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: 54580702
      num_examples: 8757
  download_size: 32173218
  dataset_size: 54580702
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   8729.000000
mean    1447.751289
std     2578.411960
min       40.000000
25%      446.000000
50%      666.000000
75%     1172.000000
max    59143.000000

Total count: 12.64 M tokens