This model was trained on several angles of the Venator so please be patient if the first generation doesn't work out. adding phrases such as "front of Venator" or "side of Venator" can help give you your desired image.
Currently the front and side are the best angles of the ship but I'm planning to improve that in future.
Training Info:
Mostly trained on 1344x768 images as well as a few portrait photo's.
Description
FAQ
Comments (3)
FOR THE REPUBLIC!
Great idea, but I have to generate 12-20 images to get 1 or 2 venators not deformed. Probably needs some more training?
I´ve tested the LORA with Jauggernautv9, JauggernautX, Realities Edge and Zavy Chroma 6.
Thanks for the feedback! When I get a chance I'll finetune the dataset to focus more on the front and side. Currently the dataset includes many angles which confuses the model.