Google on Wednesday introduced its new series of open-source large language models, Gemma, for free. It is built from the same research as one of its models, the Gemini. It comes in two sizes: Gemma 2B and Gemma 7B. These sizes can also run on Google Cloud, workstations, and developers’ laptops.
The company has stated that first-time Cloud users will get $300 credit, as this model is optimised by Google Cloud. Gemma is developed by Google DeepMind and the Google AI team. Google says that the word Gemma is derived from the Latin word ‘Gem’ which means a “precious stone.”
Also Read: well health tips in hindi wellhealth
To offer guidance and help to the developers, Google released the “Responsible Generative AI Toolkit,” which will encourage the safe and responsible building of AI applications. Google can still change the terms and conditions of use as this open source is not fully developed. Gemma can be used as a chatbot, generate content, and do much more than a language model can do.
The purpose of Gemma is to make the art of artificial intelligence accessible and to ensure the safe and responsible use of it with the help of the toolkit provided. It is designed in such a way that it can be used in a low-resource environment. Gemma can easily process a wide variety of words as it contains 250,000 vocabulary tokens. Also, it can solve math problems and perform coding as well. Therefore, the data it provides is more accurate and relevant.
Chipmaker NVIDIA stated on Wednesday that it worked with Google and made sure that the Gemma models ran smoothly on its chips. Also, they will create chatbot software to run AI models on Windows PCs to work with Gemma.
Also Read: wellhealthorganic vitamin b12
Google made a move similar to Meta platforms by launching Gemma, its new open-source language model in its family. Google aims to attract more developers to its Vertex AL Cloud platform.