A successor to AAAAutism model of mine.
Based on AutismMix, @bluvoll's 3px v1.1, @richyrich515's Richyrichmix Pony and my training that is used for AAAAutism.
Huge kudos to Bluvoll in particular for training such big style model, please go support him, he actually spent quite a bit of money.
No additional training was done to AAAA part.
This is a multi-step block merge aiming at taking some graphics from ancestors, and using fine training from AAAA to help stabilize that.
Pos: score_9, score_8_up
Neg:
Embeddings: I WILL FIND YOU IF YOU USE THEM
Loras: anything BUT PLEASE don't use expressiveh and disney princesses ffs.
Recommended settings:
Euler a | Align Your Steps (Euler a AYS on Forge)
20-28 steps
CFG: 4.5-7
P.S. all showcase images are done with just that, no adetailer is used. Hires fix performed with NomosDAT or Ultrasharp.
Features:
Could have partial retention of style knowledge from Bluvoll's 3px model, but not to full extent. You can try to use them as positive and negative prompts.
描述:
训练词语:
名称: isolicePonyXL_isolice.safetensors
大小 (KB): 6775432
类型: Model
Pickle 扫描结果: Success
Pickle 扫描信息: No Pickle imports
病毒扫描结果: Success
