Hi
I was trying to test the model on my dataset but it seems that in the code it's assumed (hard-coded) that the input always has only one sequence of frames like (1,3,3,256,256) and when I tried to input it with more than one sequences(batch size greater than 1) I got error for resizing:
File ~/MIA-VSR/archs/mia_sliding_arch.py", line 473, in forward
xh = xh.view(1,h*w,c)
^^^^^^^^^^^^^^^^
RuntimeError: shape '[1, 4096, 120]' is invalid for input of size xxxxxxx
Could you please confirm this or tell me how can I feed it with batch size greater than 1?
Hi
I was trying to test the model on my dataset but it seems that in the code it's assumed (hard-coded) that the input always has only one sequence of frames like (1,3,3,256,256) and when I tried to input it with more than one sequences(batch size greater than 1) I got error for resizing:
File ~/MIA-VSR/archs/mia_sliding_arch.py", line 473, in forward
xh = xh.view(1,h*w,c)
^^^^^^^^^^^^^^^^
RuntimeError: shape '[1, 4096, 120]' is invalid for input of size xxxxxxx
Could you please confirm this or tell me how can I feed it with batch size greater than 1?