diff --git a/.flake8 b/.flake8
index 2d854aa6b25..ea8524f372f 100644
--- a/.flake8
+++ b/.flake8
@@ -12,7 +12,7 @@ ignore =
     # to line this up with executable bit
     EXE001,
     # these ignores are from flake8-bugbear; please fix!
-    B007,B008,B017,B019,B023,B028,B903,B904,B905,B906,B907
+    B007,B008,B017,B019,B023,B028,B903,B904,B905,B906,B907,B908,B910
     # these ignores are from flake8-comprehensions; please fix!
     C407,
     # these ignores are from flake8-logging-format; please fix!
diff --git a/benchmarks/dynamo/genai_layers/utils.py b/benchmarks/dynamo/genai_layers/utils.py
index e11995ee0b5..749b9cea203 100644
--- a/benchmarks/dynamo/genai_layers/utils.py
+++ b/benchmarks/dynamo/genai_layers/utils.py
@@ -134,7 +134,7 @@ class BenchmarkKernel:
                 print(
                     f"Failed to run {backend} backend on {self.name} kernel for {setting} due to {e}"
                 )
-                self.available_backends.remove(backend)
+                self.available_backends.remove(backend)  # noqa: B909
                 continue
             mem_bytes = self.get_memory_bytes(args_ref, kwargs_ref)
             perf = Performance(setting, avg_time, mem_bytes)
diff --git a/docs/source/conf.py b/docs/source/conf.py
index 8b2112c165e..f49968b932d 100644
--- a/docs/source/conf.py
+++ b/docs/source/conf.py
@@ -181,7 +181,6 @@ html_theme_options = {
 
 theme_variables = pytorch_sphinx_theme2.get_theme_variables()
 html_context = {
-    "theme_variables": theme_variables,
     "github_url": "https://github.com",
     "github_user": "pytorch",
     "github_repo": "pytorch",
@@ -189,7 +188,7 @@ html_context = {
     "github_version": "main",
     "pytorch_project": "docs",
     "doc_path": "docs/source",
-    "theme_variables": theme_variables,  # noqa: F601
+    "theme_variables": theme_variables,
     # library links are defined in
     # pytorch_sphinx_theme2/pytorch_sphinx_theme2/links.json
     "library_links": theme_variables.get("library_links", []),
diff --git a/test/distributed/optim/test_zero_redundancy_optimizer.py b/test/distributed/optim/test_zero_redundancy_optimizer.py
index 603f671546a..5c90ad8be14 100644
--- a/test/distributed/optim/test_zero_redundancy_optimizer.py
+++ b/test/distributed/optim/test_zero_redundancy_optimizer.py
@@ -1165,27 +1165,23 @@ class TestZeroRedundancyOptimizerDistributed(TestZeroRedundancyOptimizer):
 
                 # Increased tolerances are needed to pass when using TF32
                 # See: https://github.com/pytorch/pytorch/issues/67764
-                (
-                    torch.testing.assert_close(
-                        local_loss.cpu(),
-                        ddp_loss.cpu(),
-                        rtol=1e-03,
-                        atol=1e-08,
-                    ),
-                    "Losses differ between local optimizer and ZeRO",
+                torch.testing.assert_close(
+                    local_loss.cpu(),
+                    ddp_loss.cpu(),
+                    rtol=1e-03,
+                    atol=1e-08,
+                    msg="Losses differ between local optimizer and ZeRO",
                 )
 
                 for local_p, ddp_p in zip(
                     local_model.parameters(), ddp_model.parameters()
                 ):
-                    (
-                        torch.testing.assert_close(
-                            local_p.cpu(),
-                            ddp_p.cpu(),
-                            rtol=1e-03,
-                            atol=1e-04,
-                        ),
-                        "Models differ after a step",
+                    torch.testing.assert_close(
+                        local_p.cpu(),
+                        ddp_p.cpu(),
+                        rtol=1e-03,
+                        atol=1e-04,
+                        msg="Models differ after a step",
                     )
 
     @skipIfHpu
diff --git a/test/distributions/test_distributions.py b/test/distributions/test_distributions.py
index 5fd8f71cecd..7cb8cc67813 100644
--- a/test/distributions/test_distributions.py
+++ b/test/distributions/test_distributions.py
@@ -6369,7 +6369,7 @@ class TestLazyLogitsInitialization(DistributionsTestCase):
             except NotImplementedError:
                 pass
             self.assertNotIn("probs", dist.__dict__, msg=message)
-            dist.batch_shape, dist.event_shape
+            _ = (dist.batch_shape, dist.event_shape)
             self.assertNotIn("probs", dist.__dict__, msg=message)
 
     def test_lazy_probs_initialization(self):
@@ -6386,7 +6386,7 @@ class TestLazyLogitsInitialization(DistributionsTestCase):
             except NotImplementedError:
                 pass
             self.assertNotIn("logits", dist.__dict__, msg=message)
-            dist.batch_shape, dist.event_shape
+            _ = (dist.batch_shape, dist.event_shape)
             self.assertNotIn("logits", dist.__dict__, msg=message)
 
 
diff --git a/test/inductor/test_ordered_set.py b/test/inductor/test_ordered_set.py
index cbbb9bda56c..216b8ab0f02 100644
--- a/test/inductor/test_ordered_set.py
+++ b/test/inductor/test_ordered_set.py
@@ -991,7 +991,7 @@ class TestExceptionPropagation(TestCase):
         s = OrderedSet([1, 2, 3])
         try:
             for i in s:
-                s.update([4])
+                s.update([4])  # noqa: B909
         except RuntimeError:
             pass
         else:
diff --git a/test/inductor/test_torchinductor.py b/test/inductor/test_torchinductor.py
index ffffc68939d..329785792f3 100644
--- a/test/inductor/test_torchinductor.py
+++ b/test/inductor/test_torchinductor.py
@@ -8863,7 +8863,7 @@ def forward(self, arg0_1: "Sym(s77)", arg1_1: "Sym(s27)", arg2_1: "Sym(s53)", ar
         with torch.library._scoped_library("mylib", "FRAGMENT") as m:
 
             def impl(a, b, c, d, e=2):
-                (a.add_(b[0] * c * e),)
+                a.add_(b[0] * c * e)
                 if d is not None:
                     d.add_(b[1])
 
@@ -8936,7 +8936,7 @@ def forward(self, arg0_1: "Sym(s77)", arg1_1: "Sym(s27)", arg2_1: "Sym(s53)", ar
         with torch.library._scoped_library("mylib", "FRAGMENT") as m:
 
             def impl(a, b, c, d, e=2):
-                (a.add_(b[0] * c * e),)
+                a.add_(b[0] * c * e)
                 if d is not None:
                     d.add_(b[1])
                 return b[0] + b[1]
diff --git a/test/onnx/torchlib/ops_test_common.py b/test/onnx/torchlib/ops_test_common.py
index 884b66d4e02..54ecbdc1951 100644
--- a/test/onnx/torchlib/ops_test_common.py
+++ b/test/onnx/torchlib/ops_test_common.py
@@ -246,7 +246,7 @@ def duplicate_opinfo_for_prims(
             new_opinfo = copy.deepcopy(opinfo)
             new_opinfo.name = new_name
             new_opinfo.op = getattr(torch.ops.prims, prims_name)
-            opinfos.append(new_opinfo)
+            opinfos.append(new_opinfo)  # noqa: B909
             return
     raise RuntimeError(f"OpInfo '{name}' not found in the database.")
 
diff --git a/test/profiler/test_execution_trace.py b/test/profiler/test_execution_trace.py
index 267e36ef8f2..74ab3cf6bc3 100644
--- a/test/profiler/test_execution_trace.py
+++ b/test/profiler/test_execution_trace.py
@@ -1,18 +1,5 @@
 # Owner(s): ["oncall: profiler"]
 
-# if tqdm is not shutdown properly, it will leave the monitor thread alive.
-# This causes an issue in the multithreading test because we check all events
-# in that test with their tids. The events that correspond to these lingering
-# threads all have TID of (uint64_t)(-1) which is invalid.
-# The work around is turnning off monitoring thread when tqdm is loaded.
-# Since these are unit tests, it is safe to turn off monitor thread.
-try:
-    import tqdm
-
-    tqdm.tqdm.monitor_interval = 0
-except ImportError:
-    pass
-
 import json
 import os
 import tempfile
@@ -52,6 +39,19 @@ from torch.testing._internal.common_utils import (
 from torch.utils._triton import has_triton
 
 
+# if tqdm is not shutdown properly, it will leave the monitor thread alive.
+# This causes an issue in the multithreading test because we check all events
+# in that test with their tids. The events that correspond to these lingering
+# threads all have TID of (uint64_t)(-1) which is invalid.
+# The work around is turnning off monitoring thread when tqdm is loaded.
+# Since these are unit tests, it is safe to turn off monitor thread.
+try:
+    import tqdm
+
+    tqdm.tqdm.monitor_interval = 0
+except ImportError:
+    pass
+
 Json = dict[str, Any]
 
 
diff --git a/test/profiler/test_record_function.py b/test/profiler/test_record_function.py
index 4bc4ad16acb..9ab80c9a07a 100644
--- a/test/profiler/test_record_function.py
+++ b/test/profiler/test_record_function.py
@@ -1,19 +1,6 @@
 # Owner(s): ["oncall: profiler"]
 # ruff: noqa: F841
 
-# if tqdm is not shutdown properly, it will leave the monitor thread alive.
-# This causes an issue in the multithreading test because we check all events
-# in that test with their tids. The events that correspond to these lingering
-# threads all have TID of (uint64_t)(-1) which is invalid.
-# The work around is turnning off monitoring thread when tqdm is loaded.
-# Since these are unit tests, it is safe to turn off monitor thread.
-try:
-    import tqdm
-
-    tqdm.tqdm.monitor_interval = 0
-except ImportError:
-    None
-
 from typing import Any
 
 import torch
@@ -29,6 +16,19 @@ from torch.profiler import kineto_available, record_function
 from torch.testing._internal.common_utils import run_tests, TestCase
 
 
+# if tqdm is not shutdown properly, it will leave the monitor thread alive.
+# This causes an issue in the multithreading test because we check all events
+# in that test with their tids. The events that correspond to these lingering
+# threads all have TID of (uint64_t)(-1) which is invalid.
+# The work around is turnning off monitoring thread when tqdm is loaded.
+# Since these are unit tests, it is safe to turn off monitor thread.
+try:
+    import tqdm
+
+    tqdm.tqdm.monitor_interval = 0
+except ImportError:
+    pass
+
 Json = dict[str, Any]
 
 
diff --git a/test/profiler/test_torch_tidy.py b/test/profiler/test_torch_tidy.py
index e10fe8d0d71..efbd4b8189d 100644
--- a/test/profiler/test_torch_tidy.py
+++ b/test/profiler/test_torch_tidy.py
@@ -1,18 +1,5 @@
 # Owner(s): ["oncall: profiler"]
 
-# if tqdm is not shutdown properly, it will leave the monitor thread alive.
-# This causes an issue in the multithreading test because we check all events
-# in that test with their tids. The events that correspond to these lingering
-# threads all have TID of (uint64_t)(-1) which is invalid.
-# The work around is turnning off monitoring thread when tqdm is loaded.
-# Since these are unit tests, it is safe to turn off monitor thread.
-try:
-    import tqdm
-
-    tqdm.tqdm.monitor_interval = 0
-except ImportError:
-    None
-
 import gc
 import re
 import textwrap
@@ -24,14 +11,25 @@ import torch
 import torch.nn as nn
 import torch.optim
 import torch.utils.data
-from torch._C._profiler import _TensorMetadata
+from torch._C._profiler import _ExtraFields_PyCall, _TensorMetadata
 from torch.profiler import _utils, profile
 from torch.testing._internal.common_utils import run_tests, TestCase
 
 
-Json = dict[str, Any]
+# if tqdm is not shutdown properly, it will leave the monitor thread alive.
+# This causes an issue in the multithreading test because we check all events
+# in that test with their tids. The events that correspond to these lingering
+# threads all have TID of (uint64_t)(-1) which is invalid.
+# The work around is turnning off monitoring thread when tqdm is loaded.
+# Since these are unit tests, it is safe to turn off monitor thread.
+try:
+    import tqdm
 
-from torch._C._profiler import _ExtraFields_PyCall
+    tqdm.tqdm.monitor_interval = 0
+except ImportError:
+    pass
+
+Json = dict[str, Any]
 
 
 def find_node_with_name(nodes, name):
diff --git a/test/quantization/fx/test_quantize_fx.py b/test/quantization/fx/test_quantize_fx.py
index 359767f3997..e38c56da2a7 100644
--- a/test/quantization/fx/test_quantize_fx.py
+++ b/test/quantization/fx/test_quantize_fx.py
@@ -827,7 +827,7 @@ class TestFuseFx(QuantizationTestCase):
         named_modules = dict(m.named_modules())
         for node in m.graph.nodes:
             if node.op == "call_module" and type(named_modules[node.target]) == torch.nn.Conv2d:
-                self.assertTrue(len(node.args) == 2), "Expecting the fused op to have two arguments"
+                self.assertTrue(len(node.args) == 2, msg="Expecting the fused op to have two arguments")
 
     def test_fusion_pattern_with_matchallnode(self):
         """This test tests that the node matched by MatchAllNode will be regared as an input
diff --git a/test/test_linalg.py b/test/test_linalg.py
index f49db43b4ff..cd58ee57754 100644
--- a/test/test_linalg.py
+++ b/test/test_linalg.py
@@ -8097,7 +8097,7 @@ scipy_lobpcg  | {eq_err_scipy:10.2e}  | {eq_err_general_scipy:10.2e}  | {iters2:
         with self.assertWarnsOnceRegex(
                 UserWarning, f"This overload of {func}_ is deprecated"):
             getattr(out_tensor, func + "_")(1, b1, b2)
-        self.assertEqual(out_tensor, ref * 2),
+        self.assertEqual(out_tensor, ref * 2)
         getattr(res3, func + "_")(b1, b2, beta=1)
         self.assertEqual(out_tensor, res3)
 
@@ -8113,7 +8113,7 @@ scipy_lobpcg  | {eq_err_scipy:10.2e}  | {eq_err_general_scipy:10.2e}  | {iters2:
             self.assertEqual(out_tensor, getattr(torch, func)(1, out_tensor, 0, b1, b2))
 
         res4 = getattr(torch, func)(out_tensor, b1, b2, beta=1, alpha=.5)
-        self.assertEqual(res4, ref * 3),
+        self.assertEqual(res4, ref * 3)
 
         nan = torch.full_like(out_tensor, math.nan)
         res5 = getattr(torch, func)(nan, b1, b2, beta=0, alpha=1)
diff --git a/test/test_multiprocessing_spawn.py b/test/test_multiprocessing_spawn.py
index 47c429ba3f9..25eea49b755 100644
--- a/test/test_multiprocessing_spawn.py
+++ b/test/test_multiprocessing_spawn.py
@@ -201,7 +201,7 @@ class _TestMultiProcessing:
                 try:
                     os.kill(pid, 0)
                 except ProcessLookupError:
-                    pids.remove(pid)
+                    pids.remove(pid)  # noqa: B909
                     break
 
             # This assert fails if any nested child process is still
diff --git a/test/test_sparse_semi_structured.py b/test/test_sparse_semi_structured.py
index 5078649bb00..51fb4aa48c2 100644
--- a/test/test_sparse_semi_structured.py
+++ b/test/test_sparse_semi_structured.py
@@ -714,16 +714,18 @@ class TestSparseSemiStructuredTraining(TestCase):
         max_diff = (ref_gemm - pack_gemm).abs().argmax()
         torch.testing.assert_close(
             ref_gemm, pack_gemm,
-            **atol_rtol_kw[dtype]
-        ), f"packed is wrong at pos: ({max_diff // N}, {max_diff % N})"
+            **atol_rtol_kw[dtype],
+            msg=f"packed is wrong at pos: ({max_diff // N}, {max_diff % N})",
+        )
         # Test A.t@B
         pack_gemm = torch._sparse_semi_structured_linear(b.t(), packed_t, meta_t)
         max_diff = (ref_gemm - pack_gemm).abs().argmax()
 
         torch.testing.assert_close(
             ref_gemm, pack_gemm,
-            **atol_rtol_kw[dtype]
-        ), f"packed_t is wrong at pos: ({max_diff // N}, {max_diff % N})"
+            **atol_rtol_kw[dtype],
+            msg=f"packed_t is wrong at pos: ({max_diff // N}, {max_diff % N})",
+        )
 
     @training_dtypes
     @unittest.skipIf(TEST_WITH_ROCM, "Not supported on ROCm")
diff --git a/torch/_dynamo/variables/ctx_manager.py b/torch/_dynamo/variables/ctx_manager.py
index ec5c881c415..09c9f4b7b72 100644
--- a/torch/_dynamo/variables/ctx_manager.py
+++ b/torch/_dynamo/variables/ctx_manager.py
@@ -807,10 +807,8 @@ class DeterministicAlgorithmsVariable(ContextWrappingVariable):
     def _call_func(self, tx: "InstructionTranslator", values):
         assert len(values) == 1
         value = values[0]
-        (
-            tx.output.create_node(
-                "call_function", torch._C._set_deterministic_algorithms, (value,), {}
-            ),
+        tx.output.create_node(
+            "call_function", torch._C._set_deterministic_algorithms, (value,), {}
         )
         torch._C._set_deterministic_algorithms(value)
 
diff --git a/torch/_inductor/pattern_matcher.py b/torch/_inductor/pattern_matcher.py
index fd10e643c18..67e6773cf33 100644
--- a/torch/_inductor/pattern_matcher.py
+++ b/torch/_inductor/pattern_matcher.py
@@ -545,7 +545,7 @@ class _TargetExpr(PatternExpr):
         fns = [fns] if callable(fns) or isinstance(fns, str) else list(fns)
         for fn in fns:
             if isinstance(fn, torch._ops.OpOverloadPacket):
-                fns.extend(getattr(fn, overload) for overload in fn.overloads())
+                fns.extend(getattr(fn, overload) for overload in fn.overloads())  # noqa: B909
 
         self.fns = fns
         self.fns_set = OrderedSet(fns)
diff --git a/torch/_inductor/runtime/triton_heuristics.py b/torch/_inductor/runtime/triton_heuristics.py
index 5195770dc24..196dc329e80 100644
--- a/torch/_inductor/runtime/triton_heuristics.py
+++ b/torch/_inductor/runtime/triton_heuristics.py
@@ -574,7 +574,7 @@ class CachingAutotuner(KernelInterface):
                     assert hasattr(self, "_reload_kernel")
                     assert callable(self._reload_kernel)
                     self.fn = self._reload_kernel().fn
-                self.compile_results.append(self._precompile_config(new_config))
+                self.compile_results.append(self._precompile_config(new_config))  # noqa: B909
 
             self._make_launchers()
 
diff --git a/torch/_inductor/scheduler.py b/torch/_inductor/scheduler.py
index 5d4ce3e4f58..92a404e7020 100644
--- a/torch/_inductor/scheduler.py
+++ b/torch/_inductor/scheduler.py
@@ -3937,7 +3937,7 @@ class Scheduler:
             if remaining:
                 for rd in remaining:
                     if self.fusable_read_and_write(rd, cd):
-                        remaining.remove(rd)
+                        remaining.remove(rd)  # noqa: B909
 
         remaining_deps = OrderedSet(
             dep.name
diff --git a/torch/distributed/checkpoint/default_planner.py b/torch/distributed/checkpoint/default_planner.py
index baae0b2bd49..66cdff8a6b7 100644
--- a/torch/distributed/checkpoint/default_planner.py
+++ b/torch/distributed/checkpoint/default_planner.py
@@ -408,7 +408,7 @@ class _EmptyStateDictLoadPlanner(DefaultLoadPlanner):
             return True
 
         if key in self.keys:
-            True
+            return True
 
         unflattened_keys: list[str] = []
         planner_data = metadata.planner_data.get(key)
diff --git a/torch/fx/experimental/shape_inference/infer_symbol_values.py b/torch/fx/experimental/shape_inference/infer_symbol_values.py
index d7ff154c16a..1ee6d2a939a 100644
--- a/torch/fx/experimental/shape_inference/infer_symbol_values.py
+++ b/torch/fx/experimental/shape_inference/infer_symbol_values.py
@@ -64,7 +64,7 @@ def infer_symbol_values(
         for right_var in right_vars:
             if sp.sympify(right_var) == sp.sympify("s0"):
                 right_equation = sp.cancel(right_equation / right_var)
-                right_vars.remove(right_var)
+                right_vars.remove(right_var)  # noqa: B909
 
         var = right_vars[0]
         idx = symbol_idx_dict[str(var)]
diff --git a/torch/fx/passes/split_module.py b/torch/fx/passes/split_module.py
index 413584070d1..56ab34d3a87 100644
--- a/torch/fx/passes/split_module.py
+++ b/torch/fx/passes/split_module.py
@@ -394,7 +394,7 @@ def split_module(
         root_partition = root_partitions.pop()
         sorted_partitions.append(root_partition)
         for dependent in partitions[root_partition].dependents:
-            partitions[dependent].dependencies.pop(root_partition)
+            partitions[dependent].dependencies.pop(root_partition)  # noqa: B909
             if not partitions[dependent].dependencies:
                 root_partitions.append(dependent)
     if len(sorted_partitions) != len(partitions):
diff --git a/torch/testing/_internal/common_methods_invocations.py b/torch/testing/_internal/common_methods_invocations.py
index f80da003227..41bb2b96bd9 100644
--- a/torch/testing/_internal/common_methods_invocations.py
+++ b/torch/testing/_internal/common_methods_invocations.py
@@ -11847,8 +11847,11 @@ op_db: list[OpInfo] = [
                        safe_val=2)),
     BinaryUfuncInfo('add',
                     # NumPy has no builtin reference for the alpha kwarg, but it is easy enough to emulate
-                    ref=lambda input, other, *, alpha=1: np.add(input, other) if alpha == 1 \
-                    else np.add(input, np.multiply(alpha, other)),
+                    ref=lambda input, other, *, alpha=1: (
+                        np.add(input, other)
+                        if alpha == 1
+                        else np.add(input, np.multiply(alpha, other))
+                    ),
                     dtypes=all_types_and_complex_and(torch.bool, torch.bfloat16,
                                                      torch.float16, torch.chalf),
                     dtypesIfHpu=custom_types(torch.float32, torch.bfloat16, torch.int32),
@@ -20498,8 +20501,11 @@ op_db: list[OpInfo] = [
         'jiterator_binary',
         op=torch.cuda.jiterator._create_jit_fn(
             "template <typename T> T binary(T x, T y, T alpha) { return x + alpha * y; }", alpha=1),
-        ref=lambda input, other, *, alpha=1: np.add(input, other) if alpha == 1 \
-            else np.add(input, np.multiply(alpha, other)),
+        ref=lambda input, other, *, alpha=1: (
+            np.add(input, other)
+            if alpha == 1
+            else np.add(input, np.multiply(alpha, other))
+        ),
         dtypes=all_types_and_complex_and(torch.bfloat16, torch.float16, torch.bool),
         sample_inputs_func=partial(sample_inputs_jiterator, num_inputs=2, alpha=-3.14),
         supports_out=False,
diff --git a/torch/utils/viz/_cycles.py b/torch/utils/viz/_cycles.py
index a5aebde06a3..79d8e8b8b17 100644
--- a/torch/utils/viz/_cycles.py
+++ b/torch/utils/viz/_cycles.py
@@ -336,7 +336,7 @@ def cuda_allocation_context():
 def to_dot(nodes):
     lines = ["digraph GraphName {", "node [shape=rect];", 'rankdir=LR;']
     for i, n in enumerate(nodes):
-        lines.append(f'{i} [label={escape(n.label)}, color={ "red" if n.root else "black"}];')
+        lines.append(f'{i} [label={escape(n.label)}, color={"red" if n.root else "black"}];')
 
     for i, f in enumerate(nodes):
         for label, j in f.referrents: