v1.0

🕺 Enhancements:

• Running the baselines in main.py now writes the model predictions to a TSV file before running the score function from scorer.py on them. This means that this process works more similarly to the evaluation on the Codalab page. Participants can have a look at the TSV file with predictions to see how the format of their own submission file should look like.

🔨 Breaking changes:

• Calling main.py now requires the additional flag --path_to_predictions to specify a TSV file in which to write baseline predictions.
• The function retrieve_instances_from_dataset in data.py now returns a tuple with a list of ids and a list of instance strings instead of only a list of instance strings.

🐛 Bug fixes:

• The function check_format_for_ranking_submission in format_checker_for_submission.py does no longer require predicted ranking scores to be within the range between 1 and 5 as this is a too strict assumption.