wasteland
Wasteland style model
Good keywords:
wasteland,1girl
Step 20, turn on HD repair, redraw 0.45
Sampling method: Any of the ++karras will do
If you have a full-length portrait, you need to get an ADetailer
There's no set formula. You can explore it yourself
If you like the model, you can support me a little bit for my electricity bill
It make a difference to me
Thank you very much
I'm ObitoD and I will keep to make good models for you
废土世界
废土风格模型
好用的关键词：
wasteland,1girl
步数20，开启高清修复，重绘幅度0.45
采样方法：那几个++karras的都可以
如果全身像，需要开ADetailer修脸
没有固定套路，可以自己摸索一下
喜欢的可以在咖啡支持我一下
Description
FAQ
Comments (7)
Error: Could not load the stable-diffusion model! Reason: 'Attention' object has no attribute 'to_to_k'
I'm sorry, I don't know how
@obitod does it work on Windows?
@DJ_Point Sure,That didn't happen to me
The model works great, but there is in fact something funny with it. The model toolkit identifies this: Model is 1.99 GB. Multiple model types identified: UNET-v1-BROKEN, CLIP-v1-BROKEN. Model type UNET-v1-BROKEN will be used. Model components are: UNET-v1-SD. Contains 394.28 MB of junk data! Model will be pruned to 1.60 GB.
I think I know what the issue is, it has a VAE with wrong keys. (for example: "first_stage_model.encoder.mid.attn_1.to_q.weight" instead of "first_stage_model.encoder.mid.attn_1.q.weight"). Remove the bad VAE and the issue is gone
@victorc25744 Thank you! You're right