Tensors Operations

This Notebook is divided in 4 objectives.

Creating Tensors in different ways with requires_grad property
Comparing Tensor creation memory model
Accessing Tensors
Understanding WTH is requires_grad parameters

# Import torch and other required modules
import torch
import numpy as np

Objective-1 - Creating Tensors in different ways with requires_grad property

Calculating time to perform each type of tensors

Create all_one tensor with 2 internal matrixes where each matrix has 6 rows and 3 cols:

%time 
all_one = torch.ones([2,6,3], dtype=torch.float64)
all_one

CPU times: user 3 µs, sys: 1e+03 ns, total: 4 µs
Wall time: 7.15 µs

tensor([[[1., 1., 1.],
         [1., 1., 1.],
         [1., 1., 1.],
         [1., 1., 1.],
         [1., 1., 1.],
         [1., 1., 1.]],

        [[1., 1., 1.],
         [1., 1., 1.],
         [1., 1., 1.],
         [1., 1., 1.],
         [1., 1., 1.],
         [1., 1., 1.]]], dtype=torch.float64)