Datum održavanja: 24. studenog 2022. u 14:00 sati, prostorija O-357
Predavač: Marko Robnik Šikonja, Professor of Computer Science and Informatics at University of Ljubljana, Faculty of Computer and Information Science
Naziv predavanja: Cross-lingual transfer with large language models
Sažetak predavanja:
Currently, the most successful approach to natural language processing is based on large pretrained language models using the transformer architecture of neural networks. These are typically pretrained on huge text corpora on the task of predicting masked tokens. While the majority of existing models are monolingual, multilingual variants also exist and can help in cross-lingual transfer of knowledge and models. We will present a few variants of cross-lingual transfer and show their advantages and disadvantages in text classification, summarization and question answering.