Sentence similarity with SparkNLP only works on Google DataProc with ONE sentence, FAILS when multiple sentences are provided

Question

Deployed the following colab python code(see link below) to DataProc on Google Cloud and it only works when the input_list is an array with one item, when t