The core message of this article is to develop an effective dataset and deep learning models for accurately recognizing and classifying different Sorani Kurdish subdialects.