PyTorch教程-2

• reshape

input = torch.arange(12)
# tensor([ 0,  1,  2,  3,  4,  5,  6,  7,  8,  9, 10, 11])

input = torch.reshape(input, (3, 4))
# tensor([[ 0,  1,  2,  3],
#         [ 4,  5,  6,  7],
#         [ 8,  9, 10, 11]])

• view

返回一个张量，数据与输入张量相同，大小不同

x = torch.randn(2, 2)
# tensor([[ 1.5941, -0.1812],
#         [ 0.5612,  1.5465]])
print(x.size())
# torch.Size([2, 2])

y = x.view(4)   # 数据是一样的
# tensor([ 1.5941, -0.1812,  0.5612,  1.5465])
print(y.size()) # 大小可以改变
# torch.Size([4])

• flatness

input = torch.reshape(torch.arange(6), (2, 3))
# tensor([[0, 1, 2],
#         [3, 4, 5]])

output = torch.flatten(input)
# tensor([0, 1, 2, 3, 4, 5])

• argmax

input = torch.tensor([[1, 3, 5],
                      [4, 8, 0]])

print(input.argmax(1))  # 每行最大值下标
tensor([2, 1])

• manual_seed

随机数种子，每次运行的结果都是一样的

seed = 111
torch.manual_seed(seed)
# tensor([[0.7156, 0.9140]])

• Sequential

class CNN(nn.Module):
    def __init__(self):
        super(CNN, self).__init__()
        self.model1 = Sequential(
            Conv2d(3, 32, kernel_size=(5, 5), padding=2),
            MaxPool2d(2),
            Conv2d(32, 32, kernel_size=(5, 5), padding=2),
            MaxPool2d(2),
            Conv2d(32, 64, kernel_size=(5, 5), padding=2),
            MaxPool2d(2),
            Flatten(),
            Linear(1024, 64),
            Linear(64, 10)
        )

    def forward(self, x):
        x = self.model1(x)
        return x


model = CNN()
input = torch.ones((64, 3, 32, 32))
output = model(input)
print(output.shape)
# torch.Size([64, 10])