[AMD] Enable Qwen3 FSDP hybrid-shard CI on ROCm - #2265
Merged
Conversation
Xinyu-Kang
requested review from
Rockdu,
Zhichenzzz,
fzyzcjy,
guapisolo and
yueming-yuan
as code owners
August 7, 2026 21:44
Merged
Reapply the DeviceMesh capability fallback after the FSDP backend moved out of experimental and dropped context parallelism. Keep the MI300X r2s2 case enabled on the current-main branch.
guapisolo
force-pushed
the
ci/amd-fsdp-hybrid-shard
branch
from
August 12, 2026 00:56
e8020ed to
5eec662
Compare
guapisolo
approved these changes
Aug 12, 2026
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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
Enable the Qwen3 FSDP hybrid-shard CI test on AMD GPUs.
ROCm’s PyTorch version does not provide the private
DeviceMesh._unflattenmethod used to construct the hybrid-shard mesh. This PR adds a capability-based fallback usinginit_device_mesh.Newer PyTorch releases continue using the existing
_unflattenpath, so behavior on NVIDIA and newer environments remains unchanged.Testing
Validated on 4× AMD MI355 GPUs with PyTorch 2.9.1+ROCm:
tests/e2e/fsdp/test_qwen3_4B_fsdp_hybrid_shard_r2s2.pytests/fast-gpu/test_fsdp_hybrid_shard.pyAll tests passed.