Remove unnecessary noqa suppressions (#164106)

This PR removes unused `noqa` suppressions in Python code.

Pull Request resolved: https://github.com/pytorch/pytorch/pull/164106
Approved by: https://github.com/albanD

torch/_inductor/fuzzer.py

@@ -310,7 +310,7 @@ class SamplingMethod(Enum):
                 )
             try:
                 new_default = new_type()
-            except Exception:  # noqa: E722
+            except Exception:
                 # if default constructor doesn't work, try None
                 new_default = None
 
@@ -779,7 +779,7 @@ class ConfigFuzzer:
         test_model_fn = self.test_model_fn_factory()
         try:
             test_model_fn()
-        except Exception as exc:  # noqa: E722
+        except Exception as exc:
             return handle_return(
                 "Eager exception", Status.FAILED_RUN_EAGER_EXCEPTION, True, exc
             )
@@ -788,7 +788,7 @@ class ConfigFuzzer:
         try:
             test_model_fn2 = self.test_model_fn_factory()
             comp = torch.compile(test_model_fn2, backend="inductor")
-        except Exception as exc:  # noqa: E722
+        except Exception as exc:
             return handle_return(
                 "Exception compiling", Status.FAILED_COMPILE, True, exc
             )
@@ -796,7 +796,7 @@ class ConfigFuzzer:
         # try running compiled
         try:
             compile_result = comp()
-        except Exception as exc:  # noqa: E722
+        except Exception as exc:
             return handle_return(
                 "Exception running compiled",
                 Status.FAILED_RUN_COMPILE_EXCEPTION,

torch/_logging/_internal.py

@@ -699,7 +699,7 @@ Examples:
 
   TORCH_LOGS_OUT=/tmp/output.txt will output the logs to /tmp/output.txt as
   well. This is useful when the output is long.
-"""  # flake8: noqa: B950
+"""
     msg = f"""
 TORCH_LOGS Info
 {examples}

torch/nn/intrinsic/qat/modules/conv_fused.py

@@ -1,4 +1,3 @@
-# flake8: noqa: F401
 r"""Intrinsic QAT Modules.
 
 This file is in the process of migration to `torch/ao/nn/intrinsic/qat`, and

torch/nn/intrinsic/qat/modules/linear_fused.py

@@ -1,4 +1,3 @@
-# flake8: noqa: F401
 r"""Intrinsic QAT Modules.
 
 This file is in the process of migration to `torch/ao/nn/intrinsic/qat`, and

torch/nn/intrinsic/qat/modules/linear_relu.py

@@ -1,4 +1,3 @@
-# flake8: noqa: F401
 r"""Intrinsic QAT Modules.
 
 This file is in the process of migration to `torch/ao/nn/intrinsic/qat`, and

torch/nn/modules/pooling.py

@@ -298,7 +298,7 @@ class MaxPool3d(_MaxPoolNd):
 
     .. _link:
         https://github.com/vdumoulin/conv_arithmetic/blob/master/README.md
-    """  # noqa: E501
+    """
 
     kernel_size: _size_3_t
     stride: _size_3_t

torch/nn/parallel/distributed.py

@@ -775,7 +775,7 @@ class DistributedDataParallel(Module, Joinable):
                     "DistributedDataParallel device_ids and output_device arguments "
                     "only work with single-device/multiple-device GPU modules or CPU modules, "
                     f"but got device_ids {device_ids}, output_device {output_device}, "
-                    f"and module parameters { ({p.device for p in self._module_parameters}) }.",  # noqa: E201,E202
+                    f"and module parameters { ({p.device for p in self._module_parameters}) }.",
                 )
 
             self.device_ids = None

torch/nn/utils/_named_member_accessor.py

@@ -146,7 +146,7 @@ class NamedMemberAccessor:
                     f"{module._get_name()} has no attribute `{attr}`"
                 ) from ex
             if not isinstance(submodule, torch.nn.Module):
-                raise TypeError(  # noqa: B904
+                raise TypeError(
                     f"submodule `{name}`: {submodule} is not an instance of torch.nn.Module"
                 )
             self.memo[name] = submodule

torch/utils/_triton.py

@@ -16,7 +16,7 @@ def has_triton_package() -> bool:
 @functools.cache
 def get_triton_version(fallback: tuple[int, int] = (0, 0)) -> tuple[int, int]:
     try:
-        import triton  # noqa: F401
+        import triton
 
         major, minor = tuple(int(v) for v in triton.__version__.split(".")[:2])
         return (major, minor)