metadata
dataset_info:
features:
- name: audio
dtype: audio
- name: 'Unnamed: 0'
dtype: int64
- name: transcription
dtype: string
- name: speaker_id
dtype: int64
splits:
- name: train
num_bytes: 3596434582.17
num_examples: 6765
download_size: 3089965519
dataset_size: 3596434582.17
configs:
- config_name: default
data_files:
- split: train
path: data/train-*