MAMMOTH: Massively Multilingual Modular Open Translation @ Helsinki

arXiv:2403.07544v1 Announce Type: new
Abstract: NLP in the age of monolithic large language models is approaching its limits in terms of size and information that can be handled. The trend goes to modularization, a necessary step into the direction of designing smaller sub-networks and components with specialized functionality. In this paper, we present the MAMMOTH toolkit: a framework designed for training massively multilingual modular machine translation systems at scale, initially derived from OpenNMT-py and then adapted to ensure efficient training …

abstract age arxiv components cs.cl designing helsinki information language language models large language large language models massively multilingual modular multilingual networks nlp paper terms translation trend type