Type | |
Stats | 381 521 10 |
Reviews | (100) |
Published | Oct 28, 2023 |
Base Model | |
Usage Tips | Clip Skip: 2 |
Trigger Words | suoiresnu |
Hash | AutoV2 16F6B8C03A |
Description
Artist Style Lora, Purged off the site, I use it in my workflows and didn't want the community to lose it.
Although this is an "Artist Style", I primarily use this to create very clean flat shaded images / clean line work on almost any model I work on. This LORA is extremely powerful at doing this without washing out the checkpoint completely.
Prompting Info
Overall very generous LyCORIS, works with anything basically.
A little over trained on artist signatures, the trainer seems to have this issue as they don't clean up their datasets and like to train on high values.
Original page documented a trigger word but I never use one
The author recommended a strength of 1.0
Personally I feel this is too high, it doesn't fracture the image but it washes out a little too much detail
This LyCORIS was trained on "AnyLoRA_noVae_fp16"
Upscaling and Adetailer are highly recommended
Basically required..
Negative Prompt
(worst quality, low quality:1.4), 3D
Credits
Original Artist
Suoiresnu - https://twitter.com/suoiresnu
Trainer
444Butter