metadata
dataset_info:
features:
- name: file_name
dtype: string
- name: transcription
dtype: string
- name: transcription_normalised
dtype: string
- name: audio
dtype: audio
- name: description
dtype: string
splits:
- name: train
num_bytes: 19337540605
num_examples: 76000
- name: test
num_bytes: 1017765263
num_examples: 4000
download_size: 20092004186
dataset_size: 20355305868
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*