Datasets:
Leaderboard
This is a great resource. Do you plan to maintain a leaderboard?
Thanks, Eli. We have had some initial discussions about that. Do you have any specific ideas/thoughts about what you'd like to see in the leaderboard?
I think it would be great to maintain a leaderboard to track new model performance over time, since the paper is just a snapshot.
Yes, agreed. Before we set up the leaderboard we want to release official evaluation code for the dataset to make sure the leaderboard values are all produced fairly/correctly.
We are still working on full evaluation scripts but in the meantime if you want to do your own GriTS scoring we just open sourced official code for GriTS here: https://github.com/kensho-technologies/grits
Install: pip install grits-metric
I believe we should be able to get the evaluation scripts out and set up a leaderboard very soon. Happy to have your model be the first new entry!