torch.load has an issue when loading models from wandb (PyTorch version 2.6)

[MarshallYan]

Got error loading a model from wandb with PyTorch 2.6:

UnpicklingError: Weights only load failed. This file can still be loaded, to do so you have two options, do those steps only if you trust the source of the checkpoint.
(1) In PyTorch 2.6, we changed the default value of the weights_only argument in torch.load from False to True. Re-running torch.load with weights_only set to False will likely succeed, but it can result in arbitrary code execution. Do it only if you got the file from a trusted source.
(2) Alternatively, to load with weights_only=True please check the recommended steps in the following error message.
WeightsUnpickler error: Unsupported global: GLOBAL modelforge.potential.parameters.SchNetParameters was not an allowed global by default. Please use torch.serialization.add_safe_globals([SchNetParameters]) or the torch.serialization.safe_globals([SchNetParameters]) context manager to allowlist this global if you trust this class/function.

Check the documentation of torch.load to learn more about types accepted by default with weights_only https://pytorch.org/docs/stable/generated/torch.load.html.

[chrisiacovella]

This is part of the new change to enhance security due to vulnerabilities of pickle. I'm not quite sure how to fix this yet (aside from setting weights_only=False, which might be what we need to do at this point), since this functionality is just reading in a checkpoint file that was generated by pytorch lightning.