Here we provide HuggingFace datasets for the public datasets and the Text serialization.
You can load and inspect these datasets with:
from datasets import load_from_disk
dataset = load_from_disk('/root/TabLLM/datasets_serialized/heart')
dataset['note'][0]
dataset['label'][0]