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: 18131497.022292994
num_examples: 135
download_size: 18027741
dataset_size: 18131497.022292994
configs:
- config_name: default
data_files:
- split: train
path: data/train-*