metadata
dataset_info:
features:
- name: transcription
dtype: string
- name: mel_spectrogram
dtype: string
splits:
- name: train
num_bytes: 2490065755
num_examples: 2602
download_size: 1959924391
dataset_size: 2490065755
configs:
- config_name: default
data_files:
- split: train
path: data/train-*