Pasang PyTorch dan Verifikasi Impornya

Before you train anything, you need a toolkit. PyTorch is the library you will use to build and train every model in this course. 🔧

PyTorch installs like any other Python package, so you grab it with pip. No special compiler or magic setup is needed to get started.

This single pip install line pulls down PyTorch and its companion packages so your environment is ready to build models.

The torchvision package ships datasets, image transforms, and pretrained models, so you rarely start an image project from scratch.

Once installed, you bring PyTorch into any script with one line. The convention is to import torch and use it directly, no alias needed.

To prove the install worked, print the version. If a number appears instead of an error, PyTorch is alive and ready.

A version like 2.3.0 means a successful install. The first number is the major version, which signals big, sometimes breaking, changes.

A ModuleNotFoundError means Python cannot find torch. Usually you installed into a different environment than the one running your code.

A virtual environment keeps PyTorch and its versions isolated per project, so one upgrade never breaks another app on your machine.

Always activate the environment before you pip install, so the package lands where your scripts will actually look for it.

The default CPU build runs everything in this course. You only need a GPU-specific install later, once your models grow larger.

Let's confirm you know the fastest way to test the install.

You installed PyTorch with pip in a virtual environment and confirmed it works by printing its version. Your workshop is open for business. 🎉