fiajoiwfj/tensorizer-1

Tensorizer torch_compat weights_only default bypass

This package reproduces a Tensorizer torch_compat issue where PyTorch 2.6+ default safe loading blocks a malicious checkpoint, but the normal tensorizer_loading() compatibility context executes the same checkpoint without the caller passing weights_only=False.

Run from the repository root:

reproduce.py

Expected marker:

MFV_TENSORIZER_WEIGHTS_ONLY_DEFAULT_BYPASS

The marker appears only in the explicit unsafe PyTorch baseline and the default Tensorizer compatibility context. It does not appear for plain torch.load(), plain torch.load(..., weights_only=True), or tensorizer_loading() with explicit weights_only=True.