Associative Memory Based Experience Replay for Deep Reinforcement Learning
Fuente:
arXiv
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Preprint |
| Published: |
2022
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866911788845498368 |
|---|---|
| author | Li, Mengyuan Kazemi, Arman Laguna, Ann Franchesca Hu, X. Sharon |
| author_facet | Li, Mengyuan Kazemi, Arman Laguna, Ann Franchesca Hu, X. Sharon |
| contents | Experience replay is an essential component in deep reinforcement learning (DRL), which stores the experiences and generates experiences for the agent to learn in real time. Recently, prioritized experience replay (PER) has been proven to be powerful and widely deployed in DRL agents. However, implementing PER on traditional CPU or GPU architectures incurs significant latency overhead due to its frequent and irregular memory accesses. This paper proposes a hardware-software co-design approach to design an associative memory (AM) based PER, AMPER, with an AM-friendly priority sampling operation. AMPER replaces the widely-used time-costly tree-traversal-based priority sampling in PER while preserving the learning performance. Further, we design an in-memory computing hardware architecture based on AM to support AMPER by leveraging parallel in-memory search operations. AMPER shows comparable learning performance while achieving 55x to 270x latency improvement when running on the proposed hardware compared to the state-of-the-art PER running on GPU. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2207_07791 |
| institution | arXiv |
| publishDate | 2022 |
| record_format | arxiv |
| spellingShingle | Associative Memory Based Experience Replay for Deep Reinforcement Learning Li, Mengyuan Kazemi, Arman Laguna, Ann Franchesca Hu, X. Sharon Hardware Architecture Emerging Technologies Machine Learning Experience replay is an essential component in deep reinforcement learning (DRL), which stores the experiences and generates experiences for the agent to learn in real time. Recently, prioritized experience replay (PER) has been proven to be powerful and widely deployed in DRL agents. However, implementing PER on traditional CPU or GPU architectures incurs significant latency overhead due to its frequent and irregular memory accesses. This paper proposes a hardware-software co-design approach to design an associative memory (AM) based PER, AMPER, with an AM-friendly priority sampling operation. AMPER replaces the widely-used time-costly tree-traversal-based priority sampling in PER while preserving the learning performance. Further, we design an in-memory computing hardware architecture based on AM to support AMPER by leveraging parallel in-memory search operations. AMPER shows comparable learning performance while achieving 55x to 270x latency improvement when running on the proposed hardware compared to the state-of-the-art PER running on GPU. |
| title | Associative Memory Based Experience Replay for Deep Reinforcement Learning |
| topic | Hardware Architecture Emerging Technologies Machine Learning |
| url | https://arxiv.org/abs/2207.07791 |