Skip to content

Implement sharding and device mesh debug tool #2328

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
fegin wants to merge 2 commits into
base: gh/fegin/78/base
Choose a base branch
from
+1,255 −2
Open

Implement sharding and device mesh debug tool #2328

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
bf326d2
Update
fegin Feb 5, 2026
efefd75
Update
fegin Feb 6, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
14 changes: 14 additions & 0 deletions torchtitan/config/job_config.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -987,6 +987,20 @@ class Debug:
moe_force_load_balance: bool = False
"""If True, we force each experts to get the same amount of tokens via round-robin. This option is for debugging usage only."""

log_sharding_info: bool = False
"""
If True, logs DTensor sharding/mesh info for module inputs, params,
outputs during one fwd/bwd pass. Only the first step will recorded and
this flag should only be used for debugging purpose and should not
be enabled during the real training.
"""

collapse_identical_layers: bool = False
"""
If True, and log_sharding_info is True, collapse repeated layer modules with
with identical sharding patterns in the sharding log.
"""


@dataclass
class JobConfig:
Expand Down
Loading
Loading