Submarine + MLflow

1. Build your dockerfile with your python script. Here is an example

2. Lauch a training job from submarine workbench

3.(optional) Lauch a job from command line

curl -X POST -H "Content-Type: application/json" -d '                          
{
  "meta": {
    "name": "mlflow-example",
    "namespace": "default",
    "framework": "TensorFlow",
    "cmd": "python3 /var/example/train.py",
    "envVars": {
      "MLFLOW_TRACKING_URI": "http://submarine-mlflow-service:5000"
    }
  },
  "environment": {
    "image": "pingsutw/mlflow-example:latest" 
  },
  "spec": {
    "Worker": {
      "replicas": 1,
      "resources": "cpu=4,memory=4096M"
    }
  }
}
' http://127.0.0.1:80/api/v1/experiment

4. Metrics in mlflow UI