burn-tensor 0.1.0

This library provides multiple tensor implementations hidden behind an easy to use API that supports reverse mode automatic differentiation. 
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
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139

use crate::tensor::{
    backend::tch::{TchShape, TchTensor},
    ops::*,
    Data, Distribution, Shape,
};
use rand::distributions::{uniform::SampleUniform, Standard};

impl<P, const D: usize> TensorCreationLike<P, D> for TchTensor<P, D>
where
    P: tch::kind::Element + std::fmt::Debug + SampleUniform + Default,
    Standard: rand::distributions::Distribution<P>,
{
    fn new_like_empty(&self) -> Self {
        let tensor = self.tensor.empty_like();
        let shape = self.shape.clone();
        let kind = self.kind.clone();

        Self {
            kind,
            tensor,
            shape,
        }
    }

    fn new_like_random(&self, distribution: Distribution<P>) -> Self {
        let device = self.tensor.device();
        let data = Data::<P, D>::random(self.shape.clone(), distribution);

        Self::from_data(data, device)
    }

    fn new_like_data(&self, data: Data<P, D>) -> Self {
        let device = self.tensor.device();
        Self::from_data(data, device)
    }

    fn new_like_zeros(&self) -> Self {
        self.zeros()
    }

    fn new_like_ones(&self) -> Self {
        self.ones()
    }
}

impl<P, const D: usize, const D2: usize> TensorCreationFork<P, D, D2, TchTensor<P, D2>>
    for TchTensor<P, D>
where
    P: tch::kind::Element + std::fmt::Debug + SampleUniform + Default + Copy,
    Standard: rand::distributions::Distribution<P>,
{
    fn new_fork_empty(&self, shape: Shape<D2>) -> TchTensor<P, D2> {
        let device = self.tensor.device();
        let kind = self.kind.clone();

        let tch_shape = TchShape::from(shape.clone());
        let tensor = tch::Tensor::empty(&tch_shape.dims, (kind.kind(), device));

        TchTensor {
            kind,
            tensor,
            shape,
        }
    }

    fn new_fork_random(&self, shape: Shape<D2>, distribution: Distribution<P>) -> TchTensor<P, D2> {
        let device = self.tensor.device();
        let data = Data::<P, D2>::random(shape, distribution);

        TchTensor::from_data(data, device)
    }

    fn new_fork_data(&self, data: Data<P, D2>) -> TchTensor<P, D2> {
        let device = self.tensor.device();
        TchTensor::from_data(data, device)
    }

    fn new_fork_zeros(&self, shape: Shape<D2>) -> TchTensor<P, D2> {
        let device = self.tensor.device();
        let kind = self.kind.clone();

        let tch_shape = TchShape::from(shape.clone());
        let tensor = tch::Tensor::zeros(&tch_shape.dims, (kind.kind(), device));

        TchTensor {
            kind,
            tensor,
            shape,
        }
    }

    fn new_fork_ones(&self, shape: Shape<D2>) -> TchTensor<P, D2> {
        let device = self.tensor.device();
        let kind = self.kind.clone();

        let tch_shape = TchShape::from(shape.clone());
        let tensor = tch::Tensor::ones(&tch_shape.dims, (kind.kind(), device));

        TchTensor {
            kind,
            tensor,
            shape,
        }
    }
}

impl<P, const D: usize> Zeros<TchTensor<P, D>> for TchTensor<P, D>
where
    P: tch::kind::Element,
{
    fn zeros(&self) -> TchTensor<P, D> {
        let tensor = self.tensor.zeros_like();
        let shape = self.shape.clone();
        let kind = self.kind.clone();

        Self {
            kind,
            tensor,
            shape,
        }
    }
}

impl<P, const D: usize> Ones<TchTensor<P, D>> for TchTensor<P, D>
where
    P: tch::kind::Element,
{
    fn ones(&self) -> TchTensor<P, D> {
        let tensor = self.tensor.ones_like();
        let shape = self.shape.clone();
        let kind = self.kind.clone();

        Self {
            kind,
            tensor,
            shape,
        }
    }
}