euler a with normal, or er sde with sgm uniform, 30-50 step, if using turbo 15 step fine most of the time,

thanks for the request, STXai
for the dataset i only gathered little because i found not many that colored. So the training not really need high step, especially with repeated kind of pose, so this model i think its needed only around 750 step that i found better and flexible rather than the one i make it to 2000 step, well i shared this one because i think its better, so hope you guys like it.

if you're not satisfied and want to try higher epoch, just contact me, i'll give it to you.

pls support if you like model https://ko-fi.com/rastaai. i love making those lora but since i makin it on cloud gpu it'll be really useful for training, thx.