Which model was this trained on?
1.5? 1.4? :)
This was trained on 1.5, I update the readme!
strange, when I run the lara croft example prompt with the proper seed it doesn't match the sample image. I assumed that was because I was on 1.5 and the model was trained on 1.4 -- but if that isn't the case, I will have to see what else might have gone wrong with my install. (the model is definitely working, it just doesn't match your results.)
any suggestions?
JTC
I used the highres fix in automatics repo for the images, that alters them as well as the xformers to some extent, as far as I know. I also did a batch of four and this was the second image of the batch. Here are the complete settings from the chunks:
modern disney lara croft Steps: 50, Sampler: Euler a, CFG scale: 7, Seed: 3940025417, Size: 512x768, Model hash: ccf3615f, Batch size: 4, Batch pos: 2, Denoising strength: 0.7, First pass size: 448x640
Amazing will this further be trained?
I'm always looking to improve these models. While this model turned out great in version one, I will return to it to make a few improvements. I already have some on my list.
Any updates on updateing this? :)
This was trained on 1.5, I update the readme!
You said you will but you didn't. Please update that to readme
And thank you very much for all your work :)