Run a Preconfigured ASCMO Project

A preconfigured project is an ASCMO project file that has been exported for execution in Docker.

You can export an ASCMO-STATIC project for Docker at the following locations:

Optimization menu > Single Result > Optimize/ Optimize at OPs > Export Job to Docker / Export Job to Docker (at OPs)

Single Result Optimization
Optimization menu > Multi Result > Optimize > Export Job to Docker

Multi Result Optimization
Optimization menu > Global Optimization > Optimize > Export Job to Docker

Global Optimization
Model menu > Model Properties > Start Training > Export Job to Docker

Model Properties (ASCMO-STATIC)
Model menu > Automated Machine Learning > Export Job to M Script > Export Job to Docker

Automated Machine Learning

You can export an ASCMO-DYNAMIC project for Docker at the following locations:

Model menu > Configurations/Configuration for All> Export Job to M Script > Export Job to Docker

Model Configurations (ASCMO-DYNAMIC)
Model menu > Automated Machine Learning > Export Job to M Script > Export Job to Docker

Automated Machine Learning

Prepare the project folder

Create a folder on the host machine.
Copy the prepared project file into this folder.
Copy all files required by the project into the same folder or into the required relative subfolders.

Start the computation

Run the container with the project folder mounted to /data:

Copy

docker run --rm -t \
    -e LICENSE_USER=<your_user_name> \
    --mount type=bind,source=/path/to/project_folder,target=/data \
    --mount type=bind,source=/path/to/network.lic,target=/license/network.lic,readonly \
    ascmo_image compute /data/my_project.docker.moca

The container executes the project and writes the results back to the mounted project folder on the host machine.

The path after compute is the path inside the container. In this example, the host folder /path/to/project_folder is mounted to /data. Therefore, the project file is referenced as /data/my_project.docker.moca.