Cargando…
Towards Optimal Compression: Joint Pruning and Quantization
Model compression is instrumental in optimizing deep neural network inference on resource-constrained hardware. The prevailing methods for network compression, namely quantization and pruning, have been shown to enhance efficiency at the cost of performance. Determining the most effective quantizati...
| Autores principales: | , , , , , |
|---|---|
| Lenguaje: | eng |
| Publicado: |
2023
|
| Materias: | |
| Acceso en línea: | http://cds.cern.ch/record/2856527 |