Adjust learning rate exponentially from a given epoch to the end of training.
(optimizer, base_lr, epoch, num_epochs, decay_start_epoch)
source not stored for this graph (policy: none)
no outgoing calls
no test coverage detected