Abstract: The paper deals with the problem of English-Telugu neural machine translation (NMT) in the conditions of extremely low resources. Telugu is a morphologically rich, free word order, Dravidian ...