[tensorflow] Add __slots__ (#15459)

stubs/tensorflow/tensorflow/__init__.pyi

@@ -222,6 +222,7 @@ class Operation:
     def __getattr__(self, name: str) -> Incomplete: ...
 
 class TensorShape(metaclass=ABCMeta):
+    __slots__ = ["_dims"]
     def __init__(self, dims: ShapeLike) -> None: ...
     @property
     def rank(self) -> int: ...
@@ -308,6 +309,7 @@ class UnconnectedGradients(Enum):
 _SpecProto = TypeVar("_SpecProto", bound=Message)
 
 class TypeSpec(ABC, Generic[_SpecProto]):
+    __slots__ = ["_cached_cmp_key"]
     @property
     @abstractmethod
     def value_type(self) -> Any: ...
@@ -323,6 +325,7 @@ class TypeSpec(ABC, Generic[_SpecProto]):
     def most_specific_compatible_type(self, other: Self) -> Self: ...
 
 class TensorSpec(TypeSpec[struct_pb2.TensorSpecProto]):
+    __slots__: list[str] = []
     def __init__(self, shape: ShapeLike, dtype: DTypeLike = ..., name: str | None = None) -> None: ...
     @property
     def value_type(self) -> Tensor: ...
@@ -339,6 +342,7 @@ class TensorSpec(TypeSpec[struct_pb2.TensorSpecProto]):
     def is_compatible_with(self, spec_or_tensor: Self | TensorCompatible) -> _bool: ...  # type: ignore[override]
 
 class SparseTensorSpec(TypeSpec[struct_pb2.TypeSpecProto]):
+    __slots__ = ["_shape", "_dtype"]
     def __init__(self, shape: ShapeLike | None = None, dtype: DTypeLike = ...) -> None: ...
     @property
     def value_type(self) -> SparseTensor: ...
@@ -350,6 +354,7 @@ class SparseTensorSpec(TypeSpec[struct_pb2.TypeSpecProto]):
     def from_value(cls, value: SparseTensor) -> Self: ...
 
 class RaggedTensorSpec(TypeSpec[struct_pb2.TypeSpecProto]):
+    __slots__ = ["_shape", "_dtype", "_ragged_rank", "_row_splits_dtype", "_flat_values_spec"]
     def __init__(
         self,
         shape: ShapeLike | None = None,

stubs/tensorflow/tensorflow/dtypes.pyi

@@ -10,6 +10,7 @@ from tensorflow.python.framework.dtypes import HandleData
 class _DTypeMeta(ABCMeta): ...
 
 class DType(metaclass=_DTypeMeta):
+    __slots__ = ["_handle_data"]
     def __init__(self, type_enum: int, handle_data: HandleData | None = None) -> None: ...
     @property
     def name(self) -> str: ...

stubs/tensorflow/tensorflow/saved_model/__init__.pyi

@@ -18,6 +18,13 @@ class Asset:
     def __init__(self, path: str | Path | tf.Tensor) -> None: ...
 
 class LoadOptions:
+    __slots__ = (
+        "allow_partial_checkpoint",
+        "experimental_io_device",
+        "experimental_skip_checkpoint",
+        "experimental_variable_policy",
+        "experimental_load_function_aliases",
+    )
     allow_partial_checkpoint: bool
     experimental_io_device: str | None
     experimental_skip_checkpoint: bool

stubs/tensorflow/tensorflow/train/__init__.pyi

@@ -1,3 +1,4 @@
+from _typeshed import Incomplete
 from collections.abc import Callable
 from typing import Any, TypeVar
 from typing_extensions import Self
@@ -18,10 +19,20 @@ from tensorflow.python.trackable.base import Trackable
 from tensorflow.python.training.tracking.autotrackable import AutoTrackable
 
 class CheckpointOptions:
+    __slots__ = (
+        "experimental_io_device",
+        "experimental_enable_async_checkpoint",
+        "experimental_write_callbacks",
+        "enable_async",
+        "experimental_sharding_callback",
+        "experimental_skip_slot_variables",
+    )
     experimental_io_device: None | str
     experimental_enable_async_checkpoint: bool
     experimental_write_callbacks: None | list[Callable[[str], object] | Callable[[], object]]
     enable_async: bool
+    experimental_sharding_callback: Incomplete  # should be ShardingCallback
+    experimental_skip_slot_variables: bool
 
     def __init__(
         self,