Announcing Build with Google AI release 3: A Season of Gemma!

OCT 02, 2024
Joe Fernandez AI Developer Relations

As seasons change, technology evolves, bringing us new, exciting ways to solve problems and create solutions with artificial intelligence (AI). Google's Gemma family of open generative AI models has been available less than a year, and we are seeing incredible engagement with this technology from the developer and research communities. It's definitely time for us to dig into this Google AI product offering and show you how you can put it to work.

Gemma models allow organizations and individuals to create completely custom, self-managed AI solutions without having to build generative AI models from scratch. These downloadable models are provided with open weights, meaning you can fine-tune them to solve specific tasks and achieve new capabilities. These models offer a great way for developers to create independent AI solutions and manage them on their own hardware, or host them on their preferred cloud service.

Today, we are launching Season 3 of the Build with Google AI series, featuring projects built with Google's Gemma open model technology. As always, our goal with this series is to provide you with practical applications of AI, and help you put AI capabilities to work for you and your organization.

Link to Youtube Video (visible only when JS is disabled)

This season we are showing off some new applications and bringing back a developer favorite: code generation. We've upgraded the Pipet Code Agent to work with Gemma to let you create a truly personal AI code assistant. We also explore some new applications, including using Gemma for non-English tasks, and a project for turning business emails into actionable, structured data. Here's the line up:

  • Native Language AI Assistant with Gemma (10/2) - Artificial intelligence is more useful when you can use it in your native language. We'll show you how to tune a Gemma model to accomplish tasks in Korean, or any spoken language used by your organization.

  • Personal AI Code Assistant with Gemma (10/8) - Would you like your own code generation AI assistant running on your own hardware? We show you how to set up your own Gemma instance for coding assistance, complete with a customizable Visual Studio Code extension!

  • Business Inquiry AI Assistant with Gemma (10/15) - Handling emails and inquiries is a critical task for many businesses, but it takes a lot of time and effort. We show you how to use a Gemma model to help process customer emails, and how to tune Gemma to work with your business and automation systems.

All these projects are provided with open source code and are designed to let you easily build on and extend them for your own needs. They are simple on purpose: so you can use them as a stepping stone toward your own application of AI. What's more, we provide them with guidance and advice from Google AI technology experts.


Learn from the developers

The Build with Google AI series is not just about code, it's about learning from developers who've mapped the coding road ahead and want to share what they've learned. We'll be back to the studio to talk with coders who built these projects so they can share what they learned along the way. What's the best way to tune a Gemma model? Do you always need to tune a Gemma model to get it to work for you? How can you use Gemma in your language? Get insights into coding with AI to jump start your own development project.


Get even more Gemma developer content

The Gemma model family has a full set of developer documentation where you can learn more about the official variants of Gemma and how to build with these open models. You can also find the code for this season of Build with Google AI in the Gemma cookbook code repository, along with many other useful Gemma projects.


We are excited to bring you the third season of Build with Google AI! Check out Season 3 right now! Also check out Season 1 and Season 2 if you missed them. Use those video comments to let us know what you think and tell us what you'd like to see in future episodes.


Keep learning! Keep building!