Apple has unveiled a series of four small language models, collectively known as OpenELM. These models prioritize efficient performance on local Apple devices like emails, leveraging a leaner structure with varying parameter sizes ranging from 270 million to 3 billion. This release portrays Apple’s dedication to advancing generative AI features within its ecosystem. Apple CEO Tim Cook has previously indicated substantial investments and efforts in this domain. Despite prior releases of AI models, Apple has yet to introduce an AI foundation model for commercial use, contrasting the approaches taken by competitors. In December, Apple introduced Core ML frameworks to optimize AI model performance on Apple Silicon. Additionally, the company released an AI-powered photo enhancement feature. Rumors suggest Apple is developing a dedicated AI chip. However, Apple’s plans for integrating these models into its product lineup remain undisclosed.
Apple unveils OpenELM: Small language models for swift local AI processing
