metadata

language:
  - en
license: mit
size_categories:
  - 1M<n<10M
task_categories:
  - image-to-image
  - text-classification
  - text-to-image
pretty_name: stable diffusion prompts
dataset_info:
  features:
    - name: image_id
      dtype: int64
    - name: url
      dtype: string
    - name: prompt
      dtype: string
    - name: negative_prompt
      dtype: string
    - name: size
      dtype: string
    - name: model
      dtype: string
    - name: stats
      struct:
        - name: commentCount
          dtype: int64
        - name: cryCount
          dtype: int64
        - name: dislikeCount
          dtype: int64
        - name: heartCount
          dtype: int64
        - name: laughCount
          dtype: int64
        - name: likeCount
          dtype: int64
    - name: nsfw_label
      dtype: string
    - name: nsfw_score
      dtype: float64
  splits:
    - name: train
      num_bytes: 877345095
      num_examples: 896874
  download_size: 216888972
  dataset_size: 877345095
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
tags:
  - not-for-all-audiences
  - nsfw
  - uncensored
  - art
  - stable diffusion

Dataset Card for "stable-diffusion-prompts-stats-full-uncensored"

Not SAFE for public - Definately Unfiltered with URL links being rendered

This dataset comes from prompts shared from images' metadata on Civitai. Not for the faint of heart. Thanks to Civitai.com for all the models, building a playground, allowing fine tuning of models, and generally being a good influence on model building and generation.

A reasonable attempt was made to tag unsafe prompts by adding a label column for 'NSFW' and 'SFW', with additional manual filtering, as well as adding a score column generated by a machine model.

The purpose of this dataset is to allow for analysis of prompts and feature analysis in prompts and negative prompts.

This could be for:

semantic evaluation (see stats column)
prompt quality
effective prompting
prompt alignment or misalignment
statistical research on prompts and categories
popularity of image generation approaches
mimimalism prompts with certain models
matching generated prompts to images for LLAVA purposes
mimimizing prompts for better context usage
social research on interest level and creative approaches
modeling based on prompts for automating prompt generation strategy
modeling of categorical interest and similarity
modeling of evolution of prompts based on model versioning

A seperate upload includes only prompts, negative prompts, and model name for brevity, squeamishness, and research purposes.