LNS-Madam: Low-Precision Training in Logarithmic Number System Using Multiplicative Weight Update.
| Autor: | Zhao, Jiawei1 (AUTHOR) jiawei@caltech.edu, Dai, Steve2 (AUTHOR) sdai@nvidia.com, Venkatesan, Rangharajan2 (AUTHOR) rangharajanv@nvidia.com, Zimmer, Brian2 (AUTHOR) bzimmer@nvidia.com, Ali, Mustafa3 (AUTHOR) ali102@purdue.edu, Liu, Ming-Yu2 (AUTHOR) mingyul@nvidia.com, Khailany, Brucek2 (AUTHOR) bkhailany@nvidia.com, Dally, William J.2 (AUTHOR) bdally@nvidia.com, Anandkumar, Anima1 (AUTHOR) anima@caltech.edu |
|---|---|
| Zdroj: | IEEE Transactions on Computers. Dec2022, Vol. 71 Issue 12, p3179-3190. 12p. |
| Databáze: | Academic Search Ultimate |
| Externí odkaz: |
|