[MXNET-1420] Implement RAdam

• Type: New Feature
• Status: To Do
• Priority: Major
• Resolution: Unresolved
• Component/s: Apache MXNet Backend, Gluon
• Assignee: Unassigned
• Reporter: Dheeraj M
• Created: 19/Aug/19 15:32
• Updated: 22/Aug/19 14:19

Description

Implementation of RAdam optimizer based on the new paper [ON THE VARIANCE OF THE ADAPTIVE LEARNING RATE AND BEYOND|https://arxiv.org/pdf/1908.03265v1.pdf]