rustyml 0.14.0

A high-performance machine learning & deep learning library in pure Rust, offering ML algorithms and neural network support
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

//! Regularization layers and the shared training-mode infrastructure that backs them
//!
//! Re-exports the 3 families of regularization layers. Defines the macros that generate their
//! common training-mode methods, plus a private `validation` submodule of parameter and
//! input-shape checks shared across the layers.
//!
//! The families are:
//! - dropout: [`Dropout`](crate::neural_network::layers::regularization::dropout::dropout::Dropout)
//!   and the spatial variants [`SpatialDropout1D`](crate::neural_network::layers::regularization::dropout::spatial_dropout_1d::SpatialDropout1D),
//!   [`SpatialDropout2D`](crate::neural_network::layers::regularization::dropout::spatial_dropout_2d::SpatialDropout2D),
//!   and [`SpatialDropout3D`](crate::neural_network::layers::regularization::dropout::spatial_dropout_3d::SpatialDropout3D)
//! - noise injection: [`GaussianNoise`](crate::neural_network::layers::regularization::noise_injection::gaussian_noise::GaussianNoise)
//!   and [`GaussianDropout`](crate::neural_network::layers::regularization::noise_injection::gaussian_dropout::GaussianDropout)
//! - normalization: [`BatchNormalization`](crate::neural_network::layers::regularization::normalization::batch_normalization::BatchNormalization),
//!   [`LayerNormalization`](crate::neural_network::layers::regularization::normalization::layer_normalization::LayerNormalization),
//!   [`GroupNormalization`](crate::neural_network::layers::regularization::normalization::group_normalization::GroupNormalization),
//!   and [`InstanceNormalization`](crate::neural_network::layers::regularization::normalization::instance_normalization::InstanceNormalization)
//!
//! Every layer behaves differently in training versus inference. The module defines 2 macros to
//! toggle the shared `training` field. `mode_dependent_layer_set_training` generates the inherent
//! `set_training` method. `mode_dependent_layer_trait` generates the
//! `set_training_if_mode_dependent` trait method that calls it.

/// Dropout layers for neural networks
pub mod dropout;
/// Noise injection layers for neural networks
pub mod noise_injection;
/// Normalization layers for neural networks
pub mod normalization;
/// Input validation functions for regularization layers
mod validation;

pub use dropout::*;
pub use noise_injection::*;
pub use normalization::*;

/// Defines a layer-specific `set_training` method for toggling training mode
///
/// The generated method sets the `training` field to `true` (training) or `false`
/// (inference). This drives behavior in the forward and backward passes
macro_rules! mode_dependent_layer_set_training {
    () => {
        /// Sets the training mode for the layer, updating its `training` field
        ///
        /// # Parameters
        ///
        /// - `is_training` - `true` for training mode, `false` for inference mode
        pub fn set_training(&mut self, is_training: bool) {
            self.training = is_training;
        }
    };
}
pub(in crate::neural_network::layers::regularization) use mode_dependent_layer_set_training;

/// Defines the trait method `set_training_if_mode_dependent` for a layer whose behavior
/// depends on training versus inference mode
macro_rules! mode_dependent_layer_trait {
    () => {
        fn set_training_if_mode_dependent(&mut self, is_training: bool) {
            self.set_training(is_training);
        }
    };
}
pub(in crate::neural_network::layers::regularization) use mode_dependent_layer_trait;