33
/
AIzaSyAYiBZKx7MnpbEhh9jyipgxe19OcubqV5w
August 1, 2025
3251448
288434
1
Public Timelines
Menu
Public Timelines
FAQ
Public Timelines
FAQ
For education
For educational institutions
For teachers
For students
Cabinet
For educational institutions
For teachers
For students
Open cabinet
Create
Close
Create a timeline
Public timelines
Library
FAQ
mar 8, 2019 - ChengGong: $\mu$L2Q: An Ultra-Low Loss Quantization Method for DNN
Description:
本文实现了多位宽的量化方案。相比于二值三值或者定点数量化等方案,本文首次从权值分布角度进行量化,实现了多位宽的高精度量化。实验表明再相同条件下,uL2Q可以实现更低精度下降。
Added to timeline:
Quantization timeline
By
龚成
29 Mar 2020
1
0
744
Date:
mar 8, 2019
Now
~ 6 years and 3 months ago
About & Feedback
Terms
Privacy
Library
FAQ
Support 24/7
Cabinet
Get premium
Donate
The service accepts bank transfer (ACH, Wire) or cards (Visa, MasterCard, etc). Processed by Stripe.
Secured with SSL