If you'd like to support me and do more:
https://ko-fi.com/edgonkofi
If you're looking for a >>LoRA Making Tutorial<<
These Boots Are Made For Walking
"And that's just what they'll do
One of these days these boots are gonna walk all over you"
Get your boots now!
Description
FAQ
Comments (6)
Great boots. If possible, can you do pro-wrestling and boxing boots too? Thanks.
Thank you for your kind comment.
As you can see I am publishing LoRA every day, and have a LOT of requests almost daily.
I cannot possibly keep up with them all and need to prioritize, which is why I've opened up commissions
Commissioners may choose to keep a LoRA for themselves or ask me to publish it. It's their choice.
If you would not like to commission me, there is the option to learn to make your own LoRA through this free tutorial. It's a good base to get started.
I am very grateful for your appreciating my work, but unfortunately, as things stand, it would be impolite and irresponsible to promise you that I would deliver a pro-wrestling/boxing boots LoRA in the near future.
Please do look forward to my next work!
Best,
@EDG Thanks. Totally understand. Keep up the great work.
hallo agian, I would like to ask some questions that confused me in lora training.
As you can see, i have made three lora about different types of boots. But I face some troubles when I'm trying the 4th lora.
It's a full laced-up thigh boots, but the results were not good. lace-up trings were fuzzled, and metal ring on the two side of boots were also mixed in great chaos, sometimes it's even hard to tell what they are.
But I found in your lora, the lace up trings are good, and the shoe tring holes are perfectly shown, so I was wondering, is it connected to the models you use that will influence the lora trainning result? I use the NAI final-pruned ckpt as the base model, and any your suggestions?
Forgive my poor english.
Hello @XQAnima and thank you for showing interest in my work and asking a question in such easy to understand and good English.
After reading I would give you 3 suggestions for improving your model:
- Try increasing the training rate per image (epoch)
- Try training on SD1.5 while using the most realistic images possibles to get fine detail
- Try refining your captioning, my >>LoRA Making Tutorial<< insists on the importance of good captioning
Good luck, and come back to tell me about your progress!
@EDG thx for your suggestions, I also figure out some posibilities to improve the work. Maybe I should give a shot in' REG dataset/Learning Rate/Tagging Refine/Base Model change', I'll try all of them, it will be a long big experiment though (TAT)rz.
Still, many thx to your kindness. Chao~
Details
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.