Google has unveiled a new Gemma AI model aimed at researchers who would like to run AI models on their devices instead of cloud services.
According to Google, The New Gemma model was designed by its UK-based subsidiary DeepMind in collaboration with other teams at the company.
The Gemma model shares some “technical and infrastructure components” with the larger Gemini model, and Google offers it in two models: Gemma 2B and Gemma 7B.
Google has improved its new model to work on Nvidia graphics processing cards as well as Google Cloud TPUs, the Google processors that run artificial intelligence tasks in the Google Cloud Cloud.
Researchers have access to previously trained and tuned Gemma models to receive instructions, as they can be run locally on desktops, laptops, or through the cloud.
Google says that the Gemma model with its two versions is technically smaller compared to other large language models (LLMs), but claimed that it “significantly outperforms it in key measurements,” including the llama 2 model from Meta.
Google is now making the Gemma model available to researchers for free via its Kaggle and Colab platforms, and also provides free credit for researchers to use the model in the cloud via Google Cloud.
Earlier this month, the company launched the Gemini model, and made available a free application of the same name and paid subscriptions for advanced features at a price of 20 US dollars. Later, the company announced the Gemini 1.5 Pro model with high capabilities compared to the first version.
