metadata
dataset_info:
features:
- name: audio
dtype: audio
- name: file
dtype: string
- name: transcription
dtype: string
- name: speaker
dtype: string
- name: gender
dtype: string
- name: book
dtype: string
- name: is_gold_transcript
dtype: bool
splits:
- name: train
num_bytes: 5156178186.59
num_examples: 31535
download_size: 4993612385
dataset_size: 5156178186.59
configs:
- config_name: default
data_files:
- split: train
path: data/train-*