Argument definitions for custom metric

[Benjamin Lee (bl3e967)]

I'm struggling to find an answer to the following:

As per the custom metrics docs here, a custom metric function signature should look like the following:

def custom_metric(
    X_val,
    y_val,
    estimator,
    labels,
    X_train,
    y_train,
    weight_val=None,
    weight_train=None,
    config=None,
    groups_val=None,
    groups_train=None,
):
    return metric_to_minimize, metrics_to_log

Args X_val, y_val, estimator, X_train, y_train are self-explanatory. But it's not stark clear what the other arguments are, and I can't seem to find documentation / clues anywhere in the docs or the codebase.

Can I please get some answers to this? And could we update the documentation to reflect this information?