metadata
dataset_info:
features:
- name: audio
dtype: audio
- name: text
dtype: string
- name: noisy_speech
sequence: float64
- name: noise
dtype: string
splits:
- name: train
num_bytes: 1132681520.176
num_examples: 3168
- name: test
num_bytes: 176892283
num_examples: 459
download_size: 1042211351
dataset_size: 1309573803.176
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*