Google, the company under Alphabet, has released a new family of AI models called Gemma.
These models are considered ‘open,’ meaning individuals and businesses can use them to build AI software for free. Google is making key technical data, including model weights, publicly available.
While not fully ‘open source,’ this move aims to attract software engineers to work on Google’s technology and promote the use of its profitable cloud division.
Gemma models are optimized for Google Cloud, offering first-time cloud users $300 in credits.
Introducing Gemma: a family of lightweight, state-of-the-art open models for developers and researchers to build with AI. 🌐
— Google DeepMind (@GoogleDeepMind) February 21, 2024
We’re also releasing tools to support innovation and collaboration – as well as to guide responsible use.
Get started now. → https://t.co/nsoFqfHffY pic.twitter.com/PRNIndgU4S
The company has not made its premier model, Gemini, open but disclosed that Gemma models have two billion or seven billion parameters.
This announcement follows Meta’s similar move with its Llama 2 models.
Google collaborates with Nvidia to ensure Gemma models run smoothly on Nvidia’s chips, with plans to make chatbot software compatible with Gemma.