AIAA (AI-Assisted Annotation)

MITK plugin

AIAA Installation

Official installation guide: https://docs.nvidia.com/clara/clara-train-sdk/aiaa/index.html

Create a folder for the AIAA to save models, logs and configurations:

mkdir aiaa_experiments
# change the permission because AIAA is running as non-root
chmod 777 aiaa_experiments

Install Docker Compose

Official installation guide: https://docs.docker.com/compose/install/

Download Docker Compose by running the following, replacing 1.29.2 with the latest stable release:

sudo curl -L "https://github.com/docker/compose/releases/download/1.29.2/docker-compose-$(uname -s)-$(uname -m)" -o /usr/local/bin/docker-compose

Apply executable permissions:

sudo chmod +x /usr/local/bin/docker-compose

Test installation:

docker-compose --version

Tip

If docker-compose fails, try creating a symbolic link in /usr/bin:

sudo ln -s /usr/local/bin/docker-compose /usr/bin/docker-compose

Set up Docker containers

Create files docker-compose.yml and docker-compose.env by copying the contents from here.

In docker-compose.env, change <YOURK WORKSPACE> to the absolute path of the aiaa_experiments folder.

AIAA_PORT can be changed to the port you want to use on the host machine.

You can modify the device_ids section under the deploy section of tritonserver service to change the GPU id that you want to use.

Start the AIAA Server

Start:

docker-compose --env-file docker-compose.env -p aiaa_triton up --remove-orphans -d

Stop:

docker-compose --env-file docker-compose.env -p aiaa_triton down --remove-orphans

Check log:

docker-compose --env-file docker-compose.env -p aiaa_triton logs -f -t

Start one service:

docker-compose --env-file docker-compose.env -p aiaa_triton up [the service name]

Check if AIAA and Triton is running:

docker ps

Access the server at http://127.0.0.1:$AIAA_PORT, where $AIAA_PORT is the port number in docker-compose.env.