FIRST OF ALL: I ASK THAT IF YOU USE MY LORA; PLEASE DO NOT POST NSFW CONTENT THAT IS HARMFUL, ABUSIVE, DEGRADING, INSULTING, HUMILIATING, OR OFFENSIVE TO HUMAN DIGNITY. I DID NOT CREATE MY MODELS FOR PEOPLE TO DO THOSE THINGS. PLEASE RESPECT.
This new model aims to adopt an artistic style that is characterized by the addition of art qualities of prominent brushstrokes and ink.
This LoRa has been tested with the AnimagineXL base model and has generated these images you see now.
IMPORTANT: The trigger keyword for this model is "INK". The LoRa will add it to the end of the prompt you type, if the word does not appear, include it in the prompt to activate the model's effects.
NOTE: Try this LoRa as much as you wish, and if you wish, you are totally free to write your opinions and recommendations in the comments. It is greatly appreciated.
Suggested specifications:
1. Image size: 768x1152 or custom (512x768 at 2x resolution).
2. Sampling steps: 20 to 25.
3. Vae: Pastel-waifu-diffusion.vae.pt.
4. Sampling Method: Euler a, DPM ++ SDE or DPM ++ 2M Karras.
5. CFG Scale: 6.5 to 10.
6. Scaler: R-ESRGAN 4x+ Anime6B.
7. AD detailer: optional, according to your taste and preference.
8. Base model: The best so far is AnimagineXL 3.1 and 3.0. (Anyway you are free to use the base model of your choice).
9. Extra (important): This LoRa will work only if you detail the type of ink art you want (for example: brushtrokes, ink droplets, sumi-e, ukiyo-e, colorful ink, monochrome ink, etc).
Example: You can start with a prompt like "Very pretty girl, simple background, looking at the viewer, colorful ink strokes on the background, masterpiece, INK art"; and from there you can modify it to your liking.
