Deep Neural Networks in C++

I'm created a github repository for a DNN (Deep Neural Network) that will be written in C++. Right now it requires a C++17 compatible compiler, which can be found at  Compiler Explorer . I am compiling for either the ARM Cortex-a72 (Rpi 4B) gcc 8.2 (with -Os -std=c++17 -mtune=cortex-a72 or fur the X86-64 using -O3 which may give faster code, but more assembly) or X86-64 gcc 9.1 (with -Os -std=c++17/2a -Wall -Wextra) . My repository is at Neural Network cpp . There is much to do, feel free to peruse and send along suggestions. The goal is to make it multi dimensional so that imagery can be analyzed, thanks...