Linaqruf / sd-notebook-collection

209 stars 42 forks source link

RuntimeError #22

Open Cheesedozer opened 1 year ago

Cheesedozer commented 1 year ago

Hello,

Whenever I try to generate images using a Lora I encounter this error:

RuntimeError: Expected all tensors to be on the same device, but found at least two devices, cuda:0 and cpu! (when checking argument for argument mat2 in method wrapper_CUDA_mm)

Is there any way to fix it?

Linaqruf commented 1 year ago

Hi, I believe it's because your LoRA/ControlNet has different version with your main model, for example you can't load SDv1.x LoR if your main model is Illuminati Diffusion V1.1 because that model is SDv2.1 based

Cheesedozer commented 1 year ago

So I can't use a Lora if it is not made with the same model version?