Why gradient clipping accelerates training: A theoretical justification for adaptivity (Q6319511)
From MaRDI portal
 | This is the item page for this Wikibase entity, intended for internal use and editing purposes. Please use this page instead for the normal view: Why gradient clipping accelerates training: A theoretical justification for adaptivity |
preprint article from arXiv
| Language | Label | Description | Also known as |
|---|---|---|---|
| English | Why gradient clipping accelerates training: A theoretical justification for adaptivity |
preprint article from arXiv |
Statements
28 May 2019
0 references
math.OC
0 references
cs.LG
0 references
Jingzhao Zhang
0 references
Tianxing He
0 references
Suvrit Sra
0 references
Ali Jadbabaie
0 references
