I am trying to read my data files from an Azure ML dataset. My code is as follows:
from azureml.core import Dataset
dataset = Dataset.get_by_name(aml_workspace, "mydatasetname")
dataset_mount = dataset.as_named_input("mydatasetname").as_mount(path_on_compute="dataset")
The type of dataset_mount is class DatasetConsumptionConfig. How do I get the actual directory path from that class? I can do it in a very complicated manner by passing the dataset_mount into a script as follows:
PythonScriptStep(script_name="myscript.py", arguments=["--dataset_mount", dataset_mount], ...)
Then, when that script step is run, "myscript.py" mysteriously gets the real directory path of the data in the argument "--dataset_mount", instead of it being DatasetConsumptionConfig. So, DatasetConsumptionConfig somehow gets converted into directory path under the hoods. However, that's an overcomplicated and strange approach to get the thing done. Is there any direct way to get the data path from DatasetConsumptionConfig? Or maybe I have misunderstood something here?
