Lora uses a single trigger word:
Additional trigger words : animation, art.
Lora is trained on this model: https://civitai.com/models/4823?modelVersionId=156110 (Deliberate V3)
Lora's weight tolerance ranges from:
<lora:DE_Alexander_Rostov:0.1> to <lora:DE_Alexander_Rostov:1.2> as needed.
I recommend using the following prompts to generate by style:
Painting (oil, pastel)
Illustration, cartoon, soothing tones, calm colors
from (A_Rostov_Style:0.6) to (A_Rostov_Style:0.9).
You can also take more hints from the examples.
Don't forget to experiment :)
Please note that in the examples the lore name has been changed.
Features of this version:
The model is trained on images with resolution: 768x768 and 768x512.
The dataset includes general plans, portraits, full-length paintings.
The model does well with portraits and general terrain plans.
Add a review if you liked it!