metadata
dataset_info:
features:
- name: client_id
dtype: string
- name: path
dtype: string
- name: audio
dtype:
audio:
sampling_rate: 16000
- name: transcription
dtype: string
- name: up_votes
dtype: int64
- name: down_votes
dtype: int64
- name: age
dtype: string
- name: gender
dtype: string
- name: accent
dtype: string
- name: locale
dtype: string
- name: segment
dtype: string
- name: variant
dtype: string
- name: in_spectrogram
sequence:
sequence: float32
splits:
- name: train
num_bytes: 6287282248
num_examples: 20000
- name: val
num_bytes: 629444984
num_examples: 2000
download_size: 6825396882
dataset_size: 6916727232
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: val
path: data/val-*