There appears to be an inconsistency in the documentation regarding GPU usage configuration in AutoGluon. The documentation suggests that setting num_gpus=1 within predictor.fit() should enable GPU ...
It seems like serialization and deserialization associated with python's multiprocessing limit the benefits of processing data in parallel. In the following example, I create a custom iterable that ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results
Feedback