metadata
dataset_info:
features:
- name: speaker
dtype: string
- name: file
dtype: string
- name: text
dtype: string
- name: audio
dtype:
audio:
sampling_rate: 24000
- name: mel
sequence:
sequence:
sequence: float32
splits:
- name: aew
num_bytes: 333194279.5
num_examples: 1132
download_size: 326677670
dataset_size: 333194279.5
configs:
- config_name: default
data_files:
- split: aew
path: data/aew-*