Concurrent for MLflow Free Service - Hello World¶

Get started with Concurrent for MLflow using this simple Hello World DAG

Here is a screen capture video:

Try it¶

Concurrent for MLflow Free Service UI is available at https://mlflowui.concurrent-ai.org/: click here (This will open a new tab). Perform the steps described below to create and run a simple Concurrent.

Click Concurrent tab¶

Click 'Create Concurrent' button¶

Give the parallel a name, helloworld-test1 in this example. Note that the Concurrent name has to be unique among all users in the Enterprise. This is necessary in order to allow for sharing of Concurrent. It is useful to attach your username to the Concurrent's name. After successful creation of the Concurrent, you will be taken to a page with a list of Concurrent that you have created or have access to. Click on the newly created Concurrent (helloworld-test1 in this example) and you will be taken to the page for the specific Concurrent. In this page, click on the Create Template button as shown below

Add Node to DAG¶

As you might already know, each step in the processing of a DAG (Directed Acyclic Graph) is termed a Node. In this example, we add a single node and configure it to be the helloworld MLproject in the Concurrent for MLflow git tree. Click on the Add Node button as shown below

Fill Node details¶

Here are the node details:

We have chosen a container with 256 Megabytes of RAM and 250 milli CPUs, i.e. one quarter of a virtual CPU

These are some details that you will need while trying this out:

Kubernetes Cluster: parallels-free
Cluster's Namespace: parallelsns
GIT Url: https://github.com/concurrent-ai/concurrent-examples.git
Project Path: helloworld

Click Add and then save the template. For this simple example, we will not be using any inputs or parameters. There are more sophisticated examples that will demonstrate these features.

Run DAG¶

Once you have saved the template using the Save Template button, you are ready to run the Concurrent. From the detail page for the Concurrent, press the Run Template button. You are presented with an opportunity to modify any of the parameters stored in the template. For this example, you do not need to modify anything - just press the Run Template button. Once the template run starts up, you will be taken to the MLflow experiments tab where you can follow the progress of the run.