aws-ml-neural-style

Use an MXNet neural network and a p2.xlarge (GPU) instance to transfer artistic style to photos

Getting Started

Deploy the CloudFormation stack in the template in infrastructure/. The template creates a user with the following credentials and minimal required permisisons to complete the Lab:

• Username: student
• Password: password

Instructions

• Connect to the instance using the SSH username: ubuntu.
• Change into the /home/ubuntu/src/mxnet/example/neural-sytle directory
• Run the GPU monitoring script developed by AWS: python gpumon.py
• Perform a neural style transfer by executing the following command:

python nstyle.py --content-image input/IMG_4343.jpg \
                 --style-image input/starry_night.jpg \
                 --gpu 0 \
                 --output_dir output/ \
                 --save-epochs 20 \
                 --max-num-epochs 300

Cleaning Up

Delete the CloudFormation stack to remove all the resources. No resources are created outside of those created by the template.