Lab 2 - Data

Download the dataset

1
# Create a directory
2
mkdir data
3
cd data
4
5
# Download the data and validation set
6
wget https://github.com/hnky/dataset-lego-figures/raw/master/_download/train-and-validate.zip
7
8
# Unzip the dataset
9
unzip train-and-validate.zip
10
11
# remove the zip file
12
rm train-and-validate.zip
13
14
# Go back to the root of your project
15
cd ..
Copied!

Create a dataset in your Azure Machine Learning workspace

1
code data.yml
Copied!
Add this content to the file
1
$schema: https://azuremlschemas.azureedge.net/latest/dataset.schema.json
2
name: LegoSimpsons
3
version: 1
4
datastore: azureml:workspaceblobstore
5
local_path: ./data
Copied!
Now run the CLI command to upload the data to your default datastore and create the dataset
1
az ml data create -f data.yml
Copied!
To see if the dataset is created you can list all the datasets in your workspace with the command below.
1
az ml data list --output table
Copied!

Checklist

Now you have versioned dataset of Simpson Images in your Azure Machine Learning Workspace
  • Downloaded the dataset
  • Unzipped the dataset
  • Create a dataset configuration file in YAML
  • Used the CLI to create the dataset in Azure Machine Learning