metadata
dataset_info:
features:
- name: audio_filename
dtype: string
- name: audio
dtype: audio
- name: transcription
dtype: string
splits:
- name: train
num_bytes: 31534340552.75
num_examples: 382258
- name: test
num_bytes: 1296484841.824
num_examples: 12676
- name: validation
num_bytes: 681585733.706
num_examples: 7522
download_size: 32930111240
dataset_size: 33512411128.280003
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*
- split: validation
path: data/validation-*