dgs.models

Definition of modules, loading, and further module utility functions.

Modules

alpha	Modules for predicting the values of the alpha gates given State s as input.
combine	Combine multiple similarity matrices.
dataset	Modules for loading data, including datasets and data loaders.
dgs	Use the DGS model to run multiple similarity modules (SimilarityModule) and combine them using the combine modules (CombineSimilaritiesModule).
embedding_generator	Modules for generating embeddings for Re-Identification ('ReID-Embeddings').
engine	Modules for running, training and testing multiple other modules.
loader	Load and register modules.
loss	Functions to load and manage torch and custom loss functions.
metric	Load, register, and initialize different metric functions.
module	Base model class as lowest building block for dynamic modules
modules
optimizer	Load, register, and initialize different optimizers.
scheduler	Load, register, and initialize different learning rate schedulers.
similarity	Abstraction for modules that compute the similarity between two input tensors.
submission	Modules for saving submission files in different formats.