PonyXL style LoRA for emulating laserflip's art style. Works well with base model and AutismMixSDXL.
Update 3/10/24: V1.1 release
New version trained on updated dataset.
Baked for 20 epochs for a total of ~2000 steps on a thoroughly cleaned dataset - no text or watermarks here.
Usage
Activation Tag: laserflip
Weight: 0.7~1.0Notes
This model has a heavy NSFW bias (duh). SFW is doable with
rating_safetag.
The dataset contained a not insignificant portion of censored images. Include
uncensoredin positive prompt ormosaic censoringin the negative prompt if your outputs are censored.
Description
Retrained on revised dataset (old dataset had glaring tag issues)
Improved pose variety
Fixed overly simple oil paint looking backgrounds
Increased network dim 16 --> 32 (should retain details better)
FAQ
Details
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.