luma-tensor 0.0.1

tensor crate with autograd
Documentation 
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
140
141
142
143
144
145
146
147
148
149

use luma_tensor::Device;
use super::*;

#[allow(dead_code)]
pub fn test_cat_dim0_f32(device: &impl Device) {
    let a = tensor_f32_dev(&[1.0, 2.0, 3.0], (1, 3), device);
    let b = tensor_f32_dev(&[4.0, 5.0, 6.0, 7.0, 8.0, 9.0], (2, 3), device);
    let c = Tensor::cat(&[&a, &b], 0usize).unwrap();
    assert_eq!(c.dims(), &[3, 3]);
    assert_close(&c.to_vec().unwrap(), &[1.0, 2.0, 3.0, 4.0, 5.0, 6.0, 7.0, 8.0, 9.0], 1e-5, 1e-5);
}

#[allow(dead_code)]
pub fn test_contiguous_after_transpose(device: &impl Device) {
    let t = tensor_f32_dev(&[1.0, 2.0, 3.0, 4.0, 5.0, 6.0], (2, 3), device);
    let tr = t.transpose(0usize, 1usize).unwrap();
    assert!(!tr.is_contiguous());
    let c = tr.contiguous().unwrap();
    assert!(c.is_contiguous());
    assert_close(&tr.to_vec().unwrap(), &c.to_vec().unwrap(), 1e-5, 1e-5);
}

#[allow(dead_code)]
pub fn test_reshape_f32(device: &impl Device) {
    let t = tensor_f32_dev(&[1.0, 2.0, 3.0, 4.0, 5.0, 6.0], (2, 3), device);
    let r = t.reshape((3, 2)).unwrap();
    assert_eq!(r.dims(), &[3, 2]);
    assert_close(&r.to_vec().unwrap(), &[1.0, 2.0, 3.0, 4.0, 5.0, 6.0], 1e-5, 1e-5);
}

#[allow(dead_code)]
pub fn test_transpose_f32(device: &impl Device) {
    let t = tensor_f32_dev(&[1.0, 2.0, 3.0, 4.0, 5.0, 6.0], (2, 3), device);
    let tr = t.transpose(0usize, 1usize).unwrap();
    assert_eq!(tr.dims(), &[3, 2]);
    assert_close(&tr.to_vec().unwrap(), &[1.0, 4.0, 2.0, 5.0, 3.0, 6.0], 1e-5, 1e-5);
}

#[allow(dead_code)]
pub fn test_cat_empty() {
    let arrs: &[&luma_tensor::Tensor<luma_tensor::Cpu>] = &[];
    assert!(luma_tensor::Tensor::<luma_tensor::Cpu>::cat(arrs, 0usize).is_err());
}

#[allow(dead_code)]
pub fn test_broadcast_as_f32(device: &impl Device) {
    let t = tensor_f32_dev(&[1.0, 2.0, 3.0], (1, 3), device);
    let b = t.broadcast_as((2, 3)).unwrap();
    assert_eq!(b.dims(), &[2, 3]);
    assert_close(&b.to_vec().unwrap(), &[1.0, 2.0, 3.0, 1.0, 2.0, 3.0], 1e-5, 1e-5);
}

#[allow(dead_code)]
pub fn test_narrow_dim0(device: &impl Device) {
    let t = tensor_f32_dev(&[1.0, 2.0, 3.0, 4.0, 5.0, 6.0], (2, 3), device);
    let n = t.narrow(0usize, 0, 1).unwrap();
    assert_eq!(n.dims(), &[1, 3]);
    assert_close(&n.to_vec().unwrap(), &[1.0, 2.0, 3.0], 1e-5, 1e-5);
}

#[allow(dead_code)]
pub fn test_squeeze_dim1(device: &impl Device) {
    let t = tensor_f32_dev(&[1.0, 2.0, 3.0], (1, 3), device);
    let s = t.squeeze(0usize).unwrap();
    assert_eq!(s.dims(), &[3]);
    assert_close(&s.to_vec().unwrap(), &[1.0, 2.0, 3.0], 1e-5, 1e-5);
}

#[allow(dead_code)]
pub fn test_unsqueeze(device: &impl Device) {
    let t = tensor_f32_dev(&[1.0, 2.0, 3.0], (3,), device);
    let u = t.unsqueeze(0usize).unwrap();
    assert_eq!(u.dims(), &[1, 3]);
    assert_close(&u.to_vec().unwrap(), &[1.0, 2.0, 3.0], 1e-5, 1e-5);
}

#[allow(dead_code)]
pub fn test_flatten_all(device: &impl Device) {
    let t = tensor_f32_dev(&[1.0, 2.0, 3.0, 4.0], (2, 2), device);
    let f = t.flatten_all().unwrap();
    assert_eq!(f.dims(), &[4]);
    assert_close(&f.to_vec().unwrap(), &[1.0, 2.0, 3.0, 4.0], 1e-5, 1e-5);
}

#[allow(dead_code)]
pub fn test_permute_f32(device: &impl Device) {
    let t = tensor_f32_dev(&[1.0, 2.0, 3.0, 4.0, 5.0, 6.0], (2, 3), device);
    let p = t.permute([1usize, 0]).unwrap();
    assert_eq!(p.dims(), &[3, 2]);
    assert_close(&p.to_vec().unwrap(), &[1.0, 4.0, 2.0, 5.0, 3.0, 6.0], 1e-5, 1e-5);
}

#[allow(dead_code)]
pub fn test_split_f32(device: &impl Device) {
    let t = tensor_f32_dev(&[1.0, 2.0, 3.0, 4.0], (4,), device);
    let parts = t.split(0usize).unwrap();
    assert_eq!(parts.len(), 4);
    assert_close(&parts[0].to_vec().unwrap(), &[1.0], 1e-5, 1e-5);
    assert_close(&parts[3].to_vec().unwrap(), &[4.0], 1e-5, 1e-5);
}

#[allow(dead_code)]
pub fn test_repeat_dim_f32(device: &impl Device) {
    let t = tensor_f32_dev(&[1.0, 2.0], (1, 2), device);
    let r = t.repeat_dim(0usize, 3).unwrap();
    assert_eq!(r.dims(), &[3, 2]);
    assert_close(&r.to_vec().unwrap(), &[1.0, 2.0, 1.0, 2.0, 1.0, 2.0], 1e-5, 1e-5);
}

#[allow(dead_code)]
pub fn test_transpose_last(device: &impl Device) {
    let t = tensor_f32_dev(&[1.0, 2.0, 3.0, 4.0, 5.0, 6.0], (2, 3), device);
    let tr = t.transpose_last().unwrap();
    assert_eq!(tr.dims(), &[3, 2]);
}

#[allow(dead_code)]
pub fn test_already_contiguous_is_noop(device: &impl Device) {
    let t = tensor_f32_dev(&[1.0, 2.0, 3.0], (3,), device);
    assert!(t.is_contiguous());
    let c = t.contiguous().unwrap();
    assert_eq!(t.to_vec().unwrap(), c.to_vec().unwrap());
}

#[allow(dead_code)]
pub fn test_flatten_range(device: &impl Device) {
    let vals: Vec<f64> = (1..=12).map(|x| x as f64).collect();
    let t = tensor_f32_dev(&vals, (3, 2, 2), device);
    let f = t.flatten(1usize, 2usize).unwrap();
    assert_eq!(f.dims(), &[3, 4]);
}

#[allow(dead_code)]
pub fn test_stack_f32(device: &impl Device) {
    let a = tensor_f32_dev(&[1.0, 2.0, 3.0], (3,), device);
    let b = tensor_f32_dev(&[4.0, 5.0, 6.0], (3,), device);
    let s = luma_tensor::Tensor::stack(&[&a, &b], 0usize).unwrap();
    assert_eq!(s.dims(), &[2, 3]);
    assert_close(&s.to_vec().unwrap(), &[1.0, 2.0, 3.0, 4.0, 5.0, 6.0], 1e-5, 1e-5);
}

#[allow(dead_code)]
pub fn test_chunk_f32(device: &impl Device) {
    let t = tensor_f32_dev(&[1.0, 2.0, 3.0, 4.0, 5.0], (5,), device);
    let chunks = t.chunk(2, 0usize).unwrap();
    assert_eq!(chunks.len(), 2);
    assert_eq!(chunks[0].dims(), &[3]);
    assert_eq!(chunks[1].dims(), &[2]);
}