Containerized Orchestration and Distribution Engine for Code Optimization and Deep-learning Execution (CODE-2)

ODE-2 (Containerized Orchestration and Distribution Engine for Code Optimization and Deep-learning Execution) is an infrastructure to help quickly produce high-fidelity artificial intelligence (A.I.) agents for complex Air Force relevant domains (e.g., air-to-air engagements) using large-scale computational resources. More precisely, CODE-2 is a system that integrates a set of technologies (e.g., automated container building, orchestration, and model optimization algorithms) to leverage a set of distributed resources (i.e., cloud, DoD HPCs and local resources) in order to evaluate hyper-parameter spaces associated with Deep Reinforcement Learning (DRL) agents in complex simulation environments (e.g., AFSIM).