optimizers · facebook
Adam variant that partitions parameters into blocks sharing a single learning rate. ~45% less optimizer memory.
| Weight file SHA-256 | 6215d6b5c97ffad8a2055ff4d04212f4077114a046a08a550ca5bbd3af5c5f4d |
|---|
| Publisher | |
|---|---|
| Origin | United States |
| Licence | MIT |
| Pulls recorded here | 8,602 |
Registry record as of 2026-06-06. Identity, licence and parameter count are properties of a release and do not move; tier can, and the verify link above re-checks it against the live registry rather than this page.