test: skip mfsdp_fully_shard cases when world_size < mesh size#4487
Open
wujingyue wants to merge 1 commit intoNVIDIA:mainfrom
Open
test: skip mfsdp_fully_shard cases when world_size < mesh size#4487wujingyue wants to merge 1 commit intoNVIDIA:mainfrom
wujingyue wants to merge 1 commit intoNVIDIA:mainfrom
Conversation
The tests in tests/unit_tests/distributed/megatron_fsdp/test_mfsdp_fully_shard.py construct an 8-rank device mesh unconditionally, so launching the file with fewer GPUs (e.g. 2) errors out with "Mesh should not be bigger than default world size". Compute the required world size from mesh_dim_config and pytest.skip when the launched world size is too small, mirroring the guard already in test_mcore_fully_sharded_data_parallel.py. Co-Authored-By: Claude Opus 4.7 (1M context) <[email protected]>
|
Auto-sync is disabled for draft pull requests in this repository. Workflows must be run manually. Contributors can view more details about this message here. |
Contributor
Author
|
/ok to test |
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
2 participants
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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
tests/unit_tests/distributed/megatron_fsdp/test_mfsdp_fully_shard.pybuilds an 8-rankinit_device_meshunconditionally, so launching the file with fewer GPUs (e.g. 2) errors out withMesh should not be bigger than default world size N, but found 8 ranks!instead of skipping.build_distributed_environmentthat computes the required world size frommesh_dim_configand callspytest.skip(...)when the launched world size is too small. This mirrors the existingworld_size != dp_sizeskip in the siblingtest_mcore_fully_sharded_data_parallel.py.Test plan
torch.distributed.run --nproc_per_node 2 ... pytest test_mfsdp_fully_shard.py— previously failed with mesh-size errors, now reports6 passed, 314 skipped, 6 xfailed, 0 failed.🤖 Generated with Claude Code