AInseven
AInseven
图14:12左图，坐标 2 3，应为坐标3 2
I am also very interested in reviewing these documents to understand more about the architecture. Would you be able to share them with me as well? My email is [email protected]...
**controlnet_frames and input_frames have same length, but still error, any ideas what can I do? Thank you.** I checked stable_diffusion_video.py, in line 315, ```python def add_data_to_pipeline_inputs(self, data, pipeline_inputs): pipeline_inputs["input_frames"] =...
It seems that the resolution of the output image sequence is 1x576, and the VHS video combine node can't handle this. you may check the resolution of empty latent node.
@aswordok There is a dubug node that can check the tensor size,you can use it to find where the img size goes to 1,1,n,3
> @cardenluo I also got this problem a moment ago, and I think you could just change the fold name from 'ComfyUI-fastblend-main' to 'AInseven', then it will go well. @AInseven...
You need install CUDA Toolkit. https://developer.nvidia.com/cuda-downloads
Full error message,please. It's very helpful to read if ues Python markdown format, but whatever, full message.
是的，可以适当降低 Window_size
The main code is from https://github.com/Artiprocher/sd-webui-fastblend I just made a tiny alteration, shifting to comfyUI. So, I can't answer your question.