









Just add 'world of warcraft' to your prompt!
Works best at 0.5-1.0 LoRA strength for Flux1.D. Schnell seems to have an issue with it at 1.0, trying lower strengths usually results in a better image.
A strength of 1.0 will give you an image that will likely have jpeg artifacts, due to the training dataset having been from 2001-2002. You may be able to circumvent this issue with a FLUX image enhancer LoRA, or photographic enhancement LoRA, depending on what outputs you are expecting.
Notes:
Trained on pre-alpha screenshots from World of Warcraft,
Will be used to help reconstruct the pre-alpha world terrain in our project, hosted over on https://github.com/akspa0/parp/
Thanks to https://archive.thealphaproject.eu/ for the screenshots!
1. The rights to reposted models belong to original creators.
2. Original creators should contact SeaArt.AI staff through official channels to claim their models. We are committed to protecting every creator's rights. Click to Claim
