2.0.0

Features:

• KD trainer and resnet50 recipe (81.92 % accuracy)
• Repeated augmentation sampler
• Cooldown epochs
• Beit architecture
• Lamb Optimizer
• Passing torch data loaders directly to SgModel.

Refactoring:

• Checkpoint and architecture params decoupled.

1.7.5

• STDC architectures binary segmentation support bug fix.

1.7.4

torchmetrics version requirement set to 0.7.3 to avoid metric names depreciation bug.

1.7.3

This GitHub Release was done automatically by CircleCI

1.7.2

This GitHub Release was done automatically by CircleCI

1.7.1

What's new ?

• BCE with Dice loss.
• Binary IOU metric object (I.e IOU only for target class).
• Binary segmentation visualisation callback.
• Supervisely dataset interface.
• Different lr assignment for head and backbone for RegSeg.
• Google Colab notebook for semantic segmentation quick start - Check it out in our GitHub repo README.md
• Google Colab notebook for semantic segmentation transfer learning - Check it out in our GitHub repo README.md

1.7.0

CircleCi - Changed version to 1.7.0rc3128

1.6.0

• Added RegSeg model, recipe, and pre-trained checkpoints.
• Updated EfficientNet recipe.
• Updated Resnet50 recipe + pre-trained checkpoint (Top-1=79.47)

1.5.2

• Detection visualisation callback, wrong color ordering for images in tensor board fix.

1.5.1

• Minor fixes for transfer learning example notebook support.