I found the code snippet here:
Also, according to the doc, torch.Storage.from_file
If shared is True the file will be created if needed.
Therefore,
To create:
# create.py
import torch
fname = 'tensor.pt'
storage = torch.LongStorage.from_file(fname, True, 40)
torch.LongTensor(storage).copy_(torch.arange(40))
To read:
# read.py
import torch
fname = 'tensor.pt'
storage = torch.LongStorage.from_file(fname, True, 40)
tensor = torch.LongTensor(storage)
print(tensor)
# Demo
$ python create.py
$ python read.py
tensor([ 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17,
18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35,
36, 37, 38, 39])