Apple introduces Ferret, its first open-source AI project surpassing GPT-4's power
Apple is entering the field of artificial intelligence with Ferret, its open-source multimodal language model. Capable of understanding and generating text, images, sounds, and videos, Ferret outperforms OpenAI’s GPT-4 in image analysis. Apple makes a noteworthy debut in artificial intelligence with Ferret, its open-source multimodal language model. Like Gemini, ChatGPT, or Google Bard, this model can comprehend and produce not only text but also images, sounds, and videos. Introduced in October by Zhe Gan, an AI researcher at Apple, Ferret has remained under the radar until now. Ferret is the result of collaboration between Gan, his colleagues at Apple, and researchers from Columbia University. According to Gan, Ferret surpasses OpenAI’s GPT-4—the most sophisticated language model to date—in analyzing and describing small areas of images while making fewer errors.
An AI trained using 8 Nvidia A100 GPUs
Apple optimized Ferret with the use of eight Nvidia A100 GPUs, high-end graphic components featuring 80 GB of HBM2e RAM. These GPUs are highly sought after in the generative AI space, a burgeoning technology that enables content creation from scratch. OpenAI's ChatGPT, an interactive chatbot, has brought this technology into the spotlight. The A100 GPU, capable of achieving a computing speed of 312 TeraFLOPS with Tensor Float 32 precision, is widely used for AI computations.
Ferret on smartphones soon?
Apple is taking its first steps in generative AI with Ferret, aiming to adapt this language model for smartphones. OpenAI’s GPT-4, boasting over one trillion parameters, far exceeds the current capacity of smartphones, which can only handle LLMs with about 10 billion parameters. To address this challenge, Apple researchers recently demonstrated how to utilize smartphone flash memory in addition to RAM to run larger models than would typically be feasible on the device. As a result, it’s likely that the iPhone 16 will feature an enhanced AI-powered assistant.
When you visit a website, it may store or retrieve information on your browser, mainly in the form of cookies. This information may relate to you, your preferences or your device and is primarily used to make the site work as you expect. The information generally does not directly identify you, but it may provide you with a more personalized web experience. Because we respect your right to privacy, you can choose not to allow certain types of cookies. Click on the different category sections to find out more and change our default settings. However, blocking certain types of cookies may impact your experience of the site and the services we are able to offer.
These cookies are necessary for the website to function and cannot be disabled in our systems. They are generally only set in response to actions you take that constitute a request for services, such as setting your privacy preferences, logging in, or filling out forms. You can set your browser to block or alert you about these cookies, but some parts of the site will then not work. These cookies do not store any personally identifiable information.
These cookies allow us to count visits and traffic sources so that we can measure and improve the performance of our site. They help us know which pages are the most and least popular and see how visitors move around the site. All information collected by these cookies is aggregated and therefore anonymous. If you do not allow these cookies, we will not know when you have visited our site and will not be able to monitor its performance.