This is really awesome merge! Is there a chance we can get an I2V version?
This is 100% most stable fast and cinematic ready model!
Can we get a I2V version? Is it even possible?
I am attempting to make one but the lora's that are in this one are not really meant for Image to Video. So far i have done some testing and the results are really not much different than the base. I am working on training some more lora's for image to video though so once I have a good set I can then create a fined tuned image to video merge. Stay tuned! Also, I posted this here as a placeholder for other to test but the official place for this model is here now https://civitai.com/models/1651125 I don't really use hugging face much so don't see the messages.
Just updated the Hugging Face page with all the latest links and resources — including workflows, models, GGUFs, and example videos. Check out the new description! 🙂
'vace_blocks.8.modulation' i got this error, any clue ( I use fusionx-vace model)