sinapsis-huggingface v0.2.0

We are excited to announce the upgrade of the sinapsis-huggingface mono repo, introducing significant enhancements and new tools to improve your development experience. This release includes updates to the sinapsis-huggingface-transformers package and the introduction of a new webapp for testing.

Key Features:

Enhanced sinapsis-huggingface-transformers:
    Paligemma Support: The package now includes support for paligemma, allowing for an improved experience for vision models, with the ability to take text and image as input and generate text from it
    PaliGemmaInference: Generate descriptive captions for images, enabling applications like image tagging and accessibility features.
    PaliGemmaDetection: Detect specific objects within images, useful for scenarios such as surveillance and quality control.
    Updated Pipeline: The pipeline for transformers tasks has been enhanced to allow explicit selection of a particular task when working with multi-modal models, providing greater flexibility and control over model usage.

New Webapp for Testing:
    Interactive Testing Tool: A new web-based interface has been developed to facilitate easy interaction paligemma. This tool allows users to test model functionalities, visualize outputs, and experiment with different configurations without extensive coding. It serves as an accessible platform for developers to explore and integrate models into their projects seamlessly.

Conclusion:
This upgrade represents a substantial step forward in both model capabilities and user accessibility. The addition of paligemma support, the updated pipeline, and the new webapp for testing underscore our commitment to providing cutting-edge tools and user-friendly experiences. We encourage all users to update to the latest version and explore these new features to enhance their development workflows.