Fix PPO input sharding to match pmap axis - #674
Open
lmj0209 wants to merge 1 commit into
Open
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
Pass PPO's
_PMAP_AXIS_NAMEtobcast_local_deviceswhenplacing the initial training state.
The new
jax.pmapimplementation returns arrays with aNamedShardingwhose mesh axis matches the pmap axis name(
iin PPO). The current initial state uses_device_put_shardedinstead. The shapes and devices arethe same, but the sharding objects differ, so the first pmap
result changes the input signature seen by the next host
call.
This change:
bcast_local_devicesaccept an axis name whileretaining its existing default;
iaxis used by thetraining pmap; and
Validation
brax.training.pmap_test: passedsharding both use axis
iThe default
_device_put_shardedvalue remains unchangedfor other callers.