Dataset about IMDB

[Theheavens]

Dataset IMDB has multi labels. But I am wondering how to save the file for evaluation, cause I don't see it in run_multi.py. And else, I want to know what API in sklearn to evaluate multi-label.

[Theheavens]

info.dat in IMDB, #L8160 "meaning" maybe is 'movie->actor'.
info.dat in ACM, the 'Attribute Dimension' of author is not zero, right?

[1049451037]

You can see here for reference to generate prediction file for IMDB: https://github.com/THUDM/HGB/blob/master/NC/benchmark/methods/baseline/run_multi.py#L151

[1049451037]

Yes, #L8160 is 'movie->actor'. It is a typo. The info.dat files contain some information that we haven't used so far. To load data and check the attribute dimension, you can use our data_loader. Detailed document is on the way... You can see the code in baseline folder as example to use. Sorry for the inconvenience.

[Theheavens]

Can we get more chances to submit results everyday?

[1049451037]

We recommend to tune performance on validation set, which can save your submission budget.