33
/ru/
AIzaSyB4mHJ5NPEv-XzF7P6NDYXjlkCWaeKw5bc
November 1, 2025
3909867
288434
1
Public Timelines
FAQ Получить премиум

24 март 2018 г. - QAT: Quantization and training of neural networks for efficient integer-arithmetic-only inference, CVPR

Описание:

QAT employs the affine mapping of integers to real values with two constant parameters: Scale and Zeropoint. It first subtracts the Zero-point parameter from data (weights/activation), then divides the data by scale parameter, and finally getting the quantized results with rounding operation and affine mapping.

Добавлено на ленту времени:

29 мар 2020
1
0
809

Дата:

24 март 2018 г.
Сейчас
~ 7 гг и 7 мес назад