Ecosyste.ms: Awesome

An open API service indexing awesome lists of open source software.

Awesome Lists | Featured Topics | Projects

https://github.com/a1exxd0/hadolibrary

Lightweight, header only ML library with support for OpenMP for a variety of neural networks.
https://github.com/a1exxd0/hadolibrary

cpp20 machinelearning neuralnetwork

Last synced: 24 days ago
JSON representation

Lightweight, header only ML library with support for OpenMP for a variety of neural networks.

Awesome Lists containing this project

README 

        
# A library for implementing Neural Networks in C++

Lightweight, templated, highly vectorised (through Eigen backend), and optimised library for a variety of neural networks. Tested for Linux.



Supports models for:

 - Deep Neural Networks

 - Convolutional Neural Networks



Will support soon:

 - RNN Networks

 - LSTM Networks

 - Transformer Networks



Example usages are stored in src folder to run. See "XorModel.cpp", for example. 



# Quickstart

You'll need the Eigen library inside of your project, as well as everything in the HaDo/ subdirectory. See the Makefile for a general idea of how to compile it all. Use "-fopenmp" for compilation with multithreading enabled, if you have OpenMP installed on your system. The MakeFile provided gives an example using 'make omp' as a target.



You could have for an example file structure:



```

project  

|   MakeFile

|   

├───src

|   |   main.cpp

|

├───Eigen

|   ...

|

├───HaDo

|   ...

|

└───res.png

```



Notably, you should include the Eigen and Hado files in your include flags when compiling.



To use in your main program, simply import the desired module set available in the top level of the HaDo folder, for example:



```cpp

#include 



void TwoCategoryMNIST(){

    hado::Pipeline pipeline;



    pipeline.pushLayer(

        hado::ConvolutionalLayer, hado::relu_prime>(1, 2, 28, 28, 3, 1, 0)

    );



    pipeline.pushLayer(

        hado::MaxPoolLayer(2, 26, 26, 2, 2, 0)

    );



    ...

}

```



# Contributions

Feel free to for the repository and set up a pull request, either to resolve an existing issue, or add a new feature. All help is appreciated here with the team of just 2 of us!



Code should pass tests before requesting to merge, and if you write fully new code, it would be awesome to see some tests made for that too!



In order to make and build for tests, run:

```sh

cd ${PROJECT_SOURCE_DIR}/

mkdir build && cd build

cmake ..

cmake  --build .



# FOR THIS, IT MAY BE THE CASE THAT THERE ARE SOME "POTENTIAL MEMORY LEAKS"

# After thorough research and profiling, we have deduced it is not from our

# end, but instead from the use of OpenMP. We have suppressed these errors

# such that the error code is 0. Any other (memory) error except this will

# flag up.

ctest -V -C -T memcheck

# or just ctest -V -C if you want a quick check

```

This is more of a learning project for new C++ers, so there's no real strict requirements here. Feel free to reach out to either myself or other contributors for help on this.



Happy coding!



# TODO

  - [X] Image -> Bitmap formatting ??? Maybe use PPM for raw RGB

  - [x] Dense Layers

  - [x] Activation Layers (tanh, sigmoid, ReLU, softmax so far)

  - [X] Mean Squared Error implementation

  - [X] Cross-Entropy Loss implementation

  - [X] Method to pass through and verify layer setup

  - [X] Pipeline class

  - [X] Convolutional Layers

  - [X] Pooling Layers

  - [ ] Get MNIST working

  - [ ] Get egg categorization working

  - [ ] Saving a model (JSON or binary?)

  - [ ] Getting it running on a GPU

  - [ ] Getting it running on a cluster