Releases: pytorch-labs/torchfix
Releases · pytorch-labs/torchfix
TorchFix 0.6.0
- Added
torch.utils._pytree._register_pytree_node
andtorch.backends.cuda.sdp_kernel
to the deprecated APIs rules - Enhanced rule TOR203 to support
torchvision.datasets
andtransforms
in addition tomodels
- Added rules TOR106 and TOR107 to suggest replacing
torch.log(1 + x)
andtorch.exp(x) - 1
with more numerically stable equivalents - Multiple code refactorings, bug fixes, and quality of life and documentation improvements
TorchFix 0.5.0
- Added rule TOR203 to replace 'import torchvision.models as models' with 'from torchvision import models'
- Added rules TOR104 and TOR105 for calling and importing non-public PyTorch functions that have known public aliases
- Added rules TOR004 and TOR103 for importing removed and deprecated functions (in addition to the existing rules for calling those functions)
- Fixed loading for deprecated symbols config in zipped deployments
- Done several smaller bug fixes and refactorings
TorchFix 0.4.0
- Improvements for the standalone
torchfix
command:- Added
--version
flag --select
flag now accepts specific rules, not justALL
- Fixed excessive debug output on MacOS
- Added
- Added PyTorch-internal rule TOR901
- TorchFix explicitly requires at least Python 3.9 now
- Small clean-ups and bugfixes
TorchFix 0.3.0
- Added rule TOR003 about explicitly passing
use_reentrant
totorch.utils.checkpoint
- Added
torch.nn.utils.weight_norm
to the list of deprecated functions flagged by TOR101 - Updated README with TOR0 rules description
TorchFix 0.2.1: first release for pytorch-labs/torchfix repo
This is the first release for pytorch-labs/torchfix repo, with the only differences from TorchFix 0.2.0 on PyPI are files related to repo maintenance and project metadata.