README.md

metadata

license: creativeml-openrail-m
language:
  - en
pipeline_tag: text-to-image
tags:
  - stable-diffusion
  - art
  - dataset
  - concept
  - character
  - style
  - dreambooth
  - lora
  - textual inversion

Misc Datasets

Here i will upload datasets (images + captions) of concepts/styles/characters for anyone to use on their models, as i am not able to do LoRA's myself, alongside other datasets i've used for other models.
Some are handcropped and/or handpicked, some not. If it's a big dataset it's probably automatically cropped (https://www.birme.net, 1280x1280, jpeg 95% quality) and not handpicked.

I've also included a python script for anyone that wants to use gallery-dl to download images, since its tags are pretty fucked up.
It basically fixes its main problems and also removes metatags like 'commentary', 'translated' and similar, and gives the option to change underscores with spaces and other stuff.

Characters

• Neru (Blue Archive)
• Jibril (No Game No Life)
• Fubuki (One Punch Man) Doesn't include captions! You might want to use something like WD Tagger.

Styles

• Cutesexyrobutts
• One Punch Man - Yuusuke Murata Doesn't include captions! You might want to use something like WD Tagger.
• Phantom IX Row
• Mamimi (Mamamimi)

Concepts

• Breasts On Glass Doesn't include captions! You might want to use something like WD Tagger.
• Fingering
• Oversized Breast Cup
• White Eyelashes
• Mizumizuni Fellatio
• Unaligned Breasts Doggystyle
• Milking Handjob
• Fellatio + View Between Legs