Using pretrained model to get argids produces the same output as frameid identification

[fabrahman]

Hi,

I was able to run the first two steps of annotating a corpus using pretrained model, i.e. target id and frame id identification.
But when I run the code for the argid identification, it produces the same output as its raw_input which is: logs/fn1.7-pretrained-frameid/predicted-frames.conll

Any one encouter the same issue? What am I doing wrong?

Thanks