Style-LoRA of the artist DrunkAvocado If you like their style give them a follow or otherwise support them! If you like the LoRA give it a like (or even better a review and show me your works!). This LoRA was created via Pixai LoRA onsite-creation.
Recommended Weight : 0.85 to 0.9
Recommended positive trigger words: DrunkAvocado
Description
FAQ
Comments (3)
i cant seem to get it to work, ive copied everything. although i dont know what "ADetailer" is. im fairly new to this, could you please explain how you got it to work?
well...for starters, ADetailer is an extension you get from stable diffusions web ui that you have to install. It normally fixes discrepancies such as faces or hands giving them more detail. As for my workflow, I tend to train all my loras on Anythingelse v4 and use it to showcase the finished work. I use Euler A for sampling method, upscale the image by about 2 keep the denoising strength to about .45 and use a clip skip of 2. also when using certain loras especially for art styles I recommend using a weight of about 0.8 to 0.9 to determine if the style works well. I hope this information helps.
@Mizuku243 I appreciate the response, very helpful information. i've got the extension now and its awesome. unfortunately, despite everything, i cant seem to get the image to generate in the drunk avocado artstyle. It works perfectly fine using Pixai on the website, but when i try using stable diffusion, it doesnt seem to pick the lora up. Ive spent a few hours making sure i got everything right, testing the lora on a simpler prompt on multiple checkpoints but to no avail. i checked the generation data of one of the examples and the only thing i was missing was "DrunkAvocado Style Lora: cce816609d89" sandwiched inbetween some ADetailer settings.