Fix: Explicitly set weights_only=True
in torch.load
to handle FutureWarning (#949)
#950
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Description
This PR addresses closes #949 by modifying the
default_deserialize_torch_model
function inPyTorchShim
. The change explicitly sets theweights_only=True
flag when callingtorch.load
. This resolves theFutureWarning
regarding potential security issues with implicit pickle module usage and aligns with future PyTorch updates whereweights_only
will default toTrue
.Types of change
Bug Fix
Checklist