[pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

thunder/tests/distributed/helper.py

@@ -125,6 +125,7 @@ def forward(self, x):
 
 if torch.distributed.is_available():
     from torch.testing._internal import common_distributed, common_utils
+
     # note(crcrpar): How to write a test with `DDP`
     # Just add a method to :class:`CompileDDPTest`. The class is responsible for
     #     - calling `torch.distributed.init_process_group` with NCCL backend
@@ -201,7 +202,6 @@ def _run(cls, rank, test_name, file_name, pipe, *, fake_pg=False):
                 torch.distributed.destroy_process_group()
             sys.exit(0)
 
-
     # Configures PyTorch's default process group, must be called at the start of each
     #   distributed process
     def init_per_process_distributed(
@@ -225,7 +225,6 @@ def init_per_process_distributed(
         #   so we want to pass the ProcessGroup explicitly
         return torch.distributed.distributed_c10d._get_default_group()
 
-
     # Wraps a function so that it becomes one process of several executing the test
     #   See test_native_ddp and its helper _test_native_ddp_helper below for an example
     #   of how to use this wrapper.
@@ -293,7 +292,6 @@ def error_callback(ex):
 
             return test_fn
 
-
     # Creates a dataloader for a process
     #   If sample_seed is specified then the dataloader will load tensors with the same values
     #   on each process.
@@ -325,7 +323,6 @@ def to_device(tensors: list[torch.Tensor]) -> list[torch.Tensor]:
 
         return dataloader
 
-
     def run_test_no_sync_grad_accumulation(
         test_case: DistributedParallelTestCase,
         get_model_and_optimizer: Callable[[torch.device], tuple[torch.nn.Module, torch.optim.Optimizer]],

thunder/tests/distributed/test_checkpoint.py

@@ -5,6 +5,7 @@
 
 import pytest
 import torch
+
 if not torch.distributed.is_available():
     pytest.skip(allow_module_level=True)
 from torch.distributed import distributed_c10d as c10d

thunder/tests/distributed/test_ddp.py

@@ -6,6 +6,7 @@
 import pytest
 import torch
 import torch.distributed as tdist
+
 if not tdist.is_available():
     pytest.skip(allow_module_level=True)
 import torch.nn as nn

thunder/tests/distributed/test_fsdp.py

@@ -7,6 +7,7 @@
 import pytest
 import torch
 import torch.distributed as tdist
+
 if not tdist.is_available():
     pytest.skip(allow_module_level=True)
 import torch.nn as nn

thunder/tests/distributed/test_ops.py

@@ -4,6 +4,7 @@
 
 import pytest
 import torch
+
 if not torch.distributed.is_available():
     pytest.skip(allow_module_level=True)
 from torch.testing import make_tensor

thunder/tests/distributed/test_tensor_parallel.py

@@ -3,6 +3,7 @@
 import pytest
 import torch
 import torch.nn as nn
+
 if not torch.distributed.is_available():
     pytest.skip(allow_module_level=True)