MiniTorch

MiniTorch is a diy teaching library for machine learning engineers who wish to learn about the internal concepts underlying deep learning systems. It is a pure Python re-implementation of the Torch API designed to be simple, easy-to-read, tested, and incremental. The final library can run Torch code.

MiniTorch++

MiniTorch++ extend MiniTorch by implement core functionalities in C++ backend and emitted it in Python. So, MiniTorch++ is more efficient while keeping simple and easy to use.

Note: some functionality is under development....

Besides, MiniTorch++ support multiple device, such as CPU、GPU and Atlas accelerator.

Go to core/ directory, and build with CMake:

# 1. create build directory
mkdir build
# configurate project
cmake .. -D-DASCEND_CANN_PACKAGE_PATH="Your CANN Package Path"
# build project
make -j
# try test
./tensor_test