Gemma: Desktop friendly, lightweight Open Source launched by Google

Google launched its latest lightweight laptop-friendly open-source Gemma on February 21. The blog post by Google says that Gemma is built on the same infrastructure and technology as Gemini models.

Gemma would prove super beneficial for the new generation of AI development models used to assist researchers and developers. Google Deepmind and other teams across Google develop it. Along with model weight, Google also released guides and tools to support development, innovation, and collaboration.

Also Read: Gemini Chatbot rolled out by Google for Android Users

What is Gemma?

Gemma is derived from the Latin word meaning precious stone, developed by Google Deepmind and other team members at Google. Gemma is inspired by the Gemini model’s research and technology but belongs to the family of lightweight open models. 

What are its features?

Some of the details mentioned in the blog are listed here:

• For creating safer AI applications with Gemma A new Responsible Generative AI Toolkit provides guidance and essential tools.
• Gemma 2B and Gemma 7B are the two sizes for model weights released by Google for Gemma, along with pre-trained and instruction-tuned variants.
• Along with integration with popular tools such as Hugging Face, MaxText, NVIDIA NeMo, TensorRT-LLM, Ready-to-Use Colab, and Kaggle notebooks, it’s easy to get started with Gemma.
• Pre-trained and instruction-tuned Gemma models can run on your laptop, workstation, or Google Cloud with easy deployment on Vertex AI and the Google Kubernetes Engine (GKE).
• Including NVIDIA GPUs and Google Cloud TPUs, optimization across multiple AI hardware platforms ensures industry-leading performance.
• Terms of use permit responsible commercial usage and distribution for all organizations, regardless of size.

Also Read: Gemini AI Blunder Costs Google $70 Billion Loss: Pischai Calls It Unacceptable and Warned Employee

Google also shared technical reports on dataset composition, performance, and modeling methodologies for Gemma in their blog. Here is the report for the viewers to have a look at:

Google will give free credits for research and development to empower AI innovation. Developers can start working with Gemma today using free access in Kaggle, a free tier for Colab notebooks, and $300 in credits for first-time Google Cloud users. Researchers can also apply for Google Cloud credits of up to $500,000 to accelerate their projects.