A research-grade codebase to investigate around 1_000 hyperparameters' effect on continual-learning tasks or even offline-learning problems, which consists of diffusion models, autoencoder networks, UNets, Vision Transformers, cognitive-inspired methods, etc.
tensorflow keras hyperparameter-optimization cognitive-science representation-learning consolidation modulation variational-autoencoder curriculum-learning distillation continual-learning progressive-learning conditional-generation contrastive-learning copmuter-vision generative-replay vision-transformers class-incremental-learning diffusion-transformer joint-diffusion-classifier
-
Updated
Oct 15, 2026 - Jupyter Notebook