remove F401 from ignore rules -- unused imports (Lightning-AI#2272)

crcrpar · pre-commit-ci[bot] · web-flow · commit 20efc0d06ecb · 2025-09-04T12:00:20.000+02:00
Signed-off-by: Masaki Kozuki &lt;mkozuki@nvidia.com&gt;
Co-authored-by: pre-commit-ci[bot] &lt;66853113+pre-commit-ci[bot]@users.noreply.github.com&gt;
diff --git a/pyproject.toml b/pyproject.toml
@@ -157,7 +157,6 @@ ignore = [
     # TODO(crcrpar): Resolves the following ignores as these are added while enabling ruff check in pre-commit
     "F821",  # https://docs.astral.sh/ruff/rules/undefined-name/
     "E402",  # https://docs.astral.sh/ruff/rules/module-import-not-at-top-of-file/
-    "F401",  # https://docs.astral.sh/ruff/rules/unused-import/
     "F405",  # https://docs.astral.sh/ruff/rules/undefined-local-with-import-star-usage/
     "E712",  # https://docs.astral.sh/ruff/rules/true-false-comparison/
     "E721",  # https://docs.astral.sh/ruff/rules/type-comparison/
diff --git a/thunder/__init__.py b/thunder/__init__.py
@@ -78,7 +78,26 @@
 
 # TODO RC1 Review exposed names
 __all__ = [
+    # module aliases
+    "clang",
+    "dtypes",
+    "devices",
     "transforms",
+    # function aliases
+    "get_compile_data",
+    "trace",
+    # class aliases
+    "Proxy",
+    "TensorProxy",
+    "NumberProxy",
+    "StringProxy",
+    "IntegerProxy",
+    "FloatProxy",
+    "ComplexProxy",
+    "TupleProxy",
+    "ListProxy",
+    "DictProxy",
+    "AnyProxy",
     # dtype aliases
     "bool8",
     "uint8",
diff --git a/thunder/benchmarks/benchmark_peft.py b/thunder/benchmarks/benchmark_peft.py
@@ -1,4 +1,3 @@
-import sys
 import argparse
 import os
 import random
@@ -14,7 +13,7 @@
 
 from tqdm import tqdm
 
-from transformers import AutoConfig, AutoModel, AutoModelForCausalLM, AutoTokenizer, WordpieceTokenizer
+from transformers import AutoConfig, AutoModel, AutoModelForCausalLM, AutoTokenizer
 from peft import LoraConfig, get_peft_model
 from datasets import Dataset
 
diff --git a/thunder/clang/__init__.py b/thunder/clang/__init__.py
@@ -1,5 +1,4 @@
 from __future__ import annotations
-from collections import namedtuple
 from collections.abc import Callable, Sequence
 from functools import partial, reduce
 from numbers import Number
diff --git a/thunder/core/jit_ext.py b/thunder/core/jit_ext.py
@@ -12,7 +12,6 @@
 import time
 import warnings
 from types import (
-    BuiltinFunctionType,
     BuiltinMethodType,
     CellType,
     FunctionType,
diff --git a/thunder/core/recipe.py b/thunder/core/recipe.py
@@ -131,8 +131,6 @@ def apply(self, model):
         for plugin in post_plugins:
             lookasides.extend(plugin.setup_lookasides() or [])
 
-        from thunder.core import jit_ext, interpreter
-
         if lookasides is not None:
             self._lookaside_executor = TemporaryExecutor()
             for lookaside in lookasides:
diff --git a/thunder/core/update_aliases.py b/thunder/core/update_aliases.py
@@ -5,7 +5,6 @@
 from thunder.core.pytree import tree_flatten
 from thunder.core.symbol import BoundSymbol, BoundSymbolTag, has_tags
 from thunder.core.trace import from_trace, tracectx, TraceCtx as Trace, TraceProvenance, VariableInterface
-from thunder.torch import setitem_
 
 
 def _update_swap_map(swap_map, old_alias, new_alias):
diff --git a/thunder/distributed/__init__.py b/thunder/distributed/__init__.py
@@ -15,7 +15,6 @@
 from torch.utils.weak import WeakTensorKeyDictionary
 
 import thunder.core.utils as utils
-from thunder.core.proxies import DistParallelType
 from thunder.distributed.tensor_parallel import column_parallel
 from thunder.distributed.tensor_parallel import row_parallel
 
diff --git a/thunder/dynamo/utils.py b/thunder/dynamo/utils.py
@@ -5,14 +5,12 @@
 import dataclasses
 import inspect
 import itertools
-import copy
 from types import NoneType
 from collections import defaultdict
 from collections import namedtuple
 
 import torch
 from torch.nn.modules.module import _addindent
-from torch._subclasses.fake_tensor import FakeTensor
 from torch.utils.weak import TensorWeakRef
 
 if torch.distributed.is_available():
diff --git a/thunder/examine/__init__.py b/thunder/examine/__init__.py
@@ -1,4 +1,3 @@
-from typing import Any
 from collections.abc import Callable
 import collections
 import traceback
diff --git a/thunder/executors/__init__.py b/thunder/executors/__init__.py
@@ -1,6 +1,3 @@
-from typing import Optional, Any, List, Tuple
-from collections.abc import Sequence
-
 import thunder.executors.passes as passes
 
 import thunder.extend as extend
diff --git a/thunder/executors/apex_fused_rms_norm_impl.py b/thunder/executors/apex_fused_rms_norm_impl.py
@@ -16,7 +16,7 @@
     if not is_available():
         raise ImportError
     import fused_layer_norm_cuda
-    from apex.normalization.fused_layer_norm import FusedRMSNormAffineMixedDtypesFunction
+    from apex.normalization.fused_layer_norm import FusedRMSNormAffineMixedDtypesFunction  # noqa: F401
 except ImportError:
     APEX_FUSED_NORMS_AVAILABLE = False
 
diff --git a/thunder/executors/cudnnex.py b/thunder/executors/cudnnex.py
@@ -1,6 +1,5 @@
 from collections.abc import Callable
 
-from lightning_utilities.core.imports import package_available
 from looseversion import LooseVersion
 from thunder.extend import OperatorExecutor
 
diff --git a/thunder/executors/fa3ex.py b/thunder/executors/fa3ex.py
@@ -4,7 +4,7 @@
 # is hopper and fa3 has been built
 
 try:
-    from flash_attn_interface import _flash_attn_forward, _flash_attn_backward, flash_attn_func
+    from flash_attn_interface import _flash_attn_forward, _flash_attn_backward, flash_attn_func  # noqa: F401
 
     HAS_FA3 = True
 except ImportError:
diff --git a/thunder/executors/nvfuserex_impl.py b/thunder/executors/nvfuserex_impl.py
@@ -2,10 +2,9 @@
 from functools import partial, lru_cache
 from numbers import Number
 from typing import Any
-from collections.abc import Callable, Mapping, Hashable, Sequence
+from collections.abc import Callable, Hashable, Sequence
 import os
 import time
-from copy import copy
 from itertools import chain, filterfalse
 import warnings
 from typing import cast
@@ -17,7 +16,7 @@
 IS_TORCH_DISTRIBUTED_AVAILABLE = torch.distributed.is_available()
 if IS_TORCH_DISTRIBUTED_AVAILABLE:
     from torch.distributed.tensor import DTensor
-    from torch.distributed.tensor.placement_types import Placement, Shard, Replicate
+    from torch.distributed.tensor.placement_types import Placement, Shard
     import torch.distributed as dist
 
 import thunder.core.dtypes as dtypes
@@ -54,10 +53,6 @@
     put_grads,
 )
 
-from nvfuser.pytorch_utils import (
-    torch_dtype_to_nvfuser_dtype,
-)
-
 
 from thunder.executors.utils import (
     Region,
diff --git a/thunder/extend/__init__.py b/thunder/extend/__init__.py
@@ -529,7 +529,7 @@ def register_executor(ex: Executor) -> Executor:
 
 def get_all_executors() -> tuple[Executor, ...]:
     # manually import all native executors to let them register themselves
-    from thunder.executors import (
+    from thunder.executors import (  # noqa: F401
         apexex,
         cudnn_layernormex,
         cudnnex,
diff --git a/thunder/numpy/__init__.py b/thunder/numpy/__init__.py
@@ -1,9 +1,6 @@
 from numbers import Number
-from typing import Any
 from collections.abc import Callable
 
-import numpy as np
-
 from thunder.core.langctx import langctx, Languages
 from thunder.numpy.langctx import register_method
 
diff --git a/thunder/recipes/__init__.py b/thunder/recipes/__init__.py
@@ -1,4 +1,4 @@
-from typing import Any, Type
+from typing import Any
 
 from thunder.recipes.base import BaseRecipe
 from thunder.recipes.hf_transformers import HFTransformers
diff --git a/thunder/tests/coverage_tests/test_coverage_trace.py b/thunder/tests/coverage_tests/test_coverage_trace.py
@@ -1,5 +1,4 @@
 import torch
-import traceback
 import thunder
 import os
 import pytest
@@ -11,7 +10,6 @@
 
 from transformers import (
     AutoConfig,
-    AutoTokenizer,
     AutoModel,
     AutoModelForCausalLM,
     AutoModelForSeq2SeqLM,
diff --git a/thunder/tests/distributed/helper.py b/thunder/tests/distributed/helper.py
@@ -2,6 +2,7 @@
 from functools import partial
 from functools import wraps
 from typing import ClassVar, TYPE_CHECKING
+import importlib
 import math
 import os
 import sys
@@ -13,10 +14,7 @@
 from thunder.tests.framework import TorchExecutor, nvFuserExecutor
 from thunder.tests.make_tensor import make_tensor
 
-try:
-    import expecttest
-    import hypothesis
-except ImportError:
+if importlib.util.find_spec("expecttest") is None or importlib.util.find_spec("hypothesis") is None:
     raise ImportError(
         "Required packages of `expecttest` and/or `hypothesis` are missing. "
         "Install them with `pip install expecttest hypothesis`"
diff --git a/thunder/tests/distributed/test_dtensor.py b/thunder/tests/distributed/test_dtensor.py
@@ -8,12 +8,11 @@
 if not torch.distributed.is_available():
     pytest.skip(allow_module_level=True)
 
-from thunder.dynamo import thunderfx
 import thunder
 
 from thunder.tests.distributed.helper import DistributedParallelTestCase
 from torch.distributed._tensor import DeviceMesh, distribute_tensor
-from torch.distributed.tensor.placement_types import Placement, Shard, Replicate
+from torch.distributed.tensor.placement_types import Shard
 from torch.testing._internal.distributed._tensor.common_dtensor import DTensorConverter
 
 from torch.testing._internal import common_utils
diff --git a/thunder/tests/test_core.py b/thunder/tests/test_core.py
@@ -5,7 +5,6 @@
 import dataclasses
 import re
 import weakref
-import itertools
 
 import pytest
 import torch
diff --git a/thunder/tests/test_networks.py b/thunder/tests/test_networks.py
@@ -17,8 +17,6 @@
     requiresCUDA,
     DynamoThunderExecutor,
     _all_test_executors,
-    version_between,
-    BITSANDBYTES_AVAILABLE,
     requiresDeviceMemory,
 )
 import thunder.tests.nanogpt_model as nanogpt_model
diff --git a/thunder/tests/test_recipes.py b/thunder/tests/test_recipes.py
@@ -12,7 +12,7 @@
 from thunder.recipes import HFTransformers
 from thunder.executors import nvfuser_available
 from thunder.executors.cudnnex import cudnn_available
-from thunder.tests.framework import version_between, IS_WINDOWS
+from thunder.tests.framework import IS_WINDOWS
 
 
 @pytest.mark.skipif(not cudnn_available(), reason="cuDNN is not available")
diff --git a/thunder/transforms/constant_folding.py b/thunder/transforms/constant_folding.py
@@ -1,6 +1,6 @@
 from numbers import Number
 from typing import Any
-from collections.abc import Callable, Sequence
+from collections.abc import Sequence
 
 import torch
 

Original file line number	Diff line number	Diff line change
`@@ -1,5 +1,4 @@`
`1`	`1`	`from __future__ import annotations`
`2`		`-from collections import namedtuple`
`3`	`2`	`from collections.abc import Callable, Sequence`
`4`	`3`	`from functools import partial, reduce`
`5`	`4`	`from numbers import Number`
Original file line number	Diff line number	Diff line change
`@@ -1,4 +1,3 @@`
`1`		`-from typing import Any`
`2`	`1`	`from collections.abc import Callable`
`3`	`2`	`import collections`
`4`	`3`	`import traceback`