x_train = torch.FloatTensor([[1], [2], [3]])
y_train = torch.FloatTensor([[1], [2], [3]])

print(x_train)
print(x_train.shape)

tensor([[2., 4.],
        [6., 8.]])
tensor([[1., 2.],
        [3., 4.]])
tensor([[2., 4.],
        [6., 8.]])
tensor([[2., 4.],
        [6., 8.]])

Cost or Loss

linear regression에선 mean squared error (MSE)함수로 계산함

W = torch.zeros(1, requires_grad=True)
b = torch.zeros(1, requires_grad=True)
hypothesis = x_train * W + b

cost = torch.mean((hypothesis - y_train) ** 2)
print(cost)

Gradient Descent

optimizer = optim.SGD([W, b], lr=0.01)	//3.Optimizer 정의

optimizer.zero_grad() //gradient 초기화
cost.backward()		//ackward()로 gragient 계산
optimizer.step()	//step()으로 개선

Full code

# 데이터 정의
x_train = torch.FloatTensor([[1], [2], [3]])
y_train = torch.FloatTensor([[1], [2], [3]])
# Hypothesis 정의
W = torch.zeros(1, requires_grad=True)
b = torch.zeros(1, requires_grad=True)
# optimizer 정의
optimizer = optim.SGD([W, b], lr=0.01)

nb_epochs = 1000
for epoch in range(nb_epochs + 1):
    
    # H(x) 계산
    hypothesis = x_train * W + b
    
    # cost 계산
    cost = torch.mean((hypothesis - y_train) ** 2)

    # cost로 H(x) 개선
    optimizer.zero_grad()
    cost.backward()
    optimizer.step()

    # 100번마다 로그 출력
    if epoch % 100 == 0:
        print('Epoch {:4d}/{} W: {:.3f}, b: {:.3f} Cost: {:.6f}'.format(
            epoch, nb_epochs, W.item(), b.item(), cost.item()
        ))