OpenEuroLLM is an initiative launched by a consortium of 20 European institutions and companies, spearheaded by Jan Hajič and Peter Sarlin. The project focuses on developing multilingual large language models that are not only open-source but also compliant with European regulations. While it emphasizes transparency and accessibility in AI development, experts have raised concerns about the project's feasibility, particularly regarding its budget of 56 million EUR and whether 20 collaborating institutions can effectively produce competitive models. This initiative underscores the ongoing debate around the balance between open-source principles and practical outcomes.
OpenEuroLLM's initiative represents a collective effort to develop open-source, multilingual large language models that align with European regulatory frameworks.
The project's commitment to openness includes not just model weights but datasets, training code, and evaluation metrics, fostering transparency.
Collection
[
|
...
]