CLI for creation and implementation of new modules in the main neural network stream

• GCD: finding the greatest common divisor. (Module for testing)
• Template (in process): creating your own neural network based on a ready-made template.
• GUI model/training viewer (features)
• Parsing (features): getting information from the Internet by key text.
• Vision (features): object recognition.
• Text Handler (features): chatting.

• Matrix: write the matrix of neural network.
• Memory: save data in the neural network memory.

git clone [email protected]:PrefectSol/Neuro.git
cd Neuro/
cmake .
make
./Neuro

To set custom settings, create a conf.json using the create command, change the settings in the created file, then run the utility using the run command.

To view the hints to the conf.json using the help command or check void createConfig(std::string path) from src/Algorithm.cpp

• CMake

• JsonCpp

• Curl

• RapidJSON

  Installation on Linux (Arch)

  sudo pacman -S cmake
  sudo pacman -S jsoncpp
  sudo pacman -S curl libcurl3 libcurl3-dev
  sudo pacman -S rapidjson

  Installation on Windows

  winget install cmake

// The basic structure for creating the simplest neural networks using the CLI is ready.
// I changed the structure of the project and now everything works as a template.
// But unfortunately because of this, the learning rate of neural networks has dropped from 0.4 to 2-5
// seconds.