feat: Imagen model 002 update

[katiemn]

Description

Update notebooks to latest Imagen model, add example for prompt enhancement feature and mask-free editing, and update samples to use new genai sdk.

[review-notebook-app]

Check out this pull request on 

See visual diffs & provide feedback on Jupyter Notebooks.

---

Powered by ReviewNB

[code-review-assist]

Hi @katiemn, reviewers, and future viewers,

I'm Gemini, currently reviewing this pull request. In the meantime, I'm providing a summary for everyone to quickly grasp the changes and intent.

This pull request, titled "feat: Imagen model 002 update", aims to update the Imagen model used in several Jupyter notebooks and integrate the new Google Gen AI SDK. The description mentions adding examples for prompt enhancement and mask-free editing.

Here's a breakdown of the changes:

• vision/getting-started/imagen3_editing.ipynb: This notebook received the most significant updates. The changes include:

  • Updating references to the Vertex AI SDK to the Google Gen AI SDK (lines 135-136, 146-147).
  • Removing the runtime restart steps (lines 149-186).
  • Adding mask-free editing to the tutorial description (lines 117-118).
  • Updating the model versions used for image generation and editing (lines 351-354, 364-367).
  • Refactoring code to use the new Google Gen AI SDK API calls for image generation and editing (lines 396-405, 419-429, 460-469, 534-549, 585-596, 654-664, 739-748, 769-779).
  • Updating the person_generation and safety_filter_level parameters to use the new enum values (lines 381-382).
  • Modifying image display functions to directly use PIL images (lines 335-337).
  • Changing how the initial image is loaded for background swap (lines 625-626).
  • Adding a new section for mask-free editing (lines 671-714).
  • Minor text corrections and clarifications throughout the notebook.

• vision/getting-started/imagen3_image_generation.ipynb: This notebook was also updated to use the new Google Gen AI SDK and the imagen-3.0-generate-002 model (lines 289-292, 331-362, 437-448, 475-486, 512-523).

  • Added a section demonstrating prompt enhancement (lines 370-404).
  • Updated person_generation and safety_filter_level parameters to use the new enum values (lines 418-419).

• vision/use-cases/creating_high_quality_visual_assets_with_gemini_and_imagen_vertex_ai_sdk.ipynb: This notebook was updated to use the new Google Gen AI SDK (lines 158-160).

• vision/getting-started/imagen3_editing_vertex_ai_sdk.ipynb: A new notebook was added, seemingly a previous version of vision/getting-started/imagen3_editing.ipynb.

I'll be providing a more detailed review shortly. In the meantime, please review the summary above.

Short poem:

Code flows like a river,
Clean commits, a gentle stream,
PRs merge to the sea.

[code-review-assist]

This pull request updates the notebooks to use the latest Imagen model (imagen-3.0-generate-002) and the GenAI SDK. It also adds examples for prompt enhancement and mask-free editing. The changes are well-structured and address the stated objectives. Here's a summary of my feedback, referencing elements from the provided style guide (if applicable):

• Correctness: The code changes correctly update the model names and API calls to reflect the new Imagen model and GenAI SDK. The added examples for prompt enhancement and mask-free editing are functional and demonstrate the intended features.
• Efficiency: The code is efficient and does not introduce any obvious performance bottlenecks. The use of the GenAI SDK simplifies the code and improves readability.
• Maintainability: The code is well-organized and easy to understand. The use of descriptive variable names and comments enhances readability. The update to the GenAI SDK aligns with current best practices for interacting with Google Cloud AI services.
• Security: No security vulnerabilities were identified in the code changes.
• Miscellaneous: The updated notebooks provide clear examples and explanations of how to use the new features. The documentation is comprehensive and up-to-date.