Rust Convolutional Neural Network from Scratch
This repository contains a Rust implementation of a Convolutional Neural Network (CNN) built from scratch. The CNN is designed to learn and classify the MNIST dataset.
Overview
The repository contains the following main components:
src/
├── cnn_struct.rs
├── conv_layer.rs
├── fully_connected_layer.rs
├── layer.rs
├── lib.rs
├── main.rs
├── max_pooling_layer.rs
└── run.rs
cnn_struct.rs: Defines the structure of the CNN model.conv_layer.rs: Implements the convolutional layer for the CNN.fully_connected_layer.rs: Implements the fully connected layer for the CNN.layer.rs: Defines the interface for the CNN layers.lib.rs: The Rust library file.main.rs: A demo of the CNN's use.max_pooling_layer.rs: Implements the max pooling layer for the CNN.run.rs: Contains functions to run the CNN.
Installation
To use this CNN implementation, you must have Rust and Cargo installed on your machine. After installing Rust and Cargo, you can clone this repository to your local machine and build the project with the following command:
$ cargo build
Usage
To run the demo of the CNN, place the MNIST dataset in a folder named data, and use the following command:
$ cargo run
This command will run a demo of the CNN and train it on the MNIST dataset.
Further Reading
For more information about this project, read my blog post on CNNs.
License
This project is licensed under the GNU Affero General Public License v3.0 - see the LICENSE file for details.