metadata
dataset_info:
features:
- name: audio
dtype: audio
- name: video_id
dtype: string
- name: segment_id
dtype: int64
- name: title
dtype: string
- name: transcription
dtype: string
- name: youtube_url
dtype: string
splits:
- name: train
num_bytes: 15765226460.953596
num_examples: 113204
- name: test
num_bytes: 1957249220.5622025
num_examples: 14151
- name: val
num_bytes: 1977921383.4392023
num_examples: 14151
download_size: 19638509254
dataset_size: 19700397064.955
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*
- split: val
path: data/val-*