Gemma - a family of lightweight, state-of-the art open models from Google.

Gemma models are a family of lightweight, state-of-the-art open models built using the same technology and research as the Gemini models. They incorporate comprehensive safety measures, ensuring responsible and trustworthy AI solutions through curated datasets and rigorous tuning. These models also deliver unmatched performance at their 2B and 7B sizes, even surpassing some larger open models. They are compatible with Keras 3.0, JAX, TensorFlow, and PyTorch, allowing users to easily switch between frameworks depending on their tasks.

Gemma models are available in all popular model hubs and are pre-trained on carefully curated data for safety, promoting responsible AI development. They undergo comprehensive evaluations and transparent reporting to reveal model limitations, fostering a responsible approach for each use case. The Responsible Generative AI Toolkit further supports developers in designing and implementing Responsible AI best practices.

Key takeaways:

Gemma models are lightweight, state-of-the-art open models built from the same research and technology used to create the Gemini models.
These models incorporate comprehensive safety measures to ensure responsible and trustworthy AI solutions.
Gemma models achieve exceptional benchmark results at its 2B and 7B sizes, even outperforming some larger open models.
With Keras 3.0, Gemma models offer seamless compatibility with JAX, TensorFlow, and PyTorch, allowing users to effortlessly choose and switch frameworks depending on their task.

Gemma - a family of lightweight, state-of-the art open models from Google. | Google AI for Developers

Key takeaways:

Comments (0)

Newsletter