Tensorflow Invalid Argument: Assertation Failed [Label IDs must < n_classes]

Question

I get an error implementing a DNNClassifier in Tensorflow 1.3.0 with Python 2.7. I got the sample code from the Tensorflow tf.estimator Quickstart Tutorial and

Answer 1

So the solution as Ishant Mrinal pointed out:

Tensorflow expects the integers from 0 up to the number of classes as class labels (range(0, num_classes)), not "arbitrary" numbers like in my case. Thanks!:)

...The other option I just came across is to add a label_vocabulary to the classifier-definition:

classifier = tf.estimator.DNNClassifier(feature_columns=feature_columns,
                                          hidden_units=[10, 20, 10],
                                          n_classes=numClasses,
                                          model_dir=saveAt,
                                          label_vocabulary=uniqueTrain)

With this option I can define the labels like I had before, converted to strings.