1blackbar
1blackbar
It would be much faster this way
I think SD upscaling is now limited to 1024 output,, would be nice to have options for higher res, even tho im not really sure if there will be a...
Its pretty overwhelming after awhile when we have thousands of generated images and we load one by one to upscale seeking for particular ones, So i think it would really...
https://www.youtube.com/watch?v=N913hReVxMM I think it works pretty well but i wouldnt implement it like he did (on small 512 image and synthesizing whole image) What i would do is open a...
All is working well but i was wondering can you incorporate esrgan blending with original ? So we can move slider like with gfpgan and it would blend two images...
Does it work like dreambooth with identity preservation ? OR does it synhesize derivatives /mutations like textural iversion? Did you tested it on human face or complex character like a...
Can we get defaults exacly like in this file ? https://github.com/nicolai256/Stable-textual-inversion_win/blob/main/configs/stable-diffusion/v1-finetune.yaml Learning rate is 5.0e-03 but i think other stuff was changed too I think current defaults give pretty bad...
Someone did a fork with this webui and is using xformers for speedup so now you get 512x512 in 6 seconds with P100 on colab pro Free colab takes 12...
I can only get likeness with heavy overfitting but this also heavily learns background , any way to avoid it? white background? black? or a piece of code ?
What should i set it to? I need overfitted embeddings to get 100% likeness with like 50 vectors but sometimes i have to wait for too long until it gets...