LoRA for giant, slug-like monsters pinning, absorbing, or swallowing characters.

Also available on Moescape (all versions) and PixAI (IL, Anm). AFAIK PixAI doesn't currently support MOSS-image.

Square brackets in the trigger mean use one of the options listed if there's a slash or replace the word and brackets with the term you desire if there isn't.

Krea2_v1: Pending training.

Moss_v1: Same dataset as Anima_v1 plus autocaptions. Has a noticeable style impact on anime finetunes like Hassaku and Cat tower. Mix of trigger words and natural language should work, see sidebar.

Anima_v1: Trained on Anima Preview 3. Combines both variants into one LoRA, but needs some natural language to work best.

For the mutant slug, you need to replace "slug" with "mtntslg" in the tags and natural language for best results.

IL_v1-pin & IL_v1_D4-pinned: Trained on just characters getting pinned down by large (human-ish sized) slug monster and BHM Monster Lab's mutant slug monster. Significantly more stable at that concept than the alpha version and has less style impact than v1-swallowing (but doesn't include that concept).

Large slug monster: lrgslg, slug,

Mutant slug monster: mtntslg, monster, tentacles,

Pinned: pinned, [on stomach / on back]

Mucus & slime: mucus, suggestive fluid, slime \(substance\),

IL_v1-vore & IL_v1_D4-pinned: Focused purely on the "giant slug vore" aspect. More stable at producing that specific concept, though still a bit finicky. Still has some style impact and doesn't work well with Hassaku.

Triggers: gntslg, monster, vore

Feet first: feetfirst_vore

Head first (unstable): headfirst_vore, no humans

IL_a1: Legacy alpha version, unstable and finicky, only still here to prove it's my LoRA. I tried to cram too much into one LoRA without balancing the dataset, so it doesn't work well.