metadata
dataset_info:
features:
- name: file_name
dtype: string
- name: Label
dtype: string
- name: Transcript
dtype: string
- name: text
dtype: string
- name: CanonicalOld
dtype: string
- name: Canonical
dtype: string
- name: Error
dtype: string
- name: audio
dtype:
audio:
sampling_rate: 16000
splits:
- name: train
num_bytes: 358268422.676
num_examples: 3606
- name: test
num_bytes: 67683633
num_examples: 669
- name: val
num_bytes: 65061419
num_examples: 668
download_size: 462177552
dataset_size: 491013474.676
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*
- split: val
path: data/val-*