At the Google I/O 2024 developer conference, the company announced some new updates for its Gemini series AI models and a completely new model called Gemini 1.5 Flash.
What are the Gemini 1.5 Pro Updates?
First of all, some updates to the Gemini 1.5 Pro model is taking. Here’s the description: “1.5 Pro can now follow increasingly complex and nuanced instructions, including those that specify product-level behaviors involving roles, formats, and styles. For specific uses, such as establishing the personality and response style of a chat agent or automating workflows through multiple function calls ±m states internal We have improved the control over the model’s responses. We also enabled users to direct model behavior by adjusting system instructions.â€
What Does Gemini 1.5 Flash Offer?
Gemini 1.5 Flash is described as a lightweight large language model (LLM) designed to run fast. Here’s the company’s description for the model: “1.5 Flash provides summarization, chat applications, adding image and video captions, extracting data from long documents and tables, and more. He is excellent in many subjects. This is because the 1.5 Pro uses a process called “distillation,” in which the most essential knowledge and skills from a larger model are transferred to a smaller, more efficient model. It is because he is trained.â€
While the new model is currently offered in preview form, it will be available for general use in June.
Both Flash and Pro models come with a 1 million token content window, but Google also announced today that it is testing a 2 million token content window for the Pro version. Developers who want to try this version waiting list will be able to register.
Gemma 2 Announced
Google also released the new Gemma 2 open source LLM model. announced: “Developers and researchers requested a larger model that was still in an easy-to-use size. The new Gemma 27B model should do it: It outperforms some models more than twice its size and will run efficiently on GPUs or a single TPU host on Vertex AI .â€
It is stated that the model will be launched in June.
Finally, the Gemini Nano model for on-device artificial intelligence efforts now supports visual, audible and verbal commands as well as text commands on the company’s Pixel devices. will also be able to understand language commands.
