Type | |
Stats | 576 806 248 |
Reviews | (134) |
Published | Feb 14, 2025 |
Base Model | |
Training | Steps: 530 Epochs: 10 |
Usage Tips | Clip Skip: 2 Strength: 1 |
Trigger Words | double exposure, silhouette |
Hash | AutoV2 B8FAA92A24 |
THE MODEL IS BASED ON THE WORKS OF THIS AUTHOR
The model is quite tricky to use and I can only make a couple of recommendations:
It works best at 896:1344 , 832:1344 and similar resolutions.
Simplify the prompt as much as possible! Remove everything related to character's colors (green eyes, brown hair, etc.)
You can also use tags like “red theme, green theme, blue theme”Character models based on Illustrious work with this concept, I recommend lowering their weight to 0.6-0.7
Clothing elements can be specified and it will work, but I recommend to add 2-3 tags related to clothing
The appearance of small silhouettes of people is inevitable because the model is created on the works of one particular author (I could remove them in photoshop before training, but it would take an extremely long time and I'm too lazy for that). That is why the “main silhouette” is always human
You can suggest me characters whose model you want to see! (just pm me)