What exactly are we supposed to submit in the zip file? Which data are we expected to use? Do we have to train our model on train.csv and test it on public data under development phase?
Posted by: Nisha_Ramrakhyani @ Nov. 17, 2020, 12:53 p.m.Hi Nisha,
The details of what to submit in the zip file can be found here: https://competitions.codalab.org/competitions/27446#learn_the_details-evaluation
Please use the train.csv data to train your model. You can split it into a train and dev set to run your own experiments, or you can leave the dataset as is, and submit your results to our submission page in order to check your results. You may use extra data, if required, but we request that you do not use the Kaggle Short Jokes Dataset.
Best of luck!
Posted by: jam @ Nov. 18, 2020, 8:50 p.m.