This folder contains the source code for our paper --- DRESS: Disentangled Representation-based Self-Supervised Meta-Learning for Diverse Tasks.

The code files within the repository are organized as follows:
	main.py: the main entrance point of the program.
	partition_generators.py: implementation of generating supervised and self-supervised partitions on each dataset.
	task_generator.py: implementation of generating few-shot learning tasks from any given partition.
	utils.py: implementation of helper functions.

The sub-folders within the repository are as follows:
	scripts/: the folder including the scripts to train, evaluate, and obtain visualizations.
	encoders/: the folder containing classes of encoders for obtaining the latent spaces.
	dataset_loaders/: the folder containing scripts for loading each of the dataset for experiments.
	baselines/: the folder containing implementations of baseline methods.
	analyze_results/: the folder containing scripts for post-processing results.
	visualization_results/: the folder containing visualizations on constructed tasks via DRESS.
