Downloading custom dataset for googles custom model maker - TensorFlow

I am following a tutorial on training your own custom dataset and I am using this Collab tutorial to do so.

Things go smoothly until I get to step 3 of 'Train the object Detection model'. I get this error

ValueError: The size of the train_data (0) couldn't be smaller than batch_size (2). To solve this problem, set the batch_size smaller or increase the size of the train_data.

This leads me to believe my train_data is empty which I am confused about. Am I downloading and loading the dataset correctly?

This is how I download the dataset

!git clone https://github.com/priedejm/shirts.git
!unzip -q shirts/shirts.zip

This is how I load the dataset - I am not sure why shirts/train is there twice but thats how the original does it

train_data = object_detector.DataLoader.from_pascal_voc(
    'shirts/train',
    'shirts/train',
    ['shirt', 'pants']
)

val_data = object_detector.DataLoader.from_pascal_voc(
    'shirts/validate',
    'shirts/validate',
    ['shirt', 'pants']
)

And this is how the tutorial does it

!wget https://storage.googleapis.com/download.tensorflow.org/data/android_figurine.zip
!unzip -q android_figurine.zip

train_data = object_detector.DataLoader.from_pascal_voc(
    'android_figurine/train',
    'android_figurine/train',
    ['android', 'pig_android']
)

val_data = object_detector.DataLoader.from_pascal_voc(
    'android_figurine/validate',
    'android_figurine/validate',
    ['android', 'pig_android']
)