Get started with Flower

Welcome to the Flower collaborative AI tutorial!

In this tutorial, you’ll learn how to run a Flower App directly on SuperGrid. You’ll use an existing app from Flower Hub. This means you won’t have to write any code to complete the tutorial, but you’ll still get a practical introduction to the core concepts of collaborative AI with Flower. You’ll also get a look behind the scenes to understand how Flower executes your app and how the different components interact with each other.

팁

Star Flower on GitHub ⭐️ and join the Flower community on Flower Discuss or Flower Slack to introduce yourself, ask questions, and get help.

Let’s get started! 🌼

Access SuperGrid

The only prerequisite for this tutorial is to have an account for SuperGrid. Sign up on flower.ai or log in if you already have an account.

Run a Flower App

Once you have logged into flower.ai, you are ready to start running Flower Apps. A Flower App is a Python project that defines the logic of your collaborative AI workload. More details will be covered in the next tutorial, but for now you can think of a Flower App as the code that defines what happens on the server (SuperGrid) and SuperNodes during your collaborative AI workload.

Flower Hub is a collection of open-source Flower Apps that you can use and learn from. In this tutorial, you’ll use the @flwrlabs/demo app. It is maintained by the Flower team and runs a simple collaborative AI workload in which the server aggregates NumPy arrays sent by the clients.

When you click the 🚀 Run button, you’ll be asked which federation to run the app on. Select @<your-account>/workspace (this is a federation created automatically by SuperGrid under your profile), then click Run app.

Shortly after launching your app, your run will start and you will see the run details page. This page shows you the progress of your run and gives you insights into the execution of your app.

If you click the Logs tab, you’ll see the logs from the app execution. This demo app ran for three rounds and sampled all two SuperNodes in each round. In each round, the server aggregated different metrics received from the clients. You can see these in the logs under the 'random_metric' key. At the end, the logs show the aggregated results from all rounds.

Final remarks

Congratulations, you have successfully run your first Flower App on SuperGrid! You’ve taken your first step into the world of collaborative AI with Flower. In this tutorial, you created a federation with simulated SuperNodes and ran an existing Flower App across them. You also explored the SuperGrid dashboard to monitor your app’s progress and view its logs.

If you return to the Federations page and click your federation, you’ll see the run you launched. You can click that run to open the same run details page you saw right after launching the app.

In the next tutorial, you’ll download an existing Flower App, run it from your local machine on SuperGrid, make a small customization, and learn how the main Flower App components fit together.

다음 단계

Before you continue, make sure to join the Flower community on Flower Discuss (Join Flower Discuss) and on Slack (Join Slack).

There’s a dedicated #questions Slack channel if you need help, but we’d also love to hear who you are in #introductions!

The Flower Collaborative AI Tutorial - Part 2: Write your first Flower App will walk you through the process of customizing your first Flower App and understanding how components like ServerApp and ClientApp interact with each other.