Configuring AWS with Auger
Auger is an automated machine learning platform that allows you to connect to your favorite cloud service. Auger is Kubernetes native, so it is designed from the ground up to scale with your needs. This is a very powerful way to train thousands of models in parallel to obtain an ordered leaderboard of the best performing models. You can then deploy your model of choice as a production endpoint for real time use in a very cost effective manner.
This post will focus on provisioning your Amazon Web Services account to work with Auger. This involves creating an AWS IAM Role with the necessary permissions for Auger to use. The purpose of creating a role is so you can scale your clusters through Auger’s interface, starting and stopping any number of configurations with the click of a button.
Auger follows AWS best practices to safely and securely integrate with your account.
Lets Get Started
I am assuming you already have an AWS account, but if you do not you will want to sign up first. Once signed in, go to the IAM Roles page. IAM Roles are a secure way to grant permissions to entities you trust. In this case, Auger. Click Create role.
Select Another AWS account from the top and enter the Account ID 566476274242 for Auger. Click the button Next: Permissions on the bottom right to continue.
On the next screen do not select a policy yet. We will add one later. Click Next: Review to continue.
Enter a role name such as auger-hub-role. Click Create role. A new role will be created and you will return to the Roles list.
You must use one role per Auger organization, so if you use a single AWS account for multiple organizations, you should add an identifying namespace to the end of the name. For example, you might have auger-hub-role-org1 and auger-hub-role-org2.
Click the name of the role you just created. Make sure to click the actual name of the role so you go into the detail page.
On the Permissions tab click Add inline policy. Its the link on the right with the plus icon.
From here click on the JSON tab and paste the policy document obtained in the AWS onboarding step of Auger. Click Review policy.
The policy reviewer will report syntax errors. You will need to give the policy a name such as auger-hub-policy. Click Create policy.
You will now be taken to the Role details page that you just created. Below the Summary title copy the Role ARN link to your clipboard.
You will then paste this ARN Role into the Auger AWS ARN Role form field.
Congratulations! You now can run Auger’s powerful AutoML platform on AWS.
Conclusion
You have just configured your Auger account to run on AWS! Next, you will want to create a project and deploy a cluster. If you have a prepared CSV, or Postgres table with tabular data you can begin training right away.
