PyTorch Starting - Python

PyTorch is an open source machine learning framework that was developed by Facebook's AI Research team. It is based on Torch, which is a scientific computing framework.

PyTorch provides two main features:

• An n-dimensional Tensor, which is similar to NumPy arrays, but can run on GPUs
• Automatic differentiation for building and training neural networks

Installation

To install PyTorch, you need to have Python installed on your system. You can install PyTorch using pip by running the following command:

pip install torch

You can also install PyTorch using Anaconda by running the following command:

conda install pytorch torchvision -c pytorch

Getting Started

Let's start by importing PyTorch and creating a simple tensor:

import torch

x = torch.tensor([[1, 2], [3, 4]])
print(x)

Output:

tensor([[1, 2],
        [3, 4]])

We can also create tensors of a specific size and type:

# Create a tensor of size 3x3 with random values
x = torch.rand(3, 3)
print(x)

# Create a tensor of size 2x2 with zeros
x = torch.zeros(2, 2)
print(x)

# Create a tensor of size 2x2 with ones
x = torch.ones(2, 2)
print(x)

# Create a tensor of size 2x2 with a specific value (5)
x = torch.full((2, 2), 5)
print(x)

# Create a tensor of size 2x2 with values from a normal distribution (mean=0, std=1)
x = torch.randn(2, 2)
print(x)

Output:

tensor([[0.6176, 0.5149, 0.9247],
        [0.2104, 0.3562, 0.1681],
        [0.2510, 0.1128, 0.6265]])
tensor([[0., 0.],
        [0., 0.]])
tensor([[1., 1.],
        [1., 1.]])
tensor([[5, 5],
        [5, 5]])
tensor([[-1.0983, -0.3215],
        [ 1.2306, -0.6546]])

Automatic differentiation

PyTorch provides a module called autograd, which allows for automatic differentiation of tensors. This means that it can compute the gradient of any function with respect to its inputs.

Let's create a tensor and compute its gradient:

import torch
x = torch.tensor([2.0], requires_grad=True)
y = x**2 + 3*x + 1
y.backward()
print(x.grad)

Output:

tensor([7.])

In this example, x is a tensor with a value of 2. We set requires_grad=True to enable gradient computation for x. We then create a new tensor y, which is a function of x. In this case, y is a polynomial function. Finally, we call the backward method on y to compute its gradient with respect to x. We can then use x.grad to get the gradient of y with respect to x.

Conclusion

In this tutorial, we introduced PyTorch and its main features. We demonstrated how to install PyTorch and how to create and manipulate tensors. We also showed how to use automatic differentiation to compute the gradient of a function with respect to its inputs. PyTorch is a powerful tool for building and training deep neural networks, and we hope that this tutorial has helped you get started with PyTorch!