Thursday, June 13, 2024 Banner
HomeSoftwareAIApple's Latest AI Model Offers Glimpse into Potential iPhone Integration.

Apple’s Latest AI Model Offers Glimpse into Potential iPhone Integration.

Although Apple has been mum about its generative AI plans, the company’s immediate goals seem to be to “make AI run locally on Apple devices,” as evidenced by the release of new AI models today.

On Wednesday, Apple researchers released OpenELM, a set of four incredibly tiny language models based on the Hugging Face model library. OpenELM, or “Open-source Efficient Language Models,” is touted by Apple to perform exceptionally well on text-related tasks like email composing on its Hugging Face model website. The models are available for usage by developers and are free.

There are four different sizes available: 270 million, 450 million, 1.1 billion, and 3 billion characteristics. The number of variables a model can comprehend when making decisions from its training datasets is referred to as its parameter. For instance, Google’s Gemma model has 2 billion parameters, whereas Microsoft’s recently launched Phi-3 model has a bottom out of 3.8 billion. Smaller variants are optimised to run on devices such as laptops and phones and are less expensive to operate.

When Tim Cook, the CEO of Apple, stated in February that the business is investing “a tremendous amount of time and effort” in the field, he hinted that generative AI features might be added to Apple products. Apple hasn’t disclosed any details about how it plans to use AI, though.

Although its competitors have provided AI foundation models for commercial usage, the company has already released additional AI models.

Apple introduced MLX, a machine learning framework, in December with the goal of improving the performance of AI models on Apple Silicon. Additionally, it released the MGIE model of image editing, which enables prompt-based photo correction. Smartphone navigation might also be achieved with Ferret-UI, another concept. There are also rumours that Apple is developing a code completion tool akin to Copilot on GitHub.

Nevertheless, the business apparently contacted Google and OpenAI to integrate its models into Apple goods, despite all of the model releases from Apple.


Most Popular

Recent Comments