metadata
dataset_info:
features:
- name: audio
dtype:
audio:
sampling_rate: 16000
- name: transcription
dtype: string
- name: duration
dtype: int32
splits:
- name: train
num_bytes: 34985912058.5
num_examples: 69519
- name: validation
num_bytes: 5032132659.544
num_examples: 11304
- name: test_bracarense
num_bytes: 808120688.088
num_examples: 1177
download_size: 30979292782
dataset_size: 40826165406.131996
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*
- split: test_bracarense
path: data/test_bracarense-*