metadata
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