Annaliese is based on a set of images I did quite a while ago that was pretty popular, so I thought I'd give it a shot. I make no promises as to how this is going to work, but as before, if you'd test it out and give some feedback, that'd be great. I'll put it through its paces and report back with recommended settings, checkpoints, etc.
Update: It's an epic failure. I don't understand why faces are looking horrific. I have no clue what I'm doing. I give up. Don't even bother trying this model, it sucks ass.
A couple things I've learned about this model:
1) Use an eye detailer because for some reason eyes always come out blurry. I'll try to fix this if I get around to making a V2.
2) If you're using the on-site generator, image quality is better (to my eye at least) if you have Enhanced Compatibility turned on.
Description
Comments (1)
It's not bad you gotta remember that not all checkpoints will fit the needs of a Lora and that Lora's most of the time can't carry themselves alone and will need further prompts and supportive settings and Lora's to support the created Lora's downfalls that someone makes, so don't worry, we all feel like this once our Lora's have baked.
Details
Files
Available On (2 platforms)
Same model published on other platforms. May have additional downloads or version variants.