metadata
dataset_info:
features:
- name: audio_path
dtype: audio
- name: transcription
dtype: string
- name: length
dtype: float64
splits:
- name: train
num_bytes: 3368635673.864
num_examples: 8002
- name: validation
num_bytes: 892755867.838
num_examples: 2282
- name: test
num_bytes: 509433461.506
num_examples: 1242
download_size: 4884161437
dataset_size: 4770825003.207999
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*
- split: test
path: data/test-*