File size: 1,120 Bytes
c7c4e50 c5578fe 15b8962 96c3659 97eb669 c5578fe 9a2932d 4e32b64 efea685 6b1f0bd efea685 6b1f0bd 83edf88 c5578fe 20082ec c5578fe 3e85c39 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 |
---
license: creativeml-openrail-m
---
This is a fine-tuned Stable Diffusion model (v1.5) trained on images (that I find aesthetically appealing) from various fantasy concept illustrators and designers (mostly Japanese, South Korean, or Chinese). Use the tokens **DBfantasyart style** in your prompts for the effect.
Download the checkpoint file (.ckpt) to use it.
If you want to **reproduce** the images below, I've prepared a more detailed version of the prompt settings and seed values in this **pastebin**: https://pastebin.com/2XrkSMfV
Example settings: Steps: 50, Sampler: Euler, CFG scale: 7, Size: 512x512
Example prompt: **DBfantasyart style**, masterpiece, 1girl, beautiful female portrait of sorceress
Some example images in txt2img (**very minimal editing/cleanup**):
|