Poland launches its Large Language Model

A free, open-source intelligent assistant, specifically trained in Polish, is being developed by a consortium led by Wrocław University of Science and Technology. The consortium includes State Research Institute NASK, Information Processing Center – State Research Institute (OPI PIB), Institute of Computer Science PAN, University of Łódź, and Institute of Slavic Studies PAN. The language model, named PLLuM (Polish Large Language Universal Model), aims to address the scarcity of models trained on representative Polish language datasets.
The initiative aims to provide businesses with a model featuring an expanded Polish language on a free, open-source license to meet market demands. The creators seek to overcome issues observed in popular models, such as ChatGPT, which may lack exposure to the Polish language during training, leading to errors in cultural and historical contexts. The project aims to support both businesses and citizens.
Source: forsal.pl