---
dataset_info:
  features:
  - name: text
    dtype: string
  - name: name
    dtype:
      audio:
        sampling_rate: 16000
  splits:
  - name: train
    num_bytes: 25162204991.683994
    num_examples: 34548
  - name: test
    num_bytes: 6240428802.232999
    num_examples: 8637
  download_size: 27399432987
  dataset_size: 31402633793.916992
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: test
    path: data/test-*
---

Convert [YouTube playlists](https://youtube.com/playlist?list=PLh0GSN2S41Yy2-xR3SwdI23x0KvSMcYme&si=r_7ujHwSuxzVVme_) to speech-to-text datasets