Compare word2vec models to BERT models for parliament corpora

[jgonggrijp]

Are you going to train those BERT models yourself? In that case let's meet some time, because I'm going to train BERT models in the near future, too.

[lukavdplas]

We have decided to focus on word embeddings rather than contextual embeddings / BERT - closing this.