Lightning-AI / pytorch-lightning UNCLAIMED

Pretrain, finetune ANY AI model of ANY size on 1 or 10,000+ GPUs with zero code changes.

0 0 104 Python

Fix ModelCheckpoint file_exists OOM in DDP (#21380)

* Fix ModelCheckpoint.file_exists OOM in DDP

* Document ModelCheckpoint.file_exists DDP memory fix

* Update src/lightning/pytorch/callbacks/model_checkpoint.py

---------

Co-authored-by: Justus Schock <12886177+justusschock@users.noreply.github.com>

littlebullGit committed 9mo ago

b09e96edb793507802d9702b0134f6f5ec0f3ba5

Parent: ef489f2

Committed by GitHub <noreply@github.com> on 11/25/2025, 12:22:21 PM