Data/phonemized_dataset/dataset_dict.json

{"splits": ["train", "test", "validation"]}