Explore Google Gemma, a family of powerful, open AI models. These lightweight models leverage cutting-edge Gemini technology for advanced performance.
The **Google Gemma** family represents a suite of accessible, state-of-the-art open language models designed for responsible AI development. These models, available in 2B and 7B parameter sizes, are built upon the same technology as the Google Gemini chatbot and Workspace AI tools. They utilize a decoder-only architecture and are optimized for running on developer laptops or desktops (with CPU/GPU support) or via Google Cloud Platform with GPU and TPU acceleration. Gemma models offer English language capabilities and come in pre-trained and instruction-tuned variants, ensuring flexibility across various applications, and are accessible through platforms such as Kaggle, Hugging Face, Nvidia’s NeMo, and Google’s Vertex AI.
