metadata
dataset_info:
features:
- name: audio_filename
dtype: string
- name: audio
dtype: audio
- name: transcription
dtype: string
splits:
- name: train
num_bytes: 589058948.375
num_examples: 3565
- name: test
num_bytes: 164333095.732
num_examples: 1018
- name: validation
num_bytes: 85359992
num_examples: 510
download_size: 831620409
dataset_size: 838752036.107
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*
- split: validation
path: data/validation-*