[!Note] This repository corresponds to Gemma models. You can try it out with: Google AI Edge Gallery for Android through Open Beta in the Play Store Google AI Edge Gallery for Android through GitHub Google AI Studio Google AI Edge for Web The latest Gemma 3n E2B and E4B model checkpoints support multimodal inputs, including text, vision, and audio. To test the capabilities, install Google AI Edge Gallery from the Play Store or download the app from GitHub. You can then access download the model file and navigate to the 'Ask Image' or 'Audio Scribe' for interactive demonstrations. Gemma 3n models have a novel architecture that allows them to run with a smaller number of effective parameters. They also have a Matformer architecture that allows nesting multiple models. Learn more about these techniques in the Gemma documentation. Gemma 3n model card Model Page : Gemma 3n Resources and Technical Documentation : LiteRT LM Github & Documentation Responsible Generative AI Toolkit Gemma on Kaggle Google AI Edge for Web Terms of Use : Terms\ Authors : Google DeepMind Performance Benchmarks with LiteRT LM These numbers will continue to improve while LiteRT LM is in preview. Model Device Back…
We use cookies for essential functionality and analytics. You can accept or reject analytics cookies.Cookie policy