Pytorch GPU 사용법

2023-04-05 10 분 소요

pytorch GPU 튜토리얼

GPU가 사용 가능한지 확인

import torch

torch.cuda.is_available() # 기본 CPU이기 때문에 실패

True

import torch

# ... 메뉴로 이동해서 Accelerator - GPU T4 x2로 변경한다.
torch.cuda.is_available() 

True

기본적으로 생성된 변수는 모두 CPU에 존재한다.

list0 = [1,2,3,4]
x = torch.tensor(list0)
x.is_cuda

False

x = x.to('cuda')
x.is_cuda

True

다음 코드를 cuda로 실행할 수 있도록 수정해보자.

import torch
from torch import nn
from torch.utils.data import DataLoader
from torchvision import datasets
from torchvision.transforms import ToTensor
import matplotlib.pyplot as plt
import matplotlib.cm as cm
import numpy as np
import os 

# 트레이닝 데이터셋을 다운로드한다.
training_data = datasets.FashionMNIST(
    root="data",
    train=True,
    download=True,
    transform=ToTensor()
)

# 테스트 데이터셋을 다운로드한다.
test_data = datasets.FashionMNIST(
    root="data",
    train=False,
    download=True,
    transform=ToTensor()
)

train_dataloader = DataLoader(training_data, batch_size=64)
test_dataloader = DataLoader(test_data, batch_size=64)

class DNN(nn.Module):
    def __init__(self):
        super(DNN, self).__init__()
        self.linear_relu_stack = nn.Sequential(
            nn.Flatten(),
            nn.Linear(784, 512),
            nn.ReLU(),
            nn.Linear(512, 512),
            nn.ReLU(),
            nn.Linear(512, 10),
        )
        
    # 포워드 패스    
    def forward(self, x):
        return self.linear_relu_stack(x)

Downloading http://fashion-mnist.s3-website.eu-central-1.amazonaws.com/train-images-idx3-ubyte.gz
Downloading http://fashion-mnist.s3-website.eu-central-1.amazonaws.com/train-images-idx3-ubyte.gz to data/FashionMNIST/raw/train-images-idx3-ubyte.gz

  0%|          | 0/26421880 [00:00<?, ?it/s]

Extracting data/FashionMNIST/raw/train-images-idx3-ubyte.gz to data/FashionMNIST/raw

Downloading http://fashion-mnist.s3-website.eu-central-1.amazonaws.com/train-labels-idx1-ubyte.gz
Downloading http://fashion-mnist.s3-website.eu-central-1.amazonaws.com/train-labels-idx1-ubyte.gz to data/FashionMNIST/raw/train-labels-idx1-ubyte.gz

  0%|          | 0/29515 [00:00<?, ?it/s]

Extracting data/FashionMNIST/raw/train-labels-idx1-ubyte.gz to data/FashionMNIST/raw

Downloading http://fashion-mnist.s3-website.eu-central-1.amazonaws.com/t10k-images-idx3-ubyte.gz
Downloading http://fashion-mnist.s3-website.eu-central-1.amazonaws.com/t10k-images-idx3-ubyte.gz to data/FashionMNIST/raw/t10k-images-idx3-ubyte.gz

  0%|          | 0/4422102 [00:00<?, ?it/s]

Extracting data/FashionMNIST/raw/t10k-images-idx3-ubyte.gz to data/FashionMNIST/raw

Downloading http://fashion-mnist.s3-website.eu-central-1.amazonaws.com/t10k-labels-idx1-ubyte.gz
Downloading http://fashion-mnist.s3-website.eu-central-1.amazonaws.com/t10k-labels-idx1-ubyte.gz to data/FashionMNIST/raw/t10k-labels-idx1-ubyte.gz

  0%|          | 0/5148 [00:00<?, ?it/s]

Extracting data/FashionMNIST/raw/t10k-labels-idx1-ubyte.gz to data/FashionMNIST/raw

def train_loop(dataloader, model, loss_fn, optimzer, device):
    size = len(dataloader.dataset)
    model.train() # 모델을 훈련 모드로 설정
    
    for batch, (X, y) in enumerate(dataloader):
        X = X.to(device)
        y = y.to(device)
        pred = model(X) # 포워드 패스 수행
        loss = loss_fn(pred, y) # CE 연산
        
        optimzer.zero_grad() # 0 으로 초기화
        loss.backward() # 역전파하여 그래디언트 계산
        optimzer.step() # 연산된 그래디언트를 사용해 파라미터를 업데이트
        
        if batch % 100 == 0: # 매 100회차 마다 다음 내용 출력
            loss, current = loss.item(), batch * len(X)
            #print(f'loss: {loss}, [{current:>5d}/{size:>5d}]')

def test_loop(dataloader, model, loss_fn, device):
    size = len(dataloader.dataset)
    num_batches = len(dataloader)
    test_loss, correct = 0, 0
    model.eval() # 모델을 실행 모드로 설정
    
    with torch.no_grad(): # 그래디언트 연산 안함
        for X, y in dataloader:
            X = X.to(device)
            y = y.to(device)
            
            pred = model(X) # 포워드 패스 수행
            test_loss += loss_fn(pred, y) # CE 연산
            correct += (pred.argmax(1) == y).type(torch.float).sum().item() # 결과 일치하는지 확인
    
    test_loss /= num_batches
    correct /= size
    print(f'Test Error: \n 정확도: {(100*correct):>0.1f}% 평균 Loss: {test_loss:>8f}\n')
    

def run(device):
    #device = 'cuda:0' if torch.cuda.is_available() else 'cpu'   
    #device = 'cpu'
    print(f"사용할 장치: {device}")

    model = DNN().to(device)

    learning_rate = 1e-3
    batch_size = 64
    epochs = 10


    loss_fn = nn.CrossEntropyLoss().to(device)
    optimizer = torch.optim.SGD(model.parameters(), lr=learning_rate)

    for t in range(epochs):
        print(f'Epoch {t+1}\n--------------------------------------')
        train_loop(train_dataloader, model, loss_fn, optimizer, device)
        test_loop(train_dataloader, model, loss_fn, device)
    print("Done!")

%timeit run('cpu')

사용할 장치: cpu
Epoch 1
--------------------------------------
Test Error: 
 정확도: 44.5% 평균 Loss: 2.145079

Epoch 2
--------------------------------------
Test Error: 
 정확도: 57.3% 평균 Loss: 1.859738

Epoch 3
--------------------------------------
Test Error: 
 정확도: 62.1% 평균 Loss: 1.477654

Epoch 4
--------------------------------------
Test Error: 
 정확도: 63.9% 평균 Loss: 1.217386

Epoch 5
--------------------------------------
Test Error: 
 정확도: 65.6% 평균 Loss: 1.062857

Epoch 6
--------------------------------------
Test Error: 
 정확도: 67.0% 평균 Loss: 0.964224

Epoch 7
--------------------------------------
Test Error: 
 정확도: 68.2% 평균 Loss: 0.895859

Epoch 8
--------------------------------------
Test Error: 
 정확도: 69.3% 평균 Loss: 0.845354

Epoch 9
--------------------------------------
Test Error: 
 정확도: 70.5% 평균 Loss: 0.806216

Epoch 10
--------------------------------------
Test Error: 
 정확도: 71.8% 평균 Loss: 0.774687

Done!
사용할 장치: cpu
Epoch 1
--------------------------------------
Test Error: 
 정확도: 46.2% 평균 Loss: 2.149195

Epoch 2
--------------------------------------
Test Error: 
 정확도: 58.6% 평균 Loss: 1.876602

Epoch 3
--------------------------------------
Test Error: 
 정확도: 62.0% 평균 Loss: 1.500106

Epoch 4
--------------------------------------
Test Error: 
 정확도: 64.6% 평균 Loss: 1.226242

Epoch 5
--------------------------------------
Test Error: 
 정확도: 66.0% 평균 Loss: 1.061437

Epoch 6
--------------------------------------
Test Error: 
 정확도: 67.4% 평균 Loss: 0.957544

Epoch 7
--------------------------------------
Test Error: 
 정확도: 68.8% 평균 Loss: 0.887291

Epoch 8
--------------------------------------
Test Error: 
 정확도: 69.9% 평균 Loss: 0.836570

Epoch 9
--------------------------------------
Test Error: 
 정확도: 71.1% 평균 Loss: 0.797968

Epoch 10
--------------------------------------
Test Error: 
 정확도: 72.3% 평균 Loss: 0.767037

Done!
사용할 장치: cpu
Epoch 1
--------------------------------------
Test Error: 
 정확도: 41.7% 평균 Loss: 2.130327

Epoch 2
--------------------------------------
Test Error: 
 정확도: 57.1% 평균 Loss: 1.842872

Epoch 3
--------------------------------------
Test Error: 
 정확도: 61.3% 평균 Loss: 1.480659

Epoch 4
--------------------------------------
Test Error: 
 정확도: 63.9% 평균 Loss: 1.227481

Epoch 5
--------------------------------------
Test Error: 
 정확도: 65.6% 평균 Loss: 1.070095

Epoch 6
--------------------------------------
Test Error: 
 정확도: 67.1% 평균 Loss: 0.967320

Epoch 7
--------------------------------------
Test Error: 
 정확도: 68.6% 평균 Loss: 0.896083

Epoch 8
--------------------------------------
Test Error: 
 정확도: 69.8% 평균 Loss: 0.843942

Epoch 9
--------------------------------------
Test Error: 
 정확도: 71.0% 평균 Loss: 0.803888

Epoch 10
--------------------------------------
Test Error: 
 정확도: 72.3% 평균 Loss: 0.771779

Done!
사용할 장치: cpu
Epoch 1
--------------------------------------
Test Error: 
 정확도: 43.6% 평균 Loss: 2.148021

Epoch 2
--------------------------------------
Test Error: 
 정확도: 57.7% 평균 Loss: 1.868261

Epoch 3
--------------------------------------
Test Error: 
 정확도: 61.5% 평균 Loss: 1.504941

Epoch 4
--------------------------------------
Test Error: 
 정확도: 64.1% 평균 Loss: 1.248063

Epoch 5
--------------------------------------
Test Error: 
 정확도: 66.0% 평균 Loss: 1.086045

Epoch 6
--------------------------------------
Test Error: 
 정확도: 67.4% 평균 Loss: 0.979198

Epoch 7
--------------------------------------
Test Error: 
 정확도: 68.6% 평균 Loss: 0.904965

Epoch 8
--------------------------------------
Test Error: 
 정확도: 69.7% 평균 Loss: 0.850483

Epoch 9
--------------------------------------
Test Error: 
 정확도: 70.7% 평균 Loss: 0.808730

Epoch 10
--------------------------------------
Test Error: 
 정확도: 71.8% 평균 Loss: 0.775525

Done!
사용할 장치: cpu
Epoch 1
--------------------------------------
Test Error: 
 정확도: 48.0% 평균 Loss: 2.155469

Epoch 2
--------------------------------------
Test Error: 
 정확도: 59.7% 평균 Loss: 1.883491

Epoch 3
--------------------------------------
Test Error: 
 정확도: 63.8% 평균 Loss: 1.507828

Epoch 4
--------------------------------------
Test Error: 
 정확도: 64.8% 평균 Loss: 1.237764

Epoch 5
--------------------------------------
Test Error: 
 정확도: 66.1% 평균 Loss: 1.072880

Epoch 6
--------------------------------------
Test Error: 
 정확도: 67.5% 평균 Loss: 0.967337

Epoch 7
--------------------------------------
Test Error: 
 정확도: 68.8% 평균 Loss: 0.895187

Epoch 8
--------------------------------------
Test Error: 
 정확도: 70.0% 평균 Loss: 0.842927

Epoch 9
--------------------------------------
Test Error: 
 정확도: 71.1% 평균 Loss: 0.802981

Epoch 10
--------------------------------------
Test Error: 
 정확도: 72.2% 평균 Loss: 0.771037

Done!
사용할 장치: cpu
Epoch 1
--------------------------------------
Test Error: 
 정확도: 43.3% 평균 Loss: 2.151541

Epoch 2
--------------------------------------
Test Error: 
 정확도: 55.8% 평균 Loss: 1.879372

Epoch 3
--------------------------------------
Test Error: 
 정확도: 60.4% 평균 Loss: 1.515021

Epoch 4
--------------------------------------
Test Error: 
 정확도: 63.5% 평균 Loss: 1.258509

Epoch 5
--------------------------------------
Test Error: 
 정확도: 65.3% 평균 Loss: 1.094885

Epoch 6
--------------------------------------
Test Error: 
 정확도: 66.8% 평균 Loss: 0.985699

Epoch 7
--------------------------------------
Test Error: 
 정확도: 68.1% 평균 Loss: 0.910132

Epoch 8
--------------------------------------
Test Error: 
 정확도: 69.3% 평균 Loss: 0.855560

Epoch 9
--------------------------------------
Test Error: 
 정확도: 70.6% 평균 Loss: 0.814243

Epoch 10
--------------------------------------
Test Error: 
 정확도: 71.7% 평균 Loss: 0.781519

Done!
사용할 장치: cpu
Epoch 1
--------------------------------------
Test Error: 
 정확도: 47.6% 평균 Loss: 2.164351

Epoch 2
--------------------------------------
Test Error: 
 정확도: 61.6% 평균 Loss: 1.895376

Epoch 3
--------------------------------------
Test Error: 
 정확도: 61.8% 평균 Loss: 1.531638

Epoch 4
--------------------------------------
Test Error: 
 정확도: 63.8% 평균 Loss: 1.265460

Epoch 5
--------------------------------------
Test Error: 
 정확도: 65.5% 평균 Loss: 1.096596

Epoch 6
--------------------------------------
Test Error: 
 정확도: 66.9% 평균 Loss: 0.985845

Epoch 7
--------------------------------------
Test Error: 
 정확도: 68.2% 평균 Loss: 0.909336

Epoch 8
--------------------------------------
Test Error: 
 정확도: 69.6% 평균 Loss: 0.853740

Epoch 9
--------------------------------------
Test Error: 
 정확도: 70.8% 평균 Loss: 0.811476

Epoch 10
--------------------------------------
Test Error: 
 정확도: 71.9% 평균 Loss: 0.778093

Done!
사용할 장치: cpu
Epoch 1
--------------------------------------
Test Error: 
 정확도: 21.1% 평균 Loss: 2.166270

Epoch 2
--------------------------------------
Test Error: 
 정확도: 51.2% 평균 Loss: 1.913578

Epoch 3
--------------------------------------
Test Error: 
 정확도: 58.8% 평균 Loss: 1.542653

Epoch 4
--------------------------------------
Test Error: 
 정확도: 63.4% 평균 Loss: 1.257574

Epoch 5
--------------------------------------
Test Error: 
 정확도: 65.3% 평균 Loss: 1.082608

Epoch 6
--------------------------------------
Test Error: 
 정확도: 66.8% 평균 Loss: 0.971342

Epoch 7
--------------------------------------
Test Error: 
 정확도: 68.2% 평균 Loss: 0.896464

Epoch 8
--------------------------------------
Test Error: 
 정확도: 69.3% 평균 Loss: 0.843298

Epoch 9
--------------------------------------
Test Error: 
 정확도: 70.5% 평균 Loss: 0.803402

Epoch 10
--------------------------------------
Test Error: 
 정확도: 71.8% 평균 Loss: 0.771842

Done!
2min 42s ± 1.13 s per loop (mean ± std. dev. of 7 runs, 1 loop each)

%timeit run('cuda:0')

사용할 장치: cuda:0
Epoch 1
--------------------------------------
Test Error: 
 정확도: 51.8% 평균 Loss: 2.140260

Epoch 2
--------------------------------------
Test Error: 
 정확도: 58.2% 평균 Loss: 1.856035

Epoch 3
--------------------------------------
Test Error: 
 정확도: 61.2% 평균 Loss: 1.494435

Epoch 4
--------------------------------------
Test Error: 
 정확도: 64.2% 평균 Loss: 1.241039

Epoch 5
--------------------------------------
Test Error: 
 정확도: 65.8% 평균 Loss: 1.080358

Epoch 6
--------------------------------------
Test Error: 
 정확도: 67.2% 평균 Loss: 0.973863

Epoch 7
--------------------------------------
Test Error: 
 정확도: 68.5% 평균 Loss: 0.899747

Epoch 8
--------------------------------------
Test Error: 
 정확도: 69.8% 평균 Loss: 0.845576

Epoch 9
--------------------------------------
Test Error: 
 정확도: 71.1% 평균 Loss: 0.804218

Epoch 10
--------------------------------------
Test Error: 
 정확도: 72.4% 평균 Loss: 0.771310

Done!
사용할 장치: cuda:0
Epoch 1
--------------------------------------
Test Error: 
 정확도: 54.1% 평균 Loss: 2.158655

Epoch 2
--------------------------------------
Test Error: 
 정확도: 63.4% 평균 Loss: 1.874954

Epoch 3
--------------------------------------
Test Error: 
 정확도: 63.4% 평균 Loss: 1.488790

Epoch 4
--------------------------------------
Test Error: 
 정확도: 64.3% 평균 Loss: 1.222355

Epoch 5
--------------------------------------
Test Error: 
 정확도: 65.7% 평균 Loss: 1.062467

Epoch 6
--------------------------------------
Test Error: 
 정확도: 67.0% 평균 Loss: 0.960324

Epoch 7
--------------------------------------
Test Error: 
 정확도: 68.1% 평균 Loss: 0.890408

Epoch 8
--------------------------------------
Test Error: 
 정확도: 69.3% 평균 Loss: 0.839719

Epoch 9
--------------------------------------
Test Error: 
 정확도: 70.5% 평균 Loss: 0.801025

Epoch 10
--------------------------------------
Test Error: 
 정확도: 71.8% 평균 Loss: 0.770026

Done!
사용할 장치: cuda:0
Epoch 1
--------------------------------------
Test Error: 
 정확도: 44.4% 평균 Loss: 2.158776

Epoch 2
--------------------------------------
Test Error: 
 정확도: 56.0% 평균 Loss: 1.890853

Epoch 3
--------------------------------------
Test Error: 
 정확도: 60.0% 평균 Loss: 1.520205

Epoch 4
--------------------------------------
Test Error: 
 정확도: 63.2% 평균 Loss: 1.256206

Epoch 5
--------------------------------------
Test Error: 
 정확도: 65.1% 평균 Loss: 1.092746

Epoch 6
--------------------------------------
Test Error: 
 정확도: 66.5% 평균 Loss: 0.985381

Epoch 7
--------------------------------------
Test Error: 
 정확도: 67.8% 평균 Loss: 0.911010

Epoch 8
--------------------------------------
Test Error: 
 정확도: 69.0% 평균 Loss: 0.856955

Epoch 9
--------------------------------------
Test Error: 
 정확도: 70.1% 평균 Loss: 0.815837

Epoch 10
--------------------------------------
Test Error: 
 정확도: 71.4% 평균 Loss: 0.783171

Done!
사용할 장치: cuda:0
Epoch 1
--------------------------------------
Test Error: 
 정확도: 46.0% 평균 Loss: 2.148222

Epoch 2
--------------------------------------
Test Error: 
 정확도: 57.1% 평균 Loss: 1.856956

Epoch 3
--------------------------------------
Test Error: 
 정확도: 62.2% 평균 Loss: 1.486109

Epoch 4
--------------------------------------
Test Error: 
 정확도: 64.8% 평균 Loss: 1.230731

Epoch 5
--------------------------------------
Test Error: 
 정확도: 66.1% 평균 Loss: 1.069811

Epoch 6
--------------------------------------
Test Error: 
 정확도: 67.4% 평균 Loss: 0.963705

Epoch 7
--------------------------------------
Test Error: 
 정확도: 68.7% 평균 Loss: 0.890600

Epoch 8
--------------------------------------
Test Error: 
 정확도: 69.9% 평균 Loss: 0.837781

Epoch 9
--------------------------------------
Test Error: 
 정확도: 71.2% 평균 Loss: 0.797687

Epoch 10
--------------------------------------
Test Error: 
 정확도: 72.5% 평균 Loss: 0.765782

Done!
사용할 장치: cuda:0
Epoch 1
--------------------------------------
Test Error: 
 정확도: 47.2% 평균 Loss: 2.147807

Epoch 2
--------------------------------------
Test Error: 
 정확도: 60.2% 평균 Loss: 1.871584

Epoch 3
--------------------------------------
Test Error: 
 정확도: 64.1% 평균 Loss: 1.501724

Epoch 4
--------------------------------------
Test Error: 
 정확도: 64.7% 평균 Loss: 1.233301

Epoch 5
--------------------------------------
Test Error: 
 정확도: 65.8% 평균 Loss: 1.069727

Epoch 6
--------------------------------------
Test Error: 
 정확도: 67.1% 평균 Loss: 0.965442

Epoch 7
--------------------------------------
Test Error: 
 정확도: 68.4% 평균 Loss: 0.893907

Epoch 8
--------------------------------------
Test Error: 
 정확도: 69.8% 평균 Loss: 0.841642

Epoch 9
--------------------------------------
Test Error: 
 정확도: 71.1% 평균 Loss: 0.801505

Epoch 10
--------------------------------------
Test Error: 
 정확도: 72.3% 평균 Loss: 0.769341

Done!
사용할 장치: cuda:0
Epoch 1
--------------------------------------
Test Error: 
 정확도: 55.7% 평균 Loss: 2.161969

Epoch 2
--------------------------------------
Test Error: 
 정확도: 59.0% 평균 Loss: 1.905919

Epoch 3
--------------------------------------
Test Error: 
 정확도: 61.4% 평균 Loss: 1.528095

Epoch 4
--------------------------------------
Test Error: 
 정확도: 64.1% 평균 Loss: 1.250129

Epoch 5
--------------------------------------
Test Error: 
 정확도: 65.9% 평균 Loss: 1.078414

Epoch 6
--------------------------------------
Test Error: 
 정확도: 67.3% 평균 Loss: 0.967924

Epoch 7
--------------------------------------
Test Error: 
 정확도: 68.6% 평균 Loss: 0.892969

Epoch 8
--------------------------------------
Test Error: 
 정확도: 69.8% 평균 Loss: 0.839180

Epoch 9
--------------------------------------
Test Error: 
 정확도: 71.0% 평균 Loss: 0.798609

Epoch 10
--------------------------------------
Test Error: 
 정확도: 72.3% 평균 Loss: 0.766456

Done!
사용할 장치: cuda:0
Epoch 1
--------------------------------------
Test Error: 
 정확도: 49.5% 평균 Loss: 2.137333

Epoch 2
--------------------------------------
Test Error: 
 정확도: 57.5% 평균 Loss: 1.838759

Epoch 3
--------------------------------------
Test Error: 
 정확도: 63.2% 평균 Loss: 1.474286

Epoch 4
--------------------------------------
Test Error: 
 정확도: 65.2% 평균 Loss: 1.221576

Epoch 5
--------------------------------------
Test Error: 
 정확도: 66.3% 평균 Loss: 1.061860

Epoch 6
--------------------------------------
Test Error: 
 정확도: 67.4% 평균 Loss: 0.957578

Epoch 7
--------------------------------------
Test Error: 
 정확도: 68.7% 평균 Loss: 0.886144

Epoch 8
--------------------------------------
Test Error: 
 정확도: 70.0% 평균 Loss: 0.834677

Epoch 9
--------------------------------------
Test Error: 
 정확도: 71.2% 평균 Loss: 0.795693

Epoch 10
--------------------------------------
Test Error: 
 정확도: 72.4% 평균 Loss: 0.764703

Done!
사용할 장치: cuda:0
Epoch 1
--------------------------------------
Test Error: 
 정확도: 46.9% 평균 Loss: 2.151836

Epoch 2
--------------------------------------
Test Error: 
 정확도: 58.4% 평균 Loss: 1.881667

Epoch 3
--------------------------------------
Test Error: 
 정확도: 61.5% 평균 Loss: 1.508986

Epoch 4
--------------------------------------
Test Error: 
 정확도: 64.5% 평균 Loss: 1.242306

Epoch 5
--------------------------------------
Test Error: 
 정확도: 66.3% 평균 Loss: 1.076750

Epoch 6
--------------------------------------
Test Error: 
 정확도: 67.7% 평균 Loss: 0.967935

Epoch 7
--------------------------------------
Test Error: 
 정확도: 69.1% 평균 Loss: 0.893048

Epoch 8
--------------------------------------
Test Error: 
 정확도: 70.4% 평균 Loss: 0.838853

Epoch 9
--------------------------------------
Test Error: 
 정확도: 71.6% 평균 Loss: 0.797739

Epoch 10
--------------------------------------
Test Error: 
 정확도: 72.9% 평균 Loss: 0.765048

Done!
2min 12s ± 859 ms per loop (mean ± std. dev. of 7 runs, 1 loop each)

참고자료

Pytorch-GPU 사용

Twitter Facebook LinkedIn

Pytorch GPU 사용법

pytorch GPU 튜토리얼

GPU가 사용 가능한지 확인

참고자료

공유하기

댓글남기기

참고

우잉남

[알고리즘] 그래프탐색: DFS

[알고리즘] 그래프탐색: BFS

[학회 정보] AI 주요 학회