It's a feet Lora, so expect some additional toes, morbid and mutated stuff here and there. In my experience about a half of pictures should be decent.
Tags
Feet
legs-upSoles
Trained resultions
768x768
512x768
768x512 (most unstable due to low amount of data
Params
Kinda picky on sampler, so I recomend using one of the folowing:
dpm++ 2M SDE Karras
*"euler a" provides okeish results too, but it's not that good in img2img in my opinion
And for the other parametrs
Clip skip - 2
Lora weight - 0.75 worked the best for me (photorealistic models get overburned if add more weight, but you can fix it in img2img)
Also I suggest using txt2img with 0.75 weight, and then use img2img with lower (0.3, or less) weight of the Lora. Results will be cleaner and closer to the original style of the mode
