metadata
dataset_info:
features:
- name: uuid
dtype: string
- name: audio
dtype: string
- name: transcription
dtype: string
- name: translation
dtype: string
- name: cer
dtype: float64
- name: speaker
dtype: string
splits:
- name: train
num_bytes: 586137
num_examples: 1253
download_size: 267734
dataset_size: 586137
configs:
- config_name: default
data_files:
- split: train
path: data/train-*