Federated Learning Tensorflow tutorial

Last updated: Oct 09, 2024

This tutorial demonstrates the usage of Federated Learning with the goal of training a machine learning model with data from different users without having users share their data. The steps are done in a low code environment with the UI and with a Tensorflow framework.

Note:

This is a step-by-step tutorial for running a UI driven Federated Learning experiment. To see a code sample for an API driven approach, see Federated Learning Tensorflow samples.

Tip:

In this tutorial, admin refers to the user that starts the Federated Learning experiment, and party refers to one or more users who send their model results after the experiment is started by the admin. While the tutorial can be done by the admin and multiple parties, a single user can also complete a full runthrough as both the admin and the party. For a simpler demonstrative purpose, in the following tutorial only one data set is submitted by one party. For more information on the admin and party, see Terminology.

Watch this short video tutorial of how to create a Federated Learning experiment with Watson Studio.

Video disclaimer: Some minor steps and graphical elements in this video might differ from your platform.

This video provides a visual method to learn the concepts and tasks in this documentation.

In this tutorial you will learn to:

Step 1: Start Federated Learning as the admin
Step 2: Train model as a party
Step 3: Save and deploy the model online

Step 1: Start Federated Learning as the admin

In this tutorial, you train a Federated Learning experiment with a Tensorflow framework and the MNIST data set.

Before you begin

Log in to IBM Cloud. If you don't have an account, create one with any email.
Create a Watson Machine Learning service instance if you do not have it set up in your environment.
Log in to watsonx.
Use an existing project or create a new one. You must have at least admin permission.
Associate the Watson Machine Learning service with your project.
1. In your project, click the Manage > Service & integrations.
2. Click Associate service.
3. Select your Watson Machine Learning instance from the list, and click Associate; or click New service if you do not have one to set up an instance.

Start the aggregator

Create the Federated learning experiment asset:
1. Click the Assets tab in your project.
2. Click New task > Train models on distributed data.
3. Type a Name for your experiment and optionally a description.
4. Verify the associated Watson Machine Learning instance under Select a machine learning instance. If you don't see a Watson Machine Learning instance associated, follow these steps:
  1. Click Associate a Machine Learning Service Instance.
  2. Select an existing instance and click Associate, or create a New service.
  3. Click Reload to see the associated service.
  4. Click Next.
Configure the experiment.
1. On the Configure page, select a Hardware specification.
2. Under the Machine learning framework dropdown, select Tensorflow 2.
3. Select a Model type.
4. Download the untrained model.
5. Back in the Federated Learning experiment, click Select under Model specification.
6. Drag the downloaded file named tf_mnist_model.zip onto the Upload file box.1. Select runtime-22.2-py3.10 for the Software Specification dropdown.
7. Give your model a name, and then click Add.
8. Click Weighted average for the Fusion method, and click Next.
Define the hyperparameters.
1. Accept the default hyperparameters or adjust as needed.
2. When you are finished, click Next.
Select remote training systems.
1. Click Add new systems.
1. Give your Remote Training System a name.
2. Under Allowed identities, choose the user that is your party, and then click Add. In this tutorial, you can add a dummy user or yourself, for demonstrative purposes.
  This user must be added to your project as a collaborator with Editor or higher permissions. Add additional systems by repeating this step for each remote party you intent to use.
3. When you are finished, click Add systems.
4. Return to the Select remote training systems page, verify that your system is selected, and then click Next.
Review your settings, and then click Create.
Watch the status. Your Federated Learning experiment status is Pending when it starts. When your experiment is ready for parties to connect, the status will change to Setup – Waiting for remote systems. This may take a few minutes.
Click View setup information to download the party configuration and the party connector script that can be run on the remote party.
Click the download icon besides each of the remote training systems that you created, and then click Party connector script. This gives you the party connector script. Save the script to a directory on your machine.

Step 2: Train model as the party

Follow these steps to train the model as a party:

Ensure that you are using the same Python version as the admin. Using a different Python version might cause compatibility issues. To see Python versions compatible with different frameworks, see Frameworks and Python version compatibility.
Create a new local directory, and put your party connector script in it.
Download the data handler mnist_keras_data_handler.py by right-clicking on it and click Save link as. Save it to the same directory as the party connector script.
Download the MNIST handwriting data set from our Samples. In the the same directory as the party connector script, data handler, and the rest of your files, unzip it by running the unzip command unzip MNIST-pkl.zip.
Install Watson Machine Learning.
- If you are using Linux, run pip install 'ibm-watson-machine-learning[fl-rt22.2-py3.10]'.
- If you are using Mac OS with M-series CPU and Conda, download the installation script and then run ./install_fl_rt22.2_macos.sh <name for new conda environment>.
  You now have the party connector script, mnist_keras_data_handler.py, mnist-keras-test.pkl and mnist-keras-train.pkl, data handler in the same directory.

Your party connector script looks similar to the following. Edit it by filling in the data file locations, the data handler, and API key for the user defined in the remote training system. To get your API key, go to Manage > Access(IAM) > API keys in your IBM Cloud account. If you don't have one, click Create API key, fill out the fields, and click Create.

from ibm_watson_machine_learning import APIClient
wml_credentials = {
    "url": "https://us-south.ml.cloud.ibm.com",
    "apikey": "<API KEY>"
    }
wml_client = APIClient(wml_credentials)
wml_client.set.default_project("XXX-XXX-XXX-XXX-XXX")
party_metadata = {
            wml_client.remote_training_systems.ConfigurationMetaNames.DATA_HANDLER: {
            # Supply the name of the data handler class and path to it.
            # The info section may be used to pass information to the
            # data handler.
            # For example,
            #     "name": "MnistSklearnDataHandler",
            #     "path": "example.mnist_sklearn_data_handler",
            #     "info": {
            #         "train_file": pwd + "/mnist-keras-train.pkl",
            #         "test_file": pwd + "/mnist-keras-test.pkl"
            #     }
                "name": "<data handler>",
                "path": "<path to data handler>",
                    "info": {
                        "<information to pass to data handler>"
                    }
                }
            }
party = wml_client.remote_training_systems.create_party("XXX-XXX-XXX-XXX-XXX", party_metadata)
party.monitor_logs()
party.run(aggregator_id="XXX-XXX-XXX-XXX-XXX", asynchronous=False)

Run the party connector script: python3 rts_<RTS Name>_<RTS ID>.py.
From the UI you can monitor the status of your Federated Learning experiment.

Step 3: Save and deploy the model online

In this section, you will learn to save and deploy the model that you trained.

Save your model.
1. In your completed Federated Learning experiment, click Save model to project.
2. Give your model a name and click Save.
3. Go to your project home.
Create a deployment space, if you don't have one.
1. From the navigation menu , click Deployments.
2. Click New deployment space.
3. Fill in the fields, and click Create.
Promote the model to a space.
1. Return to your project, and click the Assets tab.
2. In the Models section, click the model to view its details page.
3. Click Promote to space.
4. Choose a deployment space for your trained model.
5. Select the Go to the model in the space after promoting it option.
6. Click Promote.
When the model displays inside the deployment space, click New deployment.
1. Select Online as the Deployment type.
2. Specify a name for the deployment.
3. Click Create.
Click the Deployments tab to monitor your model's deployment status.

Next steps

Ready to create your own customized Federated Experiment? See the high level steps in Creating your Federated Learning experiment.

Parent topic: Federated Learning tutorial and samples