This is a model created to resemble a textured traditinal look of oil paint on canvas, but now it is basically a mixture of styles, with semi-realistic , textured results.
Trigger word: digioils style ( strength 0.5-1.0 recommended) - put at the beginning of your prompt.
Here are the keywords to manipulate the output:
Colours: Warm\cool\neutral.
Special - Candy colours (tries to give prismatic highlights, usually purple, pink or blue), neon colours. Both work well with abstract in prompt.
Refer to example images with different prompts that demostrate how those keywords affect the result.
Textures and level of detail: sketchy, unfinished, canvas and their combinations.
Canvas is especially noticeable, will add warm-toned, blurry, yet textured look. Combining with unfinished will produce the look of an unfinished painting.
Vague face is an attempt to add the blurry look to the subject's face, with focus on other parts of the image. Works mostly on people, but is somewhat unstable.
Description
FAQ
Comments (4)
loving this, thank you!
Trying to get it to work with my Merge Model and Exoterra Style lora no luck, the Exoterra Style LoRA of mine just overrides this, odd.
While Later: Merge Model also hates it, this is still a good LoRA. Just sadge, I can't use it with my merge model. :(
This is actually not a good lora, almost the very first one I've done. It was originally trained in a wonky way and then merged with another one that I've trained withy different parameters, making it a merged lora already. Maybe this is why it is being challenging, sorry about that :(
@OliaMiriam Well the issue may lay in the fact, I tried using it with my merge model. Which is picky about the loras but when it does use the lora right it works insanely well. It can even do certain SD LoRAs which is weird cause its a XL/Pony merge, didn't expect it to turn out that way, merge model was originally bad and I got it out of being that bad eventually. It takes time and experience for merging lol.
So... for a first time merge, it's better than what I did lol. Don't beat yourself up over it.
















