metadata
dataset_info:
features:
- name: file_name
dtype: string
- name: transcription
dtype: string
- name: transcription_normalised
dtype: string
- name: audio
dtype: audio
splits:
- name: train
num_bytes: 4983072167.73
num_examples: 20978
download_size: 3741291896
dataset_size: 4983072167.73
configs:
- config_name: default
data_files:
- split: train
path: data/train-*