gt4sd.training_pipelines.torchdrug.unpatch module

Summary

Functions:

check_openmp_availabilty	Check if OpenMP is available at runtime.
fix_datasets	Helper function to revert TorchDrug dataset handling (which breaks core pytorch functionalities).
fix_schedulers	Helper function to revert TorchDrug LR scheduler handling (which breaks core pytorch functionalities).

Reference

fix_datasets(sane_datasets)

Helper function to revert TorchDrug dataset handling (which breaks core pytorch functionalities). For details see: https://github.com/DeepGraphLearning/torchdrug/issues/96

Parameters

sane_datasets – A list of pytorch datasets.

Raises

AttributeError – If a passed dataset was not sane.

fix_schedulers(sane_schedulers)

Helper function to revert TorchDrug LR scheduler handling (which breaks core pytorch functionalities). For details see: https://github.com/DeepGraphLearning/torchdrug/issues/96

Parameters

sane_schedulers – A list of pytorch lr_schedulers.

Raises

AttributeError – If a passed lr_scheduler was not sane.

check_openmp_availabilty()

Check if OpenMP is available at runtime.

Return type

bool

Returns

True if OpenMP is available, False otherwise.