diff --git a/android/test_app/make_assets.py b/android/test_app/make_assets.py
index 98afc63125fb..a6fd9e9315d2 100644
--- a/android/test_app/make_assets.py
+++ b/android/test_app/make_assets.py
@@ -1,6 +1,7 @@
-import torch
 from torchvision import models
 
+import torch
+
 print(torch.version.__version__)
 
 resnet18 = models.resnet18(weights=models.ResNet18_Weights.IMAGENET1K_V1)
diff --git a/android/test_app/make_assets_custom.py b/android/test_app/make_assets_custom.py
index 1e0b35dae51b..e1e50f17202a 100644
--- a/android/test_app/make_assets_custom.py
+++ b/android/test_app/make_assets_custom.py
@@ -4,10 +4,11 @@ MobileNetV2 TorchScript model, and dumps root ops used by the model for custom
 build script to create a tailored build which only contains these used ops.
 """
 
-import torch
 import yaml
 from torchvision import models
 
+import torch
+
 # Download and trace the model.
 model = models.mobilenet_v2(weights=models.MobileNet_V2_Weights.IMAGENET1K_V1)
 model.eval()
diff --git a/benchmarks/distributed/ddp/benchmark.py b/benchmarks/distributed/ddp/benchmark.py
index 483f53afe082..ee7fde14a60f 100644
--- a/benchmarks/distributed/ddp/benchmark.py
+++ b/benchmarks/distributed/ddp/benchmark.py
@@ -18,11 +18,12 @@ import sys
 import time
 
 import numpy as np
+import torchvision
+
 import torch
 import torch.distributed as dist
 import torch.nn as nn
 import torch.optim as optim
-import torchvision
 
 
 def allgather_object(obj):
diff --git a/benchmarks/distributed/pipeline/pipe.py b/benchmarks/distributed/pipeline/pipe.py
index 670c1d5c1518..c465c2488565 100644
--- a/benchmarks/distributed/pipeline/pipe.py
+++ b/benchmarks/distributed/pipeline/pipe.py
@@ -3,10 +3,10 @@ import math
 import os
 import time
 
+from benchmark_dataset import BenchmarkLMDataset, collate_sentences_lm
+
 import torch
 import torch.nn as nn
-
-from benchmark_dataset import BenchmarkLMDataset, collate_sentences_lm
 from torch.distributed import rpc
 
 from torch.distributed.pipeline.sync import Pipe
diff --git a/benchmarks/distributed/rpc/parameter_server/launcher.py b/benchmarks/distributed/rpc/parameter_server/launcher.py
index 5ab8c51e2fa5..a084d1757e6e 100644
--- a/benchmarks/distributed/rpc/parameter_server/launcher.py
+++ b/benchmarks/distributed/rpc/parameter_server/launcher.py
@@ -3,18 +3,10 @@ import json
 import os
 from pathlib import Path
 
-import torch
-import torch.distributed as c10d
-import torch.distributed.rpc as rpc
-import torch.multiprocessing as mp
-
 from data import data_map
 from metrics.ProcessedMetricsPrinter import ProcessedMetricsPrinter
 from models import model_map
 from server import server_map
-from torch.distributed.rpc import TensorPipeRpcBackendOptions
-from torch.futures import wait_all
-from torch.utils.data import DataLoader
 from trainer import (
     criterion_map,
     ddp_hook_map,
@@ -25,6 +17,14 @@ from trainer import (
     trainer_map,
 )
 
+import torch
+import torch.distributed as c10d
+import torch.distributed.rpc as rpc
+import torch.multiprocessing as mp
+from torch.distributed.rpc import TensorPipeRpcBackendOptions
+from torch.futures import wait_all
+from torch.utils.data import DataLoader
+
 
 def get_name(rank, args):
     r"""
diff --git a/benchmarks/distributed/rpc/parameter_server/server/server.py b/benchmarks/distributed/rpc/parameter_server/server/server.py
index 639088a3b103..0f9a88aa48b8 100644
--- a/benchmarks/distributed/rpc/parameter_server/server/server.py
+++ b/benchmarks/distributed/rpc/parameter_server/server/server.py
@@ -3,12 +3,12 @@ import threading
 import time
 from abc import ABC, abstractmethod
 
-import torch
-import torch.distributed.rpc as rpc
-
 from metrics.MetricsLogger import MetricsLogger
 from utils import sparse_rpc_format_to_tensor, sparse_tensor_to_rpc_format
 
+import torch
+import torch.distributed.rpc as rpc
+
 
 class ParameterServerBase(ABC):
     PARAMETER_SERVER_BATCH_METRIC = "parameter_server_batch_metric"
diff --git a/benchmarks/distributed/rpc/parameter_server/trainer/hooks.py b/benchmarks/distributed/rpc/parameter_server/trainer/hooks.py
index 2e9b73084621..c4ebbb9a2ea3 100644
--- a/benchmarks/distributed/rpc/parameter_server/trainer/hooks.py
+++ b/benchmarks/distributed/rpc/parameter_server/trainer/hooks.py
@@ -1,6 +1,7 @@
+from utils import process_bucket_with_remote_server
+
 import torch
 import torch.distributed as c10d
-from utils import process_bucket_with_remote_server
 
 
 def allreduce_hook(state, bucket):
diff --git a/benchmarks/distributed/rpc/parameter_server/trainer/trainer.py b/benchmarks/distributed/rpc/parameter_server/trainer/trainer.py
index 47cae3bf64f6..d1483b7970b7 100644
--- a/benchmarks/distributed/rpc/parameter_server/trainer/trainer.py
+++ b/benchmarks/distributed/rpc/parameter_server/trainer/trainer.py
@@ -2,10 +2,10 @@ import functools
 import time
 from abc import ABC, abstractmethod
 
-import torch
-
 from metrics.MetricsLogger import MetricsLogger
 
+import torch
+
 
 class TrainerBase(ABC):
     BATCH_LEVEL_METRIC = "batch_level_metric"
diff --git a/benchmarks/distributed/rpc/rl/coordinator.py b/benchmarks/distributed/rpc/rl/coordinator.py
index 87f11407f2cb..75b1780a34f7 100644
--- a/benchmarks/distributed/rpc/rl/coordinator.py
+++ b/benchmarks/distributed/rpc/rl/coordinator.py
@@ -2,12 +2,12 @@ import time
 
 import numpy as np
 
-import torch
-import torch.distributed.rpc as rpc
-
 from agent import AgentBase
 from observer import ObserverBase
 
+import torch
+import torch.distributed.rpc as rpc
+
 COORDINATOR_NAME = "coordinator"
 AGENT_NAME = "agent"
 OBSERVER_NAME = "observer{}"
diff --git a/benchmarks/distributed/rpc/rl/launcher.py b/benchmarks/distributed/rpc/rl/launcher.py
index 0f682c16a815..22ae89629757 100644
--- a/benchmarks/distributed/rpc/rl/launcher.py
+++ b/benchmarks/distributed/rpc/rl/launcher.py
@@ -4,11 +4,11 @@ import json
 import os
 import time
 
+from coordinator import CoordinatorBase
+
 import torch.distributed.rpc as rpc
 import torch.multiprocessing as mp
 
-from coordinator import CoordinatorBase
-
 COORDINATOR_NAME = "coordinator"
 AGENT_NAME = "agent"
 OBSERVER_NAME = "observer{}"
diff --git a/benchmarks/distributed/rpc/rl/observer.py b/benchmarks/distributed/rpc/rl/observer.py
index 98ff0e59ccd5..c74c925d0478 100644
--- a/benchmarks/distributed/rpc/rl/observer.py
+++ b/benchmarks/distributed/rpc/rl/observer.py
@@ -1,10 +1,10 @@
 import random
 import time
 
+from agent import AgentBase
+
 import torch
 import torch.distributed.rpc as rpc
-
-from agent import AgentBase
 from torch.distributed.rpc import rpc_sync
 
 
diff --git a/benchmarks/dynamo/common.py b/benchmarks/dynamo/common.py
index 9bd9893ed10a..a44ec5b67338 100644
--- a/benchmarks/dynamo/common.py
+++ b/benchmarks/dynamo/common.py
@@ -36,20 +36,21 @@ from typing import (
     Type,
     TYPE_CHECKING,
 )
-
 from typing_extensions import Self
 from unittest.mock import MagicMock
 
 import numpy as np
 import pandas as pd
 import psutil
+from scipy.stats import gmean, ttest_ind
+from tqdm.auto import tqdm, trange
+
 import torch
 import torch._dynamo
 import torch._dynamo.utils
 import torch._export
 import torch.distributed
 import torch.multiprocessing as mp
-from scipy.stats import gmean, ttest_ind
 from torch._C import _has_cuda as HAS_CUDA, _has_xpu as HAS_XPU
 from torch._dynamo.profiler import fx_insert_profiling, Profiler
 from torch._dynamo.testing import (
@@ -59,8 +60,6 @@ from torch._dynamo.testing import (
     same,
 )
 
-from tqdm.auto import tqdm, trange
-
 try:
     from torch._dynamo.utils import (
         clone_inputs,
@@ -74,15 +73,14 @@ except ImportError:
         graph_break_reasons,
         maybe_enable_compiled_autograd,
     )
+
 import torch._functorch.config
 from torch._functorch.aot_autograd import set_model_name
 from torch._inductor import config as inductor_config, metrics
 from torch._subclasses.fake_tensor import FakeTensorMode
-
 from torch.utils import _pytree as pytree
 from torch.utils._pytree import tree_map, tree_map_only
 
-
 try:
     import torch_xla
     import torch_xla.core.xla_model as xm
@@ -2343,17 +2341,17 @@ class BenchmarkRunner:
 
     def get_fsdp_auto_wrap_policy(self, model_name: str):
         from diffusers.models.transformer_2d import Transformer2DModel
-
-        from torch.distributed.fsdp.wrap import (
-            ModuleWrapPolicy,
-            size_based_auto_wrap_policy,
-        )
         from torchbenchmark.models.nanogpt.model import Block
         from transformers.models.llama.modeling_llama import LlamaDecoderLayer
 
         from transformers.models.t5.modeling_t5 import T5Block
         from transformers.models.whisper.modeling_whisper import WhisperEncoderLayer
 
+        from torch.distributed.fsdp.wrap import (
+            ModuleWrapPolicy,
+            size_based_auto_wrap_policy,
+        )
+
         # handcrafted wrap policy
         MODEL_FSDP_WRAP = {
             "stable_diffusion_unet": (Transformer2DModel,),
diff --git a/benchmarks/dynamo/huggingface.py b/benchmarks/dynamo/huggingface.py
index a998d10bf33c..5e139783c196 100755
--- a/benchmarks/dynamo/huggingface.py
+++ b/benchmarks/dynamo/huggingface.py
@@ -7,9 +7,10 @@ import subprocess
 import sys
 import warnings
 
-import torch
 from common import BenchmarkRunner, download_retry_decorator, main, reset_rng_state
 
+import torch
+
 from torch._dynamo.testing import collect_results
 from torch._dynamo.utils import clone_inputs
 
diff --git a/benchmarks/dynamo/microbenchmarks/bench_mm_fusion.py b/benchmarks/dynamo/microbenchmarks/bench_mm_fusion.py
index fe1a15de9cae..424edac2b8f1 100644
--- a/benchmarks/dynamo/microbenchmarks/bench_mm_fusion.py
+++ b/benchmarks/dynamo/microbenchmarks/bench_mm_fusion.py
@@ -1,10 +1,11 @@
 # flake8: noqa
+import triton
+from prettytable import PrettyTable
+
 import torch
 
 import torch._dynamo
 import torch._inductor.config
-import triton
-from prettytable import PrettyTable
 
 # torch._inductor.config.debug = True
 torch._inductor.config.triton.dense_indexing = True
diff --git a/benchmarks/dynamo/microbenchmarks/inductor_bmm.py b/benchmarks/dynamo/microbenchmarks/inductor_bmm.py
index 7ac296a58ad8..5c5326beef62 100644
--- a/benchmarks/dynamo/microbenchmarks/inductor_bmm.py
+++ b/benchmarks/dynamo/microbenchmarks/inductor_bmm.py
@@ -1,9 +1,10 @@
+from benchmark_helper import time_with_torch_timer
+
 import torch
 
 import torch._dynamo
 import torch._dynamo.config
 import torch._inductor.config as config
-from benchmark_helper import time_with_torch_timer
 
 
 @torch._dynamo.optimize("inductor", nopython=True)
diff --git a/benchmarks/dynamo/microbenchmarks/inductor_cpu_atomic.py b/benchmarks/dynamo/microbenchmarks/inductor_cpu_atomic.py
index 679cce6d9156..7eeb6983a884 100644
--- a/benchmarks/dynamo/microbenchmarks/inductor_cpu_atomic.py
+++ b/benchmarks/dynamo/microbenchmarks/inductor_cpu_atomic.py
@@ -1,8 +1,9 @@
 import itertools
 
+from benchmark_helper import time_with_torch_timer
+
 import torch
 import torch._dynamo
-from benchmark_helper import time_with_torch_timer
 
 
 @torch._dynamo.optimize("inductor", nopython=True)
diff --git a/benchmarks/dynamo/microbenchmarks/inductor_mm.py b/benchmarks/dynamo/microbenchmarks/inductor_mm.py
index deb3d8f8b604..2db6fc0e7222 100644
--- a/benchmarks/dynamo/microbenchmarks/inductor_mm.py
+++ b/benchmarks/dynamo/microbenchmarks/inductor_mm.py
@@ -1,10 +1,11 @@
+import triton
+from benchmark_helper import time_with_torch_timer
+
 import torch
 
 import torch._dynamo
 import torch._dynamo.config
 import torch._inductor.config as config
-import triton
-from benchmark_helper import time_with_torch_timer
 
 # The flag below controls whether to allow TF32 on matmul. This flag defaults to True.
 torch.backends.cuda.matmul.allow_tf32 = True
diff --git a/benchmarks/dynamo/microbenchmarks/matmul_relu.py b/benchmarks/dynamo/microbenchmarks/matmul_relu.py
index 629b574617ec..9baf893af96d 100644
--- a/benchmarks/dynamo/microbenchmarks/matmul_relu.py
+++ b/benchmarks/dynamo/microbenchmarks/matmul_relu.py
@@ -1,8 +1,9 @@
+from benchmark_helper import time_with_torch_timer
+
 import torch
 
 import torch._dynamo
 import torch._inductor.config as inductor_config
-from benchmark_helper import time_with_torch_timer
 
 inductor_config.triton.mm = "triton"
 
diff --git a/benchmarks/dynamo/microbenchmarks/microbench.py b/benchmarks/dynamo/microbenchmarks/microbench.py
index c4fbafe4667e..f41ec9198fdf 100755
--- a/benchmarks/dynamo/microbenchmarks/microbench.py
+++ b/benchmarks/dynamo/microbenchmarks/microbench.py
@@ -5,6 +5,7 @@ import sys
 
 import numpy as np
 import tabulate
+
 import torch
 
 import torch._inductor
diff --git a/benchmarks/dynamo/microbenchmarks/operatorbench.py b/benchmarks/dynamo/microbenchmarks/operatorbench.py
index 22e267dfc187..dcc2b428d2e2 100644
--- a/benchmarks/dynamo/microbenchmarks/operatorbench.py
+++ b/benchmarks/dynamo/microbenchmarks/operatorbench.py
@@ -1,9 +1,10 @@
 #!/usr/bin/env python3
 import click
 import numpy as np
-import torch
 from operator_inp_utils import OperatorInputsLoader
 
+import torch
+
 from torch._dynamo.backends.cudagraphs import cudagraphs_inner
 from torch._dynamo.testing import same
 from torch._inductor.compile_fx import compile_fx
diff --git a/benchmarks/dynamo/runner.py b/benchmarks/dynamo/runner.py
index 2a2161387472..54cff5658257 100755
--- a/benchmarks/dynamo/runner.py
+++ b/benchmarks/dynamo/runner.py
@@ -47,13 +47,14 @@ import matplotlib.pyplot as plt
 
 import numpy as np
 import pandas as pd
-import torch
-
-import torch._dynamo
 from matplotlib import rcParams
 from scipy.stats import gmean
 from tabulate import tabulate
 
+import torch
+
+import torch._dynamo
+
 rcParams.update({"figure.autolayout": True})
 plt.rc("axes", axisbelow=True)
 
diff --git a/benchmarks/dynamo/timm_models.py b/benchmarks/dynamo/timm_models.py
index 1d291e8d1d75..db29a9bf365a 100755
--- a/benchmarks/dynamo/timm_models.py
+++ b/benchmarks/dynamo/timm_models.py
@@ -7,9 +7,10 @@ import subprocess
 import sys
 import warnings
 
-import torch
 from common import BenchmarkRunner, download_retry_decorator, main
 
+import torch
+
 from torch._dynamo.testing import collect_results, reduce_to_scalar_loss
 from torch._dynamo.utils import clone_inputs
 
diff --git a/benchmarks/dynamo/torchbench.py b/benchmarks/dynamo/torchbench.py
index bf497a2b1997..fd11e984bbdc 100755
--- a/benchmarks/dynamo/torchbench.py
+++ b/benchmarks/dynamo/torchbench.py
@@ -10,9 +10,10 @@ import warnings
 from collections import namedtuple
 from os.path import abspath, exists
 
-import torch
 import yaml
 
+import torch
+
 try:
     from .common import BenchmarkRunner, main
 except ImportError:
diff --git a/benchmarks/dynamo/training_loss.py b/benchmarks/dynamo/training_loss.py
index 8886553c9736..f90d8e5b0bd4 100644
--- a/benchmarks/dynamo/training_loss.py
+++ b/benchmarks/dynamo/training_loss.py
@@ -5,12 +5,13 @@ import sys
 import time
 from datetime import timedelta
 
+from datasets import load_dataset, load_metric
+from transformers import AutoModelForSequenceClassification, AutoTokenizer
+
 import torch
 
 import torch._dynamo
-from datasets import load_dataset, load_metric
 from torch.utils.data import DataLoader
-from transformers import AutoModelForSequenceClassification, AutoTokenizer
 
 torch.backends.cuda.matmul.allow_tf32 = True
 
diff --git a/benchmarks/fastrnns/runner.py b/benchmarks/fastrnns/runner.py
index 6f8d38be5dc3..435b4a9594bf 100644
--- a/benchmarks/fastrnns/runner.py
+++ b/benchmarks/fastrnns/runner.py
@@ -1,9 +1,10 @@
 from collections import namedtuple
 from functools import partial
 
-import torch
 import torchvision.models as cnn
 
+import torch
+
 from .factory import (
     dropoutlstm_creator,
     imagenet_cnn_creator,
diff --git a/benchmarks/fastrnns/test_bench.py b/benchmarks/fastrnns/test_bench.py
index d7fea21049b6..fe5bd310d8cc 100644
--- a/benchmarks/fastrnns/test_bench.py
+++ b/benchmarks/fastrnns/test_bench.py
@@ -1,4 +1,5 @@
 import pytest
+
 import torch
 
 from .fuser import set_fuser
diff --git a/benchmarks/framework_overhead_benchmark/SimpleAddModule.py b/benchmarks/framework_overhead_benchmark/SimpleAddModule.py
index bdf2c199f518..54453ba687f6 100644
--- a/benchmarks/framework_overhead_benchmark/SimpleAddModule.py
+++ b/benchmarks/framework_overhead_benchmark/SimpleAddModule.py
@@ -1,6 +1,7 @@
-import torch
 from utils import NUM_LOOP_ITERS
 
+import torch
+
 
 def add_tensors_loop(x, y):
     z = torch.add(x, y)
diff --git a/benchmarks/functional_autograd_benchmark/audio_text_models.py b/benchmarks/functional_autograd_benchmark/audio_text_models.py
index abe1a8a58698..3add9b7e3d55 100644
--- a/benchmarks/functional_autograd_benchmark/audio_text_models.py
+++ b/benchmarks/functional_autograd_benchmark/audio_text_models.py
@@ -1,10 +1,10 @@
-import torch
-
 import torchaudio_models as models
-from torch import nn, Tensor
 
 from utils import check_for_functorch, extract_weights, GetterReturnType, load_weights
 
+import torch
+from torch import nn, Tensor
+
 
 has_functorch = check_for_functorch()
 
diff --git a/benchmarks/functional_autograd_benchmark/ppl_models.py b/benchmarks/functional_autograd_benchmark/ppl_models.py
index aa61e43c344c..3aba795e7362 100644
--- a/benchmarks/functional_autograd_benchmark/ppl_models.py
+++ b/benchmarks/functional_autograd_benchmark/ppl_models.py
@@ -1,9 +1,9 @@
+from utils import GetterReturnType
+
 import torch
 import torch.distributions as dist
 from torch import Tensor
 
-from utils import GetterReturnType
-
 
 def get_simple_regression(device: torch.device) -> GetterReturnType:
     N = 10
diff --git a/benchmarks/functional_autograd_benchmark/vision_models.py b/benchmarks/functional_autograd_benchmark/vision_models.py
index ee919b040975..4f1050b29397 100644
--- a/benchmarks/functional_autograd_benchmark/vision_models.py
+++ b/benchmarks/functional_autograd_benchmark/vision_models.py
@@ -1,11 +1,12 @@
 from typing import cast
 
-import torch
 import torchvision_models as models
-from torch import Tensor
 
 from utils import check_for_functorch, extract_weights, GetterReturnType, load_weights
 
+import torch
+from torch import Tensor
+
 has_functorch = check_for_functorch()
 
 
diff --git a/benchmarks/fuser/run_benchmarks.py b/benchmarks/fuser/run_benchmarks.py
index cc6a84dd70f1..40d1b5e9d0f2 100644
--- a/benchmarks/fuser/run_benchmarks.py
+++ b/benchmarks/fuser/run_benchmarks.py
@@ -4,6 +4,7 @@ import sys
 import time
 
 import click
+
 import torch
 
 torch.set_num_threads(1)
diff --git a/benchmarks/gpt_fast/benchmark.py b/benchmarks/gpt_fast/benchmark.py
index efc332d110f6..083c98e4a92b 100644
--- a/benchmarks/gpt_fast/benchmark.py
+++ b/benchmarks/gpt_fast/benchmark.py
@@ -6,8 +6,6 @@ import os
 import time
 from typing import Optional, Tuple
 
-import torch
-import torch._inductor.config
 from mixtral_moe_model import Transformer as MixtralMoE
 from mixtral_moe_quantize import (
     WeightOnlyInt8QuantHandler as MixtralMoEWeightOnlyInt8QuantHandler,
@@ -15,6 +13,9 @@ from mixtral_moe_quantize import (
 from model import Transformer as LLaMA
 from quantize import WeightOnlyInt8QuantHandler as LLaMAWeightOnlyInt8QuantHandler
 
+import torch
+import torch._inductor.config
+
 torch._inductor.config.coordinate_descent_tuning = True
 torch._inductor.config.triton.unique_kernel_names = True
 torch._inductor.config.fx_graph_cache = True  # Experimental feature to reduce compilation times, will be on by default in future
diff --git a/benchmarks/gpt_fast/mixtral_moe_quantize.py b/benchmarks/gpt_fast/mixtral_moe_quantize.py
index fbf8caa8964a..6dfac4a9938d 100644
--- a/benchmarks/gpt_fast/mixtral_moe_quantize.py
+++ b/benchmarks/gpt_fast/mixtral_moe_quantize.py
@@ -1,10 +1,10 @@
 # flake8: noqa: E266, C417, B950
+from mixtral_moe_model import ConditionalFeedForward
+
 import torch
 import torch.nn as nn
 import torch.nn.functional as F
 
-from mixtral_moe_model import ConditionalFeedForward
-
 ##### Quantization Primitives ######
 
 
diff --git a/benchmarks/inference/server.py b/benchmarks/inference/server.py
index 28dfa58ed3c0..6c33980b887b 100644
--- a/benchmarks/inference/server.py
+++ b/benchmarks/inference/server.py
@@ -187,9 +187,10 @@ class BackendWorker:
     def _setup(self):
         import time
 
-        import torch
         from torchvision.models.resnet import BasicBlock, ResNet
 
+        import torch
+
         # Create ResNet18 on meta device
         with torch.device("meta"):
             m = ResNet(BasicBlock, [2, 2, 2, 2])
diff --git a/benchmarks/instruction_counts/core/utils.py b/benchmarks/instruction_counts/core/utils.py
index 484909dcd979..0602b540fcb9 100644
--- a/benchmarks/instruction_counts/core/utils.py
+++ b/benchmarks/instruction_counts/core/utils.py
@@ -4,11 +4,11 @@ import shutil
 import textwrap
 from typing import List, Optional, Tuple
 
-from torch.utils.benchmark.utils.common import _make_temp_dir
-
 from core.api import GroupedBenchmark, TimerArgs
 from core.types import Definition, FlatIntermediateDefinition, Label
 
+from torch.utils.benchmark.utils.common import _make_temp_dir
+
 
 _TEMPDIR: Optional[str] = None
 
diff --git a/benchmarks/operator_benchmark/benchmark_core.py b/benchmarks/operator_benchmark/benchmark_core.py
index ec4d244bb5a2..c315382d1538 100644
--- a/benchmarks/operator_benchmark/benchmark_core.py
+++ b/benchmarks/operator_benchmark/benchmark_core.py
@@ -7,6 +7,7 @@ from collections import namedtuple
 
 import benchmark_utils
 import numpy as np
+
 import torch
 
 # needs to be imported after torch
diff --git a/benchmarks/operator_benchmark/benchmark_pytorch.py b/benchmarks/operator_benchmark/benchmark_pytorch.py
index 8d898b2803de..40daf7ad43fe 100644
--- a/benchmarks/operator_benchmark/benchmark_pytorch.py
+++ b/benchmarks/operator_benchmark/benchmark_pytorch.py
@@ -2,6 +2,7 @@ import json
 import time
 
 import benchmark_cpp_extension  # noqa: F401
+
 import torch
 
 
diff --git a/benchmarks/operator_benchmark/common/repeat_benchmark.py b/benchmarks/operator_benchmark/common/repeat_benchmark.py
index 6e8170ce6fbc..8f8f5fffb8ac 100644
--- a/benchmarks/operator_benchmark/common/repeat_benchmark.py
+++ b/benchmarks/operator_benchmark/common/repeat_benchmark.py
@@ -1,6 +1,7 @@
 import time
 
 import numpy as np
+
 import torch
 
 """Microbenchmarks for Tensor repeat operator. Supports PyTorch."""
diff --git a/benchmarks/operator_benchmark/common/tests/add_ops_list_test.py b/benchmarks/operator_benchmark/common/tests/add_ops_list_test.py
index 5ac29ee20ec9..48ad9a044f5b 100644
--- a/benchmarks/operator_benchmark/common/tests/add_ops_list_test.py
+++ b/benchmarks/operator_benchmark/common/tests/add_ops_list_test.py
@@ -1,4 +1,5 @@
 import operator_benchmark as op_bench
+
 import torch
 
 
diff --git a/benchmarks/operator_benchmark/common/tests/jit_forward_test.py b/benchmarks/operator_benchmark/common/tests/jit_forward_test.py
index 92f766084afa..339b2856c708 100644
--- a/benchmarks/operator_benchmark/common/tests/jit_forward_test.py
+++ b/benchmarks/operator_benchmark/common/tests/jit_forward_test.py
@@ -1,4 +1,5 @@
 import operator_benchmark as op_bench
+
 import torch
 
 intraop_bench_configs = op_bench.config_list(
diff --git a/benchmarks/operator_benchmark/common/tests/pt_backward_test.py b/benchmarks/operator_benchmark/common/tests/pt_backward_test.py
index 385b87aaeb55..3870e58b861c 100644
--- a/benchmarks/operator_benchmark/common/tests/pt_backward_test.py
+++ b/benchmarks/operator_benchmark/common/tests/pt_backward_test.py
@@ -1,4 +1,5 @@
 import operator_benchmark as op_bench
+
 import torch
 
 
diff --git a/benchmarks/operator_benchmark/common/tests/pt_configs_list_test.py b/benchmarks/operator_benchmark/common/tests/pt_configs_list_test.py
index 79da6b7da0c1..8e92f5bca183 100644
--- a/benchmarks/operator_benchmark/common/tests/pt_configs_list_test.py
+++ b/benchmarks/operator_benchmark/common/tests/pt_configs_list_test.py
@@ -1,4 +1,5 @@
 import operator_benchmark as op_bench
+
 import torch
 
 """Microbenchmarks for element-wise Add operator. Supports both Caffe2/PyTorch."""
diff --git a/benchmarks/operator_benchmark/common/tests/pt_cpu_gpu_forward_backward_test.py b/benchmarks/operator_benchmark/common/tests/pt_cpu_gpu_forward_backward_test.py
index 6e685bd7d705..5c12f48f66f6 100644
--- a/benchmarks/operator_benchmark/common/tests/pt_cpu_gpu_forward_backward_test.py
+++ b/benchmarks/operator_benchmark/common/tests/pt_cpu_gpu_forward_backward_test.py
@@ -1,4 +1,5 @@
 import operator_benchmark as op_bench
+
 import torch
 
 
diff --git a/benchmarks/operator_benchmark/common/tests/random_sample_test.py b/benchmarks/operator_benchmark/common/tests/random_sample_test.py
index 92b0d96cf691..f12f64f3fc5b 100644
--- a/benchmarks/operator_benchmark/common/tests/random_sample_test.py
+++ b/benchmarks/operator_benchmark/common/tests/random_sample_test.py
@@ -1,4 +1,5 @@
 import operator_benchmark as op_bench
+
 import torch
 
 
diff --git a/benchmarks/operator_benchmark/pt/add_test.py b/benchmarks/operator_benchmark/pt/add_test.py
index 3b63187dbce2..8431e4d398de 100644
--- a/benchmarks/operator_benchmark/pt/add_test.py
+++ b/benchmarks/operator_benchmark/pt/add_test.py
@@ -1,6 +1,5 @@
-import torch
-
 import operator_benchmark as op_bench
+import torch
 
 """Microbenchmarks for add_ operator. Supports both Caffe2/PyTorch."""
 
diff --git a/benchmarks/operator_benchmark/pt/ao_sparsifier_test.py b/benchmarks/operator_benchmark/pt/ao_sparsifier_test.py
index ba1bf0c1b9e0..a00e0dc3b211 100644
--- a/benchmarks/operator_benchmark/pt/ao_sparsifier_test.py
+++ b/benchmarks/operator_benchmark/pt/ao_sparsifier_test.py
@@ -1,10 +1,9 @@
+import operator_benchmark as op_bench
 import torch
 from torch import nn
 
 from torch.ao import pruning
 
-import operator_benchmark as op_bench
-
 
 """Microbenchmarks for sparsifier."""
 
diff --git a/benchmarks/operator_benchmark/pt/as_strided_test.py b/benchmarks/operator_benchmark/pt/as_strided_test.py
index 793e82cf11b5..d699b4bde476 100644
--- a/benchmarks/operator_benchmark/pt/as_strided_test.py
+++ b/benchmarks/operator_benchmark/pt/as_strided_test.py
@@ -1,9 +1,9 @@
 from typing import List
 
-import torch
-
 import operator_benchmark as op_bench
 
+import torch
+
 
 """Microbenchmarks for as_strided operator"""
 
diff --git a/benchmarks/operator_benchmark/pt/batchnorm_test.py b/benchmarks/operator_benchmark/pt/batchnorm_test.py
index 3f0dbf99e192..4524c307797b 100644
--- a/benchmarks/operator_benchmark/pt/batchnorm_test.py
+++ b/benchmarks/operator_benchmark/pt/batchnorm_test.py
@@ -1,8 +1,7 @@
+import operator_benchmark as op_bench
 import torch
 import torch.nn.functional as F
 
-import operator_benchmark as op_bench
-
 
 """Microbenchmarks for batchnorm operator."""
 
diff --git a/benchmarks/operator_benchmark/pt/binary_test.py b/benchmarks/operator_benchmark/pt/binary_test.py
index ddc8276da0b9..a8c9256610da 100644
--- a/benchmarks/operator_benchmark/pt/binary_test.py
+++ b/benchmarks/operator_benchmark/pt/binary_test.py
@@ -1,6 +1,5 @@
-import torch
-
 import operator_benchmark as op_bench
+import torch
 
 
 """Microbenchmarks for binary operators."""
diff --git a/benchmarks/operator_benchmark/pt/bmm_test.py b/benchmarks/operator_benchmark/pt/bmm_test.py
index 61c9dc470e82..3fcdb70cc6ad 100644
--- a/benchmarks/operator_benchmark/pt/bmm_test.py
+++ b/benchmarks/operator_benchmark/pt/bmm_test.py
@@ -1,6 +1,5 @@
-import torch
-
 import operator_benchmark as op_bench
+import torch
 
 """Microbenchmarks for add_ operator. Supports both Caffe2/PyTorch."""
 
diff --git a/benchmarks/operator_benchmark/pt/cat_test.py b/benchmarks/operator_benchmark/pt/cat_test.py
index f0655b749b4c..ee376d02469f 100644
--- a/benchmarks/operator_benchmark/pt/cat_test.py
+++ b/benchmarks/operator_benchmark/pt/cat_test.py
@@ -1,10 +1,10 @@
 import random
 from typing import List
 
-import torch
-
 import operator_benchmark as op_bench
 
+import torch
+
 
 """Microbenchmarks for Cat operator"""
 
diff --git a/benchmarks/operator_benchmark/pt/channel_shuffle_test.py b/benchmarks/operator_benchmark/pt/channel_shuffle_test.py
index bbc5708d15ad..40742b45d3f0 100644
--- a/benchmarks/operator_benchmark/pt/channel_shuffle_test.py
+++ b/benchmarks/operator_benchmark/pt/channel_shuffle_test.py
@@ -1,6 +1,5 @@
-import torch
-
 import operator_benchmark as op_bench
+import torch
 
 
 """Microbenchmarks for channel_shuffle operator."""
diff --git a/benchmarks/operator_benchmark/pt/chunk_test.py b/benchmarks/operator_benchmark/pt/chunk_test.py
index e471d801db01..d6c203efe423 100644
--- a/benchmarks/operator_benchmark/pt/chunk_test.py
+++ b/benchmarks/operator_benchmark/pt/chunk_test.py
@@ -1,6 +1,5 @@
-import torch
-
 import operator_benchmark as op_bench
+import torch
 
 
 """Microbenchmarks for Chunk operator"""
diff --git a/benchmarks/operator_benchmark/pt/clip_ranges_test.py b/benchmarks/operator_benchmark/pt/clip_ranges_test.py
index 07ba18a21d3f..0e4889af44fc 100644
--- a/benchmarks/operator_benchmark/pt/clip_ranges_test.py
+++ b/benchmarks/operator_benchmark/pt/clip_ranges_test.py
@@ -1,6 +1,5 @@
-import torch
-
 import operator_benchmark as op_bench
+import torch
 
 
 """Microbenchmarks for ClipRanges operator."""
diff --git a/benchmarks/operator_benchmark/pt/conv_test.py b/benchmarks/operator_benchmark/pt/conv_test.py
index 09b5d83a62e8..e01473a04f5b 100644
--- a/benchmarks/operator_benchmark/pt/conv_test.py
+++ b/benchmarks/operator_benchmark/pt/conv_test.py
@@ -1,9 +1,8 @@
-import torch
-import torch.nn as nn
-
 from pt import configs
 
 import operator_benchmark as op_bench
+import torch
+import torch.nn as nn
 
 """
 Microbenchmarks for Conv1d and ConvTranspose1d operators.
diff --git a/benchmarks/operator_benchmark/pt/diag_test.py b/benchmarks/operator_benchmark/pt/diag_test.py
index 0f3a15a939a5..bb128a56bfa1 100644
--- a/benchmarks/operator_benchmark/pt/diag_test.py
+++ b/benchmarks/operator_benchmark/pt/diag_test.py
@@ -1,6 +1,5 @@
-import torch
-
 import operator_benchmark as op_bench
+import torch
 
 
 """Microbenchmarks for diag operator"""
diff --git a/benchmarks/operator_benchmark/pt/embeddingbag_test.py b/benchmarks/operator_benchmark/pt/embeddingbag_test.py
index e9a3ade5ce29..f73fbc5425b0 100644
--- a/benchmarks/operator_benchmark/pt/embeddingbag_test.py
+++ b/benchmarks/operator_benchmark/pt/embeddingbag_test.py
@@ -1,8 +1,8 @@
 import numpy
-import torch
 from pt import configs
 
 import operator_benchmark as op_bench
+import torch
 
 """Embedding and EmbeddingBag Operator Benchmark"""
 
diff --git a/benchmarks/operator_benchmark/pt/fill_test.py b/benchmarks/operator_benchmark/pt/fill_test.py
index 71fbee459bd2..1252e81982ca 100644
--- a/benchmarks/operator_benchmark/pt/fill_test.py
+++ b/benchmarks/operator_benchmark/pt/fill_test.py
@@ -1,9 +1,8 @@
+import operator_benchmark as op_bench
 import torch
 
 from torch.testing._internal.common_device_type import get_all_device_types
 
-import operator_benchmark as op_bench
-
 """Microbenchmark for Fill_ operator."""
 
 fill_short_configs = op_bench.config_list(
diff --git a/benchmarks/operator_benchmark/pt/gather_test.py b/benchmarks/operator_benchmark/pt/gather_test.py
index 6bedf8027320..8a4db487b75d 100644
--- a/benchmarks/operator_benchmark/pt/gather_test.py
+++ b/benchmarks/operator_benchmark/pt/gather_test.py
@@ -1,7 +1,7 @@
 import numpy
-import torch
 
 import operator_benchmark as op_bench
+import torch
 
 
 """Microbenchmarks for gather operator."""
diff --git a/benchmarks/operator_benchmark/pt/gelu_test.py b/benchmarks/operator_benchmark/pt/gelu_test.py
index 260566b00489..ab83f6659e19 100644
--- a/benchmarks/operator_benchmark/pt/gelu_test.py
+++ b/benchmarks/operator_benchmark/pt/gelu_test.py
@@ -1,6 +1,5 @@
-import torch
-
 import operator_benchmark as op_bench
+import torch
 
 
 """
diff --git a/benchmarks/operator_benchmark/pt/groupnorm_test.py b/benchmarks/operator_benchmark/pt/groupnorm_test.py
index bd555eb78330..ead09808ece3 100644
--- a/benchmarks/operator_benchmark/pt/groupnorm_test.py
+++ b/benchmarks/operator_benchmark/pt/groupnorm_test.py
@@ -1,8 +1,7 @@
+import operator_benchmark as op_bench
 import torch
 import torch.nn.functional as F
 
-import operator_benchmark as op_bench
-
 
 """Microbenchmarks for groupnorm operator."""
 
diff --git a/benchmarks/operator_benchmark/pt/hardsigmoid_test.py b/benchmarks/operator_benchmark/pt/hardsigmoid_test.py
index 6b2d038afc06..7e9b51a0c84d 100644
--- a/benchmarks/operator_benchmark/pt/hardsigmoid_test.py
+++ b/benchmarks/operator_benchmark/pt/hardsigmoid_test.py
@@ -1,8 +1,7 @@
+import operator_benchmark as op_bench
 import torch
 import torch.nn as nn
 
-import operator_benchmark as op_bench
-
 
 """
 Microbenchmarks for the hardsigmoid operator.
diff --git a/benchmarks/operator_benchmark/pt/hardswish_test.py b/benchmarks/operator_benchmark/pt/hardswish_test.py
index 8d53d34721aa..d93fe298d2e4 100644
--- a/benchmarks/operator_benchmark/pt/hardswish_test.py
+++ b/benchmarks/operator_benchmark/pt/hardswish_test.py
@@ -1,8 +1,7 @@
+import operator_benchmark as op_bench
 import torch
 import torch.nn as nn
 
-import operator_benchmark as op_bench
-
 
 """
 Microbenchmarks for the hardswish operators.
diff --git a/benchmarks/operator_benchmark/pt/index_select_test.py b/benchmarks/operator_benchmark/pt/index_select_test.py
index 304dd86bac40..c4a0c233e91d 100644
--- a/benchmarks/operator_benchmark/pt/index_select_test.py
+++ b/benchmarks/operator_benchmark/pt/index_select_test.py
@@ -1,7 +1,7 @@
 import numpy
-import torch
 
 import operator_benchmark as op_bench
+import torch
 
 
 """Microbenchmarks for index_select operator."""
diff --git a/benchmarks/operator_benchmark/pt/instancenorm_test.py b/benchmarks/operator_benchmark/pt/instancenorm_test.py
index 9edd846d586e..47f8243817ca 100644
--- a/benchmarks/operator_benchmark/pt/instancenorm_test.py
+++ b/benchmarks/operator_benchmark/pt/instancenorm_test.py
@@ -1,8 +1,7 @@
+import operator_benchmark as op_bench
 import torch
 import torch.nn.functional as F
 
-import operator_benchmark as op_bench
-
 
 """Microbenchmarks for instancenorm operator."""
 
diff --git a/benchmarks/operator_benchmark/pt/interpolate_test.py b/benchmarks/operator_benchmark/pt/interpolate_test.py
index 9679f9f363a5..0ae487bc8320 100644
--- a/benchmarks/operator_benchmark/pt/interpolate_test.py
+++ b/benchmarks/operator_benchmark/pt/interpolate_test.py
@@ -1,6 +1,5 @@
-import torch
-
 import operator_benchmark as op_bench
+import torch
 
 """Microbenchmarks for interpolate operator."""
 
diff --git a/benchmarks/operator_benchmark/pt/layernorm_test.py b/benchmarks/operator_benchmark/pt/layernorm_test.py
index 8ad3cd925c1a..82f12cbaf587 100644
--- a/benchmarks/operator_benchmark/pt/layernorm_test.py
+++ b/benchmarks/operator_benchmark/pt/layernorm_test.py
@@ -1,8 +1,7 @@
+import operator_benchmark as op_bench
 import torch
 import torch.nn.functional as F
 
-import operator_benchmark as op_bench
-
 
 """Microbenchmarks for layernorm operator."""
 
diff --git a/benchmarks/operator_benchmark/pt/linear_prepack_fp16_test.py b/benchmarks/operator_benchmark/pt/linear_prepack_fp16_test.py
index b8532b9de356..e2c72bec92a7 100644
--- a/benchmarks/operator_benchmark/pt/linear_prepack_fp16_test.py
+++ b/benchmarks/operator_benchmark/pt/linear_prepack_fp16_test.py
@@ -1,6 +1,5 @@
-import torch
-
 import operator_benchmark as op_bench
+import torch
 
 """Microbenchmarks for linear_prepack_fp16_ operator. Supports both Caffe2/PyTorch."""
 
diff --git a/benchmarks/operator_benchmark/pt/linear_test.py b/benchmarks/operator_benchmark/pt/linear_test.py
index c51949b8369b..cf7a87f0ec0a 100644
--- a/benchmarks/operator_benchmark/pt/linear_test.py
+++ b/benchmarks/operator_benchmark/pt/linear_test.py
@@ -1,9 +1,8 @@
-import torch
-import torch.nn as nn
-
 from pt import configs
 
 import operator_benchmark as op_bench
+import torch
+import torch.nn as nn
 
 
 """Microbenchmarks for Linear operator."""
diff --git a/benchmarks/operator_benchmark/pt/linear_unpack_fp16_test.py b/benchmarks/operator_benchmark/pt/linear_unpack_fp16_test.py
index f2615bfbef55..3b2ef5bc1345 100644
--- a/benchmarks/operator_benchmark/pt/linear_unpack_fp16_test.py
+++ b/benchmarks/operator_benchmark/pt/linear_unpack_fp16_test.py
@@ -1,6 +1,5 @@
-import torch
-
 import operator_benchmark as op_bench
+import torch
 
 """Microbenchmarks for linear_unpack_fp16_ operator. Supports both Caffe2/PyTorch."""
 
diff --git a/benchmarks/operator_benchmark/pt/matmul_test.py b/benchmarks/operator_benchmark/pt/matmul_test.py
index 73e6c7e20498..5a2d52e05202 100644
--- a/benchmarks/operator_benchmark/pt/matmul_test.py
+++ b/benchmarks/operator_benchmark/pt/matmul_test.py
@@ -1,6 +1,5 @@
-import torch
-
 import operator_benchmark as op_bench
+import torch
 
 """Microbenchmarks for MatMul operator"""
 
diff --git a/benchmarks/operator_benchmark/pt/matrix_mult_test.py b/benchmarks/operator_benchmark/pt/matrix_mult_test.py
index 38b069f96994..f95c28c663cf 100644
--- a/benchmarks/operator_benchmark/pt/matrix_mult_test.py
+++ b/benchmarks/operator_benchmark/pt/matrix_mult_test.py
@@ -1,6 +1,5 @@
-import torch
-
 import operator_benchmark as op_bench
+import torch
 
 """
 Microbenchmarks for batch matrix mult with einsum and torch.bmm.
diff --git a/benchmarks/operator_benchmark/pt/nan_to_num_test.py b/benchmarks/operator_benchmark/pt/nan_to_num_test.py
index 7067fc8ae8e1..77aa92464496 100644
--- a/benchmarks/operator_benchmark/pt/nan_to_num_test.py
+++ b/benchmarks/operator_benchmark/pt/nan_to_num_test.py
@@ -1,9 +1,9 @@
 import math
 
-import torch
-
 import operator_benchmark as op_bench
 
+import torch
+
 
 """Microbenchmarks for torch.nan_to_num / nan_to_num_ operators"""
 
diff --git a/benchmarks/operator_benchmark/pt/pool_test.py b/benchmarks/operator_benchmark/pt/pool_test.py
index 266aa40ea934..6ac67cece9b3 100644
--- a/benchmarks/operator_benchmark/pt/pool_test.py
+++ b/benchmarks/operator_benchmark/pt/pool_test.py
@@ -1,8 +1,7 @@
+import operator_benchmark as op_bench
 import torch
 import torch.nn as nn
 
-import operator_benchmark as op_bench
-
 """
 Microbenchmarks for MaxPool1d and AvgPool1d operators.
 """
diff --git a/benchmarks/operator_benchmark/pt/qactivation_test.py b/benchmarks/operator_benchmark/pt/qactivation_test.py
index 980c8e2d2747..65e8516a2522 100644
--- a/benchmarks/operator_benchmark/pt/qactivation_test.py
+++ b/benchmarks/operator_benchmark/pt/qactivation_test.py
@@ -1,8 +1,7 @@
+import operator_benchmark as op_bench
 import torch
 import torch.ao.nn.quantized.functional as qF
 
-import operator_benchmark as op_bench
-
 r"""Microbenchmarks for the quantized activations."""
 
 qactivation_long_configs = op_bench.cross_product_configs(
diff --git a/benchmarks/operator_benchmark/pt/qarithmetic_test.py b/benchmarks/operator_benchmark/pt/qarithmetic_test.py
index 96db3f99cd8a..7a8def81bc5a 100644
--- a/benchmarks/operator_benchmark/pt/qarithmetic_test.py
+++ b/benchmarks/operator_benchmark/pt/qarithmetic_test.py
@@ -1,8 +1,7 @@
+import operator_benchmark as op_bench
 import torch
 from torch._ops import ops
 
-import operator_benchmark as op_bench
-
 qarithmetic_binary_configs = op_bench.cross_product_configs(
     N=(2, 8, 64, 512),
     dtype=(torch.quint8, torch.qint8, torch.qint32),
diff --git a/benchmarks/operator_benchmark/pt/qatembedding_ops_test.py b/benchmarks/operator_benchmark/pt/qatembedding_ops_test.py
index b5b8e6efb4e9..2e72dbe14223 100644
--- a/benchmarks/operator_benchmark/pt/qatembedding_ops_test.py
+++ b/benchmarks/operator_benchmark/pt/qatembedding_ops_test.py
@@ -1,10 +1,10 @@
 import numpy
-import torch
-import torch.ao.nn.qat as nnqat
 from pt import configs
-from torch.ao.quantization import default_embedding_qat_qconfig
 
 import operator_benchmark as op_bench
+import torch
+import torch.ao.nn.qat as nnqat
+from torch.ao.quantization import default_embedding_qat_qconfig
 
 """
 Microbenchmarks for QAT Embedding + EmbeddingBag operators.
diff --git a/benchmarks/operator_benchmark/pt/qbatchnorm_test.py b/benchmarks/operator_benchmark/pt/qbatchnorm_test.py
index 38a11acbb588..b8cee3ae02ba 100644
--- a/benchmarks/operator_benchmark/pt/qbatchnorm_test.py
+++ b/benchmarks/operator_benchmark/pt/qbatchnorm_test.py
@@ -1,6 +1,5 @@
-import torch
-
 import operator_benchmark as op_bench
+import torch
 
 
 """Microbenchmarks for quantized batchnorm operator."""
diff --git a/benchmarks/operator_benchmark/pt/qcat_test.py b/benchmarks/operator_benchmark/pt/qcat_test.py
index 09c0f9290011..86882e05f551 100644
--- a/benchmarks/operator_benchmark/pt/qcat_test.py
+++ b/benchmarks/operator_benchmark/pt/qcat_test.py
@@ -1,10 +1,10 @@
 from typing import List
 
+import operator_benchmark as op_bench
+
 import torch
 import torch.ao.nn.quantized as nnq
 
-import operator_benchmark as op_bench
-
 
 """Microbenchmarks for quantized Cat operator"""
 
diff --git a/benchmarks/operator_benchmark/pt/qcomparators_test.py b/benchmarks/operator_benchmark/pt/qcomparators_test.py
index 73b2f4d6db56..9a251d6417aa 100644
--- a/benchmarks/operator_benchmark/pt/qcomparators_test.py
+++ b/benchmarks/operator_benchmark/pt/qcomparators_test.py
@@ -1,6 +1,5 @@
-import torch
-
 import operator_benchmark as op_bench
+import torch
 
 qcomparators_configs = op_bench.cross_product_configs(
     N=(8, 64),
diff --git a/benchmarks/operator_benchmark/pt/qconv_test.py b/benchmarks/operator_benchmark/pt/qconv_test.py
index 4d7b77ffa43a..52374acf95be 100644
--- a/benchmarks/operator_benchmark/pt/qconv_test.py
+++ b/benchmarks/operator_benchmark/pt/qconv_test.py
@@ -1,9 +1,8 @@
-import torch
-import torch.ao.nn.quantized as nnq
-
 from pt import configs
 
 import operator_benchmark as op_bench
+import torch
+import torch.ao.nn.quantized as nnq
 
 """
 Microbenchmarks for qConv operators.
diff --git a/benchmarks/operator_benchmark/pt/qembedding_bag_lookups_test.py b/benchmarks/operator_benchmark/pt/qembedding_bag_lookups_test.py
index c9b34215cc45..7ee0792d1ef6 100644
--- a/benchmarks/operator_benchmark/pt/qembedding_bag_lookups_test.py
+++ b/benchmarks/operator_benchmark/pt/qembedding_bag_lookups_test.py
@@ -1,12 +1,12 @@
 from typing import Optional
 
 import numpy as np
+
+import operator_benchmark as op_bench
 import torch
 
 from torch.testing._internal.common_quantization import lengths_to_offsets
 
-import operator_benchmark as op_bench
-
 torch.ops.load_library("//caffe2/torch/fb/sparsenn:sparsenn_operators")
 
 
diff --git a/benchmarks/operator_benchmark/pt/qembedding_pack_test.py b/benchmarks/operator_benchmark/pt/qembedding_pack_test.py
index 7bf756751250..cfc6c51024ac 100644
--- a/benchmarks/operator_benchmark/pt/qembedding_pack_test.py
+++ b/benchmarks/operator_benchmark/pt/qembedding_pack_test.py
@@ -1,6 +1,5 @@
-import torch
-
 import operator_benchmark as op_bench
+import torch
 
 embeddingbag_conversion_short_configs = op_bench.cross_product_configs(
     num_embeddings=(80,), embedding_dim=(128, 256, 512), tags=("short",)
diff --git a/benchmarks/operator_benchmark/pt/qembeddingbag_test.py b/benchmarks/operator_benchmark/pt/qembeddingbag_test.py
index 5ee6c38d787d..fe563af0bc80 100644
--- a/benchmarks/operator_benchmark/pt/qembeddingbag_test.py
+++ b/benchmarks/operator_benchmark/pt/qembeddingbag_test.py
@@ -1,9 +1,9 @@
 import numpy
-import torch
-import torch.ao.nn.quantized as nnq
 from pt import configs
 
 import operator_benchmark as op_bench
+import torch
+import torch.ao.nn.quantized as nnq
 
 """
 Microbenchmarks for qEmbeddingBag operators.
diff --git a/benchmarks/operator_benchmark/pt/qgroupnorm_test.py b/benchmarks/operator_benchmark/pt/qgroupnorm_test.py
index fc6f85c8c0e9..0148800450e8 100644
--- a/benchmarks/operator_benchmark/pt/qgroupnorm_test.py
+++ b/benchmarks/operator_benchmark/pt/qgroupnorm_test.py
@@ -1,6 +1,5 @@
-import torch
-
 import operator_benchmark as op_bench
+import torch
 
 
 """Microbenchmarks for quantized groupnorm operator."""
diff --git a/benchmarks/operator_benchmark/pt/qinstancenorm_test.py b/benchmarks/operator_benchmark/pt/qinstancenorm_test.py
index d949ec7f6379..4dcec4f2a886 100644
--- a/benchmarks/operator_benchmark/pt/qinstancenorm_test.py
+++ b/benchmarks/operator_benchmark/pt/qinstancenorm_test.py
@@ -1,6 +1,5 @@
-import torch
-
 import operator_benchmark as op_bench
+import torch
 
 
 """Microbenchmarks for quantized instancenorm operator."""
diff --git a/benchmarks/operator_benchmark/pt/qinterpolate_test.py b/benchmarks/operator_benchmark/pt/qinterpolate_test.py
index 95fa9327cb56..c3afc308f6ba 100644
--- a/benchmarks/operator_benchmark/pt/qinterpolate_test.py
+++ b/benchmarks/operator_benchmark/pt/qinterpolate_test.py
@@ -1,6 +1,5 @@
-import torch
-
 import operator_benchmark as op_bench
+import torch
 
 """Microbenchmarks for the quantized interpolate op.
 
diff --git a/benchmarks/operator_benchmark/pt/qlayernorm_test.py b/benchmarks/operator_benchmark/pt/qlayernorm_test.py
index 27855e938d65..c2015b920d67 100644
--- a/benchmarks/operator_benchmark/pt/qlayernorm_test.py
+++ b/benchmarks/operator_benchmark/pt/qlayernorm_test.py
@@ -1,6 +1,5 @@
-import torch
-
 import operator_benchmark as op_bench
+import torch
 
 
 """Microbenchmarks for quantized layernorm operator."""
diff --git a/benchmarks/operator_benchmark/pt/qlinear_test.py b/benchmarks/operator_benchmark/pt/qlinear_test.py
index bde440afe094..1e96b3a75d1f 100644
--- a/benchmarks/operator_benchmark/pt/qlinear_test.py
+++ b/benchmarks/operator_benchmark/pt/qlinear_test.py
@@ -1,10 +1,9 @@
-import torch
-import torch.ao.nn.quantized as nnq
-import torch.ao.nn.quantized.dynamic as nnqd
-
 from pt import configs
 
 import operator_benchmark as op_bench
+import torch
+import torch.ao.nn.quantized as nnq
+import torch.ao.nn.quantized.dynamic as nnqd
 
 """
 Microbenchmarks for Quantized Linear operators.
diff --git a/benchmarks/operator_benchmark/pt/qobserver_test.py b/benchmarks/operator_benchmark/pt/qobserver_test.py
index ab17e995f7f3..c4ab625f936d 100644
--- a/benchmarks/operator_benchmark/pt/qobserver_test.py
+++ b/benchmarks/operator_benchmark/pt/qobserver_test.py
@@ -1,8 +1,7 @@
+import operator_benchmark as op_bench
 import torch
 import torch.ao.quantization.observer as obs
 
-import operator_benchmark as op_bench
-
 qobserver_short_configs_dict = {
     "attr_names": ("C", "M", "N", "dtype", "device"),
     "attrs": (
diff --git a/benchmarks/operator_benchmark/pt/qpool_test.py b/benchmarks/operator_benchmark/pt/qpool_test.py
index 5fc8ee3f9e8e..b2e1118ac890 100644
--- a/benchmarks/operator_benchmark/pt/qpool_test.py
+++ b/benchmarks/operator_benchmark/pt/qpool_test.py
@@ -1,6 +1,5 @@
-import torch
-
 import operator_benchmark as op_bench
+import torch
 
 # 2D pooling will have input matrix of rank 3 or 4
 qpool2d_long_configs = op_bench.config_list(
diff --git a/benchmarks/operator_benchmark/pt/qrnn_test.py b/benchmarks/operator_benchmark/pt/qrnn_test.py
index 56cab92f6008..5c4600683f76 100644
--- a/benchmarks/operator_benchmark/pt/qrnn_test.py
+++ b/benchmarks/operator_benchmark/pt/qrnn_test.py
@@ -1,8 +1,7 @@
+import operator_benchmark as op_bench
 import torch
 from torch import nn
 
-import operator_benchmark as op_bench
-
 """
 Microbenchmarks for RNNs.
 """
diff --git a/benchmarks/operator_benchmark/pt/qtensor_method_test.py b/benchmarks/operator_benchmark/pt/qtensor_method_test.py
index cc099c91c27d..3e545c36874e 100644
--- a/benchmarks/operator_benchmark/pt/qtensor_method_test.py
+++ b/benchmarks/operator_benchmark/pt/qtensor_method_test.py
@@ -1,6 +1,5 @@
-import torch
-
 import operator_benchmark as op_bench
+import torch
 
 # Configs for pointwise and reduction unary ops
 qmethods_configs_short = op_bench.config_list(
diff --git a/benchmarks/operator_benchmark/pt/quantization_test.py b/benchmarks/operator_benchmark/pt/quantization_test.py
index b2bef9b2f1a3..73b0ef09dba9 100644
--- a/benchmarks/operator_benchmark/pt/quantization_test.py
+++ b/benchmarks/operator_benchmark/pt/quantization_test.py
@@ -1,10 +1,9 @@
+import operator_benchmark as op_bench
 import torch
 import torch.ao.nn.quantized as nnq
 import torch.ao.quantization as tq
 import torch.nn as nn
 
-import operator_benchmark as op_bench
-
 """Microbenchmarks for general quantization operations."""
 
 # mode is used to show the direction of the benchmark:
diff --git a/benchmarks/operator_benchmark/pt/qunary_test.py b/benchmarks/operator_benchmark/pt/qunary_test.py
index 452c7d851f6c..f1f92cbf72cc 100644
--- a/benchmarks/operator_benchmark/pt/qunary_test.py
+++ b/benchmarks/operator_benchmark/pt/qunary_test.py
@@ -1,6 +1,5 @@
-import torch
-
 import operator_benchmark as op_bench
+import torch
 
 
 """Microbenchmarks for quantized unary operators (point-wise and reduction)."""
diff --git a/benchmarks/operator_benchmark/pt/remainder_test.py b/benchmarks/operator_benchmark/pt/remainder_test.py
index be308ad4e917..7b069d4386b8 100644
--- a/benchmarks/operator_benchmark/pt/remainder_test.py
+++ b/benchmarks/operator_benchmark/pt/remainder_test.py
@@ -1,6 +1,5 @@
-import torch
-
 import operator_benchmark as op_bench
+import torch
 
 
 """Microbenchmarks for remainder operators."""
diff --git a/benchmarks/operator_benchmark/pt/softmax_test.py b/benchmarks/operator_benchmark/pt/softmax_test.py
index a7d3918f399a..d7c6e170f513 100644
--- a/benchmarks/operator_benchmark/pt/softmax_test.py
+++ b/benchmarks/operator_benchmark/pt/softmax_test.py
@@ -1,8 +1,7 @@
+import operator_benchmark as op_bench
 import torch
 import torch.nn as nn
 
-import operator_benchmark as op_bench
-
 
 """
 Microbenchmarks for the softmax operators.
diff --git a/benchmarks/operator_benchmark/pt/split_test.py b/benchmarks/operator_benchmark/pt/split_test.py
index 23c875fe164d..74c388e42481 100644
--- a/benchmarks/operator_benchmark/pt/split_test.py
+++ b/benchmarks/operator_benchmark/pt/split_test.py
@@ -1,6 +1,5 @@
-import torch
-
 import operator_benchmark as op_bench
+import torch
 
 
 """Microbenchmarks for Split operator"""
diff --git a/benchmarks/operator_benchmark/pt/stack_test.py b/benchmarks/operator_benchmark/pt/stack_test.py
index 222c478ca2b8..4ebe5b8c2929 100644
--- a/benchmarks/operator_benchmark/pt/stack_test.py
+++ b/benchmarks/operator_benchmark/pt/stack_test.py
@@ -1,10 +1,10 @@
 import random
 from typing import List
 
-import torch
-
 import operator_benchmark as op_bench
 
+import torch
+
 
 """Microbenchmarks for Stack operator"""
 
diff --git a/benchmarks/operator_benchmark/pt/sum_test.py b/benchmarks/operator_benchmark/pt/sum_test.py
index 90198d8207eb..493c950569c5 100644
--- a/benchmarks/operator_benchmark/pt/sum_test.py
+++ b/benchmarks/operator_benchmark/pt/sum_test.py
@@ -1,6 +1,5 @@
-import torch
-
 import operator_benchmark as op_bench
+import torch
 
 """Microbenchmarks for sum reduction operator."""
 
diff --git a/benchmarks/operator_benchmark/pt/tensor_to_test.py b/benchmarks/operator_benchmark/pt/tensor_to_test.py
index b73058df9b82..dfd682e02463 100644
--- a/benchmarks/operator_benchmark/pt/tensor_to_test.py
+++ b/benchmarks/operator_benchmark/pt/tensor_to_test.py
@@ -1,6 +1,5 @@
-import torch
-
 import operator_benchmark as op_bench
+import torch
 
 tensor_conversion_short_configs = op_bench.cross_product_configs(
     M=(
diff --git a/benchmarks/operator_benchmark/pt/unary_test.py b/benchmarks/operator_benchmark/pt/unary_test.py
index bc1257f7d2b1..a5f617fd50cf 100644
--- a/benchmarks/operator_benchmark/pt/unary_test.py
+++ b/benchmarks/operator_benchmark/pt/unary_test.py
@@ -1,6 +1,5 @@
-import torch
-
 import operator_benchmark as op_bench
+import torch
 
 
 """Microbenchmarks for point-wise unary operator."""
diff --git a/benchmarks/operator_benchmark/pt_extension/cpp_extension_test.py b/benchmarks/operator_benchmark/pt_extension/cpp_extension_test.py
index 2b20dc4ff661..25c1818b42bc 100644
--- a/benchmarks/operator_benchmark/pt_extension/cpp_extension_test.py
+++ b/benchmarks/operator_benchmark/pt_extension/cpp_extension_test.py
@@ -1,6 +1,7 @@
 import unittest
 
 import benchmark_cpp_extension  # noqa: F401
+
 import torch
 
 
diff --git a/benchmarks/operator_benchmark/pt_extension/setup.py b/benchmarks/operator_benchmark/pt_extension/setup.py
index 52492075f845..53db50f4436c 100644
--- a/benchmarks/operator_benchmark/pt_extension/setup.py
+++ b/benchmarks/operator_benchmark/pt_extension/setup.py
@@ -1,4 +1,5 @@
 from setuptools import setup
+
 from torch.utils.cpp_extension import BuildExtension, CppExtension
 
 setup(
diff --git a/benchmarks/overrides_benchmark/bench.py b/benchmarks/overrides_benchmark/bench.py
index 93c7f1b130d4..72dcf89bedb5 100644
--- a/benchmarks/overrides_benchmark/bench.py
+++ b/benchmarks/overrides_benchmark/bench.py
@@ -1,10 +1,10 @@
 import argparse
 import time
 
-import torch
-
 from common import SubTensor, SubWithTorchFunction, WithTorchFunction
 
+import torch
+
 NUM_REPEATS = 1000
 NUM_REPEAT_OF_REPEATS = 1000
 
diff --git a/benchmarks/overrides_benchmark/pyspybench.py b/benchmarks/overrides_benchmark/pyspybench.py
index 66e20c494d13..69745162dfef 100644
--- a/benchmarks/overrides_benchmark/pyspybench.py
+++ b/benchmarks/overrides_benchmark/pyspybench.py
@@ -1,8 +1,9 @@
 import argparse
 
-import torch
 from common import SubTensor, SubWithTorchFunction, WithTorchFunction  # noqa: F401
 
+import torch
+
 Tensor = torch.tensor
 
 NUM_REPEATS = 1000000
diff --git a/benchmarks/profiler_benchmark/resnet_memory_profiler.py b/benchmarks/profiler_benchmark/resnet_memory_profiler.py
index eafb8a0c8b9a..85212b982180 100644
--- a/benchmarks/profiler_benchmark/resnet_memory_profiler.py
+++ b/benchmarks/profiler_benchmark/resnet_memory_profiler.py
@@ -1,7 +1,8 @@
+import torchvision.models as models
+
 import torch
 
 import torch.autograd.profiler as profiler
-import torchvision.models as models
 
 for with_cuda in [False, True]:
     model = models.resnet18()
diff --git a/benchmarks/serialization/simple_measurement.py b/benchmarks/serialization/simple_measurement.py
index 9c7705f82fe8..9cc12ca13292 100644
--- a/benchmarks/serialization/simple_measurement.py
+++ b/benchmarks/serialization/simple_measurement.py
@@ -1,6 +1,7 @@
-import torch
 from pyarkbench import Benchmark, default_args, Timer
 
+import torch
+
 use_new = True
 
 
diff --git a/benchmarks/sparse/benchmark_semi_structured_sparsity.py b/benchmarks/sparse/benchmark_semi_structured_sparsity.py
index 59aa63915883..66311c40428f 100644
--- a/benchmarks/sparse/benchmark_semi_structured_sparsity.py
+++ b/benchmarks/sparse/benchmark_semi_structured_sparsity.py
@@ -2,11 +2,12 @@ import argparse
 import random
 
 import pandas as pd
+from tqdm import tqdm
+
 import torch
 import torch.utils.benchmark as benchmark
 from torch import nn
 from torch.sparse import SparseSemiStructuredTensor, to_sparse_semi_structured
-from tqdm import tqdm
 
 
 torch.set_printoptions(
diff --git a/benchmarks/sparse/dlmc/matmul_bench.py b/benchmarks/sparse/dlmc/matmul_bench.py
index 889125a8c44a..07a712c9baae 100644
--- a/benchmarks/sparse/dlmc/matmul_bench.py
+++ b/benchmarks/sparse/dlmc/matmul_bench.py
@@ -9,9 +9,10 @@ import argparse
 import os
 import sys
 
+from scipy.sparse import isspmatrix
+
 import torch
 import torch.utils.benchmark as benchmark_utils
-from scipy.sparse import isspmatrix
 
 from .utils import load_dlmc_dataset
 
diff --git a/benchmarks/sparse/dlmc/utils.py b/benchmarks/sparse/dlmc/utils.py
index d07a79b15e17..196b74cd4c11 100644
--- a/benchmarks/sparse/dlmc/utils.py
+++ b/benchmarks/sparse/dlmc/utils.py
@@ -1,9 +1,10 @@
 import math
 from pathlib import Path
 
-import torch
 from scipy import sparse
 
+import torch
+
 
 def to_coo_scipy(x):
     indices_1 = x._indices().numpy()
diff --git a/benchmarks/sparse/spmm.py b/benchmarks/sparse/spmm.py
index 166ced09c9c9..b707556dd7a1 100644
--- a/benchmarks/sparse/spmm.py
+++ b/benchmarks/sparse/spmm.py
@@ -1,9 +1,10 @@
 import argparse
 import sys
 
-import torch
 from utils import Event, gen_sparse_coo, gen_sparse_coo_and_csr, gen_sparse_csr
 
+import torch
+
 
 def test_sparse_csr(m, n, k, nnz, test_count):
     start_timer = Event(enable_timing=True)
diff --git a/benchmarks/sparse/triton_ops.py b/benchmarks/sparse/triton_ops.py
index d1f856b7df41..66522694d0ad 100644
--- a/benchmarks/sparse/triton_ops.py
+++ b/benchmarks/sparse/triton_ops.py
@@ -125,6 +125,7 @@ if __name__ == "__main__":
     import sys
 
     import triton
+
     from torch.testing import make_tensor
 
     torch.manual_seed(0)
diff --git a/benchmarks/sparse/utils.py b/benchmarks/sparse/utils.py
index 97e2ee48f065..a1f695d7065c 100644
--- a/benchmarks/sparse/utils.py
+++ b/benchmarks/sparse/utils.py
@@ -4,6 +4,7 @@ import random
 import time
 
 import numpy as np
+
 import torch
 
 
diff --git a/benchmarks/tensorexpr/benchmark.py b/benchmarks/tensorexpr/benchmark.py
index d6fa1485436e..38951ac5091f 100644
--- a/benchmarks/tensorexpr/benchmark.py
+++ b/benchmarks/tensorexpr/benchmark.py
@@ -4,6 +4,7 @@ import os
 import time
 
 import numpy as np
+
 import torch
 
 from . import tensor_engine
diff --git a/benchmarks/tensorexpr/broadcast.py b/benchmarks/tensorexpr/broadcast.py
index 2401416c5cdb..7350460476ec 100644
--- a/benchmarks/tensorexpr/broadcast.py
+++ b/benchmarks/tensorexpr/broadcast.py
@@ -2,6 +2,7 @@ import itertools
 import operator
 
 import numpy as np
+
 import torch
 
 from . import benchmark
diff --git a/benchmarks/tensorexpr/concat.py b/benchmarks/tensorexpr/concat.py
index 795045221a66..568764afad2e 100644
--- a/benchmarks/tensorexpr/concat.py
+++ b/benchmarks/tensorexpr/concat.py
@@ -1,4 +1,5 @@
 import numpy as np
+
 import torch
 
 from . import benchmark
diff --git a/benchmarks/tensorexpr/elementwise.py b/benchmarks/tensorexpr/elementwise.py
index eee6ef621fd6..5ea4d503e4fd 100644
--- a/benchmarks/tensorexpr/elementwise.py
+++ b/benchmarks/tensorexpr/elementwise.py
@@ -3,6 +3,7 @@ import operator
 
 import numpy as np
 import scipy.special
+
 import torch
 
 from . import benchmark
diff --git a/benchmarks/tensorexpr/microbenchmarks.py b/benchmarks/tensorexpr/microbenchmarks.py
index 6177502f7f4c..7716aeb6bf12 100644
--- a/benchmarks/tensorexpr/microbenchmarks.py
+++ b/benchmarks/tensorexpr/microbenchmarks.py
@@ -6,6 +6,7 @@ import matplotlib.pyplot as plt
 import numpy as np
 import pandas as pd
 import seaborn as sns
+
 import torch
 import torch._C._te as te
 
diff --git a/benchmarks/transformer/attention_bias_benchmarks.py b/benchmarks/transformer/attention_bias_benchmarks.py
index fbe57b540dc2..fbcf06955527 100644
--- a/benchmarks/transformer/attention_bias_benchmarks.py
+++ b/benchmarks/transformer/attention_bias_benchmarks.py
@@ -4,14 +4,15 @@ from functools import partial
 from typing import Callable, List, Union
 
 import numpy as np
+from tabulate import tabulate
+from tqdm import tqdm
+
 import torch
 import torch.nn as nn
 import torch.nn.functional as F
 import torch.utils.benchmark as benchmark
-from tabulate import tabulate
 from torch.nn.attention.bias import CausalBias, CausalVariant
 from torch.nn.parameter import Parameter
-from tqdm import tqdm
 
 
 def benchmark_torch_function_in_microseconds(func: Callable, *args, **kwargs) -> float:
diff --git a/benchmarks/transformer/better_transformer_vs_mha_functional.py b/benchmarks/transformer/better_transformer_vs_mha_functional.py
index 71be7db4562a..44e8105ef1e9 100644
--- a/benchmarks/transformer/better_transformer_vs_mha_functional.py
+++ b/benchmarks/transformer/better_transformer_vs_mha_functional.py
@@ -20,11 +20,12 @@ from pprint import pprint
 from typing import Optional
 
 import numpy as np
-import torch
 
 from prettytable import PrettyTable
 from tqdm import tqdm
 
+import torch
+
 warnings.filterwarnings("ignore")
 
 error_dict = defaultdict(int)
diff --git a/benchmarks/transformer/score_mod.py b/benchmarks/transformer/score_mod.py
index 57088c45f8a0..29d951bc1dee 100644
--- a/benchmarks/transformer/score_mod.py
+++ b/benchmarks/transformer/score_mod.py
@@ -6,11 +6,12 @@ from functools import partial
 from typing import Callable, List, Optional, Tuple
 
 import numpy as np
+from tabulate import tabulate
+from tqdm import tqdm
+
 import torch
 import torch.nn.functional as F
-from tabulate import tabulate
 from torch.nn.attention._flex_attention import _flex_attention
-from tqdm import tqdm
 
 torch._dynamo.config.automatic_dynamic_shapes = False
 # Needed since changing args to function causes recompiles
diff --git a/benchmarks/transformer/sdp.py b/benchmarks/transformer/sdp.py
index c79ab8358bdc..eacf6cf0f00d 100644
--- a/benchmarks/transformer/sdp.py
+++ b/benchmarks/transformer/sdp.py
@@ -9,11 +9,12 @@ from pprint import pprint
 from typing import List, Optional
 
 import numpy as np
+from prettytable import PrettyTable
+from tqdm import tqdm
+
 import torch
 import torch.utils.benchmark as benchmark
-from prettytable import PrettyTable
 from torch.backends.cuda import sdp_kernel
-from tqdm import tqdm
 
 warnings.filterwarnings("ignore")
 
diff --git a/benchmarks/transformer/sdpa.py b/benchmarks/transformer/sdpa.py
index 6492561b6a91..d45970213e01 100644
--- a/benchmarks/transformer/sdpa.py
+++ b/benchmarks/transformer/sdpa.py
@@ -4,12 +4,13 @@ from contextlib import nullcontext
 from dataclasses import asdict, dataclass
 from typing import Callable, List, Tuple
 
+from tabulate import tabulate
+from tqdm import tqdm
+
 import torch
 import torch.utils.benchmark as benchmark
-from tabulate import tabulate
 from torch.nn.attention import sdpa_kernel, SDPBackend
 from torch.nn.functional import scaled_dot_product_attention
-from tqdm import tqdm
 
 
 def benchmark_torch_function_in_microseconds(func: Callable, *args, **kwargs) -> float:
diff --git a/docs/source/scripts/build_activation_images.py b/docs/source/scripts/build_activation_images.py
index 2a91d929899a..21e27a0ca1a4 100644
--- a/docs/source/scripts/build_activation_images.py
+++ b/docs/source/scripts/build_activation_images.py
@@ -7,9 +7,9 @@ online tutorials.
 from pathlib import Path
 
 import matplotlib
+from matplotlib import pyplot as plt
 
 import torch
-from matplotlib import pyplot as plt
 
 matplotlib.use("Agg")
 
diff --git a/functorch/benchmarks/cse.py b/functorch/benchmarks/cse.py
index af21a3432cde..3bf5ab11514c 100644
--- a/functorch/benchmarks/cse.py
+++ b/functorch/benchmarks/cse.py
@@ -1,11 +1,11 @@
 import torch
 import torch.fx as fx
 
+from functorch import make_fx
+
 from torch._functorch.compile_utils import fx_graph_cse
 from torch.profiler import profile, ProfilerActivity
 
-from functorch import make_fx
-
 
 def profile_it(f, inp):
     for _ in range(5):
diff --git a/functorch/benchmarks/operator_authoring.py b/functorch/benchmarks/operator_authoring.py
index f5f77071b930..975311a67c56 100644
--- a/functorch/benchmarks/operator_authoring.py
+++ b/functorch/benchmarks/operator_authoring.py
@@ -3,6 +3,7 @@ from functools import partial
 
 import numpy as np
 import pandas as pd
+
 import torch
 
 from functorch.compile import pointwise_operator
diff --git a/functorch/benchmarks/per_sample_grads.py b/functorch/benchmarks/per_sample_grads.py
index 73c1e3e2ef8a..7eb99096ec72 100644
--- a/functorch/benchmarks/per_sample_grads.py
+++ b/functorch/benchmarks/per_sample_grads.py
@@ -1,11 +1,12 @@
 import time
 
-import torch
-import torch.nn as nn
 import torchvision.models as models
 from opacus import PrivacyEngine
 from opacus.utils.module_modification import convert_batchnorm_modules
 
+import torch
+import torch.nn as nn
+
 from functorch import grad, make_functional, vmap
 
 device = "cuda"
diff --git a/functorch/dim/__init__.py b/functorch/dim/__init__.py
index 519a7cb271cd..2413ed331fb2 100644
--- a/functorch/dim/__init__.py
+++ b/functorch/dim/__init__.py
@@ -2,9 +2,9 @@ import dis
 import inspect
 from typing import Sequence, Union
 
-import torch
-
 import functorch._C
+
+import torch
 from functorch._C import dim as _C
 from .tree_map import tree_flatten, tree_map
 from .wrap_type import wrap_type
diff --git a/functorch/examples/dp_cifar10/cifar10_opacus.py b/functorch/examples/dp_cifar10/cifar10_opacus.py
index bd8239b187e6..236763ac3171 100644
--- a/functorch/examples/dp_cifar10/cifar10_opacus.py
+++ b/functorch/examples/dp_cifar10/cifar10_opacus.py
@@ -12,16 +12,17 @@ import sys
 from datetime import datetime, timedelta
 
 import numpy as np
-import torch
-import torch.nn as nn
-import torch.optim as optim
-import torch.utils.data
 import torchvision.transforms as transforms
 from opacus import PrivacyEngine
 from torchvision import models
 from torchvision.datasets import CIFAR10
 from tqdm import tqdm
 
+import torch
+import torch.nn as nn
+import torch.optim as optim
+import torch.utils.data
+
 
 logging.basicConfig(
     format="%(asctime)s:%(levelname)s:%(message)s",
diff --git a/functorch/examples/dp_cifar10/cifar10_transforms.py b/functorch/examples/dp_cifar10/cifar10_transforms.py
index bbab7f46b5e2..786031bce9b8 100644
--- a/functorch/examples/dp_cifar10/cifar10_transforms.py
+++ b/functorch/examples/dp_cifar10/cifar10_transforms.py
@@ -12,16 +12,17 @@ import sys
 from datetime import datetime, timedelta
 
 import numpy as np
+import torchvision.transforms as transforms
+from torchvision import models
+from torchvision.datasets import CIFAR10
+from tqdm import tqdm
+
 import torch
 import torch.nn as nn
 import torch.optim as optim
 import torch.utils.data
-import torchvision.transforms as transforms
 
 from torch.func import functional_call, grad_and_value, vmap
-from torchvision import models
-from torchvision.datasets import CIFAR10
-from tqdm import tqdm
 
 logging.basicConfig(
     format="%(asctime)s:%(levelname)s:%(message)s",
diff --git a/functorch/examples/maml_omniglot/maml-omniglot-higher.py b/functorch/examples/maml_omniglot/maml-omniglot-higher.py
index fa38fd616ee3..db058e1f6211 100755
--- a/functorch/examples/maml_omniglot/maml-omniglot-higher.py
+++ b/functorch/examples/maml_omniglot/maml-omniglot-higher.py
@@ -36,10 +36,11 @@ import matplotlib.pyplot as plt
 import numpy as np
 
 import pandas as pd
+from support.omniglot_loaders import OmniglotNShot
+
 import torch
 import torch.nn.functional as F
 import torch.optim as optim
-from support.omniglot_loaders import OmniglotNShot
 from torch import nn
 
 mpl.use("Agg")
diff --git a/functorch/examples/maml_omniglot/maml-omniglot-ptonly.py b/functorch/examples/maml_omniglot/maml-omniglot-ptonly.py
index 4145cf86240d..436d11d15912 100755
--- a/functorch/examples/maml_omniglot/maml-omniglot-ptonly.py
+++ b/functorch/examples/maml_omniglot/maml-omniglot-ptonly.py
@@ -35,13 +35,14 @@ import matplotlib.pyplot as plt
 import numpy as np
 
 import pandas as pd
+from support.omniglot_loaders import OmniglotNShot
+
 import torch
 import torch.nn.functional as F
 import torch.optim as optim
-from support.omniglot_loaders import OmniglotNShot
-from torch import nn
 
 from functorch import make_functional_with_buffers
+from torch import nn
 
 mpl.use("Agg")
 plt.style.use("bmh")
diff --git a/functorch/examples/maml_omniglot/maml-omniglot-transforms.py b/functorch/examples/maml_omniglot/maml-omniglot-transforms.py
index ba55e5804788..2a7c1b95fb0c 100755
--- a/functorch/examples/maml_omniglot/maml-omniglot-transforms.py
+++ b/functorch/examples/maml_omniglot/maml-omniglot-transforms.py
@@ -36,10 +36,11 @@ import matplotlib.pyplot as plt
 import numpy as np
 
 import pandas as pd
+from support.omniglot_loaders import OmniglotNShot
+
 import torch
 import torch.nn.functional as F
 import torch.optim as optim
-from support.omniglot_loaders import OmniglotNShot
 from torch import nn
 from torch.func import functional_call, grad, vmap
 
diff --git a/functorch/examples/maml_omniglot/support/omniglot_loaders.py b/functorch/examples/maml_omniglot/support/omniglot_loaders.py
index 5d7f6421bbb4..7e54d3584a87 100644
--- a/functorch/examples/maml_omniglot/support/omniglot_loaders.py
+++ b/functorch/examples/maml_omniglot/support/omniglot_loaders.py
@@ -22,11 +22,11 @@ import os
 import os.path
 
 import numpy as np
+import torchvision.transforms as transforms
+from PIL import Image
 
 import torch
 import torch.utils.data as data
-import torchvision.transforms as transforms
-from PIL import Image
 
 
 class Omniglot(data.Dataset):
diff --git a/functorch/examples/maml_regression/evjang.py b/functorch/examples/maml_regression/evjang.py
index cd0ee575c495..423aaab1e169 100644
--- a/functorch/examples/maml_regression/evjang.py
+++ b/functorch/examples/maml_regression/evjang.py
@@ -7,6 +7,7 @@ import math
 import matplotlib as mpl
 import matplotlib.pyplot as plt
 import numpy as np
+
 import torch
 from torch.nn import functional as F
 
diff --git a/functorch/examples/maml_regression/evjang_transforms.py b/functorch/examples/maml_regression/evjang_transforms.py
index c70fe5e1cde9..92a8fbcdbf55 100644
--- a/functorch/examples/maml_regression/evjang_transforms.py
+++ b/functorch/examples/maml_regression/evjang_transforms.py
@@ -7,6 +7,7 @@ import math
 import matplotlib as mpl
 import matplotlib.pyplot as plt
 import numpy as np
+
 import torch
 from torch.func import grad, vmap
 from torch.nn import functional as F
diff --git a/functorch/examples/maml_regression/evjang_transforms_module.py b/functorch/examples/maml_regression/evjang_transforms_module.py
index b4c71ed19d8e..e2b8c548282e 100644
--- a/functorch/examples/maml_regression/evjang_transforms_module.py
+++ b/functorch/examples/maml_regression/evjang_transforms_module.py
@@ -7,11 +7,12 @@ import math
 import matplotlib as mpl
 import matplotlib.pyplot as plt
 import numpy as np
+
 import torch
-from torch import nn
-from torch.nn import functional as F
 
 from functorch import grad, make_functional, vmap
+from torch import nn
+from torch.nn import functional as F
 
 mpl.use("Agg")
 
diff --git a/functorch/experimental/__init__.py b/functorch/experimental/__init__.py
index 23fc8381cc23..3941f6d96e1f 100644
--- a/functorch/experimental/__init__.py
+++ b/functorch/experimental/__init__.py
@@ -1,6 +1,5 @@
 # PyTorch forward-mode is not mature yet
+from functorch import functionalize
 from torch._functorch.apis import chunk_vmap
 from torch._functorch.batch_norm_replacement import replace_all_batch_norm_modules_
 from torch._functorch.eager_transforms import hessian, jacfwd, jvp
-
-from functorch import functionalize
diff --git a/functorch/op_analysis/gen_data.py b/functorch/op_analysis/gen_data.py
index 23672071cd97..4fda25f9a4e2 100644
--- a/functorch/op_analysis/gen_data.py
+++ b/functorch/op_analysis/gen_data.py
@@ -1,9 +1,10 @@
 import csv
 from collections import defaultdict
 
-import torch
 import yaml
 
+import torch
+
 
 def get_ops_for_key(key):
     # Needs modified PyTorch C++ code to work
diff --git a/ios/TestApp/benchmark/coreml_backend.py b/ios/TestApp/benchmark/coreml_backend.py
index 3bfee6bad0eb..7b2473637ccf 100644
--- a/ios/TestApp/benchmark/coreml_backend.py
+++ b/ios/TestApp/benchmark/coreml_backend.py
@@ -1,7 +1,8 @@
+from torchvision import models
+
 import torch
 
 from torch.backends._coreml.preprocess import CompileSpec, CoreMLComputeUnit, TensorSpec
-from torchvision import models
 
 
 def mobilenetv2_spec():
diff --git a/ios/TestApp/benchmark/trace_model.py b/ios/TestApp/benchmark/trace_model.py
index 24af8d383348..8614f396c764 100644
--- a/ios/TestApp/benchmark/trace_model.py
+++ b/ios/TestApp/benchmark/trace_model.py
@@ -1,6 +1,7 @@
+from torchvision import models
+
 import torch
 from torch.utils.mobile_optimizer import optimize_for_mobile
-from torchvision import models
 
 model = models.mobilenet_v2(weights=models.MobileNet_V2_Weights.IMAGENET1K_V1)
 model.eval()
diff --git a/ios/TestApp/custom_build/custom_build.py b/ios/TestApp/custom_build/custom_build.py
index de4335d58cb6..c07a66aeda3f 100644
--- a/ios/TestApp/custom_build/custom_build.py
+++ b/ios/TestApp/custom_build/custom_build.py
@@ -1,7 +1,8 @@
-import torch
 import yaml
 from torchvision import models
 
+import torch
+
 model = models.mobilenet_v2(weights=models.MobileNet_V2_Weights.IMAGENET1K_V1)
 model.eval()
 example = torch.rand(1, 3, 224, 224)
diff --git a/pyproject.toml b/pyproject.toml
index f11314c7268f..8091c2650a83 100644
--- a/pyproject.toml
+++ b/pyproject.toml
@@ -34,12 +34,9 @@ include_trailing_comma = true
 
 
 [tool.usort.known]
-first_party = ["caffe2", "torchgen", "functorch", "test"]
+first_party = ["caffe2", "torch", "torchgen", "functorch", "test"]
 standard_library = ["typing_extensions"]
 
-[tool.usort.kown]
-first_party = ["torch"]
-
 
 [tool.ruff]
 target-version = "py38"
diff --git a/scripts/export/update_schema.py b/scripts/export/update_schema.py
index dff24c83bd8f..674e409a7347 100644
--- a/scripts/export/update_schema.py
+++ b/scripts/export/update_schema.py
@@ -1,9 +1,10 @@
 import argparse
 import os
 
-from torch._export.serde import schema_check
 from yaml import dump, Dumper
 
+from torch._export.serde import schema_check
+
 if __name__ == "__main__":
     parser = argparse.ArgumentParser(prog="update_schema")
     parser.add_argument(
diff --git a/scripts/release_notes/categorize.py b/scripts/release_notes/categorize.py
index 0187637e7b84..6ef0e0c199f7 100644
--- a/scripts/release_notes/categorize.py
+++ b/scripts/release_notes/categorize.py
@@ -4,7 +4,6 @@ import textwrap
 from pathlib import Path
 
 import common
-import torch
 
 # Imports for working with classi
 from classifier import (
@@ -18,6 +17,8 @@ from classifier import (
 from commitlist import CommitList
 from common import get_commit_data_cache, topics
 
+import torch
+
 
 class Categorizer:
     def __init__(self, path, category="Uncategorized", use_classifier: bool = False):
diff --git a/scripts/release_notes/classifier.py b/scripts/release_notes/classifier.py
index 6c0a9a02f8bc..c64bad818e4e 100644
--- a/scripts/release_notes/classifier.py
+++ b/scripts/release_notes/classifier.py
@@ -9,12 +9,13 @@ from typing import Dict, List
 
 import common
 import pandas as pd
-import torch
-import torch.nn as nn
 import torchtext
 from torchtext.functional import to_tensor
 from tqdm import tqdm
 
+import torch
+import torch.nn as nn
+
 
 XLMR_BASE = torchtext.models.XLMR_BASE_ENCODER
 # This should not be here but it works for now
diff --git a/test/cpp_extensions/no_python_abi_suffix_test/setup.py b/test/cpp_extensions/no_python_abi_suffix_test/setup.py
index 9b39cd5723c1..69a49f0e56dc 100644
--- a/test/cpp_extensions/no_python_abi_suffix_test/setup.py
+++ b/test/cpp_extensions/no_python_abi_suffix_test/setup.py
@@ -1,4 +1,5 @@
 from setuptools import setup
+
 from torch.utils.cpp_extension import BuildExtension, CppExtension
 
 setup(
diff --git a/test/cpp_extensions/setup.py b/test/cpp_extensions/setup.py
index 4d4288a3076f..e0f1f858e884 100644
--- a/test/cpp_extensions/setup.py
+++ b/test/cpp_extensions/setup.py
@@ -1,8 +1,9 @@
 import os
 import sys
 
-import torch.cuda
 from setuptools import setup
+
+import torch.cuda
 from torch.testing._internal.common_utils import IS_WINDOWS
 from torch.utils.cpp_extension import (
     BuildExtension,
diff --git a/test/custom_backend/test_custom_backend.py b/test/custom_backend/test_custom_backend.py
index 7b3d300d0342..5faf5fa94d5a 100644
--- a/test/custom_backend/test_custom_backend.py
+++ b/test/custom_backend/test_custom_backend.py
@@ -3,9 +3,9 @@
 import os
 import tempfile
 
-import torch
-
 from backend import get_custom_backend_library_path, Model, to_custom_backend
+
+import torch
 from torch.testing._internal.common_utils import run_tests, TestCase
 
 
diff --git a/test/custom_operator/my_custom_ops.py b/test/custom_operator/my_custom_ops.py
index 6fff63e5fcf6..e9928f93ceb8 100644
--- a/test/custom_operator/my_custom_ops.py
+++ b/test/custom_operator/my_custom_ops.py
@@ -1,6 +1,7 @@
-import torch
 from model import get_custom_op_library_path
 
+import torch
+
 torch.ops.load_library(get_custom_op_library_path())
 
 
diff --git a/test/custom_operator/my_custom_ops2.py b/test/custom_operator/my_custom_ops2.py
index 8a9db6e59b75..f92ab415134c 100644
--- a/test/custom_operator/my_custom_ops2.py
+++ b/test/custom_operator/my_custom_ops2.py
@@ -1,6 +1,7 @@
-import torch
 from model import get_custom_op_library_path
 
+import torch
+
 torch.ops.load_library(get_custom_op_library_path())
 
 
diff --git a/test/custom_operator/pointwise.py b/test/custom_operator/pointwise.py
index 534aa0733a90..1a26bc7385ff 100644
--- a/test/custom_operator/pointwise.py
+++ b/test/custom_operator/pointwise.py
@@ -1,6 +1,7 @@
-import torch
 from model import get_custom_op_library_path
 
+import torch
+
 torch.ops.load_library(get_custom_op_library_path())
 
 
diff --git a/test/custom_operator/test_custom_ops.py b/test/custom_operator/test_custom_ops.py
index 1b490c49e9ca..26f4bfc62f1d 100644
--- a/test/custom_operator/test_custom_ops.py
+++ b/test/custom_operator/test_custom_ops.py
@@ -5,10 +5,10 @@ import sys
 import tempfile
 import unittest
 
+from model import get_custom_op_library_path, Model
+
 import torch
 import torch._library.utils as utils
-
-from model import get_custom_op_library_path, Model
 from torch import ops
 from torch.testing._internal.common_utils import IS_WINDOWS, run_tests, TestCase
 
diff --git a/test/distributed/_tensor/test_dtensor.py b/test/distributed/_tensor/test_dtensor.py
index e2a9d33241e7..70a67b8e0b93 100644
--- a/test/distributed/_tensor/test_dtensor.py
+++ b/test/distributed/_tensor/test_dtensor.py
@@ -1,9 +1,10 @@
 # Copyright (c) Meta Platforms, Inc. and affiliates
 # Owner(s): ["oncall: distributed"]
 
+from numpy.testing import assert_array_equal
+
 import torch
 import torch.nn.functional as F
-from numpy.testing import assert_array_equal
 from torch.distributed._functional_collectives import AsyncCollectiveTensor
 
 from torch.distributed._tensor import (
diff --git a/test/distributed/_tensor/test_dtensor_compile.py b/test/distributed/_tensor/test_dtensor_compile.py
index fb89b9908b86..f40cb4999858 100644
--- a/test/distributed/_tensor/test_dtensor_compile.py
+++ b/test/distributed/_tensor/test_dtensor_compile.py
@@ -70,9 +70,8 @@ bw_graph_cell = [None]
 fw_compiler = functools.partial(extract_graph, graph_cell=fw_graph_cell)
 bw_compiler = functools.partial(extract_graph, graph_cell=bw_graph_cell)
 
-from torch._dynamo.backends.common import aot_autograd
-
 from functorch.compile import min_cut_rematerialization_partition
+from torch._dynamo.backends.common import aot_autograd
 
 aot_eager_graph = aot_autograd(
     fw_compiler=fw_compiler,
diff --git a/test/distributed/_tensor/test_xla_integration.py b/test/distributed/_tensor/test_xla_integration.py
index 682d10250f4c..f2fa0c6dd961 100644
--- a/test/distributed/_tensor/test_xla_integration.py
+++ b/test/distributed/_tensor/test_xla_integration.py
@@ -7,6 +7,7 @@ from functools import wraps
 from typing import Any, Callable, Dict, Tuple
 
 import numpy as np
+
 import torch
 from torch import nn
 from torch.distributed._tensor import (
diff --git a/test/distributed/elastic/rendezvous/etcd_server_test.py b/test/distributed/elastic/rendezvous/etcd_server_test.py
index 6e21948a38a6..05db3045d097 100644
--- a/test/distributed/elastic/rendezvous/etcd_server_test.py
+++ b/test/distributed/elastic/rendezvous/etcd_server_test.py
@@ -10,6 +10,7 @@ import sys
 import unittest
 
 import etcd
+
 from torch.distributed.elastic.rendezvous.etcd_rendezvous import (
     EtcdRendezvous,
     EtcdRendezvousHandler,
diff --git a/test/distributed/pipeline/sync/skip/test_gpipe.py b/test/distributed/pipeline/sync/skip/test_gpipe.py
index e002d65db7ff..4f433ab38941 100644
--- a/test/distributed/pipeline/sync/skip/test_gpipe.py
+++ b/test/distributed/pipeline/sync/skip/test_gpipe.py
@@ -7,6 +7,7 @@
 # This source code is licensed under the BSD license found in the
 # LICENSE file in the root directory of this source tree.
 import pytest
+
 import torch
 from torch import nn
 
diff --git a/test/distributed/pipeline/sync/skip/test_leak.py b/test/distributed/pipeline/sync/skip/test_leak.py
index 2bf797dae559..f4d1043e0549 100644
--- a/test/distributed/pipeline/sync/skip/test_leak.py
+++ b/test/distributed/pipeline/sync/skip/test_leak.py
@@ -7,6 +7,7 @@
 # This source code is licensed under the BSD license found in the
 # LICENSE file in the root directory of this source tree.
 import pytest
+
 import torch
 from torch import nn
 
diff --git a/test/distributed/pipeline/sync/skip/test_portal.py b/test/distributed/pipeline/sync/skip/test_portal.py
index e50b5e1059b8..5ad180b6f9c8 100644
--- a/test/distributed/pipeline/sync/skip/test_portal.py
+++ b/test/distributed/pipeline/sync/skip/test_portal.py
@@ -7,6 +7,7 @@
 # This source code is licensed under the BSD license found in the
 # LICENSE file in the root directory of this source tree.
 import pytest
+
 import torch
 
 from torch.distributed.pipeline.sync.dependency import fork, join
diff --git a/test/distributed/pipeline/sync/skip/test_stash_pop.py b/test/distributed/pipeline/sync/skip/test_stash_pop.py
index e67cfd47bd92..5d273860f6a6 100644
--- a/test/distributed/pipeline/sync/skip/test_stash_pop.py
+++ b/test/distributed/pipeline/sync/skip/test_stash_pop.py
@@ -7,6 +7,7 @@
 # This source code is licensed under the BSD license found in the
 # LICENSE file in the root directory of this source tree.
 import pytest
+
 import torch
 from torch import nn
 
diff --git a/test/distributed/pipeline/sync/skip/test_tracker.py b/test/distributed/pipeline/sync/skip/test_tracker.py
index 007a5a963a7c..9c3a970f7574 100644
--- a/test/distributed/pipeline/sync/skip/test_tracker.py
+++ b/test/distributed/pipeline/sync/skip/test_tracker.py
@@ -10,6 +10,7 @@ import threading
 from queue import Queue
 
 import pytest
+
 import torch
 from torch import nn
 
diff --git a/test/distributed/pipeline/sync/skip/test_verify_skippables.py b/test/distributed/pipeline/sync/skip/test_verify_skippables.py
index 265c3fee85f9..1d5941487da8 100644
--- a/test/distributed/pipeline/sync/skip/test_verify_skippables.py
+++ b/test/distributed/pipeline/sync/skip/test_verify_skippables.py
@@ -7,6 +7,7 @@
 # This source code is licensed under the BSD license found in the
 # LICENSE file in the root directory of this source tree.
 import pytest
+
 from torch import nn
 
 from torch.distributed.pipeline.sync.skip import Namespace, skippable, verify_skippables
diff --git a/test/distributed/pipeline/sync/test_balance.py b/test/distributed/pipeline/sync/test_balance.py
index 82af7545bb0f..faf09f4581ae 100644
--- a/test/distributed/pipeline/sync/test_balance.py
+++ b/test/distributed/pipeline/sync/test_balance.py
@@ -9,6 +9,7 @@
 import time
 
 import pytest
+
 import torch
 from torch import nn
 
diff --git a/test/distributed/pipeline/sync/test_bugs.py b/test/distributed/pipeline/sync/test_bugs.py
index 1cb981c6a468..928a78db6e32 100644
--- a/test/distributed/pipeline/sync/test_bugs.py
+++ b/test/distributed/pipeline/sync/test_bugs.py
@@ -7,6 +7,7 @@
 # This source code is licensed under the BSD license found in the
 # LICENSE file in the root directory of this source tree.
 import pytest
+
 import torch
 import torch.nn.functional as F
 from torch import nn
diff --git a/test/distributed/pipeline/sync/test_checkpoint.py b/test/distributed/pipeline/sync/test_checkpoint.py
index e1ae6f6754ec..7be8ddefafe9 100644
--- a/test/distributed/pipeline/sync/test_checkpoint.py
+++ b/test/distributed/pipeline/sync/test_checkpoint.py
@@ -9,6 +9,7 @@
 from functools import partial
 
 import pytest
+
 import torch
 import torch.cuda
 from torch import nn
diff --git a/test/distributed/pipeline/sync/test_copy.py b/test/distributed/pipeline/sync/test_copy.py
index 22a3a37805f2..302c3d25d53f 100644
--- a/test/distributed/pipeline/sync/test_copy.py
+++ b/test/distributed/pipeline/sync/test_copy.py
@@ -7,6 +7,7 @@
 # This source code is licensed under the BSD license found in the
 # LICENSE file in the root directory of this source tree.
 import pytest
+
 import torch
 
 from torch.distributed.pipeline.sync.copy import Copy, Wait
diff --git a/test/distributed/pipeline/sync/test_deferred_batch_norm.py b/test/distributed/pipeline/sync/test_deferred_batch_norm.py
index 2fea96e7ddb6..c3807c57d612 100644
--- a/test/distributed/pipeline/sync/test_deferred_batch_norm.py
+++ b/test/distributed/pipeline/sync/test_deferred_batch_norm.py
@@ -10,6 +10,7 @@ from copy import deepcopy
 from itertools import chain
 
 import pytest
+
 import torch
 from torch import nn, optim
 
diff --git a/test/distributed/pipeline/sync/test_dependency.py b/test/distributed/pipeline/sync/test_dependency.py
index 73283b88abb2..e966d6541bf5 100644
--- a/test/distributed/pipeline/sync/test_dependency.py
+++ b/test/distributed/pipeline/sync/test_dependency.py
@@ -9,6 +9,7 @@
 import weakref
 
 import pytest
+
 import torch
 
 from torch.distributed.pipeline.sync.dependency import Fork, fork, Join, join
diff --git a/test/distributed/pipeline/sync/test_inplace.py b/test/distributed/pipeline/sync/test_inplace.py
index eade0f43e1bd..33f31b2a52bb 100644
--- a/test/distributed/pipeline/sync/test_inplace.py
+++ b/test/distributed/pipeline/sync/test_inplace.py
@@ -7,6 +7,7 @@
 # This source code is licensed under the BSD license found in the
 # LICENSE file in the root directory of this source tree.
 import pytest
+
 import torch
 from torch import nn
 
diff --git a/test/distributed/pipeline/sync/test_microbatch.py b/test/distributed/pipeline/sync/test_microbatch.py
index 89d7a8b995b7..b5e44aa73a8d 100644
--- a/test/distributed/pipeline/sync/test_microbatch.py
+++ b/test/distributed/pipeline/sync/test_microbatch.py
@@ -7,6 +7,7 @@
 # This source code is licensed under the BSD license found in the
 # LICENSE file in the root directory of this source tree.
 import pytest
+
 import torch
 import torch.cuda
 
diff --git a/test/distributed/pipeline/sync/test_pipe.py b/test/distributed/pipeline/sync/test_pipe.py
index 848a3ed7eb90..e493b1d5a03e 100644
--- a/test/distributed/pipeline/sync/test_pipe.py
+++ b/test/distributed/pipeline/sync/test_pipe.py
@@ -12,6 +12,7 @@ from collections import OrderedDict
 from copy import deepcopy
 
 import pytest
+
 import torch
 from torch import nn, Tensor
 
diff --git a/test/distributed/pipeline/sync/test_stream.py b/test/distributed/pipeline/sync/test_stream.py
index 29281ca606d8..f9702c8e4152 100644
--- a/test/distributed/pipeline/sync/test_stream.py
+++ b/test/distributed/pipeline/sync/test_stream.py
@@ -7,6 +7,7 @@
 # This source code is licensed under the BSD license found in the
 # LICENSE file in the root directory of this source tree.
 import pytest
+
 import torch
 
 from torch.distributed.pipeline.sync.stream import (
diff --git a/test/distributed/pipeline/sync/test_worker.py b/test/distributed/pipeline/sync/test_worker.py
index ca5d99c57638..f82af2ea0067 100644
--- a/test/distributed/pipeline/sync/test_worker.py
+++ b/test/distributed/pipeline/sync/test_worker.py
@@ -9,6 +9,7 @@
 import threading
 
 import pytest
+
 import torch
 
 from torch.distributed.pipeline.sync.microbatch import Batch
diff --git a/test/distributed/pipelining/test_backward.py b/test/distributed/pipelining/test_backward.py
index 5791f40c6102..083fd12d5820 100644
--- a/test/distributed/pipelining/test_backward.py
+++ b/test/distributed/pipelining/test_backward.py
@@ -2,9 +2,9 @@
 # Owner(s): ["oncall: distributed"]
 import copy
 
-import torch
-
 from model_registry import MLPModule
+
+import torch
 from torch.distributed.pipelining._backward import stage_backward
 from torch.testing._internal.common_utils import run_tests, TestCase
 
diff --git a/test/distributed/pipelining/test_composability.py b/test/distributed/pipelining/test_composability.py
index 941d86e576e1..bbf3f1929fbc 100644
--- a/test/distributed/pipelining/test_composability.py
+++ b/test/distributed/pipelining/test_composability.py
@@ -5,10 +5,11 @@ import os
 import sys
 import tempfile
 
+from model_registry import MLPModule
+
 import torch
 import torch.distributed as dist
 import torch.nn as nn
-from model_registry import MLPModule
 from torch.distributed._composable.fsdp.fully_shard import (
     fully_shard,
     MixedPrecisionPolicy,
diff --git a/test/distributed/pipelining/test_pipe.py b/test/distributed/pipelining/test_pipe.py
index 74d13111bec7..b7ef10ebad5b 100644
--- a/test/distributed/pipelining/test_pipe.py
+++ b/test/distributed/pipelining/test_pipe.py
@@ -1,8 +1,8 @@
 # Copyright (c) Meta Platforms, Inc. and affiliates
 # Owner(s): ["oncall: distributed"]
-import torch
-
 from model_registry import MLPModule
+
+import torch
 from torch.distributed.pipelining import pipe_split, pipeline
 from torch.testing._internal.common_utils import (
     instantiate_parametrized_tests,
diff --git a/test/distributed/pipelining/test_schedule.py b/test/distributed/pipelining/test_schedule.py
index 42bb7eaf9fdc..232f69d8bcef 100644
--- a/test/distributed/pipelining/test_schedule.py
+++ b/test/distributed/pipelining/test_schedule.py
@@ -5,10 +5,10 @@ import os
 import sys
 import tempfile
 
+from model_registry import ModelWithKwargs, MultiMLP
+
 import torch
 import torch.distributed as dist
-
-from model_registry import ModelWithKwargs, MultiMLP
 from torch.distributed.pipelining import (
     ManualPipelineStage,
     pipeline,
diff --git a/test/distributed/pipelining/test_stage.py b/test/distributed/pipelining/test_stage.py
index a7c1f51b3da9..97a147cb357a 100644
--- a/test/distributed/pipelining/test_stage.py
+++ b/test/distributed/pipelining/test_stage.py
@@ -4,10 +4,10 @@ import os
 import sys
 import tempfile
 
+from model_registry import ExampleCode, ModelWithKwargs, MultiMLP
+
 import torch
 import torch.distributed as dist
-
-from model_registry import ExampleCode, ModelWithKwargs, MultiMLP
 from torch.distributed.pipelining import (
     ManualPipelineStage,
     pipeline,
diff --git a/test/distributed/test_c10d_gloo.py b/test/distributed/test_c10d_gloo.py
index 34f8849ddbf3..6db53fea64a1 100644
--- a/test/distributed/test_c10d_gloo.py
+++ b/test/distributed/test_c10d_gloo.py
@@ -20,9 +20,6 @@ if not c10d.is_available() or not c10d.is_gloo_available():
     sys.exit(0)
 
 import test_c10d_common
-import torch.distributed as dist
-import torch.nn.functional as F
-import torch.testing._internal.common_utils as common
 from test_c10d_common import (
     gpus_for_rank,
     LOOPBACK,
@@ -30,6 +27,10 @@ from test_c10d_common import (
     SparseGradientModule,
     Task,
 )
+
+import torch.distributed as dist
+import torch.nn.functional as F
+import torch.testing._internal.common_utils as common
 from torch import nn
 from torch.distributed._shard.sharded_tensor import (
     init_from_local_shards,
diff --git a/test/distributed/test_c10d_nccl.py b/test/distributed/test_c10d_nccl.py
index 4b0ea6b4dd17..4fea855a85b9 100644
--- a/test/distributed/test_c10d_nccl.py
+++ b/test/distributed/test_c10d_nccl.py
@@ -28,12 +28,13 @@ if not c10d.is_available() or not c10d.is_nccl_available():
 from typing import Dict, List
 
 import test_c10d_common
+from test_c10d_common import ConvNet, DoubleGpuNet, gpus_for_rank, ModuleForDdpCommHook
+
 import torch.distributed as dist
 import torch.distributed.algorithms.ddp_comm_hooks.default_hooks as default
 import torch.distributed.algorithms.ddp_comm_hooks.powerSGD_hook as powerSGD
 import torch.nn.functional as F
 import torch.testing._internal.common_utils as common
-from test_c10d_common import ConvNet, DoubleGpuNet, gpus_for_rank, ModuleForDdpCommHook
 from torch import nn
 from torch._C._distributed_c10d import OpType
 from torch.nn.parallel import DistributedDataParallel
diff --git a/test/distributed/test_c10d_spawn_gloo.py b/test/distributed/test_c10d_spawn_gloo.py
index 95897a2938b1..a5c5a0a05512 100644
--- a/test/distributed/test_c10d_spawn_gloo.py
+++ b/test/distributed/test_c10d_spawn_gloo.py
@@ -6,10 +6,11 @@ import sys
 import tempfile
 
 import test_c10d_spawn
+from test_c10d_spawn import _torch_dist_nn_available, TestDistributedNNFunctions
+
 import torch
 import torch.distributed as c10d
 import torch.nn as nn
-from test_c10d_spawn import _torch_dist_nn_available, TestDistributedNNFunctions
 from torch.testing._internal.common_cuda import TEST_CUDA, TEST_MULTIGPU
 from torch.testing._internal.common_distributed import (
     create_device,
diff --git a/test/distributed/test_c10d_spawn_nccl.py b/test/distributed/test_c10d_spawn_nccl.py
index 67d6fae24266..184ecd988ec2 100644
--- a/test/distributed/test_c10d_spawn_nccl.py
+++ b/test/distributed/test_c10d_spawn_nccl.py
@@ -3,9 +3,10 @@
 import sys
 
 import test_c10d_spawn
+from test_c10d_spawn import _torch_dist_nn_available, TestDistributedNNFunctions
+
 import torch
 import torch.distributed as c10d
-from test_c10d_spawn import _torch_dist_nn_available, TestDistributedNNFunctions
 from torch.testing._internal.common_cuda import TEST_MULTIGPU
 from torch.testing._internal.common_distributed import requires_nccl, skip_if_lt_x_gpu
 from torch.testing._internal.common_utils import (
diff --git a/test/distributed/test_c10d_spawn_ucc.py b/test/distributed/test_c10d_spawn_ucc.py
index 81f7ec621018..db41e9b6274d 100644
--- a/test/distributed/test_c10d_spawn_ucc.py
+++ b/test/distributed/test_c10d_spawn_ucc.py
@@ -3,9 +3,10 @@
 import sys
 
 import test_c10d_spawn
+from test_c10d_spawn import _torch_dist_nn_available, TestDistributedNNFunctions
+
 import torch
 import torch.distributed as c10d
-from test_c10d_spawn import _torch_dist_nn_available, TestDistributedNNFunctions
 from torch.testing._internal.common_cuda import TEST_MULTIGPU
 from torch.testing._internal.common_distributed import requires_ucc, skip_if_lt_x_gpu
 from torch.testing._internal.common_utils import (
diff --git a/test/distributed/test_c10d_ucc.py b/test/distributed/test_c10d_ucc.py
index 75294c90b5df..e4a1c9004d72 100644
--- a/test/distributed/test_c10d_ucc.py
+++ b/test/distributed/test_c10d_ucc.py
@@ -18,15 +18,16 @@ if not c10d.is_available() or not c10d.is_ucc_available():
     sys.exit(0)
 
 import test_c10d_common
-import torch.distributed as dist
-import torch.nn.functional as F
-import torch.testing._internal.common_utils as common
 from test_c10d_common import (
     gpus_for_rank,
     ModuleForDdpCommHook,
     SparseGradientModule,
     Task,
 )
+
+import torch.distributed as dist
+import torch.nn.functional as F
+import torch.testing._internal.common_utils as common
 from torch import nn
 from torch.nn.parallel import DistributedDataParallel
 from torch.testing._internal.common_distributed import (
diff --git a/test/distributed/test_dynamo_distributed.py b/test/distributed/test_dynamo_distributed.py
index 058118d6fa4c..b31a2f717537 100644
--- a/test/distributed/test_dynamo_distributed.py
+++ b/test/distributed/test_dynamo_distributed.py
@@ -10,6 +10,7 @@ from typing import List
 from unittest.mock import patch
 
 import numpy as np
+
 import torch
 import torch._dynamo
 import torch._dynamo.logging
diff --git a/test/distributed/test_functional_api.py b/test/distributed/test_functional_api.py
index ecbbb4e5b3ae..f2255637a691 100644
--- a/test/distributed/test_functional_api.py
+++ b/test/distributed/test_functional_api.py
@@ -10,13 +10,13 @@ import torch.distributed as dist
 import torch.distributed._functional_collectives as ft_c
 import torch.distributed._tensor as dt
 import torch.distributed.distributed_c10d as c10d
+
+from functorch import make_fx
 from torch._inductor.utils import run_and_get_code
 from torch.testing import FileCheck
 from torch.testing._internal.distributed.fake_pg import FakeStore
 from torch.utils._triton import has_triton
 
-from functorch import make_fx
-
 if not dist.is_available():
     print("Distributed not available, skipping tests", file=sys.stderr)
     sys.exit(0)
diff --git a/test/distributions/test_distributions.py b/test/distributions/test_distributions.py
index 9e271a750999..79be804d287f 100644
--- a/test/distributions/test_distributions.py
+++ b/test/distributions/test_distributions.py
@@ -35,9 +35,10 @@ from collections import namedtuple
 from itertools import product
 from random import shuffle
 
+from packaging import version
+
 import torch
 import torch.autograd.forward_ad as fwAD
-from packaging import version
 
 from torch import inf, nan
 from torch.autograd import grad
diff --git a/test/dynamo/test_activation_checkpointing.py b/test/dynamo/test_activation_checkpointing.py
index 5382b3fcf3ba..ee6d56a84ddb 100644
--- a/test/dynamo/test_activation_checkpointing.py
+++ b/test/dynamo/test_activation_checkpointing.py
@@ -11,6 +11,8 @@ import torch._dynamo.test_case
 import torch._functorch.config
 import torch.distributed as dist
 import torch.utils.checkpoint
+
+from functorch.compile import min_cut_rematerialization_partition
 from torch._dynamo.backends.common import aot_autograd
 from torch._dynamo.testing import CompileCounterWithBackend
 from torch._higher_order_ops.wrap import tag_activation_checkpoint
@@ -19,8 +21,6 @@ from torch.testing._internal.inductor_utils import HAS_CUDA
 from torch.testing._internal.two_tensor import TwoTensor
 from torch.utils.checkpoint import _pt2_selective_checkpoint_context_fn_gen, checkpoint
 
-from functorch.compile import min_cut_rematerialization_partition
-
 requires_cuda = unittest.skipUnless(HAS_CUDA, "requires cuda")
 requires_distributed = functools.partial(
     unittest.skipIf, not dist.is_available(), "requires distributed"
diff --git a/test/dynamo/test_aot_autograd.py b/test/dynamo/test_aot_autograd.py
index f21c84a55709..84b1e0a28c0e 100644
--- a/test/dynamo/test_aot_autograd.py
+++ b/test/dynamo/test_aot_autograd.py
@@ -722,9 +722,8 @@ class AotAutogradFallbackTests(torch._dynamo.test_case.TestCase):
                 return (out,)
 
         def compile_submod(input_mod, args):
-            from torch._functorch.aot_autograd import aot_module_simplified
-
             from functorch.compile import nop
+            from torch._functorch.aot_autograd import aot_module_simplified
 
             class WrapperModule(torch.nn.Module):
                 def __init__(self):
diff --git a/test/dynamo/test_autograd_function.py b/test/dynamo/test_autograd_function.py
index 5e70bd7cdb9a..eef1d7237780 100644
--- a/test/dynamo/test_autograd_function.py
+++ b/test/dynamo/test_autograd_function.py
@@ -15,6 +15,7 @@ from torch.testing._internal.triton_utils import HAS_CUDA, requires_cuda
 
 if HAS_CUDA:
     import triton
+
     from torch.testing._internal.triton_utils import add_kernel
 
 
diff --git a/test/dynamo/test_backends.py b/test/dynamo/test_backends.py
index 68e4b745147f..209b075c7612 100644
--- a/test/dynamo/test_backends.py
+++ b/test/dynamo/test_backends.py
@@ -240,9 +240,8 @@ class TestCustomBackendAPI(torch._dynamo.test_case.TestCase):
         self.assertTrue(backend_run)
 
     def test_aot_autograd_api(self):
-        from torch._dynamo.backends.common import aot_autograd
-
         from functorch.compile import make_boxed_func
+        from torch._dynamo.backends.common import aot_autograd
 
         backend_run = False
 
diff --git a/test/dynamo/test_debug_utils.py b/test/dynamo/test_debug_utils.py
index ff5c5e5b43ed..0c305f800e58 100644
--- a/test/dynamo/test_debug_utils.py
+++ b/test/dynamo/test_debug_utils.py
@@ -3,13 +3,13 @@
 import unittest
 
 import torch
+
+from functorch import make_fx
 from torch._dynamo import debug_utils
 from torch._dynamo.debug_utils import aot_graph_input_parser
 from torch._dynamo.test_case import TestCase
 from torch.testing._internal.inductor_utils import HAS_CUDA
 
-from functorch import make_fx
-
 requires_cuda = unittest.skipUnless(HAS_CUDA, "requires cuda")
 
 f32 = torch.float32
diff --git a/test/dynamo/test_export.py b/test/dynamo/test_export.py
index cf4310d5f74c..9f1417e23247 100644
--- a/test/dynamo/test_export.py
+++ b/test/dynamo/test_export.py
@@ -17,6 +17,8 @@ import torch
 import torch._dynamo
 import torch._dynamo.test_case
 import torch._dynamo.testing
+
+from functorch.experimental.control_flow import cond
 from torch._dynamo import config
 from torch._dynamo.exc import UserError
 from torch._dynamo.testing import normalize_gm
@@ -33,8 +35,6 @@ from torch.fx.experimental.symbolic_shapes import (
 from torch.testing._internal import common_utils
 from torch.testing._internal.common_cuda import TEST_CUDA
 
-from functorch.experimental.control_flow import cond
-
 
 class ExportTests(torch._dynamo.test_case.TestCase):
     # TODO(voz): Refactor to a shared test function.
diff --git a/test/dynamo/test_higher_order_ops.py b/test/dynamo/test_higher_order_ops.py
index bab0abc5f4ec..9b86a90b02f3 100644
--- a/test/dynamo/test_higher_order_ops.py
+++ b/test/dynamo/test_higher_order_ops.py
@@ -7,6 +7,8 @@ import sys
 import unittest
 import warnings
 
+import functorch.experimental.control_flow as control_flow
+
 import torch
 import torch._dynamo.config as config
 
@@ -32,8 +34,6 @@ from torch.testing._internal.common_utils import (
 from torch.testing._internal.inductor_utils import HAS_CUDA
 from torch.testing._internal.logging_utils import LoggingTestCase, make_logging_test
 
-import functorch.experimental.control_flow as control_flow
-
 
 requires_cuda = unittest.skipUnless(HAS_CUDA, "requires cuda")
 
diff --git a/test/dynamo/test_hooks.py b/test/dynamo/test_hooks.py
index f122803f9a79..a8fe7524db91 100644
--- a/test/dynamo/test_hooks.py
+++ b/test/dynamo/test_hooks.py
@@ -8,12 +8,12 @@ import torch
 import torch._dynamo
 import torch._dynamo.test_case
 import torch._dynamo.testing
+
+from functorch.compile import nop
 from torch._dynamo import compiled_autograd
 from torch._functorch.aot_autograd import aot_module_simplified
 from torch.utils.hooks import RemovableHandle
 
-from functorch.compile import nop
-
 
 def compiler_fn(gm):
     return torch._dynamo.optimize("inductor", nopython=True, dynamic=True)(gm)
diff --git a/test/dynamo/test_misc.py b/test/dynamo/test_misc.py
index e1de4ca3e814..ff496406140e 100644
--- a/test/dynamo/test_misc.py
+++ b/test/dynamo/test_misc.py
@@ -25,6 +25,7 @@ import weakref
 from unittest.mock import patch
 
 import numpy as np
+
 import torch
 import torch._dynamo.testing
 
diff --git a/test/dynamo/test_repros.py b/test/dynamo/test_repros.py
index 1812b7fe0e48..ecb5476ebd0b 100644
--- a/test/dynamo/test_repros.py
+++ b/test/dynamo/test_repros.py
@@ -22,6 +22,7 @@ from typing import Any, Dict, Iterator, List, Tuple
 from unittest import mock
 
 import numpy as np
+
 import torch
 
 import torch._dynamo.test_case
@@ -166,9 +167,8 @@ bw_graph = [None]
 
 
 def aot_graph_capture_backend(gm, args):
-    from torch._functorch.aot_autograd import aot_module_simplified
-
     from functorch.compile import min_cut_rematerialization_partition
+    from torch._functorch.aot_autograd import aot_module_simplified
 
     def fw_compiler(gm, _):
         fw_graph[0] = gm
diff --git a/test/dynamo/test_unspec.py b/test/dynamo/test_unspec.py
index cb47a0b728a3..317fd15195ba 100644
--- a/test/dynamo/test_unspec.py
+++ b/test/dynamo/test_unspec.py
@@ -4,6 +4,7 @@ import random
 import unittest
 
 import numpy as np
+
 import torch
 import torch._dynamo.test_case
 import torch._dynamo.testing
diff --git a/test/export/test_export.py b/test/export/test_export.py
index 4a21e38fffa4..ded6a59b9700 100644
--- a/test/export/test_export.py
+++ b/test/export/test_export.py
@@ -15,6 +15,8 @@ from typing import Dict, List
 import torch
 import torch._dynamo as torchdynamo
 import torch.nn.functional as F
+
+from functorch.experimental.control_flow import cond, map
 from torch import Tensor
 from torch._dynamo.test_case import TestCase
 from torch._export.pass_base import _ExportPassBaseDeprecatedDoNotUse
@@ -58,8 +60,6 @@ from torch.utils._pytree import (
     treespec_loads,
 )
 
-from functorch.experimental.control_flow import cond, map
-
 try:
     from torchrec.sparse.jagged_tensor import KeyedJaggedTensor
 
diff --git a/test/export/test_pass_infra.py b/test/export/test_pass_infra.py
index 052da499ed14..70e94c9827ef 100644
--- a/test/export/test_pass_infra.py
+++ b/test/export/test_pass_infra.py
@@ -3,14 +3,14 @@ import copy
 import unittest
 
 import torch
+
+from functorch.experimental import control_flow
 from torch._dynamo.eval_frame import is_dynamo_supported
 from torch._export.pass_base import _ExportPassBaseDeprecatedDoNotUse
 from torch.export import export
 from torch.fx.passes.infra.pass_base import PassResult
 from torch.testing._internal.common_utils import IS_WINDOWS, run_tests, TestCase
 
-from functorch.experimental import control_flow
-
 
 @unittest.skipIf(not is_dynamo_supported(), "Dynamo not supported")
 class TestPassInfra(TestCase):
diff --git a/test/export/test_passes.py b/test/export/test_passes.py
index c7b4404e6ec6..2807a2ded907 100644
--- a/test/export/test_passes.py
+++ b/test/export/test_passes.py
@@ -11,6 +11,8 @@ from re import escape
 from typing import List, Set
 
 import torch
+
+from functorch.experimental.control_flow import cond
 from torch._dynamo.eval_frame import is_dynamo_supported
 from torch._export.non_strict_utils import (
     _fakify_script_objects,
@@ -53,8 +55,6 @@ from torch.testing._internal.common_utils import (
 from torch.testing._internal.torchbind_impls import init_torchbind_implementations
 from torch.utils import _pytree as pytree
 
-from functorch.experimental.control_flow import cond
-
 
 def count_call_function(graph: torch.fx.Graph, target: torch.ops.OpOverload) -> int:
     count = 0
diff --git a/test/export/test_unflatten.py b/test/export/test_unflatten.py
index e46c948ebfa4..3ca58e8fff79 100644
--- a/test/export/test_unflatten.py
+++ b/test/export/test_unflatten.py
@@ -10,6 +10,8 @@ from typing import Any, List
 
 import torch
 import torch._dynamo as torchdynamo
+
+from functorch.experimental.control_flow import cond, map
 from torch import Tensor
 from torch._export.utils import (
     get_buffer,
@@ -51,8 +53,6 @@ from torch.utils._pytree import (
     treespec_loads,
 )
 
-from functorch.experimental.control_flow import cond, map
-
 
 @unittest.skipIf(not torchdynamo.is_dynamo_supported(), "dynamo isn't support")
 class TestUnflatten(TestCase):
diff --git a/test/export/test_verifier.py b/test/export/test_verifier.py
index e15d3f4f0bc4..c2755d22d709 100644
--- a/test/export/test_verifier.py
+++ b/test/export/test_verifier.py
@@ -2,6 +2,8 @@
 import unittest
 
 import torch
+
+from functorch.experimental import control_flow
 from torch import Tensor
 from torch._dynamo.eval_frame import is_dynamo_supported
 
@@ -10,8 +12,6 @@ from torch.export import export
 from torch.export.exported_program import InputKind, InputSpec, TensorArgument
 from torch.testing._internal.common_utils import IS_WINDOWS, run_tests, TestCase
 
-from functorch.experimental import control_flow
-
 
 @unittest.skipIf(not is_dynamo_supported(), "dynamo isn't supported")
 class TestVerifier(TestCase):
diff --git a/test/functorch/attn_ft.py b/test/functorch/attn_ft.py
index 7329c4ed4539..3a7bd7afe67b 100644
--- a/test/functorch/attn_ft.py
+++ b/test/functorch/attn_ft.py
@@ -6,9 +6,9 @@
 import math
 
 import torch
-from torch import nn
 
 from functorch.dim import cat, dimlists, dims, softmax
+from torch import nn
 
 
 class Linear(nn.Linear):
diff --git a/test/functorch/common_utils.py b/test/functorch/common_utils.py
index 8b8c8a8efc75..d59802121c93 100644
--- a/test/functorch/common_utils.py
+++ b/test/functorch/common_utils.py
@@ -9,16 +9,17 @@ import os
 import unittest
 from collections import namedtuple
 
+from functorch_additional_op_db import additional_op_db
+
 import torch
 import torch.utils._pytree as pytree
-from functorch_additional_op_db import additional_op_db
+
+from functorch import vmap
 from torch.testing._internal.autograd_function_db import autograd_function_db
 from torch.testing._internal.common_device_type import toleranceOverride
 from torch.testing._internal.common_methods_invocations import DecorateInfo, op_db
 from torch.testing._internal.common_modules import module_db
 
-from functorch import vmap
-
 IS_FBCODE = os.getenv("FUNCTORCH_TEST_FBCODE") == "1"
 
 
diff --git a/test/functorch/discover_coverage.py b/test/functorch/discover_coverage.py
index dd7e8b6c9aeb..781b601b321b 100644
--- a/test/functorch/discover_coverage.py
+++ b/test/functorch/discover_coverage.py
@@ -7,9 +7,10 @@ from enum import Enum
 # Importing these files make modifications to the op_db that we need
 import test_ops  # noqa: F401
 import test_vmap  # noqa: F401
+from functorch_additional_op_db import additional_op_db
+
 import torch
 import torch._functorch.top_operators_github_usage as top_ops
-from functorch_additional_op_db import additional_op_db
 from torch.testing._internal.common_device_type import toleranceOverride
 from torch.testing._internal.common_methods_invocations import op_db
 
diff --git a/test/functorch/test_aotdispatch.py b/test/functorch/test_aotdispatch.py
index 25152ac9af87..348af610372a 100644
--- a/test/functorch/test_aotdispatch.py
+++ b/test/functorch/test_aotdispatch.py
@@ -15,11 +15,29 @@ from functools import partial
 from typing import Any, Callable, Dict, List, Optional, Union
 from unittest.mock import patch
 
+from common_utils import decorate, decorateForModules, skip, skipOps, xfail
+
 import torch
 import torch._dynamo as torchdynamo
 import torch.nn as nn
 import torch.utils._pytree as pytree
-from common_utils import decorate, decorateForModules, skip, skipOps, xfail
+
+from functorch import grad, jacrev, make_fx, vjp, vmap
+from functorch.compile import (
+    aot_function,
+    aot_module,
+    compiled_function,
+    compiled_module,
+    default_decompositions,
+    default_partition,
+    get_aot_compilation_context,
+    make_boxed_compiler,
+    memory_efficient_fusion,
+    min_cut_rematerialization_partition,
+    nnc_jit,
+    nop,
+)
+from functorch.experimental import control_flow
 from torch._decomp import decomposition_table
 from torch._functorch.aot_autograd import (
     aot_export_joint_simple,
@@ -61,23 +79,6 @@ from torch.testing._internal.optests import (
 )
 from torch.testing._internal.two_tensor import TwoTensor, TwoTensorMode
 
-from functorch import grad, jacrev, make_fx, vjp, vmap
-from functorch.compile import (
-    aot_function,
-    aot_module,
-    compiled_function,
-    compiled_module,
-    default_decompositions,
-    default_partition,
-    get_aot_compilation_context,
-    make_boxed_compiler,
-    memory_efficient_fusion,
-    min_cut_rematerialization_partition,
-    nnc_jit,
-    nop,
-)
-from functorch.experimental import control_flow
-
 USE_TORCHVISION = False
 try:
     import torchvision
diff --git a/test/functorch/test_control_flow.py b/test/functorch/test_control_flow.py
index 2065327f3148..cfad762c54be 100644
--- a/test/functorch/test_control_flow.py
+++ b/test/functorch/test_control_flow.py
@@ -5,6 +5,9 @@ import unittest
 
 import torch
 import torch.utils._pytree as pytree
+
+from functorch.experimental import control_flow
+from functorch.experimental.control_flow import cond, UnsupportedAliasMutationException
 from torch._higher_order_ops.while_loop import while_loop
 from torch._subclasses.functional_tensor import (
     CppFunctionalizeAPI,
@@ -25,9 +28,6 @@ from torch.testing._internal.common_utils import (
     TestCase,
 )
 
-from functorch.experimental import control_flow
-from functorch.experimental.control_flow import cond, UnsupportedAliasMutationException
-
 
 # TODO: pull these helpers from AOTAutograd later
 def to_fun(t):
diff --git a/test/functorch/test_dims.py b/test/functorch/test_dims.py
index 83ed1615f12b..928fba5039e9 100644
--- a/test/functorch/test_dims.py
+++ b/test/functorch/test_dims.py
@@ -9,17 +9,10 @@ import gc
 
 from unittest import skip, skipIf
 
-import torch
-
 from attn_ft import BertSelfAttention as BertSelfAttentionA, Linear
 from attn_positional import BertSelfAttention as BertSelfAttentionB
 
-from torch.testing._internal.common_utils import (
-    run_tests,
-    skipIfTorchDynamo,
-    TEST_CUDA,
-    TestCase,
-)
+import torch
 
 from functorch._C import dim as _C
 from functorch.dim import (
@@ -32,6 +25,13 @@ from functorch.dim import (
     Tensor,
 )
 
+from torch.testing._internal.common_utils import (
+    run_tests,
+    skipIfTorchDynamo,
+    TEST_CUDA,
+    TestCase,
+)
+
 try:
     from torchvision.models import resnet18
 except ImportError:
diff --git a/test/functorch/test_eager_transforms.py b/test/functorch/test_eager_transforms.py
index 124b86627913..3084c1795560 100644
--- a/test/functorch/test_eager_transforms.py
+++ b/test/functorch/test_eager_transforms.py
@@ -17,11 +17,28 @@ from functools import partial, wraps
 
 # NB: numpy is a testing dependency!
 import numpy as np
+from common_utils import expectedFailureIf
+
+import functorch
 import torch
 import torch.autograd.forward_ad as fwAD
 import torch.nn as nn
 import torch.nn.functional as F
-from common_utils import expectedFailureIf
+from functorch import (
+    combine_state_for_ensemble,
+    grad,
+    grad_and_value,
+    hessian,
+    jacfwd,
+    jacrev,
+    jvp,
+    make_functional,
+    make_functional_with_buffers,
+    make_fx,
+    vjp,
+    vmap,
+)
+from functorch.experimental import functionalize, replace_all_batch_norm_modules_
 from torch._C import _ExcludeDispatchKeyGuard, DispatchKey, DispatchKeySet
 from torch._dynamo import allow_in_graph
 from torch._functorch.eager_transforms import _slice_argnums
@@ -64,23 +81,6 @@ from torch.testing._internal.common_utils import (
 
 from torch.utils._pytree import tree_flatten, tree_map, tree_unflatten
 
-import functorch
-from functorch import (
-    combine_state_for_ensemble,
-    grad,
-    grad_and_value,
-    hessian,
-    jacfwd,
-    jacrev,
-    jvp,
-    make_functional,
-    make_functional_with_buffers,
-    make_fx,
-    vjp,
-    vmap,
-)
-from functorch.experimental import functionalize, replace_all_batch_norm_modules_
-
 USE_TORCHVISION = False
 try:
     import torchvision  # noqa: F401
diff --git a/test/functorch/test_memory_efficient_fusion.py b/test/functorch/test_memory_efficient_fusion.py
index d31748fe0c50..297e9fb3255b 100644
--- a/test/functorch/test_memory_efficient_fusion.py
+++ b/test/functorch/test_memory_efficient_fusion.py
@@ -8,12 +8,12 @@ from typing import Callable
 import torch
 import torch.fx as fx
 import torch.nn as nn
-from torch._functorch.compile_utils import fx_graph_cse
-from torch.nn import functional as F
-from torch.testing._internal.common_utils import run_tests, TestCase
 
 from functorch import make_fx
 from functorch.compile import memory_efficient_fusion
+from torch._functorch.compile_utils import fx_graph_cse
+from torch.nn import functional as F
+from torch.testing._internal.common_utils import run_tests, TestCase
 
 HAS_CUDA = torch.cuda.is_available()
 
diff --git a/test/functorch/test_minifier.py b/test/functorch/test_minifier.py
index b90f754a90b0..61862b2ffe21 100644
--- a/test/functorch/test_minifier.py
+++ b/test/functorch/test_minifier.py
@@ -1,11 +1,11 @@
 # Owner(s): ["module: functorch"]
 
 import torch
-from torch._functorch.compile_utils import get_outputs, get_placeholders
-from torch.testing._internal.common_utils import run_tests, TestCase
 
 from functorch import make_fx
 from functorch.compile import minifier
+from torch._functorch.compile_utils import get_outputs, get_placeholders
+from torch.testing._internal.common_utils import run_tests, TestCase
 
 
 class TestMinifier(TestCase):
diff --git a/test/functorch/test_ops.py b/test/functorch/test_ops.py
index e9b7330acfe2..c8df820c7c9c 100644
--- a/test/functorch/test_ops.py
+++ b/test/functorch/test_ops.py
@@ -10,8 +10,6 @@ import functools
 import itertools
 import unittest
 
-import torch
-import torch.autograd.forward_ad as fwAD
 from common_utils import (
     check_vmap_fallback,
     decorate,
@@ -30,6 +28,11 @@ from common_utils import (
     xfail,
 )
 from functorch_additional_op_db import additional_op_db
+
+import torch
+import torch.autograd.forward_ad as fwAD
+
+from functorch import grad, jacfwd, jacrev, vjp, vmap
 from torch import Tensor
 from torch._functorch.eager_transforms import _as_tuple, jvp
 from torch.testing._internal.autograd_function_db import autograd_function_db
@@ -61,8 +64,6 @@ from torch.testing._internal.opinfo.core import SampleInput
 from torch.utils import _pytree as pytree
 from torch.utils._pytree import tree_flatten, tree_map, tree_unflatten
 
-from functorch import grad, jacfwd, jacrev, vjp, vmap
-
 aten = torch.ops.aten
 
 
diff --git a/test/functorch/test_parsing.py b/test/functorch/test_parsing.py
index 11812f355140..d98b4aa00fd2 100644
--- a/test/functorch/test_parsing.py
+++ b/test/functorch/test_parsing.py
@@ -27,8 +27,6 @@ SOFTWARE.
 from typing import Any, Callable, Dict
 from unittest import mock
 
-from torch.testing._internal.common_utils import run_tests, TestCase
-
 from functorch.einops._parsing import (
     _ellipsis,
     AnonymousAxis,
@@ -37,6 +35,8 @@ from functorch.einops._parsing import (
     validate_rearrange_expressions,
 )
 
+from torch.testing._internal.common_utils import run_tests, TestCase
+
 mock_anonymous_axis_eq: Callable[[AnonymousAxis, object], bool] = (
     lambda self, other: isinstance(other, AnonymousAxis) and self.value == other.value
 )
diff --git a/test/functorch/test_rearrange.py b/test/functorch/test_rearrange.py
index 90c9a20d8515..47414ca6d331 100644
--- a/test/functorch/test_rearrange.py
+++ b/test/functorch/test_rearrange.py
@@ -28,10 +28,11 @@ SOFTWARE.
 from typing import List, Tuple
 
 import numpy as np
+
 import torch
-from torch.testing._internal.common_utils import run_tests, TestCase
 
 from functorch.einops import rearrange
+from torch.testing._internal.common_utils import run_tests, TestCase
 
 identity_patterns: List[str] = [
     "...->...",
diff --git a/test/functorch/test_vmap.py b/test/functorch/test_vmap.py
index b45b0ddbc1c1..e5b72b5d7eb5 100644
--- a/test/functorch/test_vmap.py
+++ b/test/functorch/test_vmap.py
@@ -18,8 +18,6 @@ from collections import namedtuple
 from typing import OrderedDict
 from unittest.case import skipIf
 
-import torch
-import torch.nn.functional as F
 from common_utils import (
     check_vmap_fallback,
     compute_quantities_for_vmap_test,
@@ -36,6 +34,13 @@ from common_utils import (
     xfail,
 )
 from functorch_additional_op_db import additional_op_db
+
+import functorch
+
+import torch
+import torch.nn.functional as F
+from functorch import grad, grad_and_value, jacfwd, jvp, vjp, vmap
+from functorch.experimental import chunk_vmap
 from torch import Tensor
 from torch._C._functorch import reshape_dim_into, reshape_dim_outof
 from torch._functorch.make_functional import functional_init_with_buffers
@@ -65,10 +70,6 @@ from torch.testing._internal.common_utils import (
 )
 from torch.utils import _pytree as pytree
 
-import functorch
-from functorch import grad, grad_and_value, jacfwd, jvp, vjp, vmap
-from functorch.experimental import chunk_vmap
-
 FALLBACK_REGEX = "There is a performance drop"
 
 
diff --git a/test/fx/test_gradual_type.py b/test/fx/test_gradual_type.py
index 1a0f5d9bde47..31f28dc45ee1 100644
--- a/test/fx/test_gradual_type.py
+++ b/test/fx/test_gradual_type.py
@@ -3,6 +3,7 @@
 import unittest
 
 import sympy
+
 import torch
 from torch.fx import GraphModule, symbolic_trace
 from torch.fx.annotate import annotate
diff --git a/test/inductor/test_aot_inductor.py b/test/inductor/test_aot_inductor.py
index 27636ff26a5f..f500687596f9 100644
--- a/test/inductor/test_aot_inductor.py
+++ b/test/inductor/test_aot_inductor.py
@@ -42,6 +42,7 @@ from torch.utils import _pytree as pytree
 
 if HAS_CUDA:
     import triton
+
     from torch.testing._internal.triton_utils import (
         add_kernel,
         add_kernel_2d_autotuned,
diff --git a/test/inductor/test_codecache.py b/test/inductor/test_codecache.py
index 5b5dd5cde3c2..994786740a65 100644
--- a/test/inductor/test_codecache.py
+++ b/test/inductor/test_codecache.py
@@ -44,6 +44,7 @@ HAS_TRITON = has_triton()
 
 if HAS_TRITON:
     import triton
+
     from torch.testing._internal.triton_utils import add_kernel
 
 requires_gpu = functools.partial(unittest.skipIf, not HAS_GPU, "requires gpu")
diff --git a/test/inductor/test_cpu_repro.py b/test/inductor/test_cpu_repro.py
index 1110fde150fc..4a9a75764080 100644
--- a/test/inductor/test_cpu_repro.py
+++ b/test/inductor/test_cpu_repro.py
@@ -12,6 +12,7 @@ from unittest.mock import patch
 
 import numpy as np
 import sympy
+
 import torch
 from torch import nn
 from torch._C import FileCheck
diff --git a/test/inductor/test_efficient_conv_bn_eval.py b/test/inductor/test_efficient_conv_bn_eval.py
index 34235a9285b7..c65b7585f9f3 100644
--- a/test/inductor/test_efficient_conv_bn_eval.py
+++ b/test/inductor/test_efficient_conv_bn_eval.py
@@ -106,9 +106,8 @@ class EfficientConvBNEvalTemplate(TestCase):
         def test_conv_bn_eval(
             test_class, use_bias, module, sync_bn, decompose_nn_module
         ):
-            from torch._dispatch.python import enable_python_dispatcher
-
             from functorch import make_fx
+            from torch._dispatch.python import enable_python_dispatcher
 
             kwargs = {"kernel_size": 3, "stride": 2} if module[0] != nn.Linear else {}
             mod_eager = test_class(
diff --git a/test/inductor/test_perf.py b/test/inductor/test_perf.py
index c2d4f935f459..5e1af26f4bfa 100644
--- a/test/inductor/test_perf.py
+++ b/test/inductor/test_perf.py
@@ -2,6 +2,8 @@
 import contextlib
 from unittest.mock import patch
 
+import functorch
+
 import torch
 import torch._inductor.config as config
 import torch.autograd
@@ -28,8 +30,6 @@ from torch.testing._internal.common_utils import skipIfRocm
 # Defines all the kernels for tests
 from torch.testing._internal.triton_utils import HAS_CUDA, requires_cuda
 
-import functorch
-
 if HAS_CUDA:
     from torch.testing._internal.triton_utils import add_kernel
 
diff --git a/test/inductor/test_triton_kernels.py b/test/inductor/test_triton_kernels.py
index 9958d1fee5e6..accab8beae6b 100644
--- a/test/inductor/test_triton_kernels.py
+++ b/test/inductor/test_triton_kernels.py
@@ -107,6 +107,7 @@ class KernelTests(torch._inductor.test_case.TestCase):
     @requires_cuda
     @skipIfRocm
     def test_triton_kernel_functionalize(self):
+        from functorch import make_fx
         from torch._higher_order_ops.triton_kernel_wrap import kernel_side_table
         from torch._subclasses.functional_tensor import (
             CppFunctionalizeAPI,
@@ -114,8 +115,6 @@ class KernelTests(torch._inductor.test_case.TestCase):
             PythonFunctionalizeAPI,
         )
 
-        from functorch import make_fx
-
         kernel_side_table.reset_table()
 
         def f(x, output):
diff --git a/test/inductor/test_utils.py b/test/inductor/test_utils.py
index 667e1513d332..66150f92bbef 100644
--- a/test/inductor/test_utils.py
+++ b/test/inductor/test_utils.py
@@ -1,6 +1,7 @@
 # Owner(s): ["module: inductor"]
 
 from sympy import Symbol
+
 from torch._inductor.test_case import run_tests, TestCase
 from torch._inductor.utils import sympy_subs
 
diff --git a/test/jit/fixtures_srcs/test_upgrader_models_generation.py b/test/jit/fixtures_srcs/test_upgrader_models_generation.py
index df0368d5d865..74516785fb4b 100644
--- a/test/jit/fixtures_srcs/test_upgrader_models_generation.py
+++ b/test/jit/fixtures_srcs/test_upgrader_models_generation.py
@@ -1,9 +1,9 @@
 # Owner(s): ["oncall: mobile"]
 
 import torch
-from torch.testing._internal.common_utils import run_tests, TestCase
 
 from test.jit.fixtures_srcs.generate_models import ALL_MODULES
+from torch.testing._internal.common_utils import run_tests, TestCase
 
 
 class TestUpgraderModelGeneration(TestCase):
diff --git a/test/jit/test_dataclasses.py b/test/jit/test_dataclasses.py
index 274818336aec..4cebc2706568 100644
--- a/test/jit/test_dataclasses.py
+++ b/test/jit/test_dataclasses.py
@@ -7,8 +7,9 @@ from dataclasses import dataclass, field, InitVar
 from enum import Enum
 from typing import List, Optional
 
-import torch
 from hypothesis import given, settings, strategies as st
+
+import torch
 from torch.testing._internal.jit_utils import JitTestCase
 
 
diff --git a/test/jit/test_decorator.py b/test/jit/test_decorator.py
index 132935c37a7e..b7d4ef0f30e8 100644
--- a/test/jit/test_decorator.py
+++ b/test/jit/test_decorator.py
@@ -7,9 +7,9 @@ from enum import Enum
 from typing import List, Optional
 
 import torch
-from torch.testing._internal.jit_utils import JitTestCase
 
 from jit.myfunction_a import my_function_a
+from torch.testing._internal.jit_utils import JitTestCase
 
 
 class TestDecorator(JitTestCase):
diff --git a/test/jit/test_misc.py b/test/jit/test_misc.py
index 8d58ab860d5b..956096173f9b 100644
--- a/test/jit/test_misc.py
+++ b/test/jit/test_misc.py
@@ -8,14 +8,14 @@ from typing import Any, Dict, List, Optional, Tuple
 import torch
 import torch.nn as nn
 import torch.testing._internal.jit_utils
+
+from jit.test_module_interface import TestModuleInterface  # noqa: F401
 from torch import jit
 from torch.testing import FileCheck
 from torch.testing._internal.common_utils import freeze_rng_state
 
 from torch.testing._internal.jit_utils import JitTestCase, make_global, RUN_CUDA_HALF
 
-from jit.test_module_interface import TestModuleInterface  # noqa: F401
-
 # Make the helper files in test/ importable
 pytorch_test_dir = os.path.dirname(os.path.dirname(os.path.realpath(__file__)))
 sys.path.append(pytorch_test_dir)
diff --git a/test/jit/test_python_ir.py b/test/jit/test_python_ir.py
index ff00823f11c8..60d85dd8d76a 100644
--- a/test/jit/test_python_ir.py
+++ b/test/jit/test_python_ir.py
@@ -3,6 +3,7 @@
 import unittest
 
 import numpy as np
+
 import torch
 from torch.testing import FileCheck
 from torch.testing._internal.common_utils import IS_MACOS
diff --git a/test/jit/test_save_load_for_op_version.py b/test/jit/test_save_load_for_op_version.py
index d9eb62bc4217..63e17a7a1867 100644
--- a/test/jit/test_save_load_for_op_version.py
+++ b/test/jit/test_save_load_for_op_version.py
@@ -7,9 +7,9 @@ from itertools import product as product
 from typing import Union
 
 import hypothesis.strategies as st
+from hypothesis import example, given, settings
 
 import torch
-from hypothesis import example, given, settings
 
 # Make the helper files in test/ importable
 pytorch_test_dir = os.path.dirname(os.path.dirname(os.path.realpath(__file__)))
diff --git a/test/jit/test_types.py b/test/jit/test_types.py
index e331aad12e44..baa1e2311bad 100644
--- a/test/jit/test_types.py
+++ b/test/jit/test_types.py
@@ -9,12 +9,12 @@ from typing import Dict, Iterator, List, Optional, Tuple
 
 import torch
 import torch.testing._internal.jit_utils
+
+from jit.test_module_interface import TestModuleInterface  # noqa: F401
 from torch.testing import FileCheck
 
 from torch.testing._internal.jit_utils import JitTestCase
 
-from jit.test_module_interface import TestModuleInterface  # noqa: F401
-
 # Make the helper files in test/ importable
 pytorch_test_dir = os.path.dirname(os.path.dirname(os.path.realpath(__file__)))
 sys.path.append(pytorch_test_dir)
diff --git a/test/lazy/test_ts_opinfo.py b/test/lazy/test_ts_opinfo.py
index c880c90a8108..102dee3664be 100644
--- a/test/lazy/test_ts_opinfo.py
+++ b/test/lazy/test_ts_opinfo.py
@@ -7,13 +7,14 @@ import pathlib
 from typing import Sequence
 from unittest import skip
 
+import yaml
+
 import torch
 import torch._lazy
 import torch._lazy.config
 import torch._lazy.ir_cache
 import torch._lazy.metrics
 import torch._lazy.ts_backend
-import yaml
 from torch.testing._internal.common_device_type import (
     instantiate_device_type_tests,
     ops,
diff --git a/test/mobile/custom_build/prepare_model.py b/test/mobile/custom_build/prepare_model.py
index 6cbf1ecef2fd..1adea1795f5e 100644
--- a/test/mobile/custom_build/prepare_model.py
+++ b/test/mobile/custom_build/prepare_model.py
@@ -4,10 +4,11 @@ MobileNetV2 TorchScript model, and dumps root ops used by the model for custom
 build script to create a tailored build which only contains these used ops.
 """
 
-import torch
 import yaml
 from torchvision import models
 
+import torch
+
 # Download and trace the model.
 model = models.mobilenet_v2(weights=models.MobileNet_V2_Weights.IMAGENET1K_V1)
 model.eval()
diff --git a/test/mobile/model_test/gen_test_model.py b/test/mobile/model_test/gen_test_model.py
index 15e3c2e807cd..e93945528b60 100644
--- a/test/mobile/model_test/gen_test_model.py
+++ b/test/mobile/model_test/gen_test_model.py
@@ -1,7 +1,6 @@
 import io
 import sys
 
-import torch
 import yaml
 from android_api_module import AndroidAPIModule
 from builtin_ops import TSBuiltinOpsModule, TSCollectionOpsModule
@@ -44,13 +43,15 @@ from tensor_ops import (
     TensorTypingOpsModule,
     TensorViewOpsModule,
 )
-from torch.jit.mobile import _load_for_lite_interpreter
 from torchvision_models import (
     MobileNetV2Module,
     MobileNetV2VulkanModule,
     Resnet18Module,
 )
 
+import torch
+from torch.jit.mobile import _load_for_lite_interpreter
+
 test_path_ios = "ios/TestApp/models/"
 test_path_android = "android/pytorch_android/src/androidTest/assets/"
 
diff --git a/test/mobile/model_test/torchvision_models.py b/test/mobile/model_test/torchvision_models.py
index cdd27a408910..8f51a6d1e4ad 100644
--- a/test/mobile/model_test/torchvision_models.py
+++ b/test/mobile/model_test/torchvision_models.py
@@ -1,7 +1,8 @@
+from torchvision import models
+
 import torch
 from torch.utils.bundled_inputs import augment_model_with_bundled_inputs
 from torch.utils.mobile_optimizer import optimize_for_mobile
-from torchvision import models
 
 
 class MobileNetV2Module:
diff --git a/test/onnx/debug_embed_params.py b/test/onnx/debug_embed_params.py
index ea6db1de1e87..8f32a838a998 100644
--- a/test/onnx/debug_embed_params.py
+++ b/test/onnx/debug_embed_params.py
@@ -2,12 +2,12 @@ import sys
 
 import onnx
 import pytorch_test_common
+
+import caffe2.python.onnx.backend as c2
 import torch
 import torch.jit
 from torch.autograd import Variable
 
-import caffe2.python.onnx.backend as c2
-
 torch.set_default_tensor_type("torch.FloatTensor")
 try:
     import torch
diff --git a/test/onnx/dynamo/test_dynamo_with_onnxruntime_backend.py b/test/onnx/dynamo/test_dynamo_with_onnxruntime_backend.py
index e11cebc26b8d..1d5127b18603 100644
--- a/test/onnx/dynamo/test_dynamo_with_onnxruntime_backend.py
+++ b/test/onnx/dynamo/test_dynamo_with_onnxruntime_backend.py
@@ -10,10 +10,10 @@ import unittest
 from typing import Tuple
 
 import onnxruntime
+from parameterized import parameterized
 
 import torch
 import torch._dynamo.backends.registry
-from parameterized import parameterized
 from torch import nn
 from torch.onnx import (
     _OrtBackend as OrtBackend,
diff --git a/test/onnx/dynamo/test_exporter_api.py b/test/onnx/dynamo/test_exporter_api.py
index 8d32101e0c1b..cc60b975a5eb 100644
--- a/test/onnx/dynamo/test_exporter_api.py
+++ b/test/onnx/dynamo/test_exporter_api.py
@@ -3,8 +3,9 @@ import io
 import os
 
 import onnx
-import torch
 from beartype import roar
+
+import torch
 from torch.onnx import dynamo_export, ExportOptions, ONNXProgram
 from torch.onnx._internal import exporter, io_adapter
 from torch.onnx._internal.exporter import (
diff --git a/test/onnx/dynamo/test_registry_dispatcher.py b/test/onnx/dynamo/test_registry_dispatcher.py
index 43fec1616463..087251131e64 100644
--- a/test/onnx/dynamo/test_registry_dispatcher.py
+++ b/test/onnx/dynamo/test_registry_dispatcher.py
@@ -6,12 +6,12 @@ import operator
 from typing import TypeVar, Union
 
 import onnxscript  # type: ignore[import]
-
-import torch
-import torch.fx
 from onnxscript import BFLOAT16, DOUBLE, FLOAT, FLOAT16  # type: ignore[import]
 from onnxscript.function_libs.torch_lib import ops  # type: ignore[import]
 from onnxscript.onnx_opset import opset15 as op  # type: ignore[import]
+
+import torch
+import torch.fx
 from torch.onnx._internal.diagnostics import infra
 from torch.onnx._internal.fx import (
     analysis,
diff --git a/test/onnx/error_reproduction.py b/test/onnx/error_reproduction.py
index 470bdd7c38e4..d2af1ec86bdf 100644
--- a/test/onnx/error_reproduction.py
+++ b/test/onnx/error_reproduction.py
@@ -14,6 +14,7 @@ import numpy as np
 import onnx
 import onnxruntime as ort
 import onnxscript
+
 import torch
 
 _MISMATCH_MARKDOWN_TEMPLATE = """\
diff --git a/test/onnx/onnx_test_common.py b/test/onnx/onnx_test_common.py
index 23a98dc0df4e..0cb97f47ddb3 100644
--- a/test/onnx/onnx_test_common.py
+++ b/test/onnx/onnx_test_common.py
@@ -30,6 +30,7 @@ import numpy as np
 import onnxruntime
 import pytest
 import pytorch_test_common
+
 import torch
 from torch import export as torch_export
 from torch.onnx import _constants, verification
diff --git a/test/onnx/test_autograd_funs.py b/test/onnx/test_autograd_funs.py
index a5498f39d2da..755f811f032a 100644
--- a/test/onnx/test_autograd_funs.py
+++ b/test/onnx/test_autograd_funs.py
@@ -1,9 +1,9 @@
 # Owner(s): ["module: onnx"]
 
 import pytorch_test_common
+from onnx_test_common import run_model_test
 
 import torch
-from onnx_test_common import run_model_test
 from torch.onnx import OperatorExportTypes
 from torch.onnx._globals import GLOBALS
 from torch.onnx.utils import _model_to_graph
diff --git a/test/onnx/test_custom_ops.py b/test/onnx/test_custom_ops.py
index 9ce2c9341580..dd8b0a539766 100644
--- a/test/onnx/test_custom_ops.py
+++ b/test/onnx/test_custom_ops.py
@@ -2,6 +2,7 @@
 
 import onnx_test_common
 import pytorch_test_common
+
 import torch
 import torch.utils.cpp_extension
 from torch.onnx import symbolic_helper
diff --git a/test/onnx/test_fx_op_consistency.py b/test/onnx/test_fx_op_consistency.py
index 0a1cabf19d41..4c71aafa473e 100644
--- a/test/onnx/test_fx_op_consistency.py
+++ b/test/onnx/test_fx_op_consistency.py
@@ -56,9 +56,9 @@ import onnx_test_common
 import parameterized
 import pytest
 import pytorch_test_common
+from onnx_test_common import skip, skip_slow, xfail
 
 import torch
-from onnx_test_common import skip, skip_slow, xfail
 from torch.onnx._internal.diagnostics import _rules
 from torch.testing._internal import (
     common_device_type,
diff --git a/test/onnx/test_fx_to_onnx.py b/test/onnx/test_fx_to_onnx.py
index f5de6aca086a..b660b0525dba 100644
--- a/test/onnx/test_fx_to_onnx.py
+++ b/test/onnx/test_fx_to_onnx.py
@@ -10,8 +10,9 @@ from typing import Mapping, Tuple
 import onnx
 import onnx.inliner
 import pytorch_test_common
-import torch
 import transformers  # type: ignore[import]
+
+import torch
 from torch import nn
 from torch._subclasses import fake_tensor
 from torch.nn import functional as F
diff --git a/test/onnx/test_fx_to_onnx_with_onnxruntime.py b/test/onnx/test_fx_to_onnx_with_onnxruntime.py
index 3f8ba1de2bc3..149b9dc987bb 100644
--- a/test/onnx/test_fx_to_onnx_with_onnxruntime.py
+++ b/test/onnx/test_fx_to_onnx_with_onnxruntime.py
@@ -14,10 +14,11 @@ import onnx_test_common
 import onnxruntime  # type: ignore[import]
 import parameterized  # type: ignore[import]
 import pytorch_test_common
-import torch
-import torch.onnx
 
 import transformers  # type: ignore[import]
+
+import torch
+import torch.onnx
 from torch import nn
 
 from torch._subclasses import fake_tensor
diff --git a/test/onnx/test_fx_type_promotion.py b/test/onnx/test_fx_type_promotion.py
index ce6e9a09cac6..1e3860ad2a8e 100644
--- a/test/onnx/test_fx_type_promotion.py
+++ b/test/onnx/test_fx_type_promotion.py
@@ -1,6 +1,7 @@
 # Owner(s): ["module: onnx"]
 
 import pytorch_test_common
+
 from torch.onnx._internal.fx.passes import type_promotion
 from torch.testing._internal import common_utils
 
diff --git a/test/onnx/test_models.py b/test/onnx/test_models.py
index f56d9cd9f23c..132c812f3b08 100644
--- a/test/onnx/test_models.py
+++ b/test/onnx/test_models.py
@@ -3,8 +3,6 @@
 import unittest
 
 import pytorch_test_common
-
-import torch
 from model_defs.dcgan import _netD, _netG, bsz, imgsz, nz, weights_init
 from model_defs.emb_seq import EmbeddingNetwork1, EmbeddingNetwork2
 from model_defs.mnist import MNIST
@@ -13,11 +11,6 @@ from model_defs.squeezenet import SqueezeNet
 from model_defs.srresnet import SRResNet
 from model_defs.super_resolution import SuperResolutionNet
 from pytorch_test_common import skipIfUnsupportedMinOpsetVersion, skipScriptTest
-from torch.ao import quantization
-from torch.autograd import Variable
-from torch.onnx import OperatorExportTypes
-from torch.testing._internal import common_utils
-from torch.testing._internal.common_utils import skipIfNoLapack
 from torchvision.models import shufflenet_v2_x1_0
 from torchvision.models.alexnet import alexnet
 from torchvision.models.densenet import densenet121
@@ -31,6 +24,13 @@ from torchvision.models.vgg import vgg16, vgg16_bn, vgg19, vgg19_bn
 from torchvision.models.video import mc3_18, r2plus1d_18, r3d_18
 from verify import verify
 
+import torch
+from torch.ao import quantization
+from torch.autograd import Variable
+from torch.onnx import OperatorExportTypes
+from torch.testing._internal import common_utils
+from torch.testing._internal.common_utils import skipIfNoLapack
+
 if torch.cuda.is_available():
 
     def toC(x):
diff --git a/test/onnx/test_models_onnxruntime.py b/test/onnx/test_models_onnxruntime.py
index f9afca487f52..7c51f45b19e4 100644
--- a/test/onnx/test_models_onnxruntime.py
+++ b/test/onnx/test_models_onnxruntime.py
@@ -10,12 +10,8 @@ import parameterized
 import PIL
 import pytorch_test_common
 import test_models
-
-import torch
 import torchvision
 from pytorch_test_common import skipIfUnsupportedMinOpsetVersion, skipScriptTest
-from torch import nn
-from torch.testing._internal import common_utils
 from torchvision import ops
 from torchvision.models.detection import (
     faster_rcnn,
@@ -27,6 +23,10 @@ from torchvision.models.detection import (
     transform,
 )
 
+import torch
+from torch import nn
+from torch.testing._internal import common_utils
+
 
 def exportTest(
     self,
diff --git a/test/onnx/test_models_quantized_onnxruntime.py b/test/onnx/test_models_quantized_onnxruntime.py
index 2c582e7d447e..81a180ea01fd 100644
--- a/test/onnx/test_models_quantized_onnxruntime.py
+++ b/test/onnx/test_models_quantized_onnxruntime.py
@@ -6,9 +6,9 @@ import unittest
 import onnx_test_common
 import parameterized
 import PIL
+import torchvision
 
 import torch
-import torchvision
 from torch import nn
 
 
diff --git a/test/onnx/test_onnxscript_no_runtime.py b/test/onnx/test_onnxscript_no_runtime.py
index 4389dc53aef0..fb59e1dc8889 100644
--- a/test/onnx/test_onnxscript_no_runtime.py
+++ b/test/onnx/test_onnxscript_no_runtime.py
@@ -6,8 +6,9 @@ from typing import List
 
 import onnx
 import onnxscript
-import torch
 from onnxscript.onnx_types import FLOAT
+
+import torch
 from torch.onnx._internal import jit_utils
 from torch.testing._internal import common_utils
 
diff --git a/test/onnx/test_onnxscript_runtime.py b/test/onnx/test_onnxscript_runtime.py
index 4266de85b107..1f02726d989c 100644
--- a/test/onnx/test_onnxscript_runtime.py
+++ b/test/onnx/test_onnxscript_runtime.py
@@ -5,8 +5,9 @@ from typing import List
 
 import onnx_test_common
 import onnxscript
-import torch
 from onnxscript.onnx_types import FLOAT
+
+import torch
 from torch.onnx._internal import jit_utils
 from torch.testing._internal import common_utils
 
diff --git a/test/onnx/test_op_consistency.py b/test/onnx/test_op_consistency.py
index b6f2db246884..95e67b3221e7 100644
--- a/test/onnx/test_op_consistency.py
+++ b/test/onnx/test_op_consistency.py
@@ -30,10 +30,10 @@ from typing import Optional, Tuple
 import onnx_test_common
 import parameterized
 
-import torch
-
 # For readability, these two are allowed to be imported as function
 from onnx_test_common import skip, xfail
+
+import torch
 from torch.testing._internal import (
     common_device_type,
     common_methods_invocations,
diff --git a/test/onnx/test_operators.py b/test/onnx/test_operators.py
index 2ed17771e3ce..99f0d533a61c 100644
--- a/test/onnx/test_operators.py
+++ b/test/onnx/test_operators.py
@@ -18,11 +18,6 @@ import tempfile
 # Full diff for expect files
 import unittest
 
-import torch
-import torch.nn as nn
-import torch.nn.functional as F
-import torch.onnx
-
 from pytorch_test_common import (
     BATCH_SIZE,
     flatten,
@@ -30,6 +25,11 @@ from pytorch_test_common import (
     RNN_INPUT_SIZE,
     RNN_SEQUENCE_LENGTH,
 )
+
+import torch
+import torch.nn as nn
+import torch.nn.functional as F
+import torch.onnx
 from torch.autograd import Function, Variable
 from torch.nn import functional, Module
 from torch.onnx._internal import diagnostics
diff --git a/test/onnx/test_pytorch_jit_onnx.py b/test/onnx/test_pytorch_jit_onnx.py
index 2c02df78aabb..68f26aea8b89 100644
--- a/test/onnx/test_pytorch_jit_onnx.py
+++ b/test/onnx/test_pytorch_jit_onnx.py
@@ -1,9 +1,9 @@
 # Owner(s): ["module: onnx"]
 import onnxruntime
 import pytorch_test_common
+from pytorch_test_common import skipIfNoCuda
 
 import torch
-from pytorch_test_common import skipIfNoCuda
 from torch.onnx import verification
 from torch.onnx._globals import GLOBALS
 from torch.testing._internal import common_utils
diff --git a/test/onnx/test_pytorch_onnx_onnxruntime.py b/test/onnx/test_pytorch_onnx_onnxruntime.py
index 966e34aedd86..816bcfc3b8df 100644
--- a/test/onnx/test_pytorch_onnx_onnxruntime.py
+++ b/test/onnx/test_pytorch_onnx_onnxruntime.py
@@ -15,7 +15,6 @@ import numpy as np
 import onnx
 import onnx_test_common
 import parameterized
-import torch
 import torchvision
 from model_defs import (
     lstm_flattening_result,
@@ -38,6 +37,8 @@ from pytorch_test_common import (
     skipTraceTest,
 )
 
+import torch
+
 from torch import Tensor
 from torch.nn.utils import rnn as rnn_utils
 from torch.onnx import errors, verification
diff --git a/test/onnx/test_pytorch_onnx_onnxruntime_cuda.py b/test/onnx/test_pytorch_onnx_onnxruntime_cuda.py
index 82aab068f501..399fe8a64dc1 100644
--- a/test/onnx/test_pytorch_onnx_onnxruntime_cuda.py
+++ b/test/onnx/test_pytorch_onnx_onnxruntime_cuda.py
@@ -6,8 +6,6 @@ import onnx_test_common
 
 import onnxruntime  # noqa: F401
 import parameterized
-
-import torch
 from onnx_test_common import MAX_ONNX_OPSET_VERSION, MIN_ONNX_OPSET_VERSION
 from pytorch_test_common import (
     skipIfNoBFloat16Cuda,
@@ -16,6 +14,8 @@ from pytorch_test_common import (
     skipScriptTest,
 )
 from test_pytorch_onnx_onnxruntime import _parameterized_class_attrs_and_values
+
+import torch
 from torch.cuda.amp import autocast
 from torch.testing._internal import common_utils
 
diff --git a/test/onnx/test_pytorch_onnx_shape_inference.py b/test/onnx/test_pytorch_onnx_shape_inference.py
index 41f23f3c0a59..f01b7b5af406 100644
--- a/test/onnx/test_pytorch_onnx_shape_inference.py
+++ b/test/onnx/test_pytorch_onnx_shape_inference.py
@@ -5,8 +5,9 @@ import io
 import numpy as np
 import onnx
 import pytorch_test_common
-import torch
 from pytorch_test_common import skipIfUnsupportedMinOpsetVersion
+
+import torch
 from torch.onnx import _constants, utils
 from torch.onnx._globals import GLOBALS
 from torch.onnx._internal import jit_utils
diff --git a/test/onnx/test_utility_funs.py b/test/onnx/test_utility_funs.py
index f133b5cf149e..9ee412987965 100644
--- a/test/onnx/test_utility_funs.py
+++ b/test/onnx/test_utility_funs.py
@@ -10,10 +10,6 @@ from typing import Callable
 import onnx
 import parameterized
 import pytorch_test_common
-
-import torch
-import torch.onnx
-import torch.utils.cpp_extension
 import torchvision
 from autograd_helper import CustomFunction as CustomFunction2
 from pytorch_test_common import (
@@ -21,12 +17,16 @@ from pytorch_test_common import (
     skipIfUnsupportedMaxOpsetVersion,
     skipIfUnsupportedMinOpsetVersion,
 )
+from verify import verify
+
+import torch
+import torch.onnx
+import torch.utils.cpp_extension
 from torch.onnx import _constants, OperatorExportTypes, TrainingMode, utils
 from torch.onnx._globals import GLOBALS
 from torch.onnx.symbolic_helper import _unpack_list, parse_args
 from torch.testing._internal import common_utils
 from torch.testing._internal.common_utils import skipIfNoCaffe2, skipIfNoLapack
-from verify import verify
 
 
 def _remove_test_environment_prefix_from_scope_name(scope_name: str) -> str:
diff --git a/test/onnx/test_verification.py b/test/onnx/test_verification.py
index 04d5678081b5..f416be34aedb 100644
--- a/test/onnx/test_verification.py
+++ b/test/onnx/test_verification.py
@@ -9,9 +9,9 @@ import numpy as np
 import onnx
 import parameterized
 import pytorch_test_common
+from packaging import version
 
 import torch
-from packaging import version
 from torch.onnx import _constants, _experimental, verification
 from torch.testing._internal import common_utils
 
diff --git a/test/onnx_caffe2/export_onnx_tests_generator.py b/test/onnx_caffe2/export_onnx_tests_generator.py
index e81f1277d8ee..43ea9a22a60b 100644
--- a/test/onnx_caffe2/export_onnx_tests_generator.py
+++ b/test/onnx_caffe2/export_onnx_tests_generator.py
@@ -5,10 +5,10 @@ import traceback
 
 import onnx
 import onnx_test_common
-
-import torch
 from onnx import numpy_helper
 from test_nn import new_module_tests
+
+import torch
 from torch.autograd import Variable
 from torch.testing._internal.common_nn import module_tests
 
diff --git a/test/onnx_caffe2/test_custom_ops.py b/test/onnx_caffe2/test_custom_ops.py
index 649db9f663e3..f25ae1b43a84 100644
--- a/test/onnx_caffe2/test_custom_ops.py
+++ b/test/onnx_caffe2/test_custom_ops.py
@@ -3,12 +3,12 @@
 import numpy as np
 import onnx
 import pytorch_test_common
-import torch
-import torch.utils.cpp_extension
 from test_pytorch_onnx_caffe2 import do_export
-from torch.testing._internal import common_utils
 
 import caffe2.python.onnx.backend as c2
+import torch
+import torch.utils.cpp_extension
+from torch.testing._internal import common_utils
 
 
 class TestCaffe2CustomOps(pytorch_test_common.ExportTestCase):
diff --git a/test/onnx_caffe2/test_pytorch_helper.py b/test/onnx_caffe2/test_pytorch_helper.py
index 778b8553b65c..56a4932a9996 100644
--- a/test/onnx_caffe2/test_pytorch_helper.py
+++ b/test/onnx_caffe2/test_pytorch_helper.py
@@ -5,16 +5,16 @@ import unittest
 
 import numpy as np
 import pytorch_test_common
+from pytorch_helper import PyTorchModule
 
 import torch.nn.init as init
 import torch.onnx
-from pytorch_helper import PyTorchModule
-from torch import nn
-from torch.testing._internal import common_utils
-from torch.testing._internal.common_utils import skipIfNoLapack
 
 from caffe2.python.core import workspace
 from caffe2.python.model_helper import ModelHelper
+from torch import nn
+from torch.testing._internal import common_utils
+from torch.testing._internal.common_utils import skipIfNoLapack
 
 
 class TestCaffe2Backend(pytorch_test_common.ExportTestCase):
diff --git a/test/onnx_caffe2/test_pytorch_onnx_caffe2.py b/test/onnx_caffe2/test_pytorch_onnx_caffe2.py
index dd425ee5b70f..18cba5b73f76 100644
--- a/test/onnx_caffe2/test_pytorch_onnx_caffe2.py
+++ b/test/onnx_caffe2/test_pytorch_onnx_caffe2.py
@@ -11,9 +11,6 @@ import model_defs.word_language_model as word_language_model
 import numpy as np
 import onnx
 import pytorch_test_common
-import torch.onnx
-import torch.onnx.operators
-import torch.utils.model_zoo as model_zoo
 import verify
 from debug_embed_params import run_embed_params
 from model_defs.lstm_flattening_result import LstmFlatteningResult
@@ -33,12 +30,6 @@ from pytorch_test_common import (
     skipIfUnsupportedMinOpsetVersion,
     skipIfUnsupportedOpsetVersion,
 )
-from torch import nn
-from torch.autograd import function, Variable
-from torch.nn.utils import rnn as rnn_utils
-from torch.onnx import ExportTypes
-from torch.testing._internal import common_utils
-from torch.testing._internal.common_utils import skipIfNoLapack
 
 # Import various models for testing
 from torchvision.models.alexnet import alexnet
@@ -48,10 +39,19 @@ from torchvision.models.resnet import resnet50
 from torchvision.models.vgg import vgg16, vgg16_bn, vgg19, vgg19_bn
 
 import caffe2.python.onnx.backend as c2
+import torch.onnx
+import torch.onnx.operators
+import torch.utils.model_zoo as model_zoo
 from caffe2.python.operator_test.torch_integration_test import (
     create_bbox_transform_inputs,
     generate_rois_rotated,
 )
+from torch import nn
+from torch.autograd import function, Variable
+from torch.nn.utils import rnn as rnn_utils
+from torch.onnx import ExportTypes
+from torch.testing._internal import common_utils
+from torch.testing._internal.common_utils import skipIfNoLapack
 
 skip = unittest.skip
 
diff --git a/test/onnx_caffe2/test_pytorch_onnx_caffe2_quantized.py b/test/onnx_caffe2/test_pytorch_onnx_caffe2_quantized.py
index deb46102c877..372961c000ed 100644
--- a/test/onnx_caffe2/test_pytorch_onnx_caffe2_quantized.py
+++ b/test/onnx_caffe2/test_pytorch_onnx_caffe2_quantized.py
@@ -5,13 +5,13 @@ import io
 import numpy as np
 import onnx
 import pytorch_test_common
+
+import caffe2.python.onnx.backend as c2
 import torch.ao.nn.quantized as nnq
 import torch.nn as nn
 import torch.onnx
 from torch.testing._internal import common_utils
 
-import caffe2.python.onnx.backend as c2
-
 
 class TestQuantizedOps(pytorch_test_common.ExportTestCase):
     def generic_test(
diff --git a/test/onnx_caffe2/test_verify.py b/test/onnx_caffe2/test_verify.py
index f5b3000eb9c8..05ca80840459 100644
--- a/test/onnx_caffe2/test_verify.py
+++ b/test/onnx_caffe2/test_verify.py
@@ -1,12 +1,12 @@
 # Owner(s): ["module: onnx"]
 
+from verify import verify
+
+import caffe2.python.onnx.backend as backend
 import torch
 from torch.autograd import Function
 from torch.nn import Module, Parameter
 from torch.testing._internal import common_utils
-from verify import verify
-
-import caffe2.python.onnx.backend as backend
 
 
 class TestVerify(common_utils.TestCase):
diff --git a/test/package/test_trace_dep/__init__.py b/test/package/test_trace_dep/__init__.py
index d6fb2e821677..b3b446aeab98 100644
--- a/test/package/test_trace_dep/__init__.py
+++ b/test/package/test_trace_dep/__init__.py
@@ -1,6 +1,7 @@
-import torch
 import yaml
 
+import torch
+
 
 class SumMod(torch.nn.Module):
     def forward(self, inp):
diff --git a/test/profiler/test_profiler.py b/test/profiler/test_profiler.py
index c9591a747d77..38e83d448fdd 100644
--- a/test/profiler/test_profiler.py
+++ b/test/profiler/test_profiler.py
@@ -32,6 +32,7 @@ from typing import List, Optional
 from unittest.mock import patch
 
 import expecttest
+
 import torch
 import torch.nn as nn
 import torch.optim
diff --git a/test/scripts/run_cuda_memcheck.py b/test/scripts/run_cuda_memcheck.py
index 924048030e33..790f286e58b6 100755
--- a/test/scripts/run_cuda_memcheck.py
+++ b/test/scripts/run_cuda_memcheck.py
@@ -20,9 +20,10 @@ import subprocess
 import sys
 
 import cuda_memcheck_common as cmc
-import torch
 import tqdm
 
+import torch
+
 ALL_TESTS = []
 GPUS = torch.cuda.device_count()
 
diff --git a/test/test_ao_sparsity.py b/test/test_ao_sparsity.py
index 37a13ede6d43..ac5f1c1a4e41 100644
--- a/test/test_ao_sparsity.py
+++ b/test/test_ao_sparsity.py
@@ -28,6 +28,7 @@ from ao.sparsity.test_structured_sparsifier import (  # noqa: F401  # noqa: F401
     TestFPGMPruner,
     TestSaliencyPruner,
 )
+
 from torch.testing._internal.common_utils import IS_ARM64, run_tests
 
 # Composability
diff --git a/test/test_autograd_fallback.py b/test/test_autograd_fallback.py
index ec6f01c4c2df..b0361306443a 100644
--- a/test/test_autograd_fallback.py
+++ b/test/test_autograd_fallback.py
@@ -4,6 +4,7 @@ import contextlib
 import warnings
 
 import numpy as np
+
 import torch
 from torch.library import _scoped_library, Library
 from torch.testing._internal.common_utils import (
diff --git a/test/test_binary_ufuncs.py b/test/test_binary_ufuncs.py
index 3ed43e6b06c3..faec1b69524b 100644
--- a/test/test_binary_ufuncs.py
+++ b/test/test_binary_ufuncs.py
@@ -10,6 +10,7 @@ from itertools import chain, product
 from numbers import Number
 
 import numpy as np
+
 import torch
 
 import torch.autograd.forward_ad as fwAD
diff --git a/test/test_bundled_images.py b/test/test_bundled_images.py
index 091bad66c15e..73f51d008bb1 100644
--- a/test/test_bundled_images.py
+++ b/test/test_bundled_images.py
@@ -4,6 +4,7 @@
 import io
 
 import cv2
+
 import torch
 import torch.utils.bundled_inputs
 from torch.testing._internal.common_utils import TestCase
diff --git a/test/test_compile_benchmark_util.py b/test/test_compile_benchmark_util.py
index 7d43b5727b3b..7a3aebb4e444 100644
--- a/test/test_compile_benchmark_util.py
+++ b/test/test_compile_benchmark_util.py
@@ -8,6 +8,7 @@ from torch.testing._internal.common_utils import run_tests, TEST_CUDA, TestCase
 
 try:
     import tabulate  # noqa: F401  # type: ignore[import]
+
     from torch.utils.benchmark.utils.compile import bench_all
 
     HAS_TABULATE = True
diff --git a/test/test_cpp_api_parity.py b/test/test_cpp_api_parity.py
index 3b57b40d6224..6d2cc51fcce3 100644
--- a/test/test_cpp_api_parity.py
+++ b/test/test_cpp_api_parity.py
@@ -3,9 +3,6 @@
 
 import os
 
-import torch
-import torch.testing._internal.common_nn as common_nn
-import torch.testing._internal.common_utils as common
 from cpp_api_parity import (
     functional_impl_check,
     module_impl_check,
@@ -15,6 +12,10 @@ from cpp_api_parity import (
 from cpp_api_parity.parity_table_parser import parse_parity_tracker_table
 from cpp_api_parity.utils import is_torch_nn_functional_test
 
+import torch
+import torch.testing._internal.common_nn as common_nn
+import torch.testing._internal.common_utils as common
+
 # NOTE: turn this on if you want to print source code of all C++ tests (e.g. for debugging purpose)
 PRINT_CPP_SOURCE = False
 
diff --git a/test/test_custom_ops.py b/test/test_custom_ops.py
index 98b3042e50ac..703de349ecbe 100644
--- a/test/test_custom_ops.py
+++ b/test/test_custom_ops.py
@@ -13,14 +13,14 @@ import torch._custom_ops as custom_ops
 
 import torch.testing._internal.optests as optests
 import torch.utils.cpp_extension
+
+from functorch import make_fx
 from torch import Tensor
 from torch._custom_op.impl import custom_op, CustomOp, infer_schema
 from torch._utils_internal import get_file_path_2
 from torch.testing._internal import custom_op_db
 from torch.testing._internal.common_cuda import TEST_CUDA
 from torch.testing._internal.custom_op_db import numpy_nonzero
-
-from functorch import make_fx
 from typing import *  # noqa: F403
 import numpy as np
 
diff --git a/test/test_determination.py b/test/test_determination.py
index bf3a24a64e0c..50cc2fa9975d 100644
--- a/test/test_determination.py
+++ b/test/test_determination.py
@@ -3,6 +3,7 @@
 import os
 
 import run_test
+
 from torch.testing._internal.common_utils import run_tests, TestCase
 
 
diff --git a/test/test_dynamic_shapes.py b/test/test_dynamic_shapes.py
index 4a2ba0d9867e..d89c8e1a4f36 100644
--- a/test/test_dynamic_shapes.py
+++ b/test/test_dynamic_shapes.py
@@ -9,6 +9,7 @@ import operator
 import re
 
 import sympy
+
 import torch
 import torch.fx
 import torch.nn.functional as F
@@ -1490,6 +1491,7 @@ class TestDimConstraints(TestCase):
 
         from sympy import Symbol
         from sympy.solvers.inequalities import reduce_inequalities
+
         from torch._dynamo.source import (
             LocalSource,
             TensorProperty,
@@ -1521,6 +1523,7 @@ class TestDimConstraints(TestCase):
 
     def test_dim_constraints_solve_full(self):
         from sympy import Eq, Integer, Ne, Symbol
+
         from torch._dynamo.source import (
             LocalSource,
             TensorProperty,
diff --git a/test/test_indexing.py b/test/test_indexing.py
index f34fa4c5669b..1ffac7834555 100644
--- a/test/test_indexing.py
+++ b/test/test_indexing.py
@@ -8,6 +8,7 @@ import warnings
 from functools import reduce
 
 import numpy as np
+
 import torch
 from torch import tensor
 
diff --git a/test/test_view_ops.py b/test/test_view_ops.py
index cb019057ae9f..a88b29f62911 100644
--- a/test/test_view_ops.py
+++ b/test/test_view_ops.py
@@ -6,6 +6,7 @@ from functools import partial
 from itertools import combinations, permutations, product
 
 import numpy as np
+
 import torch
 
 from torch.testing import make_tensor
diff --git a/test/torch_np/numpy_tests/core/test_dtype.py b/test/torch_np/numpy_tests/core/test_dtype.py
index 23cee32b525d..ccff28135a1f 100644
--- a/test/torch_np/numpy_tests/core/test_dtype.py
+++ b/test/torch_np/numpy_tests/core/test_dtype.py
@@ -13,6 +13,7 @@ from unittest import skipIf as skipif
 
 import pytest
 from pytest import raises as assert_raises
+
 from torch.testing._internal.common_utils import (
     instantiate_parametrized_tests,
     parametrize,
diff --git a/test/torch_np/numpy_tests/core/test_einsum.py b/test/torch_np/numpy_tests/core/test_einsum.py
index 1dd3b264fcb8..f2dc28e98c27 100644
--- a/test/torch_np/numpy_tests/core/test_einsum.py
+++ b/test/torch_np/numpy_tests/core/test_einsum.py
@@ -5,8 +5,9 @@ import itertools
 
 from unittest import expectedFailure as xfail, skipIf as skipif, SkipTest
 
-import torch._numpy as np
 from pytest import raises as assert_raises
+
+import torch._numpy as np
 from torch._numpy.testing import (
     assert_,
     assert_allclose,
diff --git a/test/torch_np/numpy_tests/core/test_getlimits.py b/test/torch_np/numpy_tests/core/test_getlimits.py
index 4574cb07c699..ab5b08319db6 100644
--- a/test/torch_np/numpy_tests/core/test_getlimits.py
+++ b/test/torch_np/numpy_tests/core/test_getlimits.py
@@ -13,6 +13,7 @@ from unittest import skipIf
 import numpy
 
 from pytest import raises as assert_raises
+
 from torch.testing._internal.common_utils import (
     run_tests,
     TEST_WITH_TORCHDYNAMO,
diff --git a/test/torch_np/numpy_tests/core/test_indexing.py b/test/torch_np/numpy_tests/core/test_indexing.py
index 77844e77a6e0..718bd7c8a498 100644
--- a/test/torch_np/numpy_tests/core/test_indexing.py
+++ b/test/torch_np/numpy_tests/core/test_indexing.py
@@ -13,6 +13,7 @@ from unittest import expectedFailure as xfail, skipIf as skipif, SkipTest
 import pytest
 
 from pytest import raises as assert_raises
+
 from torch.testing._internal.common_utils import (
     instantiate_parametrized_tests,
     parametrize,
diff --git a/test/torch_np/numpy_tests/core/test_numeric.py b/test/torch_np/numpy_tests/core/test_numeric.py
index 5b774d962db2..a6d351b9dc04 100644
--- a/test/torch_np/numpy_tests/core/test_numeric.py
+++ b/test/torch_np/numpy_tests/core/test_numeric.py
@@ -21,6 +21,7 @@ from unittest import expectedFailure as xfail, skipIf as skipif, SkipTest
 from hypothesis import given, strategies as st
 from hypothesis.extra import numpy as hynp
 from pytest import raises as assert_raises
+
 from torch.testing._internal.common_utils import (
     instantiate_parametrized_tests,
     parametrize,
diff --git a/test/torch_np/numpy_tests/core/test_numerictypes.py b/test/torch_np/numpy_tests/core/test_numerictypes.py
index 462c51873b66..784f4b887076 100644
--- a/test/torch_np/numpy_tests/core/test_numerictypes.py
+++ b/test/torch_np/numpy_tests/core/test_numerictypes.py
@@ -7,6 +7,7 @@ import sys
 from unittest import skipIf as skipif
 
 from pytest import raises as assert_raises
+
 from torch.testing._internal.common_utils import (
     instantiate_parametrized_tests,
     parametrize,
diff --git a/test/torch_np/numpy_tests/core/test_scalar_methods.py b/test/torch_np/numpy_tests/core/test_scalar_methods.py
index 2e763c6636a8..35bec88b382b 100644
--- a/test/torch_np/numpy_tests/core/test_scalar_methods.py
+++ b/test/torch_np/numpy_tests/core/test_scalar_methods.py
@@ -14,6 +14,7 @@ from unittest import skipIf as skipif, SkipTest
 import pytest
 
 from pytest import raises as assert_raises
+
 from torch.testing._internal.common_utils import (
     instantiate_parametrized_tests,
     parametrize,
diff --git a/test/torch_np/numpy_tests/lib/test_type_check.py b/test/torch_np/numpy_tests/lib/test_type_check.py
index 96c0ddbc9672..9df29f5a7172 100644
--- a/test/torch_np/numpy_tests/lib/test_type_check.py
+++ b/test/torch_np/numpy_tests/lib/test_type_check.py
@@ -6,6 +6,7 @@ import functools
 from unittest import expectedFailure as xfail, skipIf as skipif
 
 from pytest import raises as assert_raises
+
 from torch.testing._internal.common_utils import (
     run_tests,
     TEST_WITH_TORCHDYNAMO,
diff --git a/test/torch_np/numpy_tests/linalg/test_linalg.py b/test/torch_np/numpy_tests/linalg/test_linalg.py
index 3a5c21745e24..87f30f23ac86 100644
--- a/test/torch_np/numpy_tests/linalg/test_linalg.py
+++ b/test/torch_np/numpy_tests/linalg/test_linalg.py
@@ -19,6 +19,7 @@ import pytest
 
 from numpy.linalg.linalg import _multi_dot_matrix_chain_order
 from pytest import raises as assert_raises
+
 from torch.testing._internal.common_utils import (
     instantiate_parametrized_tests,
     parametrize,
diff --git a/test/torch_np/test_basic.py b/test/torch_np/test_basic.py
index 790b40349ce0..f0603e7feb48 100644
--- a/test/torch_np/test_basic.py
+++ b/test/torch_np/test_basic.py
@@ -5,12 +5,13 @@ import inspect
 from unittest import expectedFailure as xfail, skipIf as skip
 
 import numpy as _np
+from pytest import raises as assert_raises
+
 import torch
 
 import torch._numpy as w
 import torch._numpy._ufuncs as _ufuncs
 import torch._numpy._util as _util
-from pytest import raises as assert_raises
 from torch._numpy.testing import assert_allclose, assert_equal
 
 from torch.testing._internal.common_cuda import TEST_CUDA
diff --git a/test/torch_np/test_ufuncs_basic.py b/test/torch_np/test_ufuncs_basic.py
index 274a3d75618e..6acacf6400b7 100644
--- a/test/torch_np/test_ufuncs_basic.py
+++ b/test/torch_np/test_ufuncs_basic.py
@@ -14,6 +14,7 @@ import operator
 from unittest import skipIf as skip, SkipTest
 
 from pytest import raises as assert_raises
+
 from torch.testing._internal.common_utils import (
     instantiate_parametrized_tests,
     parametrize,
diff --git a/torch/ao/quantization/experimental/adaround_loss.py b/torch/ao/quantization/experimental/adaround_loss.py
index 8080d72cc6da..6033d38e507d 100644
--- a/torch/ao/quantization/experimental/adaround_loss.py
+++ b/torch/ao/quantization/experimental/adaround_loss.py
@@ -1,6 +1,7 @@
 from typing import Tuple
 
 import numpy as np
+
 import torch
 from torch.nn import functional as F
 
diff --git a/torch/csrc/lazy/test_mnist.py b/torch/csrc/lazy/test_mnist.py
index 20b94437239b..a3a03d9844d3 100644
--- a/torch/csrc/lazy/test_mnist.py
+++ b/torch/csrc/lazy/test_mnist.py
@@ -2,6 +2,8 @@
 
 import os
 
+from torchvision import datasets, transforms
+
 import torch
 import torch._lazy
 import torch._lazy.metrics
@@ -10,7 +12,6 @@ import torch.nn as nn
 import torch.nn.functional as F
 import torch.optim as optim
 from torch.optim.lr_scheduler import StepLR
-from torchvision import datasets, transforms
 
 torch._lazy.ts_backend.init()
 
diff --git a/torch/fx/experimental/shape_inference/infer_symbol_values.py b/torch/fx/experimental/shape_inference/infer_symbol_values.py
index 81c86fe38198..fcdbb0a4aab9 100644
--- a/torch/fx/experimental/shape_inference/infer_symbol_values.py
+++ b/torch/fx/experimental/shape_inference/infer_symbol_values.py
@@ -4,6 +4,7 @@ from typing import Any, DefaultDict, Dict, List, Tuple, Union
 import numpy as np
 
 import sympy as sp
+
 import torch
 
 square_brackets_pattern = r"\[([^]]+)\]"