Another experimental Lora. I hope it gives you ideas and inspires you to make interesting images.
The trigger word is: marionette
This lora was trained with images of classic wooden carved marionettes.
There are strings on these puppets so you can play them from above.
I suspect that the lora works better if you keep the prompt short and simple.
In the Lora, the strings are often not logically linked to the puppets. If anyone has ideas on how I can train this well, please feel free to share. But the overall picture fits in my opinion.
The more weight you give the Lora, the more wooden and basic the puppets become.
But not only do the characters become puppets, but the environment takes on a toylike character.
Have fun!
Description
FAQ
Comments (17)
Great idea, it is a refreshing subject!
My Automatic1111 cannot identify this Lora
SDXL != SD - try switching around.
not working as well
@denizbalkan Are you using SD 1.5 or SDXL?
@Catalorian tried different sdxl models like jugg v3 and v6 and no one shows this lora :(
I have the same problem. Its not listed in the loras (Stable Diffusion webui)
@CyclopsGER
@quetza
I will try to fix this problem and will release a new version. I have an idea of what the problem might be. Thanks for your feedback and please be patient :)
@Catalorian I switched to 1.5 and its listed there.
@CyclopsGER
@amitaii637
@denizbalkan
@quetza
I hope the new version solved your problems :)
@Catalorian I still have same issue, my automatic not showing your lora, but if I put trigger word with lora activation in prompt - its working.
no showing here!! either
v1 and v2 ..It shows up with SD1.5 model only!
@adamscinematic832 please provide me with a screenshot of your settings (maybe via discord message?). I can't reproduce the missing lora. It is not a SD 1.5 lora.
@Catalorian HI, as other people have said its happening! lol... in Automatic1111. What settings do you want to see?
@Catalorian Someone here said that it works with the prompt text even though it does not show up when an SDXL model is loaded. So i put a dummy SDXL lora with the name of this one that will just paste the trigger text. If anyone was interested, Awkward and a pain to set up but it works, just delete the first part of the pasted text and use the trigger word text that you can add in the lora editor in Automatic1111.
I tried to get this working in Automatic1111 and also Easy Diffusion. It does not work with either.
Easy Diffusion gives this error...
"Error: Could not load the lora model! Reason: Could not find text_encoder_2.text.projection in the given object!"