Verified: 2 years ago
SafeTensor
Type | |
Stats | 794 0 |
Reviews | (146) |
Published | Apr 10, 2023 |
Base Model | |
Training | Steps: 100 |
Hash | AutoV2 8D4461A17E |
Hi,
Most Models are trained on 1:1 or 2:3 aspect Ratio and Mainly for Portrait shots of people.
As my Screen is 32:9 i tried to create some AI Wallpapers.. but mostly they repeat the themes or look somewhat else strange.
Thats why i trained a LORA with a few Panorama Pictures i had.
Best Values seem to be between 0.3 and 0.6