Verified: a year ago
SafeTensor
Type | |
Stats | 171 110 |
Reviews | (39) |
Published | Jan 20, 2024 |
Base Model | |
Trigger Words | CorruptedOutputXL |
Hash | AutoV2 556AF0EC50 |
Sometimes you just want to break stuff.
This LoRA was trained on a bunch of heavily glitched images with the express purpose of messing up your generations. It can have interesting effects at various strengths though it doesn't always work well with various concepts. Best used with simple prompts and likes portrait output.
Don't expect consistent results.