Google Gemma

Gemma by Google represents a suite of advanced, lightweight open language models (LLMs) crafted from the same foundational research as the Gemini models. Available in 2B and 7B parameter configurations, Gemma includes both base and instruction-tuned variants. This model adheres to AI safety and reliability standards, ensuring its responsible use across various devices. It is optimized for performance on Google Cloud and NVIDIA GPUs, making it a versatile choice for developers. Users can access Gemma at no cost on platforms like Kaggle and Colab, as well as through Google Cloud services like Vertex AI or GKE.

Research & Insights