for example if you have ;

input= torch.rand([0,1,1,66650])

m=nn.MaxPool2d(1,110)

output=m(input1)

why do i get the following error?

RuntimeError: non-empty 3D or 4D input tensor expected but got ndim: 4

for example if you have ;

input= torch.rand([0,1,1,66650])

m=nn.MaxPool2d(1,110)

output=m(input1)

why do i get the following error?

RuntimeError: non-empty 3D or 4D input tensor expected but got ndim: 4

You are creating an empty tensor in:

```
input = torch.rand([0,1,1,66650])
print(input)
> tensor([], size=(0, 1, 1, 66650))
```

by setting the size of `dim0`

to zero.

As the error message suggests, a non-empty tensor would be required.

thank you. but in my case i was being stupid to feed my network with more data then required number of batches while iterating. so this error kept coming.