Optimizers

Stochastic Gradient Descent

class hebel.optimizers.SGD(model, parameter_updater, train_data, validation_data=None, progress_monitor=None, learning_rate_schedule=<generator object constant_scheduler>, momentum_schedule=None, early_stopping=True)

best_validation_loss

norm_v_norm()

run(iterations=200, validation_interval=5, yaml_config=None, task_id=None)