SETimesq\_sr is a Serbian dataset annotated for morphosyntactic information and named entities.

The dataset contains 3177 training samples, 395 validation samples and 319 test samples 
across the respective data splits. Each sample represents a sentence and includes the following features:
sentence ID ('sent\_id'), sentence text ('text'), list of tokens ('tokens'), list of lemmas ('lemmas'), list of Multext-East tags ('xpos\_tags), list of UPOS tags ('upos\_tags'), list of morphological features ('feats'), and list of IOB tags ('iob\_tags'). The 'upos\_tags' and 'iob\_tags' features
are encoded as class labels.