I am struggling with one of the functionalities of running the pipeline on the test data. The pipeline is able to predict a single line but is not working properly on the dataset and give a CoNLL-U score for posdep. The error is as follows:
539 head_seqs = [chuliu_edmonds_one_root(adj[:l, :l])[1:] for adj, l in
540 zip(predicted_dep[0], sentlens)]
--> 541 deprel_seqs = [[self._config.itos[DEPREL][predicted_dep[1][i][j + 1][h]] for j, h in
542 enumerate(hs)] for
543 i, hs
I am struggling with one of the functionalities of running the pipeline on the test data. The pipeline is able to predict a single line but is not working properly on the dataset and give a CoNLL-U score for posdep. The error is as follows:
The code link is: https://colab.research.google.com/drive/1oUdBkJDIrnHR6fhdEOeIevBoTCrG_PF2?usp=sharing
The relevant data files are: