I'm going to give the ltx trick a try. I hope it works with partially blurred images also. Because using just img2img the blurry parts stays mostly blurry.
u can search for them in the comfyui models , its like missing custome nodes, u can find all needed models inside, and it will place it to the right path
You can install / download all nodes with the manager and for any models that are not available you can use 'wget' to download anything in the folders.
ты лучший чувак и спасибо за воркфлоу давай еще классных штук
It's nice to see you
Great idea, great video breakdown
When i try it with human photos the video is moving alot and im not getting that exact pose frame.
I'm going to give the ltx trick a try. I hope it works with partially blurred images also. Because using just img2img the blurry parts stays mostly blurry.
Yes the denoising in img2img somehow understands blurriness and maintains it, if you use an elaborate prompt with LTX, it does give quite good results
can you make your all workflow run on runpod
how load missing models from civitai, huggingface. some time some models cant find, cant load
u can search for them in the comfyui models , its like missing custome nodes, u can find all needed models inside, and it will place it to the right path
You can install / download all nodes with the manager and for any models that are not available you can use 'wget' to download anything in the folders.