A neural network architecture for NLP tasks, inspired in the SENNA system. It uses as input vectors built through Vector Space Models and avoid external NLP tools. Currently, it can perform POS tagging, SRL and dependency parsing. - View it on GitHub