- torch.Tensor() is the abbreviation of torch.FloatTensor() by default, and the created float32-bit data type;
- torch.tensor() is a copy of tensor data and creates Tensor according to the type of incoming data;
- Other data types, such as LongTensor, FloatTensor, etc., are created corresponding data types;
The difference between torch.Tensor and torch.tensor() and other Tensor types in Pytorch
Guess you like
Origin blog.csdn.net/qq_29027865/article/details/104298210
Ranking