Issue using your pipeline
#1
by nudelbrot - opened
I'm using a fine tuned diffusers model with your pipeline, but it errors out. Using diffusers main.
stable_diffusion_depth_pipeline.py", line 195, in __call__
latent_model_input = torch.cat([latent_model_input, depth], dim=1)
RuntimeError: Sizes of tensors must match except in dimension 1. Expected size 48 but got size 32 for tensor number 1 in the list.
any ideas? My depthmap is the output of https://huggingface.co/spaces/nielsr/dpt-depth-estimation
I’d suggest using diffusers official pipeline and model now: https://huggingface.co/stabilityai/stable-diffusion-2-depth#examples
However the most up-to-date version of my pipeline is here: https://github.com/carson-katri/dream-textures/blob/main/generator_process/actions/depth_to_image.py
Note that some of it is specific to the Dream Textures codebase.