Just an inquiry of training process

[TheoSeo93]

Hi ! I've read your paper very impressively, and I was thinking where I could see the process of training the SEQ model for CWI, but it seems I couldn't find it from the repository. In order to do so, it seems that I should follow the instruction of Sequence Labeler by Marek Rei, doesn't it?

Your paper described that SEQ model could benefit from other than word embeddings and character-level morphology, and I'd really like to run some experiments by adding some other features. In order to do so, would there be a good approach to add other features on top of your SEQ model?

Thank you !