metadata
dataset_info:
features:
- name: audio_filename
dtype: string
- name: prompt
dtype: string
- name: transcription
dtype: string
splits:
- name: train
num_bytes: 135971289
num_examples: 645979
- name: test
num_bytes: 8461
num_examples: 40
download_size: 58396583
dataset_size: 135979750
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*