Aisdc

Latest version: v1.2.0

Safety actively analyzes 690739 Python packages for vulnerabilities to keep your Python projects secure.

Scan your dependencies

Page 2 of 2

1.0.5

Changes:
* Fix a bug calculating the number of data samples in the `Data` class ([105](https://github.com/AI-SDC/AI-SDC/pull/105))
* Add a fail-fast mechanism for the worst case attack that enables the number of attack repetitions to terminate early based on a given metric and comparison operator ([105](https://github.com/AI-SDC/AI-SDC/pull/105))
* Change the logging message when attack repetitions are run to 1-10 instead of 0-9 ([105](https://github.com/AI-SDC/AI-SDC/pull/105))
* Add the ability to specify the number of worst case attack dummy repetitions on the command line ([105](https://github.com/AI-SDC/AI-SDC/pull/105))
* Add LIRA fail-fast mechanism ([118](https://github.com/AI-SDC/AI-SDC/pull/118))
* Add the ability to load LIRA attack parameters from a config file ([118](https://github.com/AI-SDC/AI-SDC/pull/118))
* Add the ability to load worst case attack parameters from a config file ([119](https://github.com/AI-SDC/AI-SDC/pull/119))
* Standardise the MIA attack output ([120](https://github.com/AI-SDC/AI-SDC/pull/120))
* Prohibit the use of white space in report file names ([154](https://github.com/AI-SDC/AI-SDC/pull/154))
* Improve the safemodel request release test ([160](https://github.com/AI-SDC/AI-SDC/pull/160))
* Refactor LIRA attack tests ([151](https://github.com/AI-SDC/AI-SDC/pull/151))
* Fix setting the number of LIRA shadow models from a config file ([165](https://github.com/AI-SDC/AI-SDC/pull/165))
* Fix OS system calls relying on calling "python" ([162](https://github.com/AI-SDC/AI-SDC/pull/162))
* Fix invalid command line argument in worst case attack example ([164](https://github.com/AI-SDC/AI-SDC/pull/164))
* Add current output JSON format documentation ([168](https://github.com/AI-SDC/AI-SDC/pull/168))
* Add current attack config format documentation ([168](https://github.com/AI-SDC/AI-SDC/pull/168))

1.0.4

Changes:
* Fixed SafeRandomForestClassifier "base estimator changed" error ([143](https://github.com/AI-SDC/AI-SDC/pull/143))

1.0.3

Changes:
* Refactored metrics ([111](https://github.com/AI-SDC/AI-SDC/pull/111))
* Fixed a bug making a report when dummy reps is 0 ([113](https://github.com/AI-SDC/AI-SDC/pull/113))
* Fixed safemodel JSON output ([115](https://github.com/AI-SDC/AI-SDC/pull/115))
* Added a module to produce recommendations from attack JSON output ([116](https://github.com/AI-SDC/AI-SDC/pull/116))
* Disabled non-default report logs ([123](https://github.com/AI-SDC/AI-SDC/pull/123))
* Fixed a minor bug in worst case example ([124](https://github.com/AI-SDC/AI-SDC/pull/124))

1.0.2

Changes:
* Added support for Python 3.8, 3.9 and 3.10 and update requirements.
* Fixed documentation links to notebooks and added SafeSVC.
* Added option to include target model error into attacks as a feature.

1.0.1

Changes:
* Increased test coverage.
* Packaged for PyPI.

1.0.0

First version.

Page 2 of 2

© 2024 Safety CLI Cybersecurity Inc. All Rights Reserved.