Horovod Mxnet how to create DistributedTrainer object

[uday-rao-aera]

trainer = Trainer(epochs=epochs,
num_batches_per_epoch=num_batches_per_epoch,
hybridize=False, # set to True if you want to hybridize the network for faster execution
ctx=context,
callbacks=callbacks
)

estimator = MQCNNEstimator(
        freq=freq,
        prediction_length=prediction_length,
        cardinality=cardinality,
        use_feat_static_cat=use_feat_static_cat,
        use_feat_dynamic_real=use_feat_dynamic_real,
        context_length=context_length,
        quantiles=quantiles,
        batch_size=batch_size,
        trainer = trainer

        )

How to complete for distributed training.
DistributedTrainer does not take estimator and trainer parameter