metadata
dataset_info:
features:
- name: chunk_id
dtype: string
- name: source
dtype: string
- name: source_detail
dtype: string
- name: audio
dtype:
audio:
sampling_rate: 16000
- name: sentence
dtype: string
- name: asr_transcription
dtype: string
- name: duration
dtype: float64
- name: quality
dtype: string
- name: wer
dtype: float64
splits:
- name: train
num_bytes: 1245659055
num_examples: 8386
download_size: 1287947444
dataset_size: 1245659055
configs:
- config_name: default
data_files:
- split: train
path: data/train-*