Skip to content

Manual Resharding given Manifold Paths #3394

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

Manual Resharding given Manifold Paths #3394

Show file tree
Hide file tree
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
169 changes: 0 additions & 169 deletions torchrec/distributed/benchmark/benchmark_resharding_handler.py
View file
Open in desktop

This file was deleted.

11 changes: 6 additions & 5 deletions torchrec/distributed/model_parallel.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -258,11 +258,12 @@ def __init__(
device = torch.device("cpu")
self.device: torch.device = device

if sharders is None:
sharders = get_default_sharders()
self.sharders: List[ModuleSharder[nn.modules.module.Module]] = (
get_default_sharders() if sharders is None else sharders
)

self._sharder_map: Dict[Type[nn.Module], ModuleSharder[nn.Module]] = {
sharder.module_type: sharder for sharder in sharders
sharder.module_type: sharder for sharder in self.sharders
}

if data_parallel_wrapper is None:
Expand All @@ -279,9 +280,9 @@ def __init__(
)
pg = self._env.process_group
if pg is not None:
plan = planner.collective_plan(module, sharders, pg)
plan = planner.collective_plan(module, self.sharders, pg)
else:
plan = planner.plan(module, sharders)
plan = planner.plan(module, self.sharders)
self._plan: ShardingPlan = plan
self._dmp_wrapped_module: nn.Module = self._init_dmp(module)
self._optim: CombinedOptimizer = self._init_optim(self._dmp_wrapped_module)
Expand Down
Loading