metadata
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: valid
path: data/valid-*
dataset_info:
features:
- name: audio
dtype:
audio:
sampling_rate: 16000
- name: text
dtype: string
- name: normalized
dtype: string
- name: unsanitized_verbatim
dtype: string
- name: unsanitized_normalized
dtype: string
splits:
- name: train
num_bytes: 40057912984
num_examples: 383004
- name: valid
num_bytes: 420723806
num_examples: 4740
download_size: 40286918872
dataset_size: 40478636790