metadata
dataset_info:
features:
- name: audio
dtype:
audio:
sampling_rate: 16000
- name: original_sentence
dtype: string
- name: preprocessed_sentence
dtype: string
splits:
- name: train
num_bytes: 11482140431.422
num_examples: 56427
download_size: 11661321581
dataset_size: 11482140431.422
configs:
- config_name: default
data_files:
- split: train
path: data/train-*