basedpyright 1.36.2 → 1.37.0-961d59bacff1958014fe4e9ed2ef3116fe2266e2

package/dist/typeshed-fallback/stubs/tensorflow/tensorflow/python/feature_column/feature_column_v2.pyi

@@ -0,0 +1,273 @@
+# The types here are all undocumented, but all feature columns are return types of the
+# public functions in tf.feature_column. As they are undocumented internals while some
+# common methods are included, they are incomplete and do not have getattr Incomplete fallback.
+
+from _typeshed import Incomplete
+from abc import ABC, ABCMeta, abstractmethod
+from collections.abc import Callable, Sequence
+from typing import Literal
+from typing_extensions import Self, TypeAlias
+
+import tensorflow as tf
+from tensorflow._aliases import ShapeLike
+
+_Combiners: TypeAlias = Literal["mean", "sqrtn", "sum"]
+_ExampleSpec: TypeAlias = dict[str, tf.io.FixedLenFeature | tf.io.VarLenFeature]
+
+class _FeatureColumn(ABC):
+    @property
+    @abstractmethod
+    def name(self) -> str: ...
+    @property
+    @abstractmethod
+    def parse_example_spec(self) -> _ExampleSpec: ...
+    def __lt__(self, other: _FeatureColumn) -> bool: ...
+    def __gt__(self, other: _FeatureColumn) -> bool: ...
+    @property
+    @abstractmethod
+    def parents(self) -> list[_FeatureColumn | str]: ...
+
+class DenseColumn(_FeatureColumn, metaclass=ABCMeta): ...
+class SequenceDenseColumn(_FeatureColumn, metaclass=ABCMeta): ...
+
+# These classes are mostly subclasses of collections.namedtuple but we can't use
+# typing.NamedTuple because they use multiple inheritance with other non namedtuple classes.
+# _cls instead of cls is because collections.namedtuple uses _cls for __new__.
+class NumericColumn(DenseColumn):
+    key: str
+    shape: ShapeLike
+    default_value: float
+    dtype: tf.DType
+    normalizer_fn: Callable[[tf.Tensor], tf.Tensor] | None
+
+    def __new__(
+        _cls,
+        key: str,
+        shape: ShapeLike,
+        default_value: float,
+        dtype: tf.DType,
+        normalizer_fn: Callable[[tf.Tensor], tf.Tensor] | None,
+    ) -> Self: ...
+    @property
+    def name(self) -> str: ...
+    @property
+    def parse_example_spec(self) -> _ExampleSpec: ...
+    @property
+    def parents(self) -> list[_FeatureColumn | str]: ...
+
+class CategoricalColumn(_FeatureColumn):
+    @property
+    @abstractmethod
+    def num_buckets(self) -> int: ...
+
+class BucketizedColumn(DenseColumn, CategoricalColumn):
+    source_column: NumericColumn
+    boundaries: list[float] | tuple[float, ...]
+
+    def __new__(_cls, source_column: NumericColumn, boundaries: list[float] | tuple[float, ...]) -> Self: ...
+    @property
+    def name(self) -> str: ...
+    @property
+    def num_buckets(self) -> int: ...
+    @property
+    def parse_example_spec(self) -> _ExampleSpec: ...
+    @property
+    def parents(self) -> list[_FeatureColumn | str]: ...
+
+class EmbeddingColumn(DenseColumn, SequenceDenseColumn):
+    categorical_column: CategoricalColumn
+    dimension: int
+    combiner: _Combiners
+    initializer: Callable[[ShapeLike], tf.Tensor] | None
+    ckpt_to_load_from: str | None
+    tensor_name_in_ckpt: str | None
+    max_norm: float | None
+    trainable: bool
+    use_safe_embedding_lookup: bool
+
+    # This one subclasses collections.namedtuple and overrides __new__.
+    def __new__(
+        cls,
+        categorical_column: CategoricalColumn,
+        dimension: int,
+        combiner: _Combiners,
+        initializer: Callable[[ShapeLike], tf.Tensor] | None,
+        ckpt_to_load_from: str | None,
+        tensor_name_in_ckpt: str | None,
+        max_norm: float | None,
+        trainable: bool,
+        use_safe_embedding_lookup: bool = True,
+    ) -> Self: ...
+    @property
+    def name(self) -> str: ...
+    @property
+    def parse_example_spec(self) -> _ExampleSpec: ...
+    @property
+    def parents(self) -> list[_FeatureColumn | str]: ...
+
+class SharedEmbeddingColumnCreator:
+    def __init__(
+        self,
+        dimension: int,
+        initializer: Callable[[ShapeLike], tf.Tensor] | None,
+        ckpt_to_load_from: str | None,
+        tensor_name_in_ckpt: str | None,
+        num_buckets: int,
+        trainable: bool,
+        name: str = "shared_embedding_column_creator",
+        use_safe_embedding_lookup: bool = True,
+    ) -> None: ...
+    def __getattr__(self, name: str) -> Incomplete: ...
+
+class SharedEmbeddingColumn(DenseColumn, SequenceDenseColumn):
+    categorical_column: CategoricalColumn
+    shared_embedding_column_creator: SharedEmbeddingColumnCreator
+    combiner: _Combiners
+    max_norm: float | None
+    use_safe_embedding_lookup: bool
+
+    def __new__(
+        cls,
+        categorical_column: CategoricalColumn,
+        shared_embedding_column_creator: SharedEmbeddingColumnCreator,
+        combiner: _Combiners,
+        max_norm: float | None,
+        use_safe_embedding_lookup: bool = True,
+    ) -> Self: ...
+    @property
+    def name(self) -> str: ...
+    @property
+    def parse_example_spec(self) -> _ExampleSpec: ...
+    @property
+    def parents(self) -> list[_FeatureColumn | str]: ...
+
+class CrossedColumn(CategoricalColumn):
+    keys: tuple[str, ...]
+    hash_bucket_size: int
+    hash_key: int | None
+
+    def __new__(_cls, keys: tuple[str, ...], hash_bucket_size: int, hash_key: int | None) -> Self: ...
+    @property
+    def name(self) -> str: ...
+    @property
+    def num_buckets(self) -> int: ...
+    @property
+    def parse_example_spec(self) -> _ExampleSpec: ...
+    @property
+    def parents(self) -> list[_FeatureColumn | str]: ...
+
+class IdentityCategoricalColumn(CategoricalColumn):
+    key: str
+    number_buckets: int
+    default_value: int | None
+
+    def __new__(_cls, key: str, number_buckets: int, default_value: int | None) -> Self: ...
+    @property
+    def name(self) -> str: ...
+    @property
+    def num_buckets(self) -> int: ...
+    @property
+    def parse_example_spec(self) -> _ExampleSpec: ...
+    @property
+    def parents(self) -> list[_FeatureColumn | str]: ...
+
+class HashedCategoricalColumn(CategoricalColumn):
+    key: str
+    hash_bucket_size: int
+    dtype: tf.DType
+
+    def __new__(_cls, key: str, hash_bucket_size: int, dtype: tf.DType) -> Self: ...
+    @property
+    def name(self) -> str: ...
+    @property
+    def num_buckets(self) -> int: ...
+    @property
+    def parse_example_spec(self) -> _ExampleSpec: ...
+    @property
+    def parents(self) -> list[_FeatureColumn | str]: ...
+
+class VocabularyFileCategoricalColumn(CategoricalColumn):
+    key: str
+    vocabulary_file: str
+    vocabulary_size: int | None
+    num_oov_buckets: int
+    dtype: tf.DType
+    default_value: str | int | None
+    file_format: str | None
+
+    def __new__(
+        cls,
+        key: str,
+        vocabulary_file: str,
+        vocabulary_size: int | None,
+        num_oov_buckets: int,
+        dtype: tf.DType,
+        default_value: str | int | None,
+        file_format: str | None = None,
+    ) -> Self: ...
+    @property
+    def name(self) -> str: ...
+    @property
+    def num_buckets(self) -> int: ...
+    @property
+    def parse_example_spec(self) -> _ExampleSpec: ...
+    @property
+    def parents(self) -> list[_FeatureColumn | str]: ...
+
+class VocabularyListCategoricalColumn(CategoricalColumn):
+    key: str
+    vocabulary_list: Sequence[str] | Sequence[int]
+    dtype: tf.DType
+    default_value: str | int | None
+    num_oov_buckets: int
+
+    def __new__(
+        _cls, key: str, vocabulary_list: Sequence[str], dtype: tf.DType, default_value: str | int | None, num_oov_buckets: int
+    ) -> Self: ...
+    @property
+    def name(self) -> str: ...
+    @property
+    def num_buckets(self) -> int: ...
+    @property
+    def parse_example_spec(self) -> _ExampleSpec: ...
+    @property
+    def parents(self) -> list[_FeatureColumn | str]: ...
+
+class WeightedCategoricalColumn(CategoricalColumn):
+    categorical_column: CategoricalColumn
+    weight_feature_key: str
+    dtype: tf.DType
+
+    def __new__(_cls, categorical_column: CategoricalColumn, weight_feature_key: str, dtype: tf.DType) -> Self: ...
+    @property
+    def name(self) -> str: ...
+    @property
+    def num_buckets(self) -> int: ...
+    @property
+    def parse_example_spec(self) -> _ExampleSpec: ...
+    @property
+    def parents(self) -> list[_FeatureColumn | str]: ...
+
+class IndicatorColumn(DenseColumn, SequenceDenseColumn):
+    categorical_column: CategoricalColumn
+
+    def __new__(_cls, categorical_column: CategoricalColumn) -> Self: ...
+    @property
+    def name(self) -> str: ...
+    @property
+    def parse_example_spec(self) -> _ExampleSpec: ...
+    @property
+    def parents(self) -> list[_FeatureColumn | str]: ...
+
+class SequenceCategoricalColumn(CategoricalColumn):
+    categorical_column: CategoricalColumn
+
+    def __new__(_cls, categorical_column: CategoricalColumn) -> Self: ...
+    @property
+    def name(self) -> str: ...
+    @property
+    def num_buckets(self) -> int: ...
+    @property
+    def parse_example_spec(self) -> _ExampleSpec: ...
+    @property
+    def parents(self) -> list[_FeatureColumn | str]: ...

package/dist/typeshed-fallback/stubs/tensorflow/tensorflow/python/feature_column/sequence_feature_column.pyi

@@ -0,0 +1,30 @@
+from collections.abc import Callable
+from typing_extensions import Self
+
+import tensorflow as tf
+from tensorflow._aliases import ShapeLike
+from tensorflow.python.feature_column.feature_column_v2 import SequenceDenseColumn, _ExampleSpec, _FeatureColumn
+
+# Strangely at runtime most of Sequence feature columns are defined in feature_column_v2 except
+# for this one.
+class SequenceNumericColumn(SequenceDenseColumn):
+    key: str
+    shape: ShapeLike
+    default_value: float
+    dtype: tf.DType
+    normalizer_fn: Callable[[tf.Tensor], tf.Tensor] | None
+
+    def __new__(
+        _cls,
+        key: str,
+        shape: ShapeLike,
+        default_value: float,
+        dtype: tf.DType,
+        normalizer_fn: Callable[[tf.Tensor], tf.Tensor] | None,
+    ) -> Self: ...
+    @property
+    def name(self) -> str: ...
+    @property
+    def parse_example_spec(self) -> _ExampleSpec: ...
+    @property
+    def parents(self) -> list[_FeatureColumn | str]: ...

package/dist/typeshed-fallback/stubs/tensorflow/tensorflow/python/framework/dtypes.pyi

@@ -0,0 +1,7 @@
+import dataclasses
+from _typeshed import Incomplete
+
+@dataclasses.dataclass(frozen=True)
+class HandleData:
+    shape_inference: Incomplete | None = None
+    alias_id: int | None = None

package/dist/typeshed-fallback/stubs/tensorflow/tensorflow/python/keras/__init__.pyi

@@ -0,0 +1 @@
+def __getattr__(name: str): ...  # incomplete module

package/dist/typeshed-fallback/stubs/tensorflow/tensorflow/python/keras/protobuf/projector_config_pb2.pyi

@@ -0,0 +1,129 @@
+"""
+@generated by mypy-protobuf.  Do not edit manually!
+isort:skip_file
+This file is a copy of the TensorBoard ProjectorConfig proto.
+Keep this file in sync with the source proto definition at
+https://github.com/tensorflow/tensorboard/blob/master/tensorboard/plugins/projector/projector_config.proto
+"""
+
+import builtins
+import collections.abc
+import typing
+
+import google.protobuf.descriptor
+import google.protobuf.internal.containers
+import google.protobuf.message
+
+DESCRIPTOR: google.protobuf.descriptor.FileDescriptor
+
+@typing.final
+class SpriteMetadata(google.protobuf.message.Message):
+    DESCRIPTOR: google.protobuf.descriptor.Descriptor
+
+    IMAGE_PATH_FIELD_NUMBER: builtins.int
+    SINGLE_IMAGE_DIM_FIELD_NUMBER: builtins.int
+    image_path: builtins.str
+    @property
+    def single_image_dim(self) -> google.protobuf.internal.containers.RepeatedScalarFieldContainer[builtins.int]:
+        """[width, height] of a single image in the sprite."""
+
+    def __init__(
+        self, *, image_path: builtins.str | None = ..., single_image_dim: collections.abc.Iterable[builtins.int] | None = ...
+    ) -> None: ...
+    def ClearField(
+        self, field_name: typing.Literal["image_path", b"image_path", "single_image_dim", b"single_image_dim"]
+    ) -> None: ...
+
+global___SpriteMetadata = SpriteMetadata
+
+@typing.final
+class EmbeddingInfo(google.protobuf.message.Message):
+    DESCRIPTOR: google.protobuf.descriptor.Descriptor
+
+    TENSOR_NAME_FIELD_NUMBER: builtins.int
+    METADATA_PATH_FIELD_NUMBER: builtins.int
+    BOOKMARKS_PATH_FIELD_NUMBER: builtins.int
+    TENSOR_SHAPE_FIELD_NUMBER: builtins.int
+    SPRITE_FIELD_NUMBER: builtins.int
+    TENSOR_PATH_FIELD_NUMBER: builtins.int
+    tensor_name: builtins.str
+    metadata_path: builtins.str
+    bookmarks_path: builtins.str
+    tensor_path: builtins.str
+    """Path to the TSV file holding the tensor values. If missing, the tensor
+    is assumed to be stored in the model checkpoint.
+    """
+    @property
+    def tensor_shape(self) -> google.protobuf.internal.containers.RepeatedScalarFieldContainer[builtins.int]:
+        """Shape of the 2D tensor [N x D]. If missing, it will be inferred from the
+        model checkpoint.
+        """
+
+    @property
+    def sprite(self) -> global___SpriteMetadata: ...
+    def __init__(
+        self,
+        *,
+        tensor_name: builtins.str | None = ...,
+        metadata_path: builtins.str | None = ...,
+        bookmarks_path: builtins.str | None = ...,
+        tensor_shape: collections.abc.Iterable[builtins.int] | None = ...,
+        sprite: global___SpriteMetadata | None = ...,
+        tensor_path: builtins.str | None = ...,
+    ) -> None: ...
+    def HasField(self, field_name: typing.Literal["sprite", b"sprite"]) -> builtins.bool: ...
+    def ClearField(
+        self,
+        field_name: typing.Literal[
+            "bookmarks_path",
+            b"bookmarks_path",
+            "metadata_path",
+            b"metadata_path",
+            "sprite",
+            b"sprite",
+            "tensor_name",
+            b"tensor_name",
+            "tensor_path",
+            b"tensor_path",
+            "tensor_shape",
+            b"tensor_shape",
+        ],
+    ) -> None: ...
+
+global___EmbeddingInfo = EmbeddingInfo
+
+@typing.final
+class ProjectorConfig(google.protobuf.message.Message):
+    DESCRIPTOR: google.protobuf.descriptor.Descriptor
+
+    MODEL_CHECKPOINT_PATH_FIELD_NUMBER: builtins.int
+    EMBEDDINGS_FIELD_NUMBER: builtins.int
+    MODEL_CHECKPOINT_DIR_FIELD_NUMBER: builtins.int
+    model_checkpoint_path: builtins.str
+    """Path to the checkpoint file. Use either this or model_checkpoint_dir."""
+    model_checkpoint_dir: builtins.str
+    """Path to the checkpoint directory. The directory will be scanned for the
+    latest checkpoint file.
+    """
+    @property
+    def embeddings(self) -> google.protobuf.internal.containers.RepeatedCompositeFieldContainer[global___EmbeddingInfo]: ...
+    def __init__(
+        self,
+        *,
+        model_checkpoint_path: builtins.str | None = ...,
+        embeddings: collections.abc.Iterable[global___EmbeddingInfo] | None = ...,
+        model_checkpoint_dir: builtins.str | None = ...,
+    ) -> None: ...
+    def ClearField(
+        self,
+        field_name: typing.Literal[
+            "embeddings",
+            b"embeddings",
+            "model_checkpoint_dir",
+            b"model_checkpoint_dir",
+            "model_checkpoint_path",
+            b"model_checkpoint_path",
+        ],
+    ) -> None: ...
+
+global___ProjectorConfig = ProjectorConfig

package/dist/typeshed-fallback/stubs/tensorflow/tensorflow/python/keras/protobuf/saved_metadata_pb2.pyi

@@ -0,0 +1,89 @@
+"""
+@generated by mypy-protobuf.  Do not edit manually!
+isort:skip_file
+Protobuf containing the metadata for each Keras object saved in a SavedModel."""
+
+import builtins
+import collections.abc
+import typing
+
+import google.protobuf.descriptor
+import google.protobuf.internal.containers
+import google.protobuf.message
+import tensorflow.python.keras.protobuf.versions_pb2
+
+DESCRIPTOR: google.protobuf.descriptor.FileDescriptor
+
+@typing.final
+class SavedMetadata(google.protobuf.message.Message):
+    DESCRIPTOR: google.protobuf.descriptor.Descriptor
+
+    NODES_FIELD_NUMBER: builtins.int
+    @property
+    def nodes(self) -> google.protobuf.internal.containers.RepeatedCompositeFieldContainer[global___SavedObject]:
+        """Nodes represent trackable objects in the SavedModel. The data for every
+        Keras object is stored.
+        """
+
+    def __init__(self, *, nodes: collections.abc.Iterable[global___SavedObject] | None = ...) -> None: ...
+    def ClearField(self, field_name: typing.Literal["nodes", b"nodes"]) -> None: ...
+
+global___SavedMetadata = SavedMetadata
+
+@typing.final
+class SavedObject(google.protobuf.message.Message):
+    """Metadata of an individual Keras object."""
+
+    DESCRIPTOR: google.protobuf.descriptor.Descriptor
+
+    NODE_ID_FIELD_NUMBER: builtins.int
+    NODE_PATH_FIELD_NUMBER: builtins.int
+    IDENTIFIER_FIELD_NUMBER: builtins.int
+    METADATA_FIELD_NUMBER: builtins.int
+    VERSION_FIELD_NUMBER: builtins.int
+    node_id: builtins.int
+    """Index of the node in the SavedModel SavedObjectGraph."""
+    node_path: builtins.str
+    """String path from root (e.g. "root.child_layer")"""
+    identifier: builtins.str
+    """Identifier to determine loading function.
+    Must be one of:
+      _tf_keras_input_layer, _tf_keras_layer, _tf_keras_metric,
+      _tf_keras_model, _tf_keras_network, _tf_keras_rnn_layer,
+      _tf_keras_sequential
+    """
+    metadata: builtins.str
+    """Metadata containing a JSON-serialized object with the non-TensorFlow
+    attributes for this Keras object.
+    """
+    @property
+    def version(self) -> tensorflow.python.keras.protobuf.versions_pb2.VersionDef:
+        """Version defined by the code serializing this Keras object."""
+
+    def __init__(
+        self,
+        *,
+        node_id: builtins.int | None = ...,
+        node_path: builtins.str | None = ...,
+        identifier: builtins.str | None = ...,
+        metadata: builtins.str | None = ...,
+        version: tensorflow.python.keras.protobuf.versions_pb2.VersionDef | None = ...,
+    ) -> None: ...
+    def HasField(self, field_name: typing.Literal["version", b"version"]) -> builtins.bool: ...
+    def ClearField(
+        self,
+        field_name: typing.Literal[
+            "identifier",
+            b"identifier",
+            "metadata",
+            b"metadata",
+            "node_id",
+            b"node_id",
+            "node_path",
+            b"node_path",
+            "version",
+            b"version",
+        ],
+    ) -> None: ...
+
+global___SavedObject = SavedObject

package/dist/typeshed-fallback/stubs/tensorflow/tensorflow/python/keras/protobuf/versions_pb2.pyi

@@ -0,0 +1,63 @@
+"""
+@generated by mypy-protobuf.  Do not edit manually!
+isort:skip_file
+"""
+
+import builtins
+import collections.abc
+import typing
+
+import google.protobuf.descriptor
+import google.protobuf.internal.containers
+import google.protobuf.message
+
+DESCRIPTOR: google.protobuf.descriptor.FileDescriptor
+
+@typing.final
+class VersionDef(google.protobuf.message.Message):
+    """This file is a copy of the TensorFlow Versions proto.
+    Keep this file in sync with the source proto definition at
+    https://github.com/tensorflow/tensorflow/blob/master/tensorflow/core/framework/versions.proto
+
+    Version information for a piece of serialized data
+
+    There are different types of versions for each type of data
+    (GraphDef, etc.), but they all have the same common shape
+    described here.
+
+    Each consumer has "consumer" and "min_producer" versions (specified
+    elsewhere).  A consumer is allowed to consume this data if
+
+      producer >= min_producer
+      consumer >= min_consumer
+      consumer not in bad_consumers
+
+    LINT.IfChange
+    """
+
+    DESCRIPTOR: google.protobuf.descriptor.Descriptor
+
+    PRODUCER_FIELD_NUMBER: builtins.int
+    MIN_CONSUMER_FIELD_NUMBER: builtins.int
+    BAD_CONSUMERS_FIELD_NUMBER: builtins.int
+    producer: builtins.int
+    """The version of the code that produced this data."""
+    min_consumer: builtins.int
+    """Any consumer below this version is not allowed to consume this data."""
+    @property
+    def bad_consumers(self) -> google.protobuf.internal.containers.RepeatedScalarFieldContainer[builtins.int]:
+        """Specific consumer versions which are disallowed (e.g. due to bugs)."""
+
+    def __init__(
+        self,
+        *,
+        producer: builtins.int | None = ...,
+        min_consumer: builtins.int | None = ...,
+        bad_consumers: collections.abc.Iterable[builtins.int] | None = ...,
+    ) -> None: ...
+    def ClearField(
+        self,
+        field_name: typing.Literal["bad_consumers", b"bad_consumers", "min_consumer", b"min_consumer", "producer", b"producer"],
+    ) -> None: ...
+
+global___VersionDef = VersionDef

package/dist/typeshed-fallback/stubs/tensorflow/tensorflow/python/trackable/autotrackable.pyi

@@ -0,0 +1,3 @@
+from tensorflow.python.trackable.base import Trackable
+
+class AutoTrackable(Trackable): ...

package/dist/typeshed-fallback/stubs/tensorflow/tensorflow/python/trackable/base.pyi

@@ -0,0 +1,5 @@
+# Internal type that is commonly used as a base class
+# and some public apis the signature needs it. As type
+# is internal exact module it lives in is unstable across
+# versions.
+class Trackable: ...

package/dist/typeshed-fallback/stubs/tensorflow/tensorflow/python/trackable/resource.pyi

@@ -0,0 +1,9 @@
+from tensorflow.python.trackable.base import Trackable
+
+class _ResourceMetaclass(type): ...
+
+# Internal type that is commonly used as a base class
+# it is needed for the public signatures of some APIs.
+class CapturableResource(Trackable, metaclass=_ResourceMetaclass): ...
+
+def __getattr__(name: str): ...  # incomplete module

package/dist/typeshed-fallback/stubs/tensorflow/tensorflow/python/trackable/ressource.pyi

@@ -0,0 +1,7 @@
+from tensorflow.python.trackable.base import Trackable
+
+class _ResourceMetaclass(type): ...
+
+# Internal type that is commonly used as a base class
+# and some public apis the signature needs it.
+class CapturableResource(Trackable, metaclass=_ResourceMetaclass): ...

package/dist/typeshed-fallback/stubs/tensorflow/tensorflow/python/training/tracking/autotrackable.pyi

@@ -0,0 +1,3 @@
+from _typeshed import Incomplete
+
+AutoTrackable = Incomplete