metadata
dataset_info:
features:
- name: audio
dtype:
audio:
sampling_rate: 16000
- name: text
dtype: string
- name: text_normalised
dtype: string
- name: phonemes
dtype: string
- name: gender
dtype: string
- name: speaker_id
dtype: int64
- name: language
dtype: string
splits:
- name: train
num_bytes: 4214232432
num_examples: 2725
download_size: 3770791962
dataset_size: 4214232432
configs:
- config_name: default
data_files:
- split: train
path: data/train-*