Pun detection with neural networks

By William Grigor

System Requirements

• An operating system running the XServer desktop environment

• Docker version 19.0 or higher

• CUDA version 10.0

  • A CUDA enabled graphics card (you can find a list of CUDA enabled GPUs here)

• CuDNN for CUDA 10.0 (install instructions here)

Installing and using

• Clone this git repository

git clone https://github.com/FrostByte266/pun_detector.git

• Build the Docker container

docker build -t puns .

• Run the Docker container

./container

OR

xhost +
docker run -u $(id -u):$(id -g) --gpus all -it --rm --net=host -e DISPLAY -v $PWD/src:/src -v $PWD/test:/test -v $PWD/data:/data -v $PWD/logs:/logs -e TF_FORCE_GPU_ALLOW_GROWTH=true puns bash

• Start the main program

python /src/gui.py