---
dataset_info:
  features:
  - name: body
    dtype: string
  - name: comment_score
    dtype: int64
  - name: prompt
    dtype: string
  - name: post_score
    dtype: int64
  splits:
  - name: train
    num_bytes: 244506795.8945573
    num_examples: 66332
  download_size: 168000074
  dataset_size: 244506795.8945573
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
license: mit
---

Data from real humans, courtesy of https://reddit.com/r/WritingPrompts