Spaces:
Running
on
Zero
Can you provide the complete app-py file
Thanks
here is a app-modify zip file You may change it if you want to https://ezbrush.in/AI-TRAINING/app-modify.zip
here is a app-modify zip file You may change it if you want to https://ezbrush.in/AI-TRAINING/app-modify.zip
As your name suggests, you are really cool. Thank you.
thanks
here is a app-modify zip file You may change it if you want to https://ezbrush.in/AI-TRAINING/app-modify.zip
I tested it locally and it seems that the generated effect is different from the demonstration. Is it a problem with referencing the model? The similarity of the generated effect is very poor, almost no similarity
They would have to fix that I don't know a whole lot about coding I'm just throwing out some ideas you can also use enhanced upscaler but if they're not looking the same it won't do you no good here's another idea came up with but quite can't get it to work where you create your own custom styles anyone who wants to try it go ahead
They would have to fix that I don't know a whole lot about coding I'm just throwing out some ideas you can also use enhanced upscaler but if they're not looking the same it won't do you no good here's another idea came up with but quite can't get it to work where you create your own custom styles anyone who wants to try it go ahead
I am trying to use ip-adapter-faceid_stxl.bin, but there are too many modifications needed and many errors are reported
Download the file again on the top post there should be two zip files try NEW-app.zip OR click here https://ezbrush.in/AI-TRAINING/app-modify.zip
who made magicface? git rid of the model your using and use realisticVisionV51_v51VAE.safetensors it has VAE Baked in, do NOT get a newer version or any other version, this model is one of the best. if you cant find it let me know.
Download the file again on the top post there should be two zip files try NEW-app.zip OR click here https://ezbrush.in/AI-TRAINING/app-modify.zip
Thank you for your contribution. I replaced the new app file, but the result it generated still doesn't look like the reference image. Besides, it's excellent, thank you again.