Google launches Gemma open AI model for AI research and developers

Google has launched another AI model Gemma. It has been specially designed for AI developers and researchers. It is an open source model available to everyone. Gemma has been developed in collaboration with Google’s team and many other teams including Google DeepMind. The company says that this has been launched as its contribution to the emerging AI sector.

The name Gemma is derived from the Latin word Gemma which means precious stone. Google has named it Gemma in view of its contribution in the sector of AI. the company has blog I have given information about this. Gemma is made available in two different sizes which are Gemma 2B and Gemma 7B. Both models have been pre-tuned as per instructions. These are quite light and can run easily on any developer’s laptop or desktop. The company says that both the models are quite powerful and also responsible. The company has removed the data from their training set, and has also done rigorous testing so that they work safely.

Google has described Gemma as a versatile model. This is because whether the user is working on a laptop, or mobile, or installing applications on other IoT, Gemma easily integrates with them. It is compatible with Vertex AI and Google Kubernetes Engine (GKE) to make application deployment easier. Thanks to the partnership with Nvidia, it works seamlessly with Google Cloud TPUs and Nvidia GPUs.

Through the launch of Gemma, the company is emphasizing that AI should be used in a responsible manner. So that it can be used commercially also. In which organizations of all sizes can be included. The company has also launched a Responsible Generative AI Toolkit for this, through which developers will be able to use Gemma very easily, and to its full potential.