TorchRayLib++: A CMake based integration of the RayLib GUI library and the PyTorch / Libtorch C++ Deep Learning Library.

https://github.com/QuantScientist/TorchRayLib

TorchRayLib++ is a CMake based integration of the well-known raylib GUI library with my favourite Deep Learning Library Libtorch: the PyTorch C++ frontend, mainly for working on the GPU.

Why??++ Because its fun, for gaming, for deep learning researchers, for AI, for reinforcement learning, because the PyTorch community does not have a dedicated UI.

RayLib

RayLib is an amazing library which has been widely adopted by the gaming community. Read more about the raylib game framework here: https://www.raylib.com/ Or look up projects using it here: https://www.google.com/search?q=raylib+site:github.com

PyTorch / Libtorch C++

PyTorch is a Python package that provides two high-level features, Tensor computation (like NumPy) with strong GPU acceleration Deep neural networks built on a tape-based autograd system. In this project we use the C++ version entitled Libtorch. https://pytorch.org/

Alpha release

Try it here: https://github.com/QuantScientist/TorchRayLib/releases Note: it is a huge file because of the included orch / NVIDIA CUDA runtime files.

These are some of the executables packaged together with the NVIDIA DLLs for testing purposes. Tested on both a GPU and a CPU on a Windows 10 machine with CUDA 10.2 and NVIDA 1080TX GPU.