CUDANet

⚠️ Work in progress

Convolutional Neural Network inference library running on CUDA.

Features

• Input layer
• Dense (fully-connected) layer
• Conv2d layer
• Max pooling
• Average pooling
• Concat layer
• Sigmoid activation
• ReLU activation
• Softmax activation
• Load weights from file

Usage

requirements

• cmake
• CUDA
• Google Test (for testing only)

build and test

mkdir build
cd build
cmake -S ..
make

Run tests

./test/test_main