Chapter 1: Introduction & Installation¶

🎓 Learning Objectives

Understand what PyTorch is and why it's popular
Install PyTorch correctly for your system
Set up your development environment
Write your first PyTorch program

What is PyTorch?¶

PyTorch is an open-source machine learning library developed by Facebook's AI Research lab (FAIR). It provides:

Dynamic Computational Graphs: Build and modify networks on-the-fly
Python-First: Pythonic API that feels natural
GPU Acceleration: Easy transfer between CPU and GPU
Strong Community: Extensive ecosystem and support

Why PyTorch?

PyTorch's dynamic computation graph makes it ideal for research and experimentation. Unlike static graph frameworks, you can modify your network architecture during runtime, making debugging much easier.

Getting Started

If you're new to deep learning, PyTorch is an excellent choice because of its intuitive API and excellent documentation. Start with simple examples and gradually build complexity.

Why PyTorch?¶

Advantages¶

✅ Intuitive and easy to learn
✅ Dynamic computation graphs (define-by-run)
✅ Excellent for research and production
✅ Strong debugging capabilities (standard Python debuggers work)
✅ Extensive pre-trained models via torchvision, torchaudio, etc.

Use Cases¶

Deep Learning Research
Computer Vision (CNNs)
Natural Language Processing (Transformers)
Reinforcement Learning
Generative Models (GANs, VAEs)

Installation¶

Prerequisites¶

# Check Python version (3.8+ recommended)
python --version

Install PyTorch¶

CPU OnlyCUDA 11.8CUDA 12.1macOS (MPS)

pip install torch torchvision torchaudio

pip install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu118

pip install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu121

pip install torch torchvision torchaudio

Installation Tip

For the best performance, use GPU-enabled PyTorch if you have an NVIDIA GPU. Check your CUDA version with nvidia-smi before installing. For Apple Silicon Macs, PyTorch automatically uses MPS (Metal Performance Shaders) for acceleration.

Virtual Environments

Always install PyTorch in a virtual environment to avoid conflicts with other packages. Use venv or conda to create isolated environments.

Verify Installation¶

import torch
import torchvision

# Check PyTorch version
print(f"PyTorch version: {torch.__version__}")

# Check CUDA availability
print(f"CUDA available: {torch.cuda.is_available()}")
if torch.cuda.is_available():
    print(f"CUDA version: {torch.version.cuda}")
    print(f"GPU device: {torch.cuda.get_device_name(0)}")

# Check MPS (Apple Silicon) availability
print(f"MPS available: {torch.backends.mps.is_available()}")

Expected Output:

PyTorch version: 2.1.0
CUDA available: True
CUDA version: 11.8
GPU device: NVIDIA GeForce RTX 3090
MPS available: False

PyTorch Ecosystem¶

Core Libraries¶

Library	Purpose	Installation
`torch`	Core library with tensors and autograd	`pip install torch`
`torchvision`	Computer vision utilities	`pip install torchvision`
`torchaudio`	Audio processing	`pip install torchaudio`
`torchtext`	NLP utilities	`pip install torchtext`

Additional Tools¶

Tool	Purpose
TensorBoard	Visualization and monitoring
ONNX	Model export for deployment
PyTorch Lightning	High-level training framework
Hugging Face Transformers	Pre-trained transformer models

First PyTorch Program¶

import torch

# Create a simple tensor
x = torch.tensor([1.0, 2.0, 3.0])
print(f"Tensor: {x}")
print(f"Shape: {x.shape}")
print(f"Data type: {x.dtype}")
print(f"Device: {x.device}")

Understanding Tensor Properties

Always check shape, dtype, and device when working with tensors. These properties determine how operations behave and where computations run.

Default Behavior

By default, tensors are created on CPU with float32 dtype. You can move them to GPU later with .to('cuda') or .cuda().

Output:

Tensor: tensor([1., 2., 3.])
Shape: torch.Size([3])
Data type: torch.float32
Device: cpu

Hello World Neural Network¶

import torch
import torch.nn as nn

# Define a simple neural network
class SimpleNet(nn.Module):
    def __init__(self):
        super(SimpleNet, self).__init__()
        self.fc = nn.Linear(10, 1)  # 10 inputs, 1 output

    def forward(self, x):
        return self.fc(x)

# Create model instance
model = SimpleNet()
print(model)

# Create random input
x = torch.randn(5, 10)  # batch_size=5, features=10

# Forward pass
output = model(x)
print(f"Input shape: {x.shape}")
print(f"Output shape: {output.shape}")

Output:

SimpleNet(
  (fc): Linear(in_features=10, out_features=1, bias=True)
)
Input shape: torch.Size([5, 10])
Output shape: torch.Size([5, 1])

Development Environment Setup¶

Recommended IDEs¶

Jupyter Notebook - Interactive development
```
pip install jupyter
jupyter notebook
```
VS Code - Full-featured editor
Install Python extension
Install Pylance for IntelliSense
PyCharm - Professional Python IDE
Configure PyTorch as a library

IDE Recommendation

For learning and experimentation, Jupyter Notebooks are excellent. For larger projects, VS Code or PyCharm provide better code organization and debugging tools.

Interactive Development

PyTorch's dynamic nature makes it perfect for interactive development. Use Jupyter notebooks to experiment with different architectures and see results immediately.

Useful Packages¶

# Scientific computing
pip install numpy pandas matplotlib seaborn

# Progress bars
pip install tqdm

# Experiment tracking
pip install tensorboard wandb

# Model summary
pip install torchsummary torchinfo

Common Imports¶

# Core PyTorch
import torch
import torch.nn as nn
import torch.nn.functional as F
import torch.optim as optim
from torch.utils.data import Dataset, DataLoader

# Vision
import torchvision
import torchvision.transforms as transforms
from torchvision import models

# Utilities
import numpy as np
import matplotlib.pyplot as plt
from tqdm import tqdm

Quick Start Template¶

import torch
import torch.nn as nn
import torch.optim as optim

# Set random seed for reproducibility
torch.manual_seed(42)

# Set device
device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')
print(f"Using device: {device}")

Reproducibility

Always set random seeds (torch.manual_seed()) for reproducibility. Also set seeds for NumPy and Python's random module if you use them. This ensures your experiments are reproducible.

Device Management

The device variable pattern is a best practice. It allows your code to work on both CPU and GPU without modification. Always move models and data to the same device.

Define model¶

class MyModel(nn.Module): def init(self): super(MyModel, self).init() # Define layers here

def forward(self, x):
    # Define forward pass
    return x

Initialize model¶

model = MyModel().to(device)

Define loss and optimizer¶

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

print("Setup complete!") ```

Next Steps¶

Now that you have PyTorch installed and understand the basics, proceed to:

Chapter 2: Tensors Basics - Learn about PyTorch tensors
Chapter 3: Tensor Operations - Master tensor manipulations