CODEWITHSUNDEEP
pythonmxnet
Seth Kitchen
Seth Kitchen

Reputation: 1566

How to silence training info in GluonTS / Apache mxnet?

When I train my DeepAREstimator in GluonTS

estimator = DeepAREstimator(freq="D", 
                              prediction_length=56, 
                              trainer=Trainer(epochs=20))

predictor = estimator.train(training_data=training_data)

I get a bunch of debug messages:

INFO:root:Epoch[12] Learning rate is 0.001
100%|██████████| 50/50 [00:09<00:00,  5.31it/s, avg_epoch_loss=2.29]
INFO:root:Epoch[12] Elapsed time 9.422 seconds
INFO:root:Epoch[12] Evaluation metric 'epoch_loss'=2.293711
INFO:root:Epoch[13] Learning rate is 0.001

How can I silence these? It might have to do with the underlaying mxnet so I tag that as well. Can someone add the gluonts tag?

Upvotes: 2

Views: 571

Answers (1)

oetzi
oetzi

Reputation: 1052

import logging

logging.getLogger("mxnet").addFilter(lamda record: False)

Upvotes: 1

Related Questions