learned_optimization
Explore a robust research codebase for designing and evaluating learned optimizers with JAX. The project features tools for meta-training dynamic systems and includes comprehensive tutorials via Colab notebooks. Understand outer-training algorithms such as ES, PES, and truncated backprop through time, and engage with practical examples tailored for deep learning professionals. Access documentation to learn about creating custom tasks, developing gradient estimators, and applying meta-training techniques with Gradient Learner, aimed at advancing research in learned optimization.
