|View source on GitHub|
Aggregates and scales per example loss and regularization losses.
tf_agents.utils.common.aggregate_losses( per_example_loss=None, sample_weight=None, global_batch_size=None, regularization_loss=None )
global_batch_size is given it would be used for scaling, otherwise it
would use the batch_dim of per_example_loss and number of replicas.
|An AggregatedLosses named tuple with scalar losses to optimize.|