Gradient-based Automatic Per-Weight Mixed Precision Quantization for Neural Networks On-Chip.
Chang SunThea Klæboe ÅrrestadVladimir LoncarJennifer NgadiubaMaria SpiropuluPublished in: CoRR (2024)
Keyphrases
- neural network
- pattern recognition
- high speed
- fully automatic
- low cost
- semi automatic
- high precision
- back propagation
- fuzzy logic
- genetic algorithm
- neural nets
- multi layer
- sufficient conditions
- precision and recall
- fuzzy systems
- quantization error
- circuit design
- fault diagnosis
- real time
- multiresolution
- high density
- hopfield neural network
- physical design