Exception: Error when converting graph to tokenized version.

[ShuyangCao]

Congrats on your great work and thanks for releasing the code.

When I tried to run the fact graph evaluation script on summaries generated by my system with

./evaluate.sh factgraph <my_file> 0

I encountered an error saying:

Exception: Error when converting graph to tokenized version.

which seems to be triggered by:

unaccented_string = unidecode.unidecode(graph_string[idx_graph_string].lower())

IndexError: list index out of range

Any thoughts on the issue?

Thanks.