NAS-SE: Designing A Highly-Efficient In-Situ Neural Architecture Search Engine for Large-Scale Deployment.
Qiyu WanLening WangJing WangShuaiwen Leon SongXin FuPublished in: MICRO (2023)
Keyphrases
- highly efficient
- neural architecture
- search engine
- feed forward
- neural network
- multilayer perceptron
- low cost
- low complexity
- information retrieval
- web search engines
- basis functions
- activation function
- control system
- processing units
- artificial neural networks
- information retrieval systems
- computational complexity
- message passing
- shared memory
- multi layer perceptron
- real time