Verified: 2 years ago
SafeTensor
Type | |
Stats | 2,274 |
Reviews | (399) |
Published | May 24, 2023 |
Base Model | |
Hash | AutoV2 549C60E621 |
It is LoRA, a technique that increases the number of annotations in an image. Normally, it blurs the image, but by inputting a negative value, the number of annotations can be increased.
When using "highres latent," it is recommended to apply it gently.
画像の書き込みを増やすLoRAです。 通常使用ではボケますが、マイナスの値を入れることによって書き込みが増えます
"highres latent"を使用する際は弱めにかけることをおすすめします