Trained for PonyXL with anime screencaps - thanks to https://fancaps.net/ - using kohya_ss. In my tests it works well with other styles with a weight of ~0.9. The metadata of the safetensors file is preserved so you can see the training settings.
I'm still figuring out the best training settings and especially image tagging which seems to have a huge effect by adding/removing only a single tag so there is certainly room for improvement.