Dumping collective metadata to csv for each rank #154
Closed
Conversation
This file contains 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
adding flag description
|
MLCommons CLA bot All contributors have signed the MLCommons CLA ✍️ ✅ |
|
recheck |
Sign up for free
to subscribe to this conversation on GitHub.
Already have an account?
Sign in.
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Outdated suggestions cannot be applied.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
Adding a function that loops for the nodes and for each collective node dump basic data into csv file (data like the node id, the node name that includes the collective name, the collective msg size, the pg_name, root rank if needed, etc.
the dump allows to further processing the trace for example if we want to create collective implementation of those high-level collective commands.
Test Plan
Running it on several et_{rank}.json files, and see the dump created.
Additional Notes
if, for each collective node, one of the following attributes is missing: pg_name, coll_size, root_rank, the dump will indicate that in separate field so it will not fail on that. The user will need to check of it's OK. for example, not every node needs root_rank, so it's OK if it's missing for AllReduce for example, but not for Broadcase.