Scribd
Upload
28 views4 pages

PyTorch Guide With Code

This document is a comprehensive guide to PyTorch, covering its installation, tensor operations, autograd, neural network module, loss functions, optimizers, training loops, and data handling with Dataset and DataLoader. It includes code snippets for practical understanding and demonstrates how to save and load models. PyTorch is highlighted as an open-source framework for building and training neural networks, with GPU support for enhanced performance.

Uploaded by

Aslam
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
28 views4 pages

PyTorch Guide With Code

This document is a comprehensive guide to PyTorch, covering its installation, tensor operations, autograd, neural network module, loss functions, optimizers, training loops, and data handling with Dataset and DataLoader. It includes code snippets for practical understanding and demonstrates how to save and load models. PyTorch is highlighted as an open-source framework for building and training neural networks, with GPU support for enhanced performance.

Uploaded by

Aslam
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 4

PyTorch Guide with Code and Explanations

Introduction to PyTorch

PyTorch is an open-source deep learning framework developed by Facebook AI Research. It is widely used

for building and training neural networks.

# Installation: pip install torch torchvision

import torch

import torch.nn as nn

import torch.optim as optim

import torchvision

import torchvision.transforms as transforms

Tensors

Tensors are the fundamental data structures in PyTorch, similar to NumPy arrays but with GPU support.

# Tensors

a = torch.tensor([1, 2, 3])

b = torch.tensor([4, 5, 6])

c = a + b

print(c)

print(a.device)

CUDA Support (GPU)

Move tensors to GPU using `.to('cuda')` or `.cuda()` if available.

# CUDA

a = torch.tensor([1.0, 2.0]).to('cuda') if torch.cuda.is_available() else

torch.tensor([1.0, 2.0])

Tensor Operations

Includes indexing, slicing, reshaping, and arithmetic.

# Operations

t = torch.randn(3, 3)
PyTorch Guide with Code and Explanations

print(t[0])

print(t.view(-1))

Autograd and Backpropagation

Autograd automatically computes gradients for tensor operations.

# Autograd

x = torch.ones(2, 2, requires_grad=True)

y = x + 2

z = y * y * 3

out = z.mean()

out.backward()

print(x.grad)

Neural Network Module

Use `nn.Module` to define a neural network architecture.

# Neural Net

class Net(nn.Module):

def __init__(self):

super(Net, self).__init__()

self.fc1 = nn.Linear(784, 128)

self.relu = nn.ReLU()

self.fc2 = nn.Linear(128, 10)

def forward(self, x):

x = self.relu(self.fc1(x))

x = self.fc2(x)

return x

Loss Functions

PyTorch provides multiple loss functions like MSE, CrossEntropy, etc.


PyTorch Guide with Code and Explanations

# Loss

criterion = nn.CrossEntropyLoss()

Optimizers

Used to update weights using gradients (SGD, Adam, etc.).

# Optimizer

optimizer = optim.Adam(model.parameters(), lr=0.001)

Training Loop

Training involves forward pass, loss calculation, backward pass, and optimizer step.

# Training Loop

for epoch in range(10):

for data, target in dataloader:

optimizer.zero_grad()

output = model(data)

loss = criterion(output, target)

loss.backward()

optimizer.step()

Dataset and DataLoader

PyTorch provides `Dataset` and `DataLoader` for loading and batching data.

# DataLoader

transform = transforms.ToTensor()

trainset = torchvision.datasets.MNIST(root='./data', train=True, download=True,

transform=transform)

dataloader = torch.utils.data.DataLoader(trainset, batch_size=64, shuffle=True)

Saving and Loading Models

Save and load model weights using `torch.save` and `torch.load`.


PyTorch Guide with Code and Explanations

# Save

torch.save(model.state_dict(), 'model.pth')

# Load

model.load_state_dict(torch.load('model.pth'))

model.eval()

You might also like