
A LECO trained on the concept "speedpaint". This model is nice for producing images that look like rough digital art. The trigger word is "speedpaint".
I recommend a low denoise (like .35) for hires fix. Just don't use adetailer at all, because there's really nothing for it to sharpen and the results look cruddy.
The strength you need to run this lora at depends a lot on the checkpoint you're using. I tried to scale it so 1 is a good starting point, but you may need to go down below 0.5 or above 2.0.
描述:
训练词语: speedpaint
名称: EnvySeedPaintXL02.safetensors
大小 (KB): 45606
类型: Model
Pickle 扫描结果: Success
Pickle 扫描信息: No Pickle imports
病毒扫描结果: Success