Skip to content

[full_dtensor] Allow sharding config to wrap declared plain inputs #3860

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
Matrix-Z97 wants to merge 1 commit into
base: main
Choose a base branch
from
Open

[full_dtensor] Allow sharding config to wrap declared plain inputs #3860

Changes from all commits
Commits
Show all changes
1 commit
Select commit
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
27 changes: 13 additions & 14 deletions torchtitan/protocols/module.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -590,20 +590,19 @@ def _redistribute_inputs(
if mesh is None:
continue

if (
not isinstance(value, DTensor)
and parallel_dims.spmd_backend == "full_dtensor"
):
raise ValueError("Got a plain Tensor under the full_dtensor mode.")

if not isinstance(value, DTensor) and src_spmd_layout is not None:
layout = resolve_placements(src_spmd_layout, mesh)
value = DTensor.from_local(
value,
mesh,
layout,
run_check=False,
)
if not isinstance(value, DTensor):
if src_spmd_layout is not None:
layout = resolve_placements(src_spmd_layout, mesh)
value = DTensor.from_local(
value,
mesh,
layout,
run_check=False,
)
elif parallel_dims.spmd_backend == "full_dtensor":
raise ValueError(
"Got a plain Tensor under the full_dtensor mode."
)

if isinstance(value, DTensor) and src_spmd_layout is not None:
expected = resolve_placements(src_spmd_layout, mesh)
Expand Down
Loading