Tutorial: PyTorch Tutorial List

Preparation for PyTorch Tutorial

PyTorch Installation

Install PyTorch

Follow the installatin instruction: See here for more detail

You should install pytorch in a virtual environment

Check PyTorch Installation and GPU

In the Anaconda Promt, type

conda activate py39
python
import torch
torch.__version__
print("cuda" if torch.cuda.is_available() else "cpu")

Watch PyTorch Intro Video

Introduction to PyTorch (20min)

You need to know 'What is Tensor in Pytorch'

Follow Quick-Start Tutorial:

• Pytorch Tutorial(ENG)
• Pytorch Tutorial(KOR)

---

DLIP Course Tutorials

MLP

• T1: Train a simple MLP and Test with loading trained model (MNIST)

CNN- Classification

Create a simple CNN model

• T2-1: Create LeNeT CNN model, Train with opendataset, and Test with loading trained model (CIFAR10)
• T2-2-1: Create a CNN model(VGG-16) for ImageNet
• T2-2-2: Create, Train and Test a CNN model(VGG-16) for CIFAR10

Using Popular CNN models from torchvision.models

• T3-1: Test using Pretrained Model (VGG, Inception, ResNet)
• T3-2: Train Opendataset with Transfer Learning of Pretrained model

Assignment: Classification

• T3-3: (Assignment) Classification with Custom Dataset
• T3-4: (Assignment) Create ResNet-50 model for ImageNet

CNN- Object Detection

YOLO v8 in PyTorch

• T4-1: Install and Inference using YOLOv8
• T4-2: Train and Test using Custom Dataset

YOLO v5 in PyTorch

• T4-1(option1): Pretrained YOLOv5 with COCO dataset (in CoLab)
• T4-1(option2): Pretrained YOLOv5 with COCO dataset (in VS Code, Local PC)
• T4-2: Train YOLOv5 with a Custom Dataset (in VS Code, Local PC)

---

Useful Sites

Pytorch tutorial codes: Pytorch-Tutorial

Pytorch Tutorial List: PyTorch Tutorial List