create_pretraining_data.py is writing 0 records to tf_examples.tfrecord while training custom BERT model

Question

I am writing a custom BERT model on my own corpus, I generated the vocab file using BertWordPieceTokenizer and then running below code

!python create_pretraining_