Neural is an open source library designed to easily work with neural networks in the language C++ programming. It provides a simple and clear interface that allows both novices and experienced it is easy for developers to create and train neural networks.
Thanks to the open source code, any developer can study the source code of the library, as well as contribute their own changes to improve its functionality or fix bugs. This makes it possible to make the library more efficient and easy to work with.
What's more, Neural can also be used to conduct research in the field of neural networks. Thanks to simplicity use and flexibility of the library, developers can quickly create different architectures of neural networks and check their effectiveness in solving various problems.
- Multilayer Perceptron
- Four activation functions
- Auto fetching dependent DLLs (Experimental)
- Multithreading
- GPU multithreading
- Documentation
- Cross-platform
- More models
-
Install CMake
-
Create CMake project
-
Add
set(CMAKE_PREFIX_PATH path/to/NeuralConfig.cmake)
-
Add
include_directories(path/to/include)
-
Add
find_package(Neural REQUIRED)
-
Congratulations!
If you're having trouble, you can always refer to examples.