metadata
dataset_info:
features:
- name: id
dtype: string
- name: audio
dtype:
audio:
decode: false
- name: duration_ms
dtype: int32
- name: transcription
dtype: string
- name: transcription_null
dtype: string
- name: audio_null
dtype:
audio:
decode: false
splits:
- name: train
num_bytes: 2825587518.9
num_examples: 12300
download_size: 2751146385
dataset_size: 2825587518.9
configs:
- config_name: default
data_files:
- split: train
path: data/train-*