rumus 0.1.0

A native-Rust deep learning framework with explicit memory safety and hardware acceleration
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

//! Neural network building blocks: parameters, modules, layers, losses, and IO.

mod activations;
mod conv;
mod dropout;
mod flatten;
mod io;
mod linear;
mod loss;
mod module;
mod parameter;
mod pool;

pub use activations::relu;
pub use conv::Conv2d;
pub use dropout::Dropout;
pub use flatten::Flatten;
pub use io::{load_safetensors, save_safetensors};
pub use linear::Linear;
pub use loss::{cross_entropy_loss, mse_loss};
pub use pool::MaxPool2d;
pub use module::Module;
#[cfg(feature = "gpu")]
pub use module::ModuleToGpu;
pub use parameter::Parameter;

// Re-export the derive macro so users can write `use rumus::nn::Module;`
// and `#[derive(Module)]` in the same scope.
pub use rumus_macros::Module;