Add Phi-4-multimodal-instruct

samples/cpp/visual_language_chat/visual_language_chat.cpp

@@ -9,7 +9,7 @@ bool print_subword(std::string&& subword) {
     return !(std::cout << subword << std::flush);
 }
 
-int main(int argc, char* argv[]) try {
+int main(int argc, char* argv[]) {
     if (argc < 3 || argc > 4) {
         throw std::runtime_error(std::string{"Usage "} + argv[0] + " <MODEL_DIR> <IMAGE_FILE OR DIR_WITH_IMAGES> <DEVICE>");
     }
@@ -37,7 +37,7 @@ int main(int argc, char* argv[]) try {
 
     std::getline(std::cin, prompt);
     pipe.generate(prompt,
-                  ov::genai::images(rgbs),
+                //   ov::genai::images(rgbs),
                   ov::genai::generation_config(generation_config),
                   ov::genai::streamer(print_subword));
     std::cout << "\n----------\n"
@@ -50,14 +50,4 @@ int main(int argc, char* argv[]) try {
             "question:\n";
     }
     pipe.finish_chat();
-} catch (const std::exception& error) {
-    try {
-        std::cerr << error.what() << '\n';
-    } catch (const std::ios_base::failure&) {}
-    return EXIT_FAILURE;
-} catch (...) {
-    try {
-        std::cerr << "Non-exception object thrown\n";
-    } catch (const std::ios_base::failure&) {}
-    return EXIT_FAILURE;
 }

src/cpp/src/visual_language/inputs_embedder.cpp

@@ -11,6 +11,7 @@
 #include "visual_language/qwen2vl/classes.hpp"
 #include "visual_language/qwen2_5_vl/classes.hpp"
 #include "visual_language/phi3_vision/classes.hpp"
+#include "visual_language/phi4mm/classes.hpp"
 #include "visual_language/minicpm/classes.hpp"
 #include "visual_language/llava/classes.hpp"
 #include "visual_language/llava_next/classes.hpp"
@@ -210,6 +211,8 @@ InputsEmbedder::InputsEmbedder(const std::filesystem::path& model_dir,
         m_impl = std::make_shared<InputsEmbedderInternVLChat>(vlm_config, model_dir, device, device_config);
     } else if (vlm_config.model_type == VLMModelType::PHI3_V) {
         m_impl = std::make_shared<InputsEmbedderPhi3V>(vlm_config, model_dir, device, device_config);
+    } else if (vlm_config.model_type == VLMModelType::PHI4MM) {
+        m_impl = std::make_shared<InputsEmbedderPhi4MM>(vlm_config, model_dir, device, device_config);
     } else if (vlm_config.model_type == VLMModelType::QWEN2_VL) {
         m_impl = std::make_shared<InputsEmbedderQwen2VL>(vlm_config, model_dir, device, device_config);
     } else if (vlm_config.model_type == VLMModelType::QWEN2_5_VL) {
@@ -236,6 +239,8 @@ InputsEmbedder::InputsEmbedder(const ModelsMap& models_map,
         m_impl = std::make_shared<InputsEmbedderInternVLChat>(vlm_config, models_map, tokenizer, config_dir_path, device, device_config);
     } else if (vlm_config.model_type == VLMModelType::PHI3_V) {
         m_impl = std::make_shared<InputsEmbedderPhi3V>(vlm_config, models_map, tokenizer, config_dir_path, device, device_config);
+    } else if (vlm_config.model_type == VLMModelType::PHI4MM) {
+        m_impl = std::make_shared<InputsEmbedderPhi4MM>(vlm_config, models_map, tokenizer, config_dir_path, device, device_config);
     } else if (vlm_config.model_type == VLMModelType::QWEN2_VL) {
         m_impl = std::make_shared<InputsEmbedderQwen2VL>(vlm_config, models_map, tokenizer, config_dir_path, device, device_config);
     } else if (vlm_config.model_type == VLMModelType::QWEN2_5_VL) {

src/cpp/src/visual_language/inputs_embedder.hpp

@@ -167,6 +167,7 @@ class InputsEmbedder {
     friend class InputsEmbedderLLaVANext;
     friend class InputsEmbedderInternVLChat;
     friend class InputsEmbedderPhi3V;
+    friend class InputsEmbedderPhi4MM;
     friend class InputsEmbedderQwen2VL;
     friend class InputsEmbedderQwen2_5_VL;
 };