Skip to content

Fix Special Tests #7841

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

Merged
merged 6 commits into from
Jun 16, 2021
Merged

Fix Special Tests #7841

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
6 commits
Select commit Hold shift + click to select a range
40dc000
Remove port setting
Jun 5, 2021
cdbf6af
Drop one of the params to see what happens
Jun 7, 2021
9454f6e
Split tests into two
Jun 16, 2021
284272e
Merge branch 'master' into fix/deepspeed_ci
Jun 16, 2021
a2a9ef0
Try using port setting
Jun 16, 2021
f806a4a
Revert back to splitting the tests
Jun 16, 2021
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: 0 additions & 1 deletion tests/plugins/test_deepspeed_plugin.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -568,7 +568,6 @@ def test_deepspeed_multigpu_stage_2_accumulated_grad_batches(tmpdir, cpu_offload
"""
Test to ensure with Stage 2 and multiple GPUs, accumulated grad batches works.
"""
os.environ['MASTER_PORT'] = "29500"
seed_everything(42)

class VerificationCallback(Callback):
Expand Down
18 changes: 15 additions & 3 deletions tests/plugins/test_sharded_plugin.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -297,12 +297,24 @@ def training_step(self, batch, batch_idx):


@RunIf(skip_windows=True, special=True, fairscale=True, min_gpus=2)
@pytest.mark.parametrize("accelerator", ["ddp_sharded", "ddp_sharded_spawn"])
def test_ddp_sharded_plugin_manual_optimization(tmpdir, accelerator):
def test_ddp_sharded_plugin_manual_optimization_spawn(tmpdir):
# todo (sean): this test has been split out as running both tests using parametrize causes "Address in use"
model = ManualBoringModel()
trainer = Trainer(
default_root_dir=tmpdir,
accelerator=accelerator,
accelerator='ddp_sharded_spawn',
fast_dev_run=2,
gpus=2,
)
trainer.fit(model)


@RunIf(skip_windows=True, special=True, fairscale=True, min_gpus=2)
def test_ddp_sharded_plugin_manual_optimization(tmpdir):
model = ManualBoringModel()
trainer = Trainer(
default_root_dir=tmpdir,
accelerator='ddp_sharded',
fast_dev_run=2,
gpus=2,
)
Expand Down