import datasets | |
ds = datasets.load_dataset('./dataset_code.py', 'notebook_defaults') | |
ds_real = ds['everything'] # there is no such thing as a train/test split here | |
one_item = ds_real[0] # grab first story, and truncuate the text of it to first 1000 characters | |
one_item_truncuated = one_item | {'story': one_item['story'][:1000]} | |
print(ds) | |
print(one_item_truncuated) | |