Just after dropping Gemini 1.5 Google has introduced Gemma, a new family of open models based on the same technology used to create Gemini models. Gemma models are text-to-text, decoder-only large language models that can handle various tasks, such as text summarization, text generation, reasoning, and Q&A. They are also designed with Google’s AI Principles in mind, and come with a Responsible Generative AI Toolkit to help developers create safer and more reliable AI applications.
– Gemma models provide **best-in-class performance** for their sizes compared to other open models.
– Gemma models outperform **significantly larger models** on key benchmarks while adhering to Google’s rigorous standards for safe and responsible outputs.
– Gemma models achieved state-of-the-art results across several key capabilities, such as language understanding, generation, and instruction following.
– Gemma models surpassed popular open models like **Llama-2** and **Mistral** on language understanding and generation performance.
Gemma models are freely available for individual developers, researchers, and commercial users for access and redistribution.