Datasets:

mrfakename
/

podcast_pile_1

dataset_info:
  features:
    - name: episode_title
      dtype: string
    - name: episode_url
      dtype: string
    - name: episode_duration
      dtype: int64
    - name: episode_description
      dtype: string
    - name: tag_title
      dtype: string
    - name: tag_language
      dtype: string
  splits:
    - name: train
      num_bytes: 18788195944
      num_examples: 17794044
  download_size: 7459263938
  dataset_size: 18788195944
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*

The big dataset

2: https://huggingface.co/datasets/mrfakename/podcast_pile_2