Run Your First Experiment in HPE Machine Learning Development Environment¶

In this tutorial, we’ll show you how to integrate a training example with the HPE Machine Learning Development Environment environment. We’ll run our experiment on a local training environment requiring only a single CPU or GPU.

Note

This tutorial is recommended as an introduction for model developers who are new to HPE.

Objective

Our goal is to integrate the PyTorch MNIST training example into HPE Machine Learning Development Environment in four steps:

Download and extract the files
Set up our training environment
Run the experiment
View the experiment in our browser

Prerequisites

Installation Requirements

Download the Files¶

To get started, we’ll first download and extract the files we need and cd into the directory.

Download the mnist_pytorch.tgz file.
Open a terminal window, extract the files, and cd into the mnist_pytorch directory:

tar xzvf mnist_pytorch.tgz
cd mnist_pytorch

Set Up Your Training Environment¶

To start your experiment, you’ll need an HPE Machine Learning Development Environment cluster. If you are new to HPE (HPE Machine Learning Development Environment), you can install the HPE Machine Learning Development Environment library and start a cluster locally:

pip install determined

# If your machine has GPUs:
det deploy local cluster-up

# If your machine does not have GPUs:
det deploy local cluster-up --no-gpu

Run the Experiment¶

To run the experiment, enter the following command:

det experiment create const.yaml . -f

A notification displays letting you know the experiment has started.

Preparing files (.../mnist_pytorch) to send to master...
Created experiment xxx

View the Experiment¶

To view the experiment progress in your browser:

Enter the following URL: http://localhost:8080/.

This is the cluster address for your local training environment.

Accept the default determined username, leave the password empty, and click Sign In.

Next Steps¶

In four simple steps, we’ve successfully configured our training environment in HPE Machine Learning Development Environment to start training the PyTorch MNIST example.

In this article, we learned how to run an experiment on a local, single CPU or GPU. To learn how to change your configuration settings, including how to run a distributed training job on multiple GPUs, visit the Quickstart for Model Developers.