Osculari

Latest version: v0.0.4

Safety actively analyzes 681790 Python packages for vulnerabilities to keep your Python projects secure.

Scan your dependencies

0.0.4

1. [New feature] Activation loader for all supported networks/layers.
2. [Bug fixes] Fixed the models whose outputs are not a tensor and are dictionary or named tuple (e.g., GoogLeNet, Inception-V3 and segmentation models).
3. [Bug fixe] Fixed minor bugs detected by unit tests.

0.0.3

1. Added odd-one-out paradigm (generic for any number of images).
* Merging strategy concatenation or difference.
2. Added MaxVits and SwinTransformers
3. FeatureExtractor class.
4. Fixed the bug with mobilenet family.

0.0.2

1. Extracting features from the following pretrained models:
* All classification and segmentation networks from [PyTorch's official website](https://pytorch.org/vision/stable/models.html).
* All OpenAI [CLIP](https://github.com/openai/CLIP) language-vision models.
* All [Taskonomy](http://taskonomy.stanford.edu/) networks.
2. Reading out the network's feature at any depth (layer).
3. Supporting both convolution and transformer architectures.
4. Adding a linear classifier on top of the frozen pretrained networks.
5. Generic shape appearance dataset to train linear classifiers.
6. Psychophysical paradigms:
* Two-alternative-forced-choice (2AFC).
* Staircase procedure.

Links

Releases

Has known vulnerabilities

© 2024 Safety CLI Cybersecurity Inc. All Rights Reserved.