[Feature-based] Word Embedding Extractor
Extractor class which uses pre-trained word embedding models to encode the lemmata of a given input sequence. This should be done using flair (https://github.com/flairNLP/) and all different embedding models this framework supports should also be supported by the extractor. One must be able to pass respective models to the constructor of this class. Again, this extractor should be able to either return the embeddings for single tokens/lemmata or all of them.