Skip to content
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

Jump to bottom

Fix num_nodes not set for FSDPStrategy #17438

Merged
merged 2 commits into from
Apr 24, 2023
Merged

Fix num_nodes not set for FSDPStrategy #17438

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
3b71dc3
port fix
awaelchli Apr 22, 2023
cff39ec
add changelog
awaelchli Apr 22, 2023
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
1 change: 1 addition & 0 deletions src/lightning/pytorch/CHANGELOG.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -66,6 +66,7 @@ The format is based on [Keep a Changelog](http://keepachangelog.com/en/1.0.0/).

- Fixed issue where `Model.load_from_checkpoint("checkpoint.ckpt", map_location=map_location)` would always return model on CPU ([#17308](https://github.com/Lightning-AI/lightning/pull/17308))

- Fixed an issue that caused `num_nodes` not to be set correctly for `FSDPStrategy` ([#17438](https://github.com/Lightning-AI/lightning/pull/17438))


## [2.0.1.post0] - 2023-04-11
Expand Down
2 changes: 1 addition & 1 deletion src/lightning/pytorch/trainer/connectors/accelerator_connector.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -591,7 +591,7 @@ def _lazy_init_strategy(self) -> None:
else:
self.strategy.parallel_devices = self._parallel_devices
if hasattr(self.strategy, "num_nodes"):
self.strategy._num_nodes = self._num_nodes_flag
self.strategy.num_nodes = self._num_nodes_flag
if hasattr(self.strategy, "_layer_sync"):
self.strategy._layer_sync = self._layer_sync
if hasattr(self.strategy, "set_world_ranks"):
Expand Down
14 changes: 14 additions & 0 deletions tests/tests_pytorch/trainer/connectors/test_accelerator_connector.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -992,3 +992,17 @@ def _mock_tpu_available(value):
assert isinstance(connector.strategy.cluster_environment, XLAEnvironment)
assert connector.strategy._start_method == "fork"
assert connector.strategy.launcher.is_interactive_compatible


@pytest.mark.parametrize(
"strategy",
[
"ddp",
"ddp_spawn",
pytest.param("deepspeed", marks=RunIf(deepspeed=True)),
pytest.param("fsdp", marks=RunIf(min_torch="1.12.0")),
],
)
def test_connector_sets_num_nodes(strategy, cuda_count_2):
trainer = Trainer(accelerator="cuda", strategy=strategy, devices=2, num_nodes=2)
assert trainer.strategy.num_nodes == 2