:sparkles: Features - Add the updated DaNE dataset - Add spaCy model for DEP, NER and POS - Add finetuned M-BERT and Danish BERT for NER - Add DSL Reddit wordembeddings
:bug: Bugfixes - Fix bug 13 loading wikiann as spaCy dataset. - Add a new Flair POS model for support for newest Flair and PyTorch
0.0.5
:sparkles: Features
- Add of the word similarity datasets DSD and the danish WS353 - Add Sketchengine fasttext wordembeddings
0.0.4
:sparkles: Features
- Loading functions for Flair NER model NER trained on the NER tagged DDT
0.0.3
:sparkles: Features
- Inclusion of the new NER annotated DDT dataset - Support for automatically downloading danish datasets - Loading functions for DDT with NER and the WikiANN dataset in both spaCy and Flair
0.0.2
:bug: Bugfixes - 55155282fdcd060af94a0775de8b7e9178c3a02d Fixes bug with always converting to spaCy format for loading embeddings with spaCy.
0.0.1
✨ Features - Load Danish word embeddings with spaCy, Gensim, PyTorch, Keras and Flair - Pretrained Danish POS tagging model trained with Flair - Automatic download of models when using load functions