I like doing some abstract stuff so I made a lora with my old sd1.5 gens again.
You don't need to use the trigger words for this lora, I just added the keywords that was mainly trained with.
This lora have a lot of bias and have difficulties focusing on humans, so you should reduce the strength a little if you are not able to have a human. Or you can describe them more and move it at the start of the prompt.
You can also use it with low strength to make more colorful background.
Description
FAQ
Comments (9)
This lora is very cool, but I am having great difficulty in calming it enough so my prompt will show through. I even lowered weight to .0001 and put more emphasis on my prompt. Doing this did work sometimes, but not consistently. It's a very cool lora but it's very VERY strong.
you don't need to use all trigger words to use this lora. You can try to add them one by one and see what you like.
@Nicows Thanks for the reply. I did actually try only using "abstract" but still had basically the same problem. I'll play with it some more. Appreciate the advice.
Great Lora. I keep coming back it and trying it on older work and being amazed by what comes out. Very interesting stuff. Nice work.
thank you <3
Et dire que je cherchais un Lora comme celui la depuis longtemps, Cimer beaucoup
Will there ever be a version of this lora made for illustrious?
I want to make a version for Illustrious but I'm afraid of the low quality of my generated pics I made. We can see with this lora at high strength that we lose a lot of quality on the image.
Such a good LoRa! I'm only starting with AI gen but it give really good results even not knowing what I'm doing. Tuned down to 0.5 and I get consistent good result despite my poor prompts. Love it!
Details
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.