metadata
dataset_info:
features:
- name: audio_id
dtype: string
- name: seed_transcript
dtype: string
- name: input
dtype: string
- name: transcription
dtype: string
- name: target
dtype: string
- name: dataset
dtype: string
- name: duration
dtype: float64
- name: audio_unit
sequence: int64
splits:
- name: train
num_bytes: 167131361
num_examples: 124088
download_size: 57600379
dataset_size: 167131361
configs:
- config_name: default
data_files:
- split: train
path: data/train-*