"Copy weights from old conv layer" occurred error

[YSL0226]

Hi I run into this error.
Should I just comment the line "conv_new.weight.data[:, :4] = unet_vton.conv_in.weight.data"?

File "/OOTDiffusion-train/ootd_train.py", line 208, in
conv_new.weight.data[:, :4] = unet_vton.conv_in.weight.data # Copy weights from old conv layer
RuntimeError: The expanded size of the tensor (4) must match the existing size (8) at non-singleton dimension 1. Target sizes: [320, 4, 3, 3]. Tensor sizes: [320, 8, 3, 3]

[GeneralWhite]

#5
I've encountered this problem before
: )

[edith-wq]

I put the config and safetensor files for the sd1.5 model in the unet_vton_train folder and the unet_garm_train folder, is this correct?

[YSL0226]

#5 I've encountered this problem before : )

@GeneralWhite could you share the sd1.5 url? I download form huggingface, but it still have some problems

[GeneralWhite]

#5 I've encountered this problem before : )

@GeneralWhite could you share the sd1.5 url? I download form huggingface, but it still have some problems

I download the sd1.5 from huggingface too.
I tried to solve this problem last week, but I have never trained a diffusion model, so I still have some problems. I'm not sure if it's my problem, and I still can't run his code.

[GeneralWhite]

#5 I've encountered this problem before : )

@GeneralWhite could you share the sd1.5 url? I download form huggingface, but it still have some problems

Although I changed the model to sd1.5, this error also occurs