Labelmodel takes up a huge amount of memory when training, and it also wouldn't release the memory

[topwasu]

During the training, labelmodel takes up a lot of memory causing an out-of-memory- error, and reducing the batch size doesn't help. Apart from that, labelmodel seems to still take up memory even at the start of a new checkpoint. For now, in the master branch, the labelmodel is replaced with majority vote. In memory-debug branch, it is still using labelmodel, and it contains different memory profilers to help debug this bug, so debugging this error should probably happen in this branch.

[stephenbach]

Have we tried, instead of setting the model variable to none, actually calling delete? i.e.,

del model

[topwasu]

Yes, I've tried that. Strangely, it didn't work.

[stephenbach]

Perhaps this is related: BatsResearch/labelmodels#4

[topwasu]

That could be the cause of OOM error!