What are the common operations of Tensor in Pytorch
This article mainly shows you "what are the common operations of Tensor in Pytorch", the content is easy to understand, clear, hope to help you solve your doubts, the following let the editor lead you to study and learn "what are the common operations of Tensor in Pytorch" this article.
1. Create the classic way of Tensorimport torch#: device = torch.device ("cuda:0") x = torch.tensor ([1w.backward 2], dtype = torch.float32,device = device,requires_grad=True) w = sum (2 * x) w.backward () print (x.device) print (x.dtype) print (x.grad) # Tensory = torch.Tensor ([1jue 2jue 3]) # equivalent to y = torch.FloatTensor ([1jue 2] 3]) # 32-bit floating-point type # the latter declares that there are other types of open gradient y.requires_grad = True#, commonly used torch.LongTensor (2 * y) torch.shortTensor (2) torch.IntTensor (2) torch.IntTensor (2 * y) w.backward () print (y.grad) print (y.dtype)
Output:
Cuda:0
Torch.float32
Tensor ([2, 2.], device='cuda:0')
Tensor ([2.,2.2,2.])
Torch.float32
Creation method similar to numpy
X = torch.linspace (1mighe10dtype = torch.float32,requires_grad = True) y = torch.ones (10) z = torch.zeros ((2jing4)) w = torch.randn ((2p3) # randomly adopted from the standard normal distribution (mean is 0, variance is 1), Gaussian noise points, while rand is equivalent to random sampling # torch.normal ()? print (x) print (y) print (z) print (w)
Output
Tensor ([1, 2, 3, 4, 5, 6, 7, 8.9, 10.], requires_grad=True)
Tensor ([1.1,1.1.1.1.1.1.1.1.1.1.1])
Tensor ([[0.,0.,0.,0.]
[0.,0.,0.,0.])
Tensor ([- 0.6505, 1.3897, 2.2265]
[- 1.7815,-1.8194,-0.4143])
Convert from numpy
Np_data = np.arange (2 numpy',np_data 13). Reshape ((2)) torch_data = torch.from_numpy (np_data) # numpy to tensorprint ('numpy',np_data) print (' torch',torch_data)
Output
Numpy [[2 4 6]
[8 10 12]]
Torch tensor ([2,4,6]
[8, 10, 12], dtype=torch.int32)
2. The combination import torchx = torch.arange (0meme 10Power1). Reshape # size= (2meme 5) y = torch.ones (10) .reshape (2maimei 1) # size= (2pime 5) print (x) print (y) w = torch.cat ((xMague y), dim = 0) # starts from the far left of size by default, where the result is: (2mem2prime5) z = torch.cat ((xmemy), dim = 1) # (2) 5) print (wMagne.Size ()) print (zmemez.size ()) # and stack ()
Output:
Tensor ([0,1,2,3,4]
[5, 6, 7, 8, 9])
Tensor ([[1.1,1.1.1.1.1.]
[1, 1, 1.])
Tensor ([[0.,1.,2.,3.4.]
[5., 6., 7., 8., 9.]
[1., 1., 1., 1.]
[1, 1, 1.]) Torch.Size ([4,5])
Tensor ([0, 1, 2, 3, 4, 1, 1, 1, 1, 1.]
[5, 6, 7, 8, 9, 1, 1.]) Torch.Size ([2,10])
3. Data type conversion
Law one
X = torch.rand ((2Yu2), dtype = torch.float32) print (x.dtype) x = x.double () print (x.dtype) x = x.int () print (x)
Output:
Torch.float32
Torch.float64
Tensor ([0,0]
[0,0]], dtype=torch.int32)
Law two
X = torch.LongTensor ((2)) print (x.dtype) x = x.type (torch.float32) print (x.dtype)
Output:
Torch.int64
Torch.float32
4. Matrix calculation x = torch.arange (0Power4). Reshape (2) print (x) print (x * x) # Direct multiplication print (torch.mm (x)) # Matrix multiplication print (x + 1) # broadcast print (x.numpy ()) # convert to numpy
Output:
Tensor ([0,1]
[2, 3]])
Tensor ([0,1]
[4, 9]])
Tensor ([2,3]
[6,11]])
Tensor ([1,2]
[3, 4]])
[[0 1]
[2 3]]
5. Dimension change
It is mainly for the up and down dimension operation with dimension size 1.
Torch.squeeze (input) # remove the dimension of dimension 1 torch.unsqueeze (input,dim) # add more than one dimension to the specified location is "what are the common operations of Tensor in Pytorch" all the content of this article, thank you for reading! I believe we all have a certain understanding, hope to share the content to help you, if you want to learn more knowledge, welcome to follow the industry information channel!
