Gemma (base)

Gemma 2 and Gemma are the core models of the Gemma family of open models. These generative artificial intelligence (AI) models are built from the same research and technology used to create the Gemini models.

Gemma base models are well-suited for a variety of text generation tasks, including question answering, summarization, reasoning, and can be further tuned for specific use cases. The base Gemma models are provided in a range of sizes, from 2 billion up to 27 billion parameters, and provide the following advantages:

  • Gemma models are provided with open weights and permit responsible commercial use, allowing you to tune and deploy them in your own projects and applications.
  • Gemma models provide state-of-the-art performance on AI tasks compared to other open models of similar size.
  • Gemma is supported on a wide variety of tools and systems, including Keras 3.0, native PyTorch, JAX, and Hugging Face Transformers.

More resources

The Gemma 2 model card contains detailed information about the model.
View more code, Colab notebooks, information, and discussions about Gemma.
Run example Colab notebooks with Gemma and a variety of development frameworks on GitHub.