metadata
dataset_info:
features:
- name: logits
sequence: float32
- name: mfcc
sequence:
sequence: float64
splits:
- name: train
num_bytes: 996254892
num_examples: 195651
download_size: 1014945287
dataset_size: 996254892
configs:
- config_name: default
data_files:
- split: train
path: data/train-*