Releases: BloodAxe/pytorch-toolbelt
Releases · BloodAxe/pytorch-toolbelt
0.1.3
PyTorch Toolbelt 0.1.1
New functionality / breaking changes
- Added visualization functions to render best/worst batches for binary and semantic segmentation.
- JaccardScoreCallback now is a single callback for computing IoU for binary/multiclass/multilabel segmentation.
- Added HFF module (Hierarchical feature fusion).
- Added
set_trainable
function to enable/disabled training and batch-norm on module and it's childs. - RLE encoding/decoding (Hi, Kaggle)
API changes
rgb_image_from_tensor
now acceptsdtype
parameters for returned image
Bugfixes
- Fixed wrong implementation of UpsampleAddConv (There was extra residual connection)
PyTorch Toolbel 0.1.0
New stuff:
- EfficientNet
- Multiscale TTA module
- New activations: Swish, HardSwish, HardSigmoid
- AGN module (Activated Group Norm), mimicks ABN
Changes:
SpatialGate2d
now acceptssqueeze_channels
for explicit number of squeeze channels.
Misc
- Code formatting
Pytorch Toolbelt 0.0.9
- Refactoring of activation functions factory method (for upcoming model builder)
- Cosmetic changes in logging
PyTorch Toolbelt 0.0.8
- Global pooling, SCSE module and MobileNetV3 encoders are not ONNX and CoreML friendly.
- Refactored FPN module for more flexible
interpolate_add
tuning (can use any module with two inputs)
Pytorch Toolbelt 0.0.7
Added MobileNetV3 encoder (implementation credits to https://github.com/Randl/MobileNetV3-pytorch)
PyTorch Toolbelt 0.0.6
New features
- Added WiderResNet & WiderResNetA2 encoders (https://github.com/mapillary/inplace_abn)
- Added implementation of reduced focal loss (https://arxiv.org/abs/1903.01347)
PyTorch Toolbelt 0.0.5
Changes
- Added 10-Crop TTA (#4)
- Added unit tests for TTA functions
- Added
freeze_bn
function to freeze all BN layers in a model - Rename
unpad_tensor
tounpad_image_tensor
to mimickpad_image_tensor
Bugfixes
- Fixed bug in
d4_image2mask
PyTorch Toolbelt 0.0.4
API Changes
- Refactored TTA interface