Open Model Radar · Dossier

少把盐 Anima/moss

Style/art-style LoRA by FEEEEF, built on the Anima base model, with no fixed trigger words.

LoRAAnima

Highlights

  • No trigger words needed; a plain description is enough
  • 9.1k total downloads · 496 likes · 1 version
  • File size roughly 87.6 MB
  • Flagged by the platform as general content

Scope

  • Suited to: style/art-style work, used with the Anima base model
  • Suited to: stacking with other LoRAs on the same base model (lower each weight by 0.1–0.2)
  • Not suited to: direct use across base models — feeding an Anima LoRA to a different base model degrades noticeably

Tips

  • Start testing at weight 0.8; going higher causes overfitting
  • If output turns gray or overly stylized, lower the weight first rather than immediately switching samplers
  • The base model must be Anima; choosing the wrong base model is the most common cause of failure

Facts

Type
LoRA
Base model
Anima
Org
FEEEEF
Downloads
9,146

Links