Accidently made this instead of LTX2.3
Whoops
I intended to make an LTX2.3 lora but forgot to actually click the correct model to train for .
I posted it anyway since it might be useful, but I'm actually unsure yet if it even works. I might get around to it eventually.
But I currently just dont want to spend buzz to test or fill out this model with examples since my computer won't generate videos, and it's so much more buzz expensive than LTX
Description
Comments (5)
And what does the accident do? :)
Perhaps I could be more clear. I'll edit the post
The point is that I intended to make an LTX2.3 lora but forgot to actually click the correct model to train for .
I posted it anyway since it might be useful, but I'm actually unsure yet if it even works.
I just didn't want to spend buzz to test or fill out this model with examples since my computer won't generate videos, and it's so much more expensive than LTX
@Eloc_Starwind
It does not appear to work in FramePack at least, which is what I'd assume most people use since it's the tuned version of HunyuanVideo that works with 6GB cards, the original HunyuanVideo model is too huge for normal people with a normal wallet :P
@punkbuzter340 I'm having a lot of issues training
My ltx lora would just crash out the generation . Had to take it down. On third training attempt and now civitai seems to be borked.
I know this one CAN make a video, I made a most unusable blurry one in there generator. Otherwise I'd take this down like I did to my ltx lora post.
@Eloc_Starwind Making progress then :) But it doesn't work in the local FramePack-Studio.
I can't speak for CivitAI's generators but I for one are using my own hardware ;D