Update gemma3 to use openvino genai (#3075)

This PR updates the Gemma3 notebook to use OpenVINO GenAI instead of the
transformers library.

## Changes Made
- **Replaced transformers with OpenVINO GenAI** 
- **Updated model loading** to use OpenVINO GenAI pipeline
- **Modified gradio helper** to work with OpenVINO GenAI streaming
capabilities

## Files Changed
- `notebooks/gemma3/gemma3.ipynb` - Main notebook updates
- `notebooks/gemma3/gradio_helper.py` - Gradio helper updates
- `notebooks/gemma3/README.md` - Documentation updates

## Testing
- [x] Notebook runs successfully with OpenVINO GenAI
- [x] Gradio interface works with streaming
- [x] Image processing functions correctly
- [x] All dependencies are properly handled

---------

Co-authored-by: Aleksandr Mokrov <[email protected]>

Author: ezelanza

notebooks/gemma3/README.md

@@ -1,4 +1,4 @@
-# Visual-language assistant with Gemma3 and OpenVINO
+# Visual-language assistant with Gemma3 and OpenVINO GenAI
 ![](https://github.com/user-attachments/assets/2540a58e-c242-4439-b151-0fd1e6938af1)
 
 Gemma 3 is Google's new iteration of open weight LLMs. It comes in four sizes, 1 billion, 4 billion, 12 billion, and 27 billion parameters with base (pre-trained) and instruction-tuned versions. The 4, 12, and 27 billion parameter models can process both images and text, while the 1B variant is text only.
@@ -19,7 +19,7 @@ The tutorial consists from following steps:
 
 - Install requirements
 - Convert and Optimize model
-- Run OpenVINO model inference
+- Run OpenVINO GenAI model inference
 - Launch Interactive demo
 
 In this demonstration, you'll create interactive chatbot that can answer questions about provided image's content.