Type | |
Stats | 1,465 2,690 |
Reviews | (256) |
Published | May 13, 2023 |
Base Model | |
Training | Steps: 4,500 Epochs: 30 |
Hash | AutoV2 3337CCAF73 |
¡Hello!
This Lora has been created to be able to make images with Suoiresnu drawing style (it is not the same, it just looks similar)
I would appreciate anyone who can give me feedback since it is my first Lora and can easily have a problem
It has not been used with any other Lora, Model or Configuration that I name below.
█▀█ ▄▀█ █▀█ ▄▀█ █▀▄▀█ █▀▀ ▀█▀ █▀▀ █▀█ █▀
█▀▀ █▀█ █▀▄ █▀█ █░▀░█ ██▄ ░█░ ██▄ █▀▄ ▄█
The Lora is activated simply by placing its corresponding Lora Tag <lora:SuoiresnuStyle-Rech44:1.0>, a keyword is not necessary for its activation.
LORA Weight: 0.9
CFG Scale: 7-6
Prompt: In positive prompts: ((plain lighting):1.0) to get more exact lighting to the original (You can change the value to test) and in negative ones (Monochromatic)
Clip Skip: 2
Sampling Method: DPM++ 2M Karras
512x512 images were used for your training and AnyLora for your training and sample images.
¡Thank you very much for everything!
PD: In case they are guided by the Prompts in the examples, they don't, due to my little experience, the prompts are horrible and the results can be 10 times better than those shown.
-Rech44