How to represent more than one item_features? The parameter "item_features" in lightFM.fit looks like to an embedding of only singe item feature. How shoud I handle more than one features? Should I handle this in Dataset().build_item_features?

Thanks for opening an issue!

Please include as much detail as possible: what does your dataset look like, what hyperparameters you are using (and have you tried other ones?).

When including code snippets, make sure you use appropriate code formatting using backtics. Have a look at the markdown reference for details: https://guides.github.com/features/mastering-markdown/.

Start by reading the behavior of lightFM.fit and Dataset().build_item_features, then check the repository's existing documentation or tests for item feature examples. The issue does not provide a dataset, hyperparameters, or a requested code change; done would require a clear, documented explanation of how multiple item features are represented and handled.

Written by the indexing model from the issue text.