We read every piece of feedback, and take your input very seriously.
To see all available qualifiers, see our documentation.
A neural net based shift-reduce dependency parser (LSTMs and MLP combiner networks, no pretrained embedding)