Browsing School of Electronic Engineering and Computer Science by Author "Repar, A"
Now showing items 1-1 of 1
-
Evaluation of contextual embeddings on less-resourced languages
Ulčar, M; Žagar, A; Armendariz, CS; Repar, A; Pollak, S; Purver, M; Robnik-Šikonja, M (2021)The current dominance of deep neural networks in natural language processing is based on contextual embeddings such as ELMo, BERT, and BERT derivatives. Most existing work focuses on English; in contrast, we present here ...