How to avoid low-performance memory layout?

I got the warning from following code.

long-video-gan/torch_utils/ops/filtered_lrelu.py

Lines 208 to 211 in 4151a8b

    
           # Warn if input storage strides are not in decreasing order due to e.g. channels-last layout. 
        
           strides = [x.stride(i) for i in range(x.ndim) if x.size(i) > 1] 
        
           if any(a < b for a, b in zip(strides[:-1], strides[1:])): 
        
               warnings.warn("low-performance memory layout detected in filtered_lrelu input", RuntimeWarning)

I wonder how to debug the problem in proper way. e.g. swapping before filteredLRelu?, adjusting tensor size at the begging of model?
Thank you:)

Hi @kingsj0405, I am not sure, since the filtered lrelu code was written by StyleGAN3 authors and was copied for this project. I recommend asking on the StyleGAN3 repo.

	# Warn if input storage strides are not in decreasing order due to e.g. channels-last layout.
	strides = [x.stride(i) for i in range(x.ndim) if x.size(i) > 1]
	if any(a < b for a, b in zip(strides[:-1], strides[1:])):
	warnings.warn("low-performance memory layout detected in filtered_lrelu input", RuntimeWarning)