In Keras iris classification model getting error process failed.

RBSB61RBSB61 Member Posts: 3 Contributor I
  • Exception: java.lang.NullPointerException
  • Message: null
  • Stack trace:
  • java.lang.ProcessBuilder.start(
  • com.rapidminer.operator.scripting.AbstractScriptRunner.getProcessWithLogging(
  • com.rapidminer.operator.scripting.python.PythonScriptRunner.start(
  • com.rapidminer.extension.keras.general.Sequential.doWork(
  • com.rapidminer.operator.Operator.execute(
  • com.rapidminer.operator.execution.SimpleUnitExecutor.execute(
  • com.rapidminer.operator.ExecutionUnit$
  • com.rapidminer.operator.ExecutionUnit$
  • Method)
  • com.rapidminer.operator.ExecutionUnit.execute(
  • com.rapidminer.operator.OperatorChain.doWork(
  • com.rapidminer.operator.Operator.execute(
  • com.rapidminer.Process.executeRoot(
  • com.rapidminer.Process.lambda$executeRootInPool$5(
  • java.util.concurrent.ForkJoinTask.doExec(
  • java.util.concurrent.ForkJoinPool$WorkQueue.runTask(
  • java.util.concurrent.ForkJoinPool.runWorker(
Also potential problem which was detected Mandatory input at port missing Apply Keras Model.model


  • ceaperezceaperez Member Posts: 473 Unicorn
    Hi @RBSB61,

    Are you trying to run a model using the Keras extension?
    The Keras extension is so specific about the version as Varunm1 said in this post
    The best option is to work with python with the Execute Python operator. 

  • jacobcybulskijacobcybulski Member, University Professor Posts: 391 Unicorn
    Yes, Keras app has not been updated for a long while - since 2017. The last time I used it it worked with Python 3.6 and Tensorflow 1.4, with Keras installed separately on top. As the world moved on and now Keras is part of Tensorflow 2+ distribution, this extension has slipped into obsolescence. I think it is time to move into Deep Learning extension, which is being developed and supported by RapidMiner! 
Sign In or Register to comment.