Verified: a year ago
SafeTensor
Type | |
Stats | 539 1,457 |
Reviews | (97) |
Published | Feb 7, 2024 |
Base Model | |
Training | Steps: 500 Epochs: 12 |
Usage Tips | Clip Skip: 2 Strength: 0.8 |
Hash | AutoV2 142626DBB6 |
Companion LoRA to the shoggoth embed. Download both for best results.
The LoRA has no trigger word. Works best to get the outfit and tentacle legs; embed will help with overall facial features and skin texture/color.
Made to work in Anime strictly! Showcase images use both embed and LoRa for the most case.
Strength: 0.7-1; At 1 strength it may nuke your background as it was trained on blank backgrounds, so for more complex scenes, you'll have to dial it down.