ComfyUI Conditioning Average ControlNet workflow

ComfyUI Conditioning Average ControlNet workflow
In this workflow we will use node conditioning average
we will see how with its help you can combine and mix prompts and conditions
which allows us, for example, to change the season of the year, or the amount of blemishes in a face, while maintaining the main characteristics of the image
#comfyui #stablediffusion #ipadapter #controlnet #depthanything #img2img
follow me @ / pixeleasel
workflow:
drive.google.com/file/d/1Dix5...
video how to install and use DepthAnyThing
• ComfyUI Architectural ...
DepthAnyThing Model
huggingface.co/spaces/LiheYou...

[@97BuckeyeNut]

Very interesting use of that node. Thank you.

[@PixelEasel]

Very welcome! thanks

[@user-kk2ve1un4u]

what about conditioning concat and the others ?

[@PixelEasel]

soon ...

[@Halfgawd_Halfdevil]

can't get it to pass k sampler. any reason? it works in every other workflow I tried to see if there was a problem. but this is the only one that causes an error, will not generate images.

[@PixelEasel]

I'm sure there is some reason , what message do you get?

[@Halfgawd_Halfdevil]

not a message but an extensive list in red of failures. Is there something that could be missing. I have all the latest updates within comfy and IO do not get anything when running the wf that suggests a problem until it hits the ksampler then it dies. freezes and floods the page with errors. the same k sampler works in every other workflow I run@@PixelEasel

[@Halfgawd_Halfdevil]

@@PixelEasel i have tried to attach a copy of the errors but I can't attach a file here and I can paste it in the text because it is so large it wont allow me to send it

[@Halfgawd_Halfdevil]

@@PixelEasel (mat1 and mat2 shapes cannot be multiplied (154x2048 and 768x320)) this is the first line of the error. all my settings are at 512 but I changed that from 768. however it looks like the example in the video is at 512 so I don't know why it would be a problem. do you have to use a specific model checkpoint or something like that? do both of the control net models have to be the same in the aio preprocessor and advanced control net block as in to they both have to depthanythingpreprocessor? it does say on huggingface that there are errors int that model but this does run for me past those points and fails at the k sampler. so I am just trying to figure out why My comfy Install works with everything else I try but not this workflow

[@PixelEasel]

what checkpoint do you use? it should be 1.5