Skip to content

File Model Output


Compatible Spark only

This node should subscribe from a mllib pipeline. By doing so, a binary file of the resulting model can be stored on a specified location.


A model is actually a serialized object, more precisely a Transfomer.


    type: punchline
    version: "6.0"
    runtime: spark
    tenant: default
    dag: [
            type: file_model_output
            component: file_model_output
            settings: {
                file_path: ./model.bin
            subscribe: [
                    component: mllib
                    stream: model


  • file_path: String

    Description: [Required] The name of the file specified within spark.files parameter.