Crate iron_learn

Crate iron_learn 

Source

Re-exports§

pub use crate::tensor::Tensor;
pub use crate::neural_network::ActivationFn;
pub use crate::neural_network::ActivationLayer;
pub use crate::neural_network::Layer;
pub use crate::neural_network::LinearLayer;
pub use crate::neural_network::LossFunction;
pub use crate::neural_network::MeanSquaredErrorLoss;
pub use crate::neural_network::NeuralNet;
pub use crate::neural_network::NeuralNetBuilder;

Modules§

init
neural_network
read_file
tensor
Tensor Module - Linear Algebra Core

Structs§

AppContext
Global application context with training configuration and GPU capabilities
Complex
Dataset container for double-precision (f64) examples.
CpuTensor
The CpuTensor structure is the backend implementaion of the Tensor trait. The implementation uses CPU for calcualations. Although, Auto Vectorization has been used whereever possible for parallel execeution.
Data
DataDoublePrecision
Dataset container for single-precision (f32) examples.

Statics§

GLOBAL_CONTEXT
Global singleton instance of application context

Traits§

Numeric
The Numeric trait defines a set of operations that numeric types must support. It includes basic arithmetic operations and the ability to return special values like zero and one.
SignedNumeric
The SignedNumeric defines all the Numeric types that can be signed like i32, i64 etc.

Functions§

gradient_descent
Perform a single gradient descent update step.
init_context
Initialize the global application context
linear_regression
Train a linear regression model using gradient descent.
logistic_regression
Train a logistic regression model using gradient descent.
normalize_features
Normalize features using provided mean and std per feature.
predict_linear
Predict outputs for x using linear model weights w.
predict_logistic
Predict binary labels for x using logistic model weights w.
run_linear
Run linear regression using configuration from the global context.
run_logistic
Run logistic regression using configuration from the global context.
run_neural_net
Train and evaluate a neural network using configuration from the global context.