tsai_models 0.1.2

Model zoo for tsai-rs: CNN, Transformer, ROCKET, RNN, and Tabular architectures
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

//! Trait implementations for model training.
//!
//! Implements `TSClassificationModel` trait for all classification models.

use burn::prelude::*;
use burn::tensor::backend::AutodiffBackend;
use tsai_core::TSClassificationModel;

use crate::cnn::{FCN, InceptionTimePlus, OmniScaleCNN, ResNetPlus, XCMPlus, XceptionTime};
use crate::rnn::{RNNAttention, RNNPlus};
use crate::rocket::HydraPlus;
use crate::transformer::{PatchTST, TSPerceiver, TSiTPlus, TSTPlus};

// CNN Models

impl<B: AutodiffBackend> TSClassificationModel<B> for InceptionTimePlus<B> {
    fn forward(&self, x: Tensor<B, 3>) -> Tensor<B, 2> {
        self.forward(x)
    }
}

impl<B: AutodiffBackend> TSClassificationModel<B> for XceptionTime<B> {
    fn forward(&self, x: Tensor<B, 3>) -> Tensor<B, 2> {
        self.forward(x)
    }
}

impl<B: AutodiffBackend> TSClassificationModel<B> for FCN<B> {
    fn forward(&self, x: Tensor<B, 3>) -> Tensor<B, 2> {
        self.forward(x)
    }
}

impl<B: AutodiffBackend> TSClassificationModel<B> for ResNetPlus<B> {
    fn forward(&self, x: Tensor<B, 3>) -> Tensor<B, 2> {
        self.forward(x)
    }
}

impl<B: AutodiffBackend> TSClassificationModel<B> for OmniScaleCNN<B> {
    fn forward(&self, x: Tensor<B, 3>) -> Tensor<B, 2> {
        self.forward(x)
    }
}

impl<B: AutodiffBackend> TSClassificationModel<B> for XCMPlus<B> {
    fn forward(&self, x: Tensor<B, 3>) -> Tensor<B, 2> {
        self.forward(x)
    }
}

// Transformer Models

impl<B: AutodiffBackend> TSClassificationModel<B> for TSTPlus<B> {
    fn forward(&self, x: Tensor<B, 3>) -> Tensor<B, 2> {
        self.forward(x)
    }
}

impl<B: AutodiffBackend> TSClassificationModel<B> for TSiTPlus<B> {
    fn forward(&self, x: Tensor<B, 3>) -> Tensor<B, 2> {
        self.forward(x)
    }
}

impl<B: AutodiffBackend> TSClassificationModel<B> for TSPerceiver<B> {
    fn forward(&self, x: Tensor<B, 3>) -> Tensor<B, 2> {
        self.forward(x)
    }
}

impl<B: AutodiffBackend> TSClassificationModel<B> for PatchTST<B> {
    fn forward(&self, x: Tensor<B, 3>) -> Tensor<B, 2> {
        self.forward(x)
    }
}

// ROCKET Models
// Note: Rocket, MiniRocket, MultiRocket take pre-extracted 2D features, not raw 3D time series
// They should use a different trait (TSFeatureModel) or be wrapped with feature extraction

impl<B: AutodiffBackend> TSClassificationModel<B> for HydraPlus<B> {
    fn forward(&self, x: Tensor<B, 3>) -> Tensor<B, 2> {
        self.forward(x)
    }
}

// RNN Models

impl<B: AutodiffBackend> TSClassificationModel<B> for RNNPlus<B> {
    fn forward(&self, x: Tensor<B, 3>) -> Tensor<B, 2> {
        self.forward(x)
    }
}

impl<B: AutodiffBackend> TSClassificationModel<B> for RNNAttention<B> {
    fn forward(&self, x: Tensor<B, 3>) -> Tensor<B, 2> {
        self.forward(x)
    }
}

#[cfg(test)]
mod tests {
    #[test]
    fn test_traits_compile() {
        // Trait implementations compile - actual tests in integration tests
    }
}