Fix Mesh setup for multiprocess CPUs. #723
Merged
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Outdated suggestions cannot be applied.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.
The following condition will evaluate to true due to base.yaml defaults, unless overridden from command line.
get_num_slices(raw_keys):
has some new logic that computes number of slices and number of devices in a slice. Adapting this logic to suit CPUs ( slices have no meaning for CPUs, because they do not support hierarchical network. Hence setting num_slices to 1 and allowing existing ICI parallelism logic in max_utils).Testing on multiprocess CPUs -
I tested standalone_checkpointer.py end-to-end (on 2 nodepools, with 2 hosts each) to verify this change.