DL-basic

Installation

https://pytorch.kr/get-started/locally/

## 아나콘다 가상환경, 파이토치 설치.
conda create --name DL python=3.8
conda install pytorch==2.0.1 torchvision==0.15.2 torchtext==0.15.2 torchaudio==2.0.2 pytorch-cuda=11.7 -c pytorch -c nvidia

conda create --name DL python=3.9
conda install pytorch torchvision torchaudio torchtext=0.18.0 pytorch-cuda=11.8 -c pytorch -c nvidia

conda install pytorch-lightning==2.0.0 hydra-core==1.3.2 -c conda-forge
conda install tensorboard tensorboardX

## 주피터 노트북 커널
conda install pexpect jupyter
pip install pexpect jupyter ipykernel
pip uninstall pyzmq
pip install pyzmq

## 라이브러리 설치
pip install timm transformers

설치가 정상적으로 되었는지 검사

import torch

torch.__version__ ## '2.3.1'
torch.cuda.is_available() ## True

Introduction

가장 좋은 공부 방법은 가르치는 것이다. 가르치는 것은 강력한 학습 방법이다

딥러닝을 독학하면서 구체적이고 명확한 개념서를 만들어 다른 사람에게 설명하고 있다고 가정하며 만든 기본 개념서입니다.

• 제 블로그https://pervin0527.notion.site/DL-Basic 에서 주제별로 정리한 포스트를 볼 수 있습니다.
• Repository에 있는 코드는 Python과 Pytorch로 작성되어 있습니다.

Contents

1 & 2. Forward, BackPropagation

• Post
  • Forward Propagation
  • Derivative For Backpropagation
  • BackPropagation
• Code
  • 01-Basic & Logistic Regression
  • 02-Multi layer logistic regression

3.Activation functions & Optimizer

• Post
• Code

4.Data Pre-Processing, Weight Initialization

• Post
• Code

5. Validation, Test dataset / Overfitting & Underfitting / Regularization

• Post
• Code

6.Convolutional Neural Network

• Post
• Code

7.BatchNormalization

• Post
• Code

8.[VGGNet]Very Deep Convolutional Neural Network

• Post
• Code

9.[ResNet]Residual Learning for Image Recongnition

• Post
• Code

10.[MobileNet]Efficient Convolutional Neural Networks for Mobile Vision Applications

• Post
• Code

11.[MobileNetV2]Inverted Residuals and Linear Bottlenecks

• Post
• Code

12.[EfficientNet]Rethinking Model Scaling for Convolutional Neural Networks

• Post
• Code

13.[RNN]Sequence Data & Recurrent Neural Network

• Post
• Code

14.[Seq2Seq & Attention]Neural Machine Translation by Jointly Learning to Align and Translate

• Post
• Code

15.[Transformer]Attention is all you need.

• Post
• Code

16.[Vision Transformer]AN IMAGE IS WORTH 16X16 WORDS: TRANSFORMERS FOR IMAGE RECOGNITION AT SCALE

• Post
• Code

17.[DETR]End-to-End Object Detection with Transformers

• Post
• Code

18.[GAN]Generative Adversarial Networks

• Post
• Code

19.[DCGAN]UNSUPERVISED REPRESENTATION LEARNING WITH DEEP CONVOLUTIONAL GENERATIVE ADVERSARIAL NETWORKS

• Post
• Code

20.[cGAN]Conditional Generative Adversarial Nets

• Post
• Code

21.[WGAN]Wasserstein GAN

• Posrt
• Code