Quantization is one of the most popular and widely used methods of speeding up a neural network. At the moment, the standard is 8-bit uniform quantization. Nevertheless, the use of uniform low-bit quantization (4- and 6-bit quantization) has significant advantages in speed and resource requirements for inference. We present our quantization algorithm that offers advantages when using uniform low-bit quantization. It is faster than quantization-aware training from scratch and more accurate than methods aimed only at selecting thresholds and reducing noise from quantization. We also investigated quantization noise in neural networks for low-bit quantization and concluded that quantization noise is not always a good metric for quantization quality.
|Журнал||Journal of Physics: Conference Series|
|Состояние||Опубликовано - 20 дек 2021|
|Событие||8th International Young Scientists Conference on Information Technologies, Telecommunications and Control Systems, ITTCS 2021 - Innopolis, Российская Федерация|
Продолжительность: 16 дек 2021 → 17 дек 2021
Предметные области OECD FOS+WOS
- 1.03 ФИЗИЧЕСКИЕ НАУКИ И АСТРОНОМИЯ