search

smartschat / cort

A toolkit for coreference resolution and error analysis.
MIT License
129 stars 34 forks source link

Memory requirements #1

Open smartschat opened 9 years ago

smartschat commented 9 years ago

cort currently needs a lot of RAM, predicting with the latent ranking model on the CoNLL-2012 development data takes ~8GB, mainly due to multiprocessing during feature extraction.

ghost commented 6 years ago

Since Cort has Cython as a dependency anyways, maybe that code should be rewritten in Cython format (.pyx), which may reduce memory consumption.