tensor_frame 0.0.3-alpha

A PyTorch-like tensor library for Rust with CPU, WGPU, and CUDA backends
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
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206

#[cfg(not(feature = "wgpu"))]
mod tests {
    use tensor_frame::{Tensor, TensorOps};
    #[test]
    fn test_matmul_basic() {
        let a = Tensor::from_vec(vec![1.0, 2.0, 3.0, 4.0], vec![2, 2]).unwrap();
        let b = Tensor::from_vec(vec![5.0, 6.0, 7.0, 8.0], vec![2, 2]).unwrap();

        let result = a.matmul(&b).unwrap();
        assert_eq!(result.shape().dims(), &[2, 2]);

        let expected = vec![19.0, 22.0, 43.0, 50.0]; // Manual calculation
        let actual = result.to_vec().unwrap();

        for (a, b) in expected.iter().zip(actual.iter()) {
            assert!((a - b).abs() < 1e-6, "Expected {}, got {}", a, b);
        }
    }

    #[test]
    fn test_matmul_different_shapes() {
        let a = Tensor::from_vec(vec![1.0, 2.0, 3.0, 4.0, 5.0, 6.0], vec![2, 3]).unwrap(); // 2x3
        let b = Tensor::from_vec(vec![7.0, 8.0, 9.0, 10.0, 11.0, 12.0], vec![3, 2]).unwrap(); // 3x2

        let result = a.matmul(&b).unwrap();
        assert_eq!(result.shape().dims(), &[2, 2]);

        // Expected: [[58, 64], [139, 154]]
        let expected = vec![58.0, 64.0, 139.0, 154.0];
        let actual = result.to_vec().unwrap();

        for (a, b) in expected.iter().zip(actual.iter()) {
            assert!((a - b).abs() < 1e-6, "Expected {}, got {}", a, b);
        }
    }

    #[test]
    fn test_matmul_incompatible_shapes() {
        let a = Tensor::from_vec(vec![1.0, 2.0, 3.0, 4.0], vec![2, 2]).unwrap();
        let b = Tensor::from_vec(vec![1.0, 2.0, 3.0], vec![3, 1]).unwrap();

        let result = a.matmul(&b);
        assert!(result.is_err());
    }

    #[test]
    fn test_matmul_non_2d() {
        let a = Tensor::from_vec(vec![1.0, 2.0, 3.0, 4.0, 5.0, 6.0], vec![2, 3]).unwrap();
        let b = Tensor::from_vec(vec![1.0, 2.0, 3.0, 4.0, 5.0, 6.0], vec![6]).unwrap(); // 1D

        let result = a.matmul(&b);
        assert!(result.is_err());
    }

    #[test]
    fn test_bmm_basic() {
        // Create 2 batches of 2x2 matrices
        let a =
            Tensor::from_vec(vec![1.0, 2.0, 3.0, 4.0, 5.0, 6.0, 7.0, 8.0], vec![2, 2, 2]).unwrap();
        let b =
            Tensor::from_vec(vec![1.0, 0.0, 0.0, 1.0, 2.0, 0.0, 0.0, 2.0], vec![2, 2, 2]).unwrap();

        let result = a.bmm(&b).unwrap();
        assert_eq!(result.shape().dims(), &[2, 2, 2]);

        // First batch: identity multiplication, second batch: scaling by 2
        let expected = vec![1.0, 2.0, 3.0, 4.0, 10.0, 12.0, 14.0, 16.0];
        let actual = result.to_vec().unwrap();

        for (a, b) in expected.iter().zip(actual.iter()) {
            assert!((a - b).abs() < 1e-6, "Expected {}, got {}", a, b);
        }
    }

    #[test]
    fn test_bmm_incompatible_batch_size() {
        let a = Tensor::ones(vec![2, 3, 4]).unwrap();
        let b = Tensor::ones(vec![3, 4, 5]).unwrap(); // Different batch size

        let result = a.bmm(&b);
        assert!(result.is_err());
    }

    #[test]
    fn test_exp_basic() {
        let tensor = Tensor::from_vec(vec![0.0, 1.0, 2.0], vec![3]).unwrap();
        let result = tensor.exp().unwrap();

        let actual = result.to_vec().unwrap();
        assert!((actual[0] - 1.0).abs() < 1e-6); // e^0 = 1
        assert!((actual[1] - std::f32::consts::E).abs() < 1e-6); // e^1 = e
        assert!((actual[2] - (std::f32::consts::E * std::f32::consts::E)).abs() < 1e-5);
        // e^2 = e^2
    }

    #[test]
    fn test_log_basic() {
        let tensor = Tensor::from_vec(
            vec![1.0, std::f32::consts::E, std::f32::consts::E.powi(2)],
            vec![3],
        )
        .unwrap();
        let result = tensor.log().unwrap();

        let actual = result.to_vec().unwrap();
        assert!((actual[0] - 0.0).abs() < 1e-6); // ln(1) = 0
        assert!((actual[1] - 1.0).abs() < 1e-6); // ln(e) = 1
        assert!((actual[2] - 2.0).abs() < 1e-5); // ln(e^2) = 2
    }

    #[test]
    fn test_sqrt_basic() {
        let tensor = Tensor::from_vec(vec![1.0, 4.0, 9.0, 16.0], vec![4]).unwrap();
        let result = tensor.sqrt().unwrap();

        let expected = vec![1.0, 2.0, 3.0, 4.0];
        let actual = result.to_vec().unwrap();

        for (a, b) in expected.iter().zip(actual.iter()) {
            assert!((a - b).abs() < 1e-6, "Expected {}, got {}", a, b);
        }
    }

    #[test]
    fn test_pow_basic() {
        let tensor = Tensor::from_vec(vec![2.0, 3.0, 4.0], vec![3]).unwrap();
        let result = tensor.pow(2.0).unwrap();

        let expected = vec![4.0, 9.0, 16.0];
        let actual = result.to_vec().unwrap();

        for (a, b) in expected.iter().zip(actual.iter()) {
            assert!((a - b).abs() < 1e-6, "Expected {}, got {}", a, b);
        }
    }

    #[test]
    fn test_sin_cos_basic() {
        use std::f32::consts::PI;

        let tensor = Tensor::from_vec(vec![0.0, PI / 2.0, PI], vec![3]).unwrap();

        let sin_result = tensor.sin().unwrap();
        let sin_actual = sin_result.to_vec().unwrap();
        assert!((sin_actual[0] - 0.0).abs() < 1e-6); // sin(0) = 0
        assert!((sin_actual[1] - 1.0).abs() < 1e-6); // sin(π/2) = 1
        assert!((sin_actual[2] - 0.0).abs() < 1e-6); // sin(π) = 0

        let cos_result = tensor.cos().unwrap();
        let cos_actual = cos_result.to_vec().unwrap();
        assert!((cos_actual[0] - 1.0).abs() < 1e-6); // cos(0) = 1
        assert!((cos_actual[1] - 0.0).abs() < 1e-6); // cos(π/2) = 0
        assert!((cos_actual[2] - (-1.0)).abs() < 1e-6); // cos(π) = -1
    }

    #[test]
    fn test_relu_basic() {
        let tensor = Tensor::from_vec(vec![-2.0, -1.0, 0.0, 1.0, 2.0], vec![5]).unwrap();
        let result = tensor.relu().unwrap();

        let expected = vec![0.0, 0.0, 0.0, 1.0, 2.0];
        let actual = result.to_vec().unwrap();

        for (a, b) in expected.iter().zip(actual.iter()) {
            assert!((a - b).abs() < 1e-6, "Expected {}, got {}", a, b);
        }
    }

    #[test]
    fn test_sigmoid_basic() {
        let tensor = Tensor::from_vec(vec![-10.0, 0.0, 10.0], vec![3]).unwrap();
        let result = tensor.sigmoid().unwrap();

        let actual = result.to_vec().unwrap();
        assert!(actual[0] < 0.001); // sigmoid(-10) ≈ 0
        assert!((actual[1] - 0.5).abs() < 1e-6); // sigmoid(0) = 0.5
        assert!(actual[2] > 0.999); // sigmoid(10) ≈ 1
    }

    #[test]
    fn test_tanh_basic() {
        let tensor = Tensor::from_vec(vec![-10.0, 0.0, 10.0], vec![3]).unwrap();
        let result = tensor.tanh().unwrap();

        let actual = result.to_vec().unwrap();
        assert!(actual[0] < -0.999); // tanh(-10) ≈ -1
        assert!((actual[1] - 0.0).abs() < 1e-6); // tanh(0) = 0
        assert!(actual[2] > 0.999); // tanh(10) ≈ 1
    }

    #[test]
    fn test_chain_math_operations() {
        let tensor = Tensor::from_vec(vec![1.0, 4.0, 9.0], vec![3]).unwrap();

        // Chain operations: sqrt -> exp -> log
        let result = tensor.sqrt().unwrap().exp().unwrap().log().unwrap();

        // Should return approximately the original sqrt values
        let expected = vec![1.0, 2.0, 3.0]; // sqrt of original values
        let actual = result.to_vec().unwrap();

        for (a, b) in expected.iter().zip(actual.iter()) {
            assert!((a - b).abs() < 1e-5, "Expected {}, got {}", a, b);
        }
    }
}