Apple uses 20 open-source Core ML models on Hugging Face platform

Apple has contributed 20 new core machine learning models to the Hugging Face open source repository to improve text and image processing. These models include image classification and semantic segmentation, designed to identify objects in images and text.

Here's What We Know

This major update includes four effective open-source LLMs (OpenELMs) that were published in April 2024 through the Hugging Face platform for collaboration and development of artificial intelligence models.

Apple has also published a large language model, Ferret, on GitHub that allows for image queries.

These moves demonstrate Apple's increased interest in developing artificial intelligence, particularly in machine learning, to improve the functionality of its products in the future, including AI tools for avatar creation and generative animation.

While this is the first public release since the announcement of Apple Intelligence at WWDC, the company has previously released research papers and other models.

Source: VentureBeat