Keras Conv2D layer to PyTorch Conv2D layer

1

I was wondering if:

Conv2D(32, (3, 3), padding='same',input_shape=(32,32,3))

is equivalent to

nn.Conv2d(3,32,kernel_size=(3,3), padding=1)

Thank you in advance.

2

Should be the same besides probably the parameter initialization.

3

Can you pick unique names for the input args instead of numbers? Many duplicates makes it a bit unclear the mappings.