Remove usage of deprecated functions (#3011)

Details:
Remove usage of deprecated functions from ov::Allocator and ov::Model

Tickets:
[CVS-174874](https://jira.devtools.intel.com/browse/CVS-174874)
[CVS-174873](https://jira.devtools.intel.com/browse/CVS-174873)

---------

Co-authored-by: Michal Lukaszewski <michal.lukaszewski@intel.com>

Author: olpipi

samples/cpp/image_generation/load_image.cpp

@@ -31,10 +31,7 @@ ov::Tensor utils::load_image(const std::filesystem::path& image_path) {
             }
             throw std::runtime_error{"Unexpected number of bytes was requested to allocate."};
         }
-        void deallocate(void*, size_t bytes, size_t) {
-            if (channels * height * width != bytes) {
-                throw std::runtime_error{"Unexpected number of bytes was requested to deallocate."};
-            }
+        void deallocate(void*, size_t, size_t) noexcept {
             stbi_image_free(image);
             image = nullptr;
         }

samples/cpp/visual_language_chat/load_image.cpp

@@ -45,10 +45,7 @@ ov::Tensor utils::load_image(const std::filesystem::path& image_path) {
             }
             throw std::runtime_error{"Unexpected number of bytes was requested to allocate."};
         }
-        void deallocate(void*, size_t bytes, size_t) {
-            if (channels * height * width != bytes) {
-                throw std::runtime_error{"Unexpected number of bytes was requested to deallocate."};
-            }
+        void deallocate(void*, size_t, size_t) noexcept {
             stbi_image_free(image);
             image = nullptr;
         }

src/python/py_image_generation_pipelines.cpp

@@ -208,10 +208,7 @@ class TorchGenerator : public ov::genai::CppStdGenerator {
                 throw std::runtime_error{"Unexpected number of bytes was requested to allocate."};
             }
 
-            void deallocate(void*, size_t bytes, size_t) {
-                if (m_total_size != bytes) {
-                    throw std::runtime_error{"Unexpected number of bytes was requested to deallocate."};
-                }
+            void deallocate(void*, size_t, size_t) noexcept {
             }
 
             bool is_equal(const TorchTensorAllocator& other) const noexcept {

tests/cpp/helper.cpp

@@ -26,6 +26,5 @@ std::shared_ptr<ov::Model> get_dummy_model(ov::Core core, size_t num_layers) {
     }
     const auto& concat1 = std::make_shared<ov::op::v0::Concat>(keys, 1);
     const auto& concat2 = std::make_shared<ov::op::v0::Concat>(values, 1);
-    auto model = std::make_shared<ov::Model>(ov::NodeVector{concat1, concat2}, params);
-    return std::make_shared<ov::Model>(ov::NodeVector{concat1, concat2}, params);
+    return std::make_shared<ov::Model>(ov::OutputVector{concat1, concat2}, params);
 }