metadata
dataset_info:
features:
- name: sid
dtype: string
- name: audio_id
dtype: string
- name: text_id
dtype: string
- name: audio
dtype:
audio:
sampling_rate: 24000
- name: text
dtype: string
- name: metadata
struct:
- name: provider
dtype: string
- name: model
dtype: string
- name: voice
dtype: string
- name: lang
dtype: string
- name: sample_rate
dtype: int32
- name: duration
dtype: float32
- name: codes
list: int64
splits:
- name: train
num_bytes: 3439950032
num_examples: 8430
download_size: 3379928537
dataset_size: 3439950032
configs:
- config_name: default
data_files:
- split: train
path: data/train-*