metadata

language:
  - en
license: odc-by
size_categories:
  - 1K<n<10K
task_categories:
  - text-generation
dataset_info:
  features:
    - name: relative_path
      dtype: string
    - name: section
      dtype: string
    - name: filename
      dtype: string
    - name: text
      dtype: string
  splits:
    - name: train
      num_bytes: 34731571
      num_examples: 5412
  download_size: 20355508
  dataset_size: 34731571
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*

The Sarcasm Scrolls

Paragraph-segmented audio transcriptions of various satirical media.

token count

GPT-4 tiktoken token count:

        token_count
count   4937.000000
mean    1495.166295
std     2787.318839
min       44.000000
25%      419.000000
50%      599.000000
75%      955.000000
max    59143.000000

Total count: 7.38 M tokens