Abstract | ||
---|---|---|
Similar to convolution neural networks, recurrent neural networks (RNNs) typically suffer from over-parameterization. Quantizing bit-widths of weights and activations results in runtime efficiency on hardware, yet it often comes at the cost of reduced accuracy. This paper proposes a quantization approach that increases model size with bit-width reduction. This approach will allow networks to perform at their baseline accuracy while still maintaining the benefits of reduced precision and overall model size reduction. |
Year | Venue | Field |
---|---|---|
2017 | arXiv: Learning | Convolution,Recurrent neural network,Size reduction,Artificial intelligence,Quantization (signal processing),Artificial neural network,Machine learning,Mathematics |
DocType | Volume | Citations |
Journal | abs/1710.07706 | 1 |
PageRank | References | Authors |
0.36 | 5 | 3 |
Name | Order | Citations | PageRank |
---|---|---|---|
Supriya Kapur | 1 | 1 | 0.36 |
Asit K. Mishra | 2 | 1216 | 46.21 |
Debbie Marr | 3 | 175 | 12.39 |