Pytorch 入门学习数据操作之算术操作-加法

算术操作-加法

在 PyTorch 中，同一种操作方式可能有很多种形式

官方文档相关信息地址：官方文档相关信息地址

方式一

def add_one():
    random_a = torch.rand(4, 2)
    random_b = torch.rand(4, 2)
    print(random_a + random_b)



if __name__ == '__main__':
    add_one()

• print result:

  Connected to pydev debugger (build 193.7288.30)
  tensor([[0.4214, 0.9930],
          [1.8052, 1.3819],
          [0.3334, 1.6448],
          [0.2293, 1.3217]])
  

方式二

def add_two():
    random_a = torch.rand(7, 5)
    random_b = torch.rand(7, 5)
    print(torch.add(random_a,random_b))
    
if __name__ == '__main__':
    add_two()

• print result:

  Connected to pydev debugger (build 193.7288.30)
  tensor([[0.9120, 0.6793, 1.8955, 1.0702, 0.5505],
          [0.2064, 0.1667, 1.3587, 1.6728, 1.2151],
          [1.3669, 0.5354, 0.8769, 1.1060, 1.7966],
          [0.6240, 1.2662, 1.1661, 0.9901, 1.1229],
          [1.1226, 1.2311, 0.8754, 0.4553, 1.7378],
          [0.7544, 0.3557, 0.7631, 0.4962, 1.1400],
          [0.5023, 0.3433, 0.9610, 0.9195, 1.8071]])
  

方式三

def add_three():
    
    """
    info: In-place version of add()
    """
    
    random_a = torch.rand(6, 3)
    random_b = torch.rand(6, 3)
    print(random_a.add_(random_b))


if __name__ == '__main__':
    add_three()

• print result:

  Connected to pydev debugger (build 193.7288.30)
  tensor([[0.6409, 0.8297, 1.8872],
          [1.4139, 0.8349, 1.4194],
          [1.0500, 1.1814, 0.8922],
          [0.2682, 0.8406, 1.3403],
          [0.8939, 1.1223, 0.8175],
          [1.0493, 0.4294, 0.9734]])