Is there any reason why gold summaries for validation data is not released, as this would be extremely useful in the Test Phase.
And to be clear, there are only 3 submissions possible for test data right? The submission platform mentions
- For every subtask, maximum 10 submissions per day are allowed.
Please clarify this point.
Regarding paper publication, will gold summaries or evaluation script for test set be released? This would be useful to carry out experiments which are not entirely focused on improving ROUGE accuracy.