Pretrain, finetune ANY AI model of ANY size on 1 or 10,000+ GPUs with zero code changes.
Support fused Adam with mixed precision (#15555)
Co-authored-by: Carlos Mocholí <[email protected]>
A
Adrian Wälchli committed
e4611ef98eba8702420a4b585fa94a3bb0ca463d
Parent: bf4653e
Committed by GitHub <[email protected]>
on 11/8/2022, 2:53:30 PM