This is just a simple demo as how to use Spark NLP in a SBT project in Scala. Usually, once you have your application you want to package it and run it via spark-submit.

After your executed sbt assembly to get a Fat JAR (without Apache Spark since your environment has Apache Spark already), you can use spark-submit like this:

sbt assembly

~/spark-3.1.1-bin-hadoop3.2/bin/spark-submit \
--class "Main" \
target/scala-2.12/spark-nlp-starter-assembly-1.1.jar

This will execute the code in Main class and finish successfully.