How to retrieve inputs from a Vertex AI custom job?

Question

I created a custom job on Google Cloud Platform Vertex AI based on a Python distribution with some configuration inputs. For example:

The list of these inputs are available in the Training menu in a JSON format:

{
  "workerPoolSpecs": [
    {
      "machineSpec": {
        "machineType": "n1-standard-4"
      },
      "replicaCount": "1",
      "diskSpec": {
        "bootDiskType": "pd-standard",
        "bootDiskSizeGb": 100
      },
      "pythonPackageSpec": {
        "executorImageUri": "europe-docker.pkg.dev/vertex-ai/training/sklearn-cpu.1-0:latest",
        "packageUris": [
          "gs://my_bucket_for_vertexai/my_package_directory/trainer-0.1.tar.gz"
        ],
        "pythonModule": "trainer.task"
      }
    }
  ],
  "baseOutputDirectory": {
    "outputUriPrefix": "gs://my_bucket_for_vertexai/my_output_directory/"
  }
}

How can my job automatically retrieve these parameters without their value being written in the Python code ?

Answer 1

You can retrieve the JSON you show with aiplatform.CustomJob.get(job_id). This answer shows how you can get the job ID. The only thing is, I am not sure if these resources are created before the container is run so that it can access them..