Hello,
i’m working on a zero-shot classifaction model/pipeline. I have a problem when loading my data into a datset_dict. It does not happen everytime but quite often.
The moment i run follwing code cell, the cell will stuck or remain in the load_dataset line and will not further proceed.
# Loading data into DatasetDict and preparing
features = Features({"nli_label": ClassLabel(num_classes=3, names=['entailment','neutral','contradiction']), "intent_description": Value("string"), "intent": Value("string"), "text": Value("string"), "zlabel": Value("string"), "intent_dn": Value("string")})
dataset_dict = load_dataset('json', data_files={'train': 'train_data.json', 'test': 'test_data.json'})
dataset_dict = dataset_dict.map(features.encode_example, features=features)
dataset_dict = dataset_dict.rename_column("nli_label", "labels")
dataset_dict = dataset_dict.rename_column("zlabel", "hypothesis")
dataset_dict = dataset_dict.rename_column("text", "premise")
Im working in a google colab env. Im not sure if the jupyter env is the root of the problem. Below you can see the output of the cell. In order to proceed, i have to restart the runtime and hope that it works (what happens sometimes)
