NAL It is intended to be a replacement for my lost model “Nalgotic Dreams” oriented to a new model, “illustrious”. It aims to be a bright and detailed anime style, but may contain some errors when creating illustrations.
Recommended settings:
Steps: 15-30
CFG scale: 3-7
Sampler: Euler a
The VAE is already integrated. If you use another one, it is at your own risk.
As for the resolutions, you can experiment with them, the default resolutions are recommended as well.
Hires upscale: 1.5
Hires steps: 30
Denoising strength: 0.35~0.44
Positive Prompt
masterpiece, best quality, amazing quality, very aesthetic, absurdres, newest, gradient colors, bright, background, masterpiece, volumetric lighting, highres, ultradetailed,Negative Prompt
bad quality,worst quality,worst detail,sketch,lowres, bad anatomy, sketch, jpeg artifacts, signature, watermark, old, oldest, censored, bar_censor,Description
6.- Bug fixes.
FAQ
Comments (10)
Idk if I'm doing something wrong, but nal v2.0 seems to work the best. Is there a trick to it?
Work better in what sense?
@Nalgotica like looks better. When I use v6.0 it comes out weird. I've tried every model except v1.0 and v2.0 looks the best visually. is v6.0 trained on a different size or something? cause i mostly use 1024 x 1024
@Rav_Lav What tool do you use?
@Rav_Lav I'll try the same thing to see if it plays back the same way.
@Nalgotica Like interface? I'm currently using a Comfy UI workflow. I'll try and replicate the workflow and send it. I'm having the problems off and on. nal v2.0 just seems more reliable for me.
@Nalgotica want me to add the workflow in the gallery when sent?
@Rav_Lav sure :3
@Nalgotica by the way love the model the style is very nice.
got any recommendations for multicolored hair and better hands?