Recommendation: 0.5 and 0.7 strength!
this negative prompt:
lowres, bad anatomy, bad hands, bar censor, text, error, missing fingers, extra digit, fewer digits, cropped, worst quality, low quality, normal quality, jpeg artifacts, signature, watermark, username, blurry, out of focus, censorship, censored, ugly, old, deformed, amateur drawing, odd, fat, morphing, black and white, anime, ((simple background)), (((depth of field)))
Description
FAQ
Comments (4)
looks amazing, but I can not use in my google colab :c, all the embeddings of 5 megas or more not work here, only the small embeddings of 1 mega or less maybe. Please do a light version :D
I've tried just about everything to get this hypernetwork to look anything like the examples. Everything from the sampler, prompt, model, clip skip. Not even close. Used every model I've access to including Lucidmix. Anyone else had the same issue?
I tried and its clearly trying to mimic the shadman style, but it looks nothing like the example I copied. So idk
@jurplepacket definitely need more experiment and testing vs Checkpoint-specific TI vs LoRA (nice)
P.S. Daring that multi-artist LoRA and TI would make style mixing easy