convolutions-rs is a crate that provides a fast, well-tested convolutions library for machine learning

Last update: Jun 28, 2022

Related tags

Overview

convolutions-rs

convolutions-rs is a crate that provides a fast, well-tested convolutions library for machine learning written entirely in Rust with minimal dependencies. In particular, this is the first convolutions crate that has absolutely no dependencies on native C libraries. We provide both transposed convolutions (also called deconvolutions), as well as normal convolutions.

This crate has been developed in the course of the ZipNet project (https://github.com/Conzel/zipnet), where we required a C-free implementation of convolutions in order to compile our code to WebASM.

Features

Minimal dependencies, especially no C-dependencies
Extensively tested through randomly generated unit tests
100% compatible with Tensorflow and Pytorch implementations
Generics to ensure smooth usage
Speed verified by benchmarking

Usage

As mentioned, this package provides normal convolutions as well as transposed convolutions. We provide both in the form of free functions as well as something resembling a neural network layer. This crate also requires ndarray to use the functions, as input and output are in the form of ndarrays.

Example:

use convolutions_rs::convolutions::*;
use ndarray::*;
use convolutions_rs::Padding;

// Input has shape (channels, height, width)
let input = Array::from_shape_vec(
    (1, 4, 4),
    vec![1.,2.,3.,4.,5.,6.,7.,8.,9.,10.,11.,12.,13.,14.,15.,16.]
)
.unwrap();

// Kernel has shape (channels out, channels in, height, width)
let kernel: Array4
   
     = Array::from_shape_vec(
    (2,1,2,2),
    vec![1.,1.,1.,1.,1.,1.,1.,1.]
)
.unwrap();

let conv_layer = ConvolutionLayer::new(kernel.clone(), 1, Padding::Valid);
let output_layer: Array3
    
      = conv_layer.convolve(&input);
let output_free = conv2d(&kernel, &input, Padding::Valid, 1);

println!("Layer: {:?}", output_layer);
println!("Free: {:?}", output_free);

You might also like...

Xaynet represents an agnostic Federated Machine Learning framework to build privacy-preserving AI applications.

Comments

Support for multiple images

Hi, I am wondering whether convolve function support for many images?

For example, I have 5 images. But if I want to compute convolve values, I need to write a for loop to iterate each image.

Is there any other way to do conv2d like pytorch? Thanks!

By the way, can convolve support i64 data?

opened by amyzx 2