metadata
dataset_info:
features:
- name: audio
dtype:
audio:
sampling_rate: 16000
- name: transcription
dtype: string
- name: dialect
dtype:
class_label:
names:
'0': adamawa
'1': borgu
'2': bororro
'3': caka
'4': liptako
'5': maacina
'6': pulaar
'7': pular
splits:
- name: train
num_bytes: 11309076022.27005
num_examples: 46550
- name: validation
num_bytes: 297607255.1689487
num_examples: 1225
- name: test
num_bytes: 297850034.3570458
num_examples: 1226
download_size: 11738974484
dataset_size: 11904533311.796045
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*
- split: test
path: data/test-*