AutoPhase: Compiler Phase-Ordering for High Level Synthesis with Deep Reinforcement Learning.
Ameer Haj AliQijing HuangWilliam S. MosesJohn XiangIon StoicaKrste AsanovicJohn WawrzynekPublished in: CoRR (2019)
Keyphrases
- high level synthesis
- reinforcement learning
- function approximation
- parallel architecture
- state space
- programming language
- reinforcement learning algorithms
- partial order
- design space exploration
- machine learning
- multi agent
- search algorithm
- pairwise
- model free
- distributed memory machines
- temporal difference
- markov decision processes
- learning algorithm