********************************************************************** * * *                          Einladung * * * *                     Informatik-Oberseminar * * * *********************************************************************** Zeit: Mittwoch, 28 August 2024, 15:30 Uhr Ort: Raum 9222, E3, Informatikzentrum Zoom: https://rwth.zoom-x.de/j/66261562709?pwd=2driMW8j0cRJLiYFWNJRqRp4Lya80Z.1 Meeting-ID: 662 6156 2709 Kenncode: 337848 Referent: Yingbo Gao, M.Sc. (Lehrstuhl Informatik 6) Thema: Language Modeling and Machine Translation: Improvements in Training and Modeling Abstract: Substantial improvements in language modeling and machine translation have been achieved since the wide adoption of artificial neural networks. In this talk, we discuss three directions related to training and modeling in neural language modeling and neural machine translation. First, sampling-based training criteria are investigated in order to speed up the training of neural language models with large vocabularies. Second, label smoothing, input smoothing as well as multi-agent training are studied to improve the generalization of neural machine translation models. Finally, a language modeling approach for machine translation is proposed to simplify the architecture of existing translation models. Es laden ein: die Dozentinnen und Dozenten der Informatik -- Stephanie Jansen Faculty of Mathematics, Computer Science and Natural Sciences Chair of Computer Science 6 ML - Machine Learning and Reasoning RWTH Aachen University Theaterstraße 35-39 D-52062 Aachen Tel: +49 241 80-21601 sek@ml.rwth-aachen.de www.hltpr.rwth-aachen.de