Quick Start

Create a Tree-based Tensor

You can create a tree-based tensor or a native tensor like the following example code.

 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
import builtins
import os
from functools import partial

import treetensor.torch as torch

print = partial(builtins.print, sep=os.linesep)

if __name__ == '__main__':
    t1 = torch.tensor([[1, 2, 3],
                       [4, 5, 6]])
    print('new native tensor:', t1)

    t2 = torch.tensor({
        'a': [1, 2, 3],
        'b': {'x': [[4, 5], [6, 7]]},
    })
    print('new tree tensor:', t2)

    t3 = torch.randn(2, 3)
    print('new random native tensor:', t3)

    t4 = torch.randn({
        'a': (2, 3),
        'b': {'x': (3, 4)},
    })
    print('new random tree tensor:', t4)

The output should be like below.

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
new native tensor:
tensor([[1, 2, 3],
        [4, 5, 6]])
new tree tensor:
<Tensor 0x7fcc36ac51d0>
├── 'a' --> tensor([1, 2, 3])
└── 'b' --> <Tensor 0x7fcc36ac5210>
    └── 'x' --> tensor([[4, 5],
                        [6, 7]])

new random native tensor:
tensor([[-0.0983,  2.2656, -1.1562],
        [ 0.5748, -0.3720, -0.4088]])
new random tree tensor:
<Tensor 0x7fcbd5380dd0>
├── 'a' --> tensor([[-0.4048, -1.6665,  1.5289],
│                   [-0.0297, -1.5519, -2.3092]])
└── 'b' --> <Tensor 0x7fcbd53808d0>
    └── 'x' --> tensor([[ 0.0310,  1.6242, -1.4545,  0.9043],
                        [-0.7733, -0.2163,  0.6049, -1.0371],
                        [ 0.6074, -1.7845,  0.3861,  0.9783]])