ${{ env.INSTALL_DIR }}
@@ -216,7 +217,7 @@ jobs:
     defaults:
       run:
         shell: bash
-    runs-on: macos-13
+    runs-on: macos-14
     env:
       CCACHE_DIR: ${{ github.workspace }}/ccache
       CMAKE_CXX_COMPILER_LAUNCHER: ccache
@@ -265,7 +266,7 @@ jobs:
         with:
           repos: ${{ env.SRC_DIR }}
           product_type: ${{ env.BASE_PRODUCT_TYPE }}_Release
-          target_arch: 'x86_64'
+          target_arch: 'arm64'
           build_type: Release
           save_to: ${{ github.workspace }}
 
@@ -274,6 +275,7 @@ jobs:
 
       - name: Build Tokenizers Wheel
         run: |
+          export CMAKE_BUILD_PARALLEL_LEVEL=$(( $(nproc) - 1 ))
           python -m pip wheel -v --no-deps --wheel-dir ${{ env.WHEELS_DIR }} \
               --config-settings='override=wheel.build_tag="${{ github.run_number }}"' \
               ${{ needs.openvino_download.outputs.ov_wheel_source }} \
@@ -282,13 +284,16 @@ jobs:
 
       - name: Build GenAI Wheel
         run: |
+          export CMAKE_BUILD_PARALLEL_LEVEL=$(( $(nproc) - 1 ))
           python -m pip wheel -v --no-deps --wheel-dir ${{ env.WHEELS_DIR }} \
               ${{ needs.openvino_download.outputs.ov_wheel_source }} \
               ${{ env.SRC_DIR }}
         working-directory: ${{ env.OV_INSTALL_DIR }}
 
       - name: Build WWB Wheel
-        run: python -m pip wheel -v --no-deps --wheel-dir ${{ env.WHEELS_DIR }} ${{ env.SRC_DIR }}/tools/who_what_benchmark
+        run: |
+          export CMAKE_BUILD_PARALLEL_LEVEL=$(( $(nproc) - 1 ))
+          python -m pip wheel -v --no-deps --wheel-dir ${{ env.WHEELS_DIR }} ${{ env.SRC_DIR }}/tools/who_what_benchmark
         working-directory: ${{ env.OV_INSTALL_DIR }}
 
       - name: Save ccache
@@ -317,7 +322,7 @@ jobs:
     defaults:
       run:
         shell: bash
-    runs-on: macos-13
+    runs-on: macos-14
     env:
       OV_INSTALL_DIR: ${{ github.workspace }}/ov
       INSTALL_DIR: ${{ github.workspace }}/install
@@ -369,11 +374,11 @@ jobs:
       matrix:
         build-type: [Release]
     needs: [ openvino_download ]
-    timeout-minutes: 30
+    timeout-minutes: 50
     defaults:
       run:
         shell: bash
-    runs-on: macos-13
+    runs-on: macos-14
 
     env:
       SRC_DIR: ${{ github.workspace }}/openvino.genai
@@ -405,7 +410,7 @@ jobs:
             -DENABLE_JS=ON -DCPACK_GENERATOR=NPM \
             -DENABLE_PYTHON=OFF -DENABLE_WHEEL=OFF \
             -S ${{ env.SRC_DIR }} -B ${{ env.BUILD_DIR }}
-          cmake --build ${{ env.BUILD_DIR }} --config ${{ matrix.build-type }} --parallel --verbose
+          cmake --build ${{ env.BUILD_DIR }} --config ${{ matrix.build-type }} --parallel $(nproc) --verbose
           cmake --install ${{ env.BUILD_DIR }} --config ${{ matrix.build-type }} --prefix ${{ env.INSTALL_DIR }}
 
       - name: Upload Node.js bindings Build Package
@@ -429,14 +434,16 @@ jobs:
             cmd: 'tests/python_tests/test_whisper_pipeline.py -k "not test_smoke[sample_from_dataset0 and not test_whisper_constructors[sample_from_dataset0 and not test_max_new_tokens[sample_from_dataset0 and not test_language_mode[language and not test_task_mode[sample_from_dataset0 and not test_language_autodetect[sample_from_dataset0 and not test_whisper_config_constructor and not test_language_autodetect[sample_from_dataset1 and not test_language_autodetect[sample_from_dataset2 and not test_initial_prompt_hotwords[sample_from_dataset0 and not test_random_sampling[sample_from_dataset0"'
             run_condition: ${{ fromJSON(needs.smart_ci.outputs.affected_components).whisper.test }}
             timeout: 120
-          - name: 'Cacheopt E2E'
-            cmd: 'tests/python_tests/test_kv_cache_eviction.py'
-            run_condition: ${{ fromJSON(needs.smart_ci.outputs.affected_components).continuous_batching.test }}
-            timeout: 240
-          - name: 'LLM & VLM'
-            cmd: 'tests/python_tests/test_llm_pipeline.py tests/python_tests/test_llm_pipeline_static.py tests/python_tests/test_vlm_pipeline.py tests/python_tests/test_structured_output.py'
-            run_condition: ${{ fromJSON(needs.smart_ci.outputs.affected_components).visual_language.test || fromJSON(needs.smart_ci.outputs.affected_components).LLM.test }}
-            timeout: 180
+          # Only supported on X64 or ARM with SVE support
+          # - name: 'Cacheopt E2E'
+          #   cmd: 'tests/python_tests/test_kv_cache_eviction.py'
+          #   run_condition: ${{ fromJSON(needs.smart_ci.outputs.affected_components).continuous_batching.test }}
+          #   timeout: 240
+          # Only supported on X64 or ARM with SVE support
+          # - name: 'LLM & VLM'
+          #   cmd: 'tests/python_tests/test_llm_pipeline.py tests/python_tests/test_llm_pipeline_static.py tests/python_tests/test_vlm_pipeline.py tests/python_tests/test_structured_output.py'
+          #   run_condition: ${{ fromJSON(needs.smart_ci.outputs.affected_components).visual_language.test || fromJSON(needs.smart_ci.outputs.affected_components).LLM.test }}
+          #   timeout: 180
           - name: 'GGUF Reader tests'
             cmd: 'tests/python_tests/test_gguf_reader.py'
             run_condition: ${{ fromJSON(needs.smart_ci.outputs.affected_components).GGUF.test || fromJSON(needs.smart_ci.outputs.affected_components).LLM.test }}
@@ -445,18 +452,19 @@ jobs:
             cmd: 'tests/python_tests/test_tokenizer.py'
             run_condition: ${{ fromJSON(needs.smart_ci.outputs.affected_components).tokenizers.test }}
             timeout: 60
-          - name: 'API tests'
-            cmd: 'tests/python_tests/test_continuous_batching.py tests/python_tests/test_generation_config.py tests/python_tests/test_sampling.py tests/python_tests/test_text_streamer.py'
-            run_condition: ${{ fromJSON(needs.smart_ci.outputs.affected_components).continuous_batching.test || fromJSON(needs.smart_ci.outputs.affected_components).sampling.test || fromJSON(needs.smart_ci.outputs.affected_components).text_streamer.test }}
-            timeout: 60
+          # Only supported on X64 or ARM with SVE support
+          # - name: 'API tests'
+          #   cmd: 'tests/python_tests/test_continuous_batching.py tests/python_tests/test_generation_config.py tests/python_tests/test_sampling.py tests/python_tests/test_text_streamer.py'
+          #   run_condition: ${{ fromJSON(needs.smart_ci.outputs.affected_components).continuous_batching.test || fromJSON(needs.smart_ci.outputs.affected_components).sampling.test || fromJSON(needs.smart_ci.outputs.affected_components).text_streamer.test }}
+          #   timeout: 60
           - name: 'Rag tests'
             cmd: 'tests/python_tests/test_rag.py'
             run_condition: ${{ fromJSON(needs.smart_ci.outputs.affected_components).RAG.test }}
             timeout: 30
     defaults:
       run:
         shell: bash
-    runs-on: macos-13
+    runs-on: macos-14
     env:
       INSTALL_DIR: ${{ github.workspace }}/install
       SRC_DIR: ${{ github.workspace }}/src
@@ -535,7 +543,7 @@ jobs:
     defaults:
       run:
         shell: bash
-    runs-on: macos-13
+    runs-on: macos-14
     env:
       INSTALL_DIR: ${{ github.workspace }}/ov
       SRC_DIR: ${{ github.workspace }}/src
@@ -635,7 +643,7 @@ jobs:
     defaults:
       run:
         shell: bash
-    runs-on: macos-13
+    runs-on: macos-14
     env:
       INSTALL_DIR: ${{ github.workspace }}/ov
       SRC_DIR: ${{ github.workspace }}/src
@@ -692,7 +700,7 @@ jobs:
     defaults:
       run:
         shell: bash
-    runs-on: macos-13
+    runs-on: macos-14
 
     env:
       SRC_DIR: ${{ github.workspace }}/openvino.genai

tests/python_tests/test_gguf_reader.py

@@ -3,6 +3,7 @@
 
 
 import pytest
+import platform
 import torch
 import gc
 import sys
@@ -112,6 +113,8 @@ def test_full_gguf_pipeline(pipeline_type, model_ids, enable_save_ov_model):
 
 @pytest.mark.parametrize("pipeline_type", get_gguf_pipeline_types())
 @pytest.mark.parametrize("model_ids", [{"gguf_model_id": "Qwen/Qwen3-0.6B-GGUF", "gguf_filename": "Qwen3-0.6B-Q8_0.gguf"}])
+@pytest.mark.xfail(condition=(sys.platform == "darwin" and platform.machine() in ('arm', 'armv7l',
+                                                                                  'aarch64', 'arm64', 'ARM64')), reason="Ticket - 172335")
 @pytest.mark.precommit
 def test_full_gguf_qwen3_pipeline(pipeline_type, model_ids):
     # Temporal testing solution until transformers starts to support qwen3 in GGUF format