Gemini Nano Now Available On Android For Experimentation Via AI Edge SDK Via AICore
Google officially launched its AI Gemini models- Gemini Ultra, Gemini Pro, and Gemini Nano. Among these Gemini Nano was said to be the most efficient model for on-device tasks. Gemini Nano is now available for experimentation to all Android Developers with the AI Edge SDK via AICore.
Developers will initially have access to experiment with text-to-text prompts on Pixel 9 series devices. As per the Android Developers blog, support for more devices and modalities will be added in the future. This experimental access is for development purposes only and not for productional usage at this time.
In the blog comparison between Nano 1 and Nano 2 is also shown that per which Nano 2 delivers a substantial improvement in quality while it excels in both academic benchmarks and real-world applications.
Gemini Nanos text and image are already leveraged by Pixel Screenshots, Talkback, Recorder, and many more. To run an inference with Gemini Nano in AI core, you use AI Edge SDK. To experiment with AI Edge SDK, add the following to your app dependency:
implementation(“com.google.ai.edge.aicore:aicore:0.0.1-exp01”)
AI Edge SDK allows you to customize inference parameters. Some of the commonly used parameters include- Temperature, Top K, Candidate count, and Max output tokens. Gemini Nano can be tried in use cases like Rephrasing, Smart Reply, Proofreading, and Summarization.
You can check out the below video to learn more details.
