mykelismyname / label-context-aware-attention-model Goto Github PK

Hi

Where can I find the code that converts the json files into the txt file?

I wanted to try using your architecture to detect and classify spans as well, on a new dataset wherein I have spans of text, with labels that could overlap. So, I was wondering If I could just convert such spans to a similar format as that of e.g. ebm_comet_multilabels.txt , Then I could run the pre_process methods, followed by training without changing much of the code, am I right ?

Thanks