Type | |
Stats | 1,801 676 |
Reviews | (286) |
Published | May 6, 2023 |
Base Model | |
Training | Epochs: 10 |
Trigger Words | hand holding umbrella wearing kimono |
Hash | AutoV2 27C46D92B8 |
Want me to create custom models for you? please visit my Fiverr page:
https://www.fiverr.com/s/Xar15P
It's not necessary but If you like my creations please consider to support me to create more projects by donate in here:
https://ko-fi.com/gorillamonsooniii
https://www.patreon.com/GorillaMonsoon
The Umbrella LoRA
Having a hard time generate umbrella model? maybe this LoRA can help you.
Legal & Risk
In principle, I do not support the use of this model to generate any adult content. When using this model, users must abide by laws and regulations, and do not infringe other people's reputation, privacy or portrait rights.
Changelogs
v1.0: Experimental, do not expect too much. Got a lot of problem with hand position and like always, extra or fewer fingers problem, but trigger word seems working to reduce the effect. Or just use ControlNet.
v2.0: Experimental+++, retrain LoRA, adding ability to generate variations of umbrella, such as Wagasa & Janomegasa (Japanese traditional oil-paper umbrella). To generate this variations, see Usage. it's just that the variations for this model are very limited. v1.0 is much better in terms of poses variations.
Features
good-shaped umbrella geometry
multicolored umbrella with adding color prompt
compatible with 2D and photorealistic models
Usage
compatible with all sampling methods
for checkpoints, i’m using my models: https://civitai.com/models/44418/eclipse and https://civitai.com/models/41141/halcyon , just because i’m trained this LoRA with it.
trigger word is a must for better umbrella and hand pose
Use 15-30 sampling steps
CFG: 4.5-7
Less complicated, short negative prompts are better
v2.0: to generate wagasa/janomegasa, just adding prompt: wearing kimono, hand holding umbrella
and write in negative prompt: rain
v2.0: to deactivate, write rain on prompt, and write wearing kimono on negative prompt.
If using Hires:
Hires step: <20
R-ESRGAN 4x+
Denoising: 0.3-0.7
After download this LoRA, try to share with us about your findings.
Please give star reviews with your generated picture if you like it.
I'm also accept commisioning for LoRA/Others