Release v0.11.1

README.md

@@ -217,6 +217,19 @@ classification. It was successfully applied in the paper
 
 * Add new transforms: `Mix`, `Padding`, `RandomCrop` and `SpliceOut`
 
+## [v0.11.1] - 2024-02-07
+
+### Changed
+
+* Add support for constant cutoff frequency in `LowPassFilter` and `HighPassFilter`
+* Add support for min_f_decay==max_f_decay in `AddColoredNoise`
+* Bump torchaudio dependency from >=0.7.0 to >=0.9.0
+
+### Fixed
+
+* Fix inaccurate type hints in `Shift`
+* Remove `set_backend` to avoid `UserWarning` from torchaudio
+
 ## [v0.11.0] - 2022-06-29
 
 ### Added

scripts/demo.py

@@ -23,9 +23,9 @@
     PitchShift,
     BandStopFilter,
     TimeInversion,
-    Padding,
     Identity,
 )
+from torch_audiomentations.augmentations.padding import Padding
 from torch_audiomentations.augmentations.shuffle_channels import ShuffleChannels
 from torch_audiomentations.augmentations.splice_out import SpliceOut
 from torch_audiomentations.core.transforms_interface import ModeNotSupportedException

setup.py

@@ -40,7 +40,7 @@ def find_version(*file_paths):
         "julius>=0.2.3,<0.3",
         "librosa>=0.6.0",
         "torch>=1.7.0",
-        "torchaudio>=0.7.0",
+        "torchaudio>=0.9.0",
         "torch-pitch-shift>=1.2.2",
     ],
     extras_require={"extras": ["PyYAML"]},

torch_audiomentations/__init__.py

@@ -18,4 +18,4 @@
 from .utils.config import from_dict, from_yaml
 from .utils.convolution import convolve
 
-__version__ = "0.11.0"
+__version__ = "0.11.1"