This version of the LoRA was trained on upscaled images of only the headware implant concept from the version 1. Also the network size was increased. Resulting in improved quality images with more detail and sharpness at the expense of twice the amount of disc space on your hard drive.
Usage Tips:
Start at a strength of 0.8.
trigger is 'mad-2cbrrm' and 'mad-1cbrrm'
I tried to train separate trigger for one and two cyberarms. But thats not really working, unfortunately.
This version of the LoRA was trained on upscaled images of only the headware implant concept from the version 1. Also the network size was increased. Resulting in improved quality images with more detail and sharpness at the expense of twice the amount of disc space on your hard drive.
Usage Tips:
Start at a strength of 0.8.
trigger is 'mad-2cbrrm' and 'mad-1cbrrm'
I tried to train separate trigger for one and two cyberarms. But thats not really working, unfortunately.