Skip to content

training is very verbose

Usually, training an EM model using bob.learn.em.train should not emit info messages, or only few info messages. However, when training, for example, a KMeans, I get tons of info messages. Running several such trainings in parallel simply kills my console.

The reason is here: https://gitlab.idiap.ch/bob/bob.learn.em/blob/master/bob/learn/em/train.py#L61, where info messages are emitted, where debug level messages should be emitted, if any.