metadata
dataset_info:
features:
- name: audio
dtype:
audio:
sampling_rate: 16000
- name: text
dtype: string
- name: speaker_id
dtype: string
- name: corpus
dtype: string
- name: gender
dtype: string
- name: condition
dtype: string
- name: diagnosis
dtype: string
- name: severity
dtype: string
- name: intelligibility
dtype: string
- name: duration
dtype: float64
splits:
- name: train
num_bytes: 3896503197
num_examples: 38741
- name: test
num_bytes: 150891887
num_examples: 1500
download_size: 3876813372
dataset_size: 4047395084
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*