Stage datasets

Once you find a suitable dataset you can stage it for training your models.

CLI

You can stage a dataset with following CLI command:

eotdl datasets get "dataset name"

Your datasets will be staged to a default folder, but you can specify a different folder with the --path option or the EOTDL_DOWNLOAD_PATH environment variable. For example, to stage the dataset to the current directory:

eotdl datasets get "dataset name" --path .

In order to overwrite a dataset that you already staged, you can use the --force option.

eotdl datasets get "dataset name" --force

If you know the specific version of the dataset to stage, use the --version option.

eotdl datasets get "dataset name" --version 1

Library

You can stage datasets using the following Python code:

from eotdl.datasets import download_dataset

download_dataset("dataset-name")

And use the same options seen before.

download_dataset("dataset-name", force=True, path="data", version=1)