metadata
dataset_info:
features:
- name: audio
dtype:
audio:
sampling_rate: 16000
- name: transcription
dtype: string
- name: w2v2_transcription
dtype: string
- name: WER
dtype: int64
splits:
- name: train
num_bytes: 89394264
num_examples: 1077
download_size: 89354930
dataset_size: 89394264