PyTorch

PyTorch

Original author(s)	Adam Paszke Sam Gross Soumith Chintala Gregory Chanan
Developer(s)	Meta AI
Initial release	September 2016(2016-09)
Stable release	2.4.0 / 24 July 2024(24 July 2024)
Repository	github/pytorch/pytorch
Written in	Python C++ CUDA
Operating system	Linux macOS Windows
Platform	IA-32,x86-64,ARM64
Available in	English
Type	Libraryformachine learninganddeep learning
License	BSD-3
Website	pytorch.org

PyTorchis amachine learninglibrarybased on theTorchlibrary,used for applications such ascomputer visionandnatural language processing,originally developed byMeta AIand now part of theLinux Foundationumbrella.It is recognized as one of the two most popular machine learning libraries alongsideTensorFlow,offeringfree and open-source softwarereleased under themodified BSD license.Although thePythoninterface is more polished and the primary focus of development, PyTorch also has aC++interface.

A number of pieces ofdeep learningsoftware are built on top of PyTorch, includingTesla Autopilot,Uber's Pyro,Hugging Face's Transformers,PyTorch Lightning,and Catalyst.

PyTorch provides two high-level features:

• Tensor computing (likeNumPy) with strong acceleration viagraphics processing units(GPU)
• Deep neural networksbuilt on a tape-basedautomatic differentiationsystem

History

Meta (formerly known as Facebook) operates bothPyTorchandConvolutional Architecture for Fast Feature Embedding(Caffe2), but models defined by the two frameworks were mutually incompatible. The Open Neural Network Exchange (ONNX) project was created by Meta andMicrosoftin September 2017 for converting models between frameworks. Caffe2 was merged into PyTorch at the end of March 2018.In September 2022, Meta announced thatPyTorchwould be governed by PyTorch Foundation, a newly created independent organization – a subsidiary ofLinux Foundation.

PyTorch 2.0 was released on 15 March 2023.

PyTorch tensors

Main article:Tensor (machine learning)

PyTorch defines a class called Tensor (torch.Tensor) to store and operate on homogeneous multidimensional rectangular arrays of numbers. PyTorch Tensors are similar toNumPyArrays, but can also be operated on aCUDA-capableNVIDIAGPU. PyTorch has also been developing support for other GPU platforms, for example, AMD'sROCmand Apple'sMetal Framework.

PyTorch supports various sub-types of Tensors.

Note that the term "tensor" here does not carry the same meaning as tensor in mathematics or physics. The meaning of the word in machine learning is only superficially related to its original meaning as a certain kind of object inlinear algebra.Tensors in PyTorch are simply multi-dimensional arrays.

PyTorch neural networks

Main article:Neural network (machine learning)

PyTorch defines a module called nn (torch.nn) to describe neural networks and to support training. This module offers a comprehensive collection of building blocks for neural networks, including various layers and activation functions, enabling the construction of complex models. Networks are built by inheriting from thetorch.nnmodule and defining the sequence of operations in theforward()function.

Example

The following program shows the low-level functionality of the library with a simple example

importtorch
dtype=torch.float
device=torch.device("cpu")# Execute all calculations on the CPU
# device = torch.device( "cuda:0" ) # Executes all calculations on the GPU

# Create a tensor and fill it with random numbers
a=torch.randn(2,3,device=device,dtype=dtype)
print(a)
# Output: tensor([[-1.1884, 0.8498, -1.7129],
# [-0.8816, 0.1944, 0.5847]])

b=torch.randn(2,3,device=device,dtype=dtype)
print(b)
# Output: tensor([[ 0.7178, -0.8453, -1.3403],
# [ 1.3262, 1.1512, -1.7070]])

print(a*b)
# Output: tensor([[-0.8530, -0.7183, 2.58],
# [-1.1692, 0.2238, -0.9981]])

print(a.sum())
# Output: tensor(-2.1540)

print(a[1,2])# Output of the element in the third column of the second row (zero based)
# Output: tensor(0.5847)

print(a.max())
# Output: tensor(0.8498)

The following code-block shows an example of the higher level functionality providednnmodule. A neural network with linear layers is defined in the example.

importtorch
fromtorchimportnn# Import the nn sub-module from PyTorch

classNeuralNetwork(nn.Module):# Neural networks are defined as classes
def__init__(self):# Layers and variables are defined in the __init__ method
super().__init__()# Must be in every network.
self.flatten=nn.Flatten()# Construct a flattening layer.
self.linear_relu_stack=nn.Sequential(# Construct a stack of layers.
nn.Linear(28*28,512),# Linear Layers have an input and output shape
nn.ReLU(),# ReLU is one of many activation functions provided by nn
nn.Linear(512,512),
nn.ReLU(),
nn.Linear(512,10),
)

defforward(self,x):# This function defines the forward pass.
x=self.flatten(x)
logits=self.linear_relu_stack(x)
returnlogits

See also

• Comparison of deep learning software
• Differentiable programming
• DeepSpeed

External links

• Official website