metadata
dataset_info:
features:
- name: audio
struct:
- name: array
sequence: float64
- name: path
dtype: 'null'
- name: sampling_rate
dtype: int64
- name: sentence_norm
dtype: string
- name: wer
dtype: int64
- name: id
dtype: string
splits:
- name: train
num_bytes: 534713143
num_examples: 1000
download_size: 126390777
dataset_size: 534713143
configs:
- config_name: default
data_files:
- split: train
path: data/train-*