Add Phi-3.5-vision-instruct model

notebooks/phi-3-vision/gradio_helper.py

@@ -7,6 +7,8 @@
 
 
 def make_demo(model, processor):
+    model_name = Path(model.config._name_or_path).parent.name
+
     example_image_urls = [
         ("https://github.com/openvinotoolkit/openvino_notebooks/assets/29454499/dd5105d6-6a64-4935-8a34-3058a82c8d5d", "small.png"),
         ("https://github.com/openvinotoolkit/openvino_notebooks/assets/29454499/1221e2a8-a6da-413a-9af6-f04d56af3754", "chart.png"),
@@ -92,12 +94,12 @@ def bot_streaming(message, history):
 
     demo = gr.ChatInterface(
         fn=bot_streaming,
-        title="Phi3 Vision 128K Instruct with OpenVINO",
+        title=f"{model_name} with OpenVINO",
         examples=[
             {"text": "What is the text saying?", "files": ["./small.png"]},
             {"text": "What does the chart display?", "files": ["./chart.png"]},
         ],
-        description="Try the [Phi3-Vision model](https://huggingface.co/microsoft/Phi-3-vision-128k-instruct) from Microsoft wiht OpenVINO. Upload an image and start chatting about it, or simply try one of the examples below. If you won't upload an image, you will receive an error.",
+        description=f"Try the [{model_name} model](https://huggingface.co/microsoft/{model_name}) from Microsoft with OpenVINO. Upload an image and start chatting about it, or simply try one of the examples below. If you won't upload an image, you will receive an error.",
         stop_btn="Stop Generation",
         multimodal=True,
     )

notebooks/phi-3-vision/ov_phi3_vision.py → notebooks/phi-3-vision/ov_phi3_vision_helper.py

@@ -192,6 +192,7 @@ def cleanup_torchscript_cache():
 
 
 def convert_phi3_model(model_id, output_dir, quantization_config):
+    model_name = Path(model_id).name
     output_dir = Path(output_dir)
 
     lang_model_path = output_dir / "language_model.xml"
@@ -207,9 +208,9 @@ def convert_phi3_model(model_id, output_dir, quantization_config):
             embed_token_path.exists(),
         ]
     ):
-        print(f"✅ Phi-3-vision model already converted. You can find results in {output_dir}")
+        print(f"✅ {model_name} model already converted. You can find results in {output_dir}")
         return
-    print("⌛ Phi-3-vision conversion started. Be patient, it may takes some time.")
+    print(f"⌛ {model_name} conversion started. Be patient, it may takes some time.")
     print("⌛ Load Original model")
     model = AutoModelForCausalLM.from_pretrained(model_id, trust_remote_code=True, _attn_implementation="eager")
     processor = AutoProcessor.from_pretrained(model_id, trust_remote_code=True)
@@ -313,7 +314,7 @@ def forward_wrap(
         cleanup_torchscript_cache()
         del model
         gc.collect()
-        print(f"✅ Phi-3-vision model conversion finished. You can find results in {output_dir}")
+        print(f"✅ {model_name} model conversion finished. You can find results in {output_dir}")
 
 
 class OvPhi3Vision(GenerationMixin):