WaifuInvoker 1.1/1.2 = Newer style, 1.0 = Older style
V1.2
Note: Updated to Base Anima.
Recommended settings
Strength: 1.0
CFG: 3-5
Steps: 20-40
Sampler: euler a
Scheduler: beta
Trained on/best with: Anima Base v1.0
Positive:
1girl, solo, Negative:
lowres, worst quality, low quality, V1.1
Recommended settings
Strength: 1.0
CFG: 3-8
Steps: 20-30
Sampler: euler a
Scheduler: karras
Trained on/best with: Illustrious XL v0.1 or any Finetune or Merge (Wai, Hassaku, Nova,)
Positive:
1girl, solo, Negative:
lowres, worst quality, low quality, artist name, V1.0
Recommended settings
Strength: 1.0
CFG: 3-8
Steps: 20-30
Sampler: dpm++ 2m sde
Scheduler: karras
Trained on/best with: Illustrious XL v0.1 or any Finetune or Merge (Wai, Hassaku, Nova,)
Positive:
1girl, solo, large breasts,Negative:
artist name, twitter tag, text, patreon username, patreon logoDescription
FAQ
Comments (5)
This LoRA does a great job of enhancing image quality, but it messes up the eyes and limbs, and it seems to reduce prompt adherence
Might be an issue on your part. Eyes and limbs have been fine for me, hands are a little unstable. Prompt adherence likely is partly due to Anima and the dataset I trained lora on
I have the same problems with it too. Faces, hands and other, are bad. All the waifuinvoker loras have those problems. I'm in search of a WI lora that work well and this is the fourth I'm trying but.. still not a good lora
I have problems with it. Faces, hands and other, are bad. All the waifuinvoker loras have those problems. I'm in search of a WI lora that work well and this is the fourth I'm trying but.. still not a good lora
Are you able to make this lora for anima?