Remove OVClassModelTestP

.github/workflows/llama_cpp_plugin_build_and_test.yml

@@ -4,6 +4,7 @@ on:
   pull_request:
     paths:
       - 'modules/llama_cpp_plugin/**'
+      - '.github/workflows/llama_cpp_plugin_build_and_test.yml'
 
 permissions: read-all
 
@@ -46,6 +47,11 @@ jobs:
     needs: build_ubuntu20
     runs-on: ubuntu-20.04
     steps:
+      - name: Set up Python 3.9
+        uses: actions/setup-python@39cd149  # v5.1.1
+        with:
+          python-version: "3.9"
+
       - name: Download build artifacts
         uses: actions/download-artifact@65a9edc5881444af0b9093a5e628f2fe47ea3b2e # v4.1.7
         with:
@@ -60,7 +66,7 @@ jobs:
 
       - name: Prepare test data - convert test model files
         run: |
-          pip install -r llama.cpp/requirements/requirements-convert-hf-to-gguf.txt
+          pip install -r llama.cpp/requirements/requirements-convert_hf_to_gguf.txt
           huggingface-cli download gpt2 model.safetensors tokenizer.json tokenizer_config.json vocab.json config.json merges.txt --local-dir hf_gpt2
           mkdir -p ${{ github.workspace }}/test_data
           python3 llama.cpp/convert-hf-to-gguf.py hf_gpt2 --outtype f32 --outfile ${{ github.workspace }}/test_data/gpt2.gguf

modules/custom_operations/tests/requirements.txt

@@ -2,4 +2,6 @@ torch
 onnx
 tensorboard
 pytest
+# WA CVS-150813
+numpy<2.0.0
 # open3d==0.16.0 - need to update with new release

modules/llama_cpp_plugin/src/compiled_model.cpp

@@ -27,11 +27,11 @@ LlamaCppModel::LlamaCppModel(const std::string& gguf_fname,
     : ICompiledModel(nullptr, plugin),
       m_gguf_fname(gguf_fname),
       m_num_threads(num_threads) {
-    OPENVINO_DEBUG << "llama_cpp_plugin: loading llama model directly from GGUF... " << std::endl;
+    OPENVINO_DEBUG("llama_cpp_plugin: loading llama model directly from GGUF... \n");
     llama_model_params mparams = llama_model_default_params();
     mparams.n_gpu_layers = 99;
     m_llama_model_ptr = llama_load_model_from_file(gguf_fname.c_str(), mparams);
-    OPENVINO_DEBUG << "llama_cpp_plugin: llama model loaded successfully from GGUF..." << std::endl;
+    OPENVINO_DEBUG("llama_cpp_plugin: llama model loaded successfully from GGUF...\n");
 
     auto input_ids = std::make_shared<ov::opset13::Parameter>(ov::element::Type_t::i64, ov::PartialShape({-1, -1}));
     auto fake_convert = std::make_shared<ov::opset13::Convert>(input_ids->output(0), ov::element::Type_t::f32);
@@ -71,7 +71,7 @@ std::shared_ptr<const ov::Model> LlamaCppModel::get_runtime_model() const {
 }
 
 void LlamaCppModel::set_property(const ov::AnyMap& properties) {
-    OPENVINO_DEBUG << "llama_cpp_plugin: attempted to set_property (did nothing)";
+    OPENVINO_DEBUG("llama_cpp_plugin: attempted to set_property (did nothing)");
 }
 
 ov::Any LlamaCppModel::get_property(const std::string& name) const {

modules/llama_cpp_plugin/src/infer_request.cpp

@@ -28,7 +28,7 @@ void allocate_tensor_impl(ov::SoPtr<ov::ITensor>& tensor,
 LlamaCppSyncInferRequest::LlamaCppSyncInferRequest(const std::shared_ptr<const LlamaCppModel>& compiled_model,
                                                    size_t num_threads)
     : ov::ISyncInferRequest(compiled_model) {
-    OPENVINO_DEBUG << "llama_cpp_plugin: infer request ctor called\n";
+    OPENVINO_DEBUG("llama_cpp_plugin: infer request ctor called\n");
     llama_context_params cparams = llama_context_default_params();
     cparams.n_threads = num_threads ? num_threads : std::thread::hardware_concurrency();
     cparams.n_ctx = 0;  // this means that the actual n_ctx will be taken equal to the model's train-time value
@@ -51,7 +51,7 @@ LlamaCppSyncInferRequest::LlamaCppSyncInferRequest(const std::shared_ptr<const L
 }
 void LlamaCppSyncInferRequest::set_tensors_impl(const ov::Output<const ov::Node> port,
                                                 const std::vector<ov::SoPtr<ov::ITensor>>& tensors) {
-    OPENVINO_DEBUG << "llama_cpp_plugin: set_tensors_impl called\n";
+    OPENVINO_DEBUG("llama_cpp_plugin: set_tensors_impl called\n");
 }
 
 void llama_batch_add_reimpl(struct llama_batch& batch,
@@ -131,12 +131,12 @@ void LlamaCppSyncInferRequest::infer() {
     llama_batch_free(batch);
 };
 std::vector<ov::ProfilingInfo> LlamaCppSyncInferRequest::get_profiling_info() const {
-    OPENVINO_DEBUG << "llama_cpp_plugin: get_profiling_info() called\n";
+    OPENVINO_DEBUG("llama_cpp_plugin: get_profiling_info() called\n");
     return std::vector<ov::ProfilingInfo>{};
 };
 
 std::vector<ov::SoPtr<ov::IVariableState>> LlamaCppSyncInferRequest::query_state() const {
-    OPENVINO_DEBUG << "llama_cpp_plugin: query_state() called\n";
+    OPENVINO_DEBUG("llama_cpp_plugin: query_state() called\n");
     return {std::static_pointer_cast<ov::IVariableState>(std::make_shared<LlamaCppState>(m_llama_ctx))};
 }
 

modules/nvidia_plugin/tests/functional/shared_tests_instances/behavior/ov_plugin/core_integration.cpp

@@ -14,14 +14,6 @@ using namespace ov::test::behavior;
 
 namespace {
 
-//
-// OV Class Common tests with <pluginName, device_name params>
-//
-
-INSTANTIATE_TEST_SUITE_P(smoke_OVClassNetworkTestP,
-                         OVClassModelTestP,
-                         ::testing::Values(ov::test::utils::DEVICE_NVIDIA));
-
 //
 // OV Class GetMetric
 //

modules/nvidia_plugin/tests/functional/skip_tests_config.cpp

@@ -39,8 +39,6 @@ std::vector<std::string> disabledTestPatterns() {
         R"(.*InferRequestIOBBlobTest.*canProcessDeallocatedOutputBlobAfterGetAndSetBlob.*)",
         // 119703
         R"(.*smoke_GroupConvolutionBias(Add|AddAdd)_2D_ExplicitPaddingSymmetric2.*FP16*.*)",
-        // Issue: 128924
-        R"(.*smoke_OVClassNetworkTestP/OVClassModelTestP.ImportModelWithNullContextThrows.*)",
     };
 
 #ifdef _WIN32