README.md

metadata

license: apache-2.0
task_categories:
  - text-classification
language:
  - en
tags:
  - prompt
  - nsfw
size_categories:
  - 10K<n<100K

Prompt samples for doing text moderation

The label columns are auto-generated instead of done by human, use with care.

Classifications

• safe - meaning no adult nor underage info detected
• underage_safe - meaning safe but involves underage descriptions
• adult - meaning explicit and nsfw but does not involve underage descriptions
• cp - meaning explicity and nsfw and also involves underage descriptions

Columns

There are 3 columns:

• prompt the original text prompt
• rule_label is labelled using OpenAI moderation API + some key word regex matching
• llm_label is labelled using Meta LLaMa 3.1 8B with some prompt engineering