Sign In

ColorExplosion

87
545
6
Updated: Oct 5, 2024
tool
Verified:
SafeTensor
Type
LoRA
Stats
545
0
Reviews
Published
Jun 17, 2023
Base Model
SD 1.5
Training
Steps: 37,100
Epochs: 10
Hash
AutoV2
95407F40BF

First attempts at creating a LoRA. Tested to work pretty well on anime style models. I find a weight of 0.4 to be my personal preference and a full weight of 1 will massively distort the image. I also prefer to use it in img2img vs originally in the txt2img with around 0.4 denoising strength.