metadata
dataset_info:
features:
- name: path
dtype: string
- name: audio
dtype: audio
- name: speaker
dtype: string
- name: l1
dtype:
class_label:
names:
'0': Arabic
'1': Hindi
'2': Korean
'3': Mandarin
'4': Spanish
'5': Vietnamese
- name: gender
dtype:
class_label:
names:
'0': F
'1': M
- name: transcript
dtype: string
- name: duration
dtype: float32
- name: file_id
dtype: string
- name: sample_rate
dtype: int32
splits:
- name: train
num_bytes: 706831046
num_examples: 2250
download_size: 713322343
dataset_size: 706831046
configs:
- config_name: default
data_files:
- split: train
path: data/train-*