1. Analyzing the effects of domain-adaptive pretraining on propaganda technique detection in a low-resource language