Visual Recognition Environment Setup and Assignments

Requirements

• Nvidia GPU Drivers (For employing GPU in ML)
• docker
• nvidia-docker

Local Setup

1. Run docker build -t <image_name> <Dockerfile location> or docker pull shivnshu/pytorchwithcuda
2. Run nvidia-docker run -v <folder_location>:/root/Project -p 8888:8888 --rm -ti --ipc=host <image_name>
3. Run jupyter notebook in container and access web interface on host machine on url http://localhost:8888 (default password: shivnshu)