Create a timeline

Public timelines

2 gen 2018 anni - Deep Neural Network Compression with Single and Multiple Level Quantization

Descrizione:

本文首次同时在宽度和深度上进行量化，SLQ/MLQ。首先在宽度上，使用SLQ，将权值划分为量化部分和重训练部分，通过重训练部分降低量化损失。在深度上，使用MLQ，通过将损失累计到下一层迭代，以降低当前层的量化损失。

Aggiunto al nastro di tempo:

Quantization timeline

29 mar 2020

1

0

744

Data:

2 gen 2018 anni

Adesso

~ 7 years and 5 months ago

About & Feedback Accordo Privatezza Biblioteca FAQ

Cabinet Get premium

Donate

The service accepts bank transfer (ACH, Wire) or cards (Visa, MasterCard, etc). Processed by Stripe.

Secured with SSL