Training a model on a dataset that is not ace04, ace05, or scierc

[serenalotreck]

I'm trying to train a set of models on my own dataset. When I provide the shorthand name of my own dataset for the --task parameters, I get an error that I have to choose either ace04 | ace05| scierc.

However, when I used scierc as the value for the --task parameter, I get a Segmentation Fault when I try and train the model.

Wondering how to train a model on a new dataset?

[a3616001]

Hi @serenalotreck , to add a new task, you may also want to modify shared/const.py. See #1 . Let me know if you have further questions!