Dataset Card for Falling through the Skies

Dataset Summary

Falling through the Skies is an unfiltered dump of 592k stories literotica.

Supported Tasks and Leaderboards

[More Information Needed]

Languages

The dataset is mainly targetted at english. However, there could be chances where the language may not be english

Dataset Structure

Data Instances

Within the 7z file, contains a folder containing each chapter/short story. Each file is formatted in json and should be intuitive in terms of what are the values stand for.

The important keys are:

content: Contains the story/chapter content. If the story is paged, it has been collapsed and seperated with 2 new line characters: <Page 1>\n\n<Page 2>

meta: Consider using this field instead of related field in the root. "view_count" > meta > view_count for example

Data Fields

There is too many to list but there is a few key fields that you might be intrested (They are located in meta key):

tag: Author tags for the work.
views: Number of views at the time of scrape
words: Number of words for the work
category: Category where the story/chapter is slotted in
writers_pick: If the work has been selected, it will be set to true
comment_count: The number of comments for this story/chapter. do note however that comments content are not captured.
desc: The short blub to introduce the story/chapter.
author: Author for the story/chapter.

The following keys are located in the root:

url: The url slug. Example: https://<Site Redacted>/s/<url>.
title: The title of the story/chapter.
content: The contents for the story/chapter.

Data Splits

No datasplits as the total size can be stored within a filesystem.

Dataset Creation

Curation Rationale

Falling through the Skies dataset was concieved due to the lack of public dataset focused on erotic literature.

Source Data

Initial Data Collection and Normalization

An inital scrape based on categories was done to scrape all the public stories. After that, we used the public json api present on the beta version of the site to further scrape all the pages and stories.

Who are the source language producers?

The respective authors of each chapter/short story.

Annotations

Annotation process

All the data presented are parsed by scripts provided via the beta api.

Who are the annotators?

No human annotators were present.

Personal and Sensitive Information

The dataset contains only works of fiction, and we do not believe it contains any PII. However, usernames and author biographies could leak PII infomation.

Considerations for Using the Data

Social Impact of Dataset

This dataset is intended to be useful for anyone who wishes to train a model to generate "more entertaining" content. It may also be useful for other languages depending on your language model.

Discussion of Biases

This dataset is composed of fictional works by various authors. Because of this fact, the contents of this dataset will reflect the biases of those authors. This dataset is only on NSFW material and was not filtered. Beware of stereotypes.

Other Known Limitations

Not Applicable

Additional Information

Dataset Curators

KaraKaraWitch

Licensing Information

[More Information Needed]

Citation Information

Apache 2.0, for all parts of which Ronsor Labs or the Ryoko AI Production Committee may be considered authors. All other material is distributed under fair use principles.

Contributions

@misc{kkr-ftts,
  title         = {Falling through the Skies: Literature on unseen unique data},
  author        = {KaraKaraWitch},
  year          = {2023},
  howpublished  = {\url{https://huggingface.co/datasets/RyokoAI/FallingThroughTheSkies-592k}},
}