This is a new approach to models. This is a finetune on Nai with a focus on artist styles and the ability to do what I'm calling "Style Blending". A Lunar Dream can recall over 160 art styles with a high degree of accuracy. Why fiddle with various style lora when you can just ask for the artist and get them! Even better, you can mix styles in ways lora can't. You can legit use as many artists in your prompt as you wish and it will yield new results without any of the problems lora might have with burn-in. You can also weigh them! I want to repeat, that this is not a model merge, or lora merge, this is a fully trained new model! HuggingFace

Link for more info and the wildcards/artistnames (README.md · lunachan/ALunarDream at main (huggingface.co))

Artist list: styles.txt · lunachan/ALunarDream at main (huggingface.co)

Have a style dataset I could use? Send me it! I'll add it to my next version! This cost me around $100 in electricity to create, so I probably won't be able to make a larger scale version without some assistance.

Donate to help pay for the creation of a SD2.1 version along with a V2 in general with more styles and features

All Art Styles in Alphabetical order:

A Lunar DreamBase Model: Nai

Total images: 30k anime images, 8000 realism pictures

Total steps: 430k + about 300k with realism @ batch 4

Tags: Anime, anime style, lunachan, realistic, chibi, outside, inside, scenery, town, city, background, bg, interior, indoors, (korean korea), (japanese japan), (Chinese china), (Ebony dark skin blackgirl blackwoman african), (desi indian), (Caucasian American European)

These tags will affect your images, there was a dedicated chibi dataset so the model can properly create chibi's.

To find the artist list, check for the .txt file on my Huggingface, which doubles as a wildcard prompt, this allows you to randomize artists, and combo rando artists. How to use wildcard: download Umi-ai extension. Place styles.txt in the wildcard folder, and then prompt using __ styles __, 2 underscores.

# DisclaimerThe use of this learning model is entirely at the discretion of the user, and they have the freedom to choose whether or not to create NSFW content. NSFW content has been gated with the NSFW tag. It should do very well at not producing nsfw content if you add it to the negative. If you choose one of the more nsfw focused artists it is advised to include clothes in the prompt, or to negative out ??????. The purpose of sharing this model is not to showcase obscene material in a public forum, but rather to provide a tool for users to utilize as they see fit. The decision of whether to engage with SFW or NSFW content lies with the user and their own personal preferences.