Submission

Participating teams are expected to submit the following by the submission deadline:

1. Codes for (i) training the model, and (ii) generating predictions on the test set using the model. Instructions for running the code should be given in a README file.

2. One or two (max) prediction file(s) for each track in which the team is participating. Exact format for the test predictions files will be announced closer to the submission deadline. It must be clearly stated which prediction file is for which track.

3. One report (2-3 pages) describing all the models, highlighting any novel steps taken to improve the models, hyperparameters / settings for reproducibility (e.g., batch size, learning rate), etc.