Wan Xinyi
Update description
29c9647
|
raw
history blame
783 Bytes

Optimizer Post Validation

In most practices of PP there's an all-reduce cross all pipeline stages for numerical robustness, e.g. global gradient norm for gradient clipping. INF/NAN check for mixed precision training, etc. This all-reduce breaks parallelogram and makes zero bubble impossible. Under the observation that during a stable training both the gradient clipping and INF/NAN rarely triggers, we replace the before-hand synchronizations with a post update validation.

image

We eagerly step the optimizers assuming the grad cliping, INF/NAN conditions are not triggered. In case an amendment to the gradient is required, a rollback will be issued and then we redo the optimizer step based on the fully reduced global state.