Pretrain, finetune ANY AI model of ANY size on 1 or 10,000+ GPUs with zero code changes.
Activation checkpointing in FSDP without boilerplate (#15826)
* initial * input type * checkpointing * fsdp in pl * all_close Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
A
Adrian Wälchli committed
05dbf48ad0b6f9eed79cee8ce8a61839c219f330
Parent: 2992002
Committed by GitHub <noreply@github.com>
on 12/6/2022, 3:45:33 PM