learned_optimization reference documentation

learned_optimization is a research codebase for training learned optimizers. It implements hand designed and learned optimizers, tasks to meta-train and meta-test them on, and outer-training algorithms such as ES and PES.

Getting Started

• Part 1: Introduction
  • Prerequisites
  • Tasks
  • Optimizers
    • Defining a custom Optimizer
  • Learned Optimizers
• Part 2: Custom Tasks, Task Families, and Performance Improvements
  • Prerequisites
  • Defining a custom Dataset
  • Defining a custom Task
  • Meta-training on multiple tasks: TaskFamily
  • Limitations of TaskFamily

Modules

• Tasks
  • QuadraticTasks
• Optimizers
  • API
• Learned Optimizers
  • API
• Outer Trainers
  • Base
  • FullES
  • TruncatedPES
• Continuous Eval
  • Evaluation Chief
  • Evaluation configuration
  • Evaluation Worker
• Population: Online hparam adjustment for multiple workers
  • Quick Start Example
  • How it works
  • Available Mutators
    • Winner take all genetic algorithms
    • Fixed schedule
    • Single worker Explore
  • Examples
    • Synthetic
    • Simple_CNN
    • Complex_CNN

Indices and tables

• Index

• Module Index

• Search Page