A social fashion style Lora.
The first version is trained without captioning.
Videos takes time to generate even on 4090, i would love if you guys and girls test the model with more dynamic motion actions prompts.
You also can use this Lora on "Flow" for ComfyUI.
https://github.com/diStyApps/ComfyUI-disty-Flow
If You like my work consider to support
https://www.patreon.com/c/distyx
Don't forget to leave a like!
Description
No caption Lora , while captaining the images trained this Lora, soon version with captions will be released.
FAQ
Comments (7)
Wow, thanks for sharing your hard work. I am gonna try it soon.
Thank you, share your thoughts.
Can you give any notes on training the LORA? Is there a specific repo I can start looking into?
The repo is the diffusion-pipe to train https://github.com/tdrussell/diffusion-pipe
What alissonerdx said.
I love the fashion this lora imparts, but users should be aware that it WILL add significant makeup. If you prefer a more natural look you may have difficulties achieving it with this lora.
Thank you for your feedback!
Yes i agree. i also prefer more natural look. maybe that will be next version.