[ENH] option to turn off printing in prediction/inference

[dara1400]

I want to use model.predict in a loop. It keeps printing this:

GPU available: True (cuda), used: True TPU available: False, using: 0 TPU cores IPU available: False, using: 0 IPUs HPU available: False, using: 0 HPUs LOCAL_RANK: 0 - CUDA_VISIBLE_DEVICES: [0]

Is there way to stop it from printing?

[IlIlllIIllIIlll]

First, import the logging module:

import logging

Then, add the following line to suppress the message:

logging.getLogger("lightning.pytorch.utilities.rank_zero").setLevel(logging.WARNING)

This will prevent the following output from being printed:

GPU available: True (cuda), used: True
TPU available: False, using: 0 TPU cores
IPU available: False, using: 0 IPUs
HPU available: False, using: 0 HPUs

Next, add this line to suppress another message:

logging.getLogger("lightning.pytorch.accelerators.cuda").setLevel(logging.WARNING)

This will stop the following output from appearing:

LOCAL_RANK: 0 - CUDA_VISIBLE_DEVICES: [0]

Good luck!

[fkiraly]

Related: https://github.com/sktime/sktime/pull/6891 - should we perhaps address this at the source, and add a verbosity option, @XinyuWuu?

[XinyuWuu]

Related: sktime/sktime#6891 - should we perhaps address this at the source, and add a verbosity option, @XinyuWuu?

Good idea.