Supporting Blobs with num_axes() != 4 in pycaffe interface

[shaibagon]

Currently, Blobs (especially inputs to net) with num_axes() != 4 are not supported by pycaffe interface.
This is because pycaffe uses .num property of Blob that is only supported for 4D Blobs.

This PR attempts to correct this by removing all references to Blob.num property (that assumes Blob is 4D).
Replacing it with accessing Blob.shape[0] that is good also for Blobs with num_axes() != 4

This is a very minor change code-wise

[shelhamer]

Thanks for this small patch to make the pycaffe keyword args work.

More precisely, this is only an issue for num_axes > 4 and when passing blobs= or diffs= keyword args to the pycaffe forward(), backward() respectively. The pycaffe blob interface, the Net.{blobs,params} properties, and Net.{forward(),backward()} methods all work in N-D.

[shaibagon]

@shelhamer
Thanks for accepting this PR.