# Changelog

## Release 0.1.10(Jan 15 2023)

  * Major Features and Improvements
    * show progress bar while a model file is downloading.
    * added examples.

## Release 0.1.9(Dec 5 2022)

  * Major Features and Improvements
    * feed back from AxonInterp project.
    * added demo_yolov4.
    * revised documents.

## Release 0.1.8(Nov 13 2022)

  * Major Features and Improvements
    * added the ability to download model files from the url.
    * added the ability to give hints about I/O tensor of the model.
    * added example of ViT image classification.

## Release 0.1.7(Sep 5 2022)

  * Bug Fixes and Other Changes
    * Fixed compilers config in mix.exs.

## Release 0.1.6(Aug 30 2022)

  * Bug Fixes and Other Changes
    * Fixed a miscalculation in the size of the input tensor.

## Release 0.1.5(Aug 2 2022)

  * Breaking Changes
    * the previous download source of nlohmann-json is gone, so changed to another one.

## Release 0.1.4(Aug 2 2022)

  * Bug Fixes and Other Changes
    * added two more ways to specify box in NMS.
    * include NanoDet plus object detection demo.
    * rename the folder of Candy demo.

## Release 0.1.3(Jul 09 2022)

  * Bug Fixes and Other Changes
    * fixed command message of non_max_suppression_multi_class/7.
    * include livebook demo: Semantic Segmentation for Self Driving Cars by UNet.
    * include YOLOv7 object detection demo.

## Release 0.1.2(May 20 2022)

  * Bug Fixes and Other Changes
    * confirmed it works in Linux Mint 20.

## Release 0.1.1(May 18 2022)

  * Experimental version