The final parser model can now be save to a flow file after training. The model is written in a form that is compatible with the format used by the PyTorch-based trainer and the C++ parser runtime.
I have also deprecated the frame-to-mention mapping in Document class since it is no longer being used, except for one place in the LEX parser, which I have fixed so it no longer uses the mention map.
The final parser model can now be save to a flow file after training. The model is written in a form that is compatible with the format used by the PyTorch-based trainer and the C++ parser runtime.
I have also deprecated the frame-to-mention mapping in Document class since it is no longer being used, except for one place in the LEX parser, which I have fixed so it no longer uses the mention map.