Construyendo Machine Learning de manera Open Source y en Comunidad

[David-Lazaro-Fernandez]

Ponente: Omar Sanseviero

"In this session, Omar will talk about how Hugging Face has been able to build collaborative Machine Learning through the usage and creation of Open Source tools.

Some areas which will be covered:

• Hugging Face created some of the most popular OS libraries, such as transformers, which has over 65k stars and 1300 contributors.

• BigScience is a one-year workshop on large multilingual models and datasets which involves over 900 researchers from 250 institutions to train an open-science large language model.

• The NLP in Spanish community organized a Hackathon in which the participants trained over 30 models, obtained 20 datasets and built 20 demos for different ML applications in Spanish (https://huggingface.co/hackathon-pln-es)"