metadata
license: apache-2.0
dataset_info:
features:
- name: original_id
dtype: string
- name: original_text
dtype: string
- name: original_speech
dtype: audio
- name: target_id
dtype: string
- name: target_text
dtype: string
- name: target_speech
dtype: audio
splits:
- name: train
num_bytes: 133366716594.308
num_examples: 504261
- name: validation
num_bytes: 2267891576.283
num_examples: 5143
download_size: 124755101537
dataset_size: 135634608170.591
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*