Quick Start

In this section, we would like to show an overview to give a quick start.

Operation

There are some operations supported in tednet, and it is convinient to use them.

import tednet as tdt

Create matrix whose diagonal elements are ones

diag_matrix = tdt.eye(5, 5)
print(diag_matrix)

tensor([[1., 0., 0., 0., 0.],
        [0., 1., 0., 0., 0.],
        [0., 0., 1., 0., 0.],
        [0., 0., 0., 1., 0.],
        [0., 0., 0., 0., 1.]])

Take Pytorch tensor to Numpy narray

print(type(diag_matrix))

<class 'torch.Tensor'>

diag_matrix = tdt.to_numpy(diag_matrix)

print(type(diag_matrix))

<class 'numpy.ndarray'>

Take Numpy narray to Pytorch tensor

diag_matrix = tdt.to_tensor(diag_matrix)

print(type(diag_matrix))

<class 'torch.Tensor'>

Tensor Decomposition Networks (Tensor Ring for Sample)

To use tensor ring decomposition models, simply calling the tensor ring module is enough.

import tednet.tnn.tensor_ring as tr

Here, we would like to give a case of building the TR-LeNet5.

# Define a TR-LeNet5
model = tr.TRLeNet5(10, [6, 6, 6, 6])

compression_ration is:  0.3968253968253968
compression_ration is:  14.17233560090703
compression_ration is:  241.54589371980677
compression_ration is:  2.867383512544803