Datasets:
from datasets import load_dataset, Audio | |
dataset = load_dataset('json', data_files="train.jsonl") | |
print(dataset) | |
dataset_audio = dataset.cast_column( | |
"audio", Audio(sampling_rate=48_000, mono=True, decode=True) | |
) | |
print(dataset_audio) | |
train_set = dataset_audio['train'] | |
train_set.to_parquet("train.parquet") | |
print(train_set[1]) | |