PyTorch Tutorial - Part 1

This code is provided as supplementary material of the lecture Machine Learning and Optimization in Communications (MLOC).

This code illustrates

Get accustomed to the basics of pytorch
Do simple operations



In [2]:

    
import torch
import numpy as np

device = 'cuda' if torch.cuda.is_available() else 'cpu'
print("We are using the following device for learning:",device)









    



We are using the following device for learning: cpu

Generate some tensors, convert to float, and copy them to the device (possibly the GPU). Then carry out an operation with the tensors and print the result.



In [3]:

    
a = torch.tensor([1,2,3]).float().to(device)
b = torch.tensor([5,6,7]).float().to(device)
result = a * b
print(a)
print(b)
print(result)









    



tensor([1., 2., 3.])
tensor([5., 6., 7.])
tensor([ 5., 12., 21.])

We can see that the objects a, b, and c are pyTorch tensors that also carry the result as such. We have built a computation graph linking variables a and b to the result via the (element-wise) multiplication and have the result immediately ready.



In [6]:

    
# generate a random number vector
d = np.random.randn(2,4)

# construct a pytorch tensor from numpy array
e = torch.from_numpy(d).float().to(device)
print(e)









    



tensor([[-0.2350, -0.0184, -0.1093,  0.8630],
        [ 0.7927,  1.0190,  0.9225,  0.0544]])



In [ ]: