Commits

Commits on Mar 6, 2025

Commits on Feb 27, 2025

Validate inputs to _nested_view_from_buffer to prevent overflows (#147356)

mikaylagawarecki
authored and
pytorchmergebot
committed
Make Tensor.set_ validate storage_offset when sizes/strides are unchanged (#147354)

mikaylagawarecki
authored and
pytorchmergebot
committed
Fix overflow in checkInBoundsForStorage (#147352)

mikaylagawarecki
authored and
pytorchmergebot
committed

Commits on Feb 25, 2025

Add sparse tensors constructed via legacy constructor to _sparse_tensors_to_validate (#147759)

mikaylagawarecki
authored and
pytorchmergebot
committed

Commits on Feb 14, 2025

Make torch.cuda.gds APIs public (#147120)

mikaylagawarecki
authored and
pytorchmergebot
committed

Commits on Feb 11, 2025

Set USE_CUFILE=1 by default and add pypi package to binary build matrix (#145748)

mikaylagawarecki
authored and
pytorchmergebot
committed

Commits on Jan 31, 2025

Commits on Jan 27, 2025

Commits on Jan 17, 2025

Prevent legacy_load when weights_only=True (correctly) (#145020)

mikaylagawarecki
authored and
pytorchmergebot
committed

Commits on Jan 16, 2025

Prevent _legacy_load with weights_only=True (#144914)

mikaylagawarecki
authored and
pytorchmergebot
committed

Commits on Jan 9, 2025

Link to transformer tutorial in transformer docs (#144425)

mikaylagawarecki
authored and
pytorchmergebot
committed

Commits on Dec 20, 2024

Commits on Dec 18, 2024

Add warning to torch.jit.load (#143403)

mikaylagawarecki
authored and
pytorchmergebot
committed
Prevent torch.jit.load path in torch.load when weights_only=True (#143326)

mikaylagawarecki
authored and
pytorchmergebot
committed

Commits on Dec 6, 2024

Allow user to manually pass module.name associated with global in {add}_safe_global (#142153)

mikaylagawarecki
authored and
pytorchmergebot
committed

Commits on Dec 3, 2024

Mention version of flip in weights_only error message (#141304)

mikaylagawarecki
authored and
pytorchmergebot
committed

Commits on Nov 22, 2024

Fix tests in test/test_serialization that were failing if run individually (#141300)

mikaylagawarecki
authored and
pytorchmergebot
committed

Commits on Nov 19, 2024

Add small test case for #140230 (#140850)

mikaylagawarecki
authored and
pytorchmergebot
committed
Fix condition for weights_only unpickler for DTensor (#140740)

mikaylagawarecki
authored and
pytorchmergebot
committed
Allow NJT by default for weights_only torch.load (take 2) (#140739)

mikaylagawarecki
authored and
pytorchmergebot
committed

Commits on Nov 15, 2024

Fix get_unsafe_globals_in_checkpoint to account for user allowed globals per docstring (#140738)

mikaylagawarecki
authored and
pytorchmergebot
committed

Commits on Nov 12, 2024

Allow NJT by default for weights_only torch.load (#140304)

mikaylagawarecki
authored and
pytorchmergebot
committed
Allow NJT by default for weights_only torch.load (#140304)

mikaylagawarecki
authored and
pytorchmergebot
committed

Commits on Nov 7, 2024

Add APIs to separate norm calculation and gradient scaling in `nn.utils.clip_grad_norm_` (#139662)

mikaylagawarecki
authored and
pytorchmergebot
committed

Commits on Nov 5, 2024

Forward fix D65441551 for T206731737 (#139767)

mikaylagawarecki
authored and
pytorchmergebot
committed

Commits on Nov 4, 2024

[BE] Change _marked_safe_globals_list to set (#139303)

mikaylagawarecki
authored and
pytorchmergebot
committed