用戶LoRA

SD
其他

Husbando LoCon V3.2

Husbando LoCon V3.2 - AI Model封面圖片
上傳時間
2023年7月18日 中午12:32
使用數
28.4k
評論
不滿意
許可權
  • 允許圖片生成與分享
  • 允許使用者下載此模型
  • 允許生成圖片用於商業用途

描述

https://civitai.com/models/19263?modelVersionId=28307

Weight Guide For general purpose: 0.5 - 0.6. You can also use a lower weight such as 0.3 if you just want to enhance the sexiness of your character.

For explicit art: Set the text encoder weight to 1.0 and U-Net weight to 0.5 - 0.6 with Additional Networks extension. (You need to toggle the "Separate UNet/Text Encoder weights" checkbox.)

How to use Select any checkpoint that works with Danbooru tags. Recommended models: AnyLoRA or AOM series.

Select a VAE if the checkpoint model you're using doesn't have a baked-in VAE, or else you'll find your generated images have dull colors. Recommended VAEs: vae-ft-mse-840000-ema-pruned or kl-f8-anime2.

Make sure to follow the steps in How to use a LoCon. I suggest you start by setting the weight to 0.5.

Generate at 512 base resolution and use hires fix to improve the details. Recommended upscaler: R-ESRGAN 4x+ Anime6B; denoising strength: 0.5.

留言1