How to do it...

1. First, we install PyTorch in our Anaconda environment, as follows:

conda install pytorch torchvision cuda80 -c soumith

If you want to install PyTorch on another platform, you can have a look at the PyTorch website for clear guidance: http://pytorch.org/.

1. Let's import PyTorch into our Python environment:

import torch

1. While Keras provides higher-level abstraction for building neural networks, PyTorch has this feature built in. This means one can build with higher-level building blocks or can even build the forward and backward pass manually. In this introduction, we will use the higher-level abstraction. First, we need to set the size of our random training data:

 batch_size = 32
 input_shape = 5
 output_shape = 10

1. To make use of GPUs, we will cast the tensors as follows:

torch.set_default_tensor_type('torch.cuda.FloatTensor') 

This ensures that all computations will use the attached GPU. 

1. We can use this to generate random training data:

from torch.autograd import Variable
X = Variable(torch.randn(batch_size, input_shape))
y = Variable(torch.randn(batch_size, output_shape), requires_grad=False)

1. We will use a simple neural network having one hidden layer with 32 units and an output layer:

model = torch.nn.Sequential(
          torch.nn.Linear(input_shape, 32),
          torch.nn.Linear(32, output_shape),
        ).cuda()

We use the .cuda() extension to make sure the model runs on the GPU. 

1. Next, we define the MSE loss function:

loss_function = torch.nn.MSELoss()

1. We are now ready to start training our model for 10 epochs with the following code:

learning_rate = 0.001
for i in range(10):
    y_pred = model(x)
    loss = loss_function(y_pred, y)
    print(loss.data[0])
    # Zero gradients
    model.zero_grad()
    loss.backward()

    # Update weights
    for param in model.parameters():
        param.data -= learning_rate * param.grad.data