Models

We provide the users numerous models implementation, ranging from a very simple logistic regression to a more complex Transformer-based model with exiting strategies. Most of the models are either direct papers implementation or code taken from a repository and rewritten into a more concise way.

If you feel like a model is missing and would be a great addition to the repository, please open an issue!

Lightning Models

The following models are LightningModule wrapping up PyTorch models and defining the corresponding training logic.

Custom Transformers

The following models are PyTorch implementations of some specific Transformer-based models.

Model Layers

The following sub-package contains implementation of multiple layers that can be reused across different model architectures.