TensorFlow
Install Learn
  • Introduction
    New to TensorFlow?
  • TensorFlow
    The core open source ML library
  • For JavaScript
    TensorFlow.js for ML using JavaScript
  • For Mobile & Edge
    TensorFlow Lite for mobile and edge devices
  • For Production
    TensorFlow Extended for end-to-end ML components
API
  • TensorFlow (v2.11.1)
  • Versions…
  • TensorFlow.js
  • TensorFlow Lite
  • TFX
Resources
  • Models & datasets
    Pre-trained models and datasets built by Google and the community
  • Tools
    Ecosystem of tools to help you use TensorFlow
  • Libraries & extensions
    Libraries and extensions built on TensorFlow
  • TensorFlow Certificate program
    Differentiate yourself by demonstrating your ML proficiency
  • Learn ML
    Educational resources to learn the fundamentals of ML with TensorFlow
  • Responsible AI
    Resources and tools to integrate Responsible AI practices into your ML workflow
Community
  • Join
    Explore ways to get involved
  • Blog
    Stay up to date with all things TensorFlow
  • Forum ↗
    Discussion platform for the TensorFlow community
  • Groups
    User groups, interest groups and mailing lists
  • Contribute
    Guide for contributing to code and documentation
Why TensorFlow
  • About
  • Case studies
  • English
  • 中文 – 简体
GitHub Sign in
  • TensorFlow v2.8.4
Overview Python C++ Java More
TensorFlow
  • Install
  • Learn
    • More
  • API
    • More
    • Overview
    • Python
    • C++
    • Java
    • More
  • Resources
    • More
  • Community
    • More
  • Why TensorFlow
    • More
  • GitHub
  • Overview
  • All Symbols
  • Python v2.8.4
  • tf
    • Overview
    • AggregationMethod
    • CriticalSection
    • DeviceSpec
    • GradientTape
    • Graph
    • IndexedSlices
    • IndexedSlicesSpec
    • Module
    • Operation
    • OptionalSpec
    • RaggedTensor
    • RaggedTensorSpec
    • RegisterGradient
    • SparseTensorSpec
    • Tensor
    • TensorArray
    • TensorArraySpec
    • TensorShape
    • TensorSpec
    • TypeSpec
    • UnconnectedGradients
    • Variable
    • Variable.SaveSliceInfo
    • VariableAggregation
    • VariableSynchronization
    • argsort
    • batch_to_space
    • bitcast
    • boolean_mask
    • broadcast_dynamic_shape
    • broadcast_static_shape
    • broadcast_to
    • case
    • cast
    • clip_by_global_norm
    • clip_by_norm
    • clip_by_value
    • concat
    • cond
    • constant
    • constant_initializer
    • control_dependencies
    • convert_to_tensor
    • custom_gradient
    • device
    • dynamic_partition
    • dynamic_stitch
    • edit_distance
    • einsum
    • ensure_shape
    • executing_eagerly
    • expand_dims
    • extract_volume_patches
    • eye
    • fill
    • fingerprint
    • foldl
    • foldr
    • function
    • gather
    • gather_nd
    • get_current_name_scope
    • get_logger
    • get_static_value
    • grad_pass_through
    • gradients
    • group
    • guarantee_const
    • hessians
    • histogram_fixed_width
    • histogram_fixed_width_bins
    • identity
    • identity_n
    • init_scope
    • inside_function
    • is_tensor
    • linspace
    • load_library
    • load_op_library
    • make_ndarray
    • make_tensor_proto
    • map_fn
    • meshgrid
    • name_scope
    • no_gradient
    • no_op
    • nondifferentiable_batch_function
    • norm
    • numpy_function
    • one_hot
    • ones
    • ones_initializer
    • ones_like
    • pad
    • parallel_stack
    • print
    • py_function
    • random_normal_initializer
    • random_uniform_initializer
    • range
    • rank
    • realdiv
    • recompute_grad
    • register_tensor_conversion_function
    • repeat
    • required_space_to_batch_paddings
    • reshape
    • reverse
    • reverse_sequence
    • roll
    • scan
    • scatter_nd
    • searchsorted
    • sequence_mask
    • shape
    • shape_n
    • size
    • slice
    • sort
    • space_to_batch
    • space_to_batch_nd
    • split
    • squeeze
    • stack
    • stop_gradient
    • strided_slice
    • switch_case
    • tensor_scatter_nd_add
    • tensor_scatter_nd_max
    • tensor_scatter_nd_min
    • tensor_scatter_nd_sub
    • tensor_scatter_nd_update
    • tensordot
    • tile
    • timestamp
    • transpose
    • truncatediv
    • truncatemod
    • tuple
    • type_spec_from_value
    • unique
    • unique_with_counts
    • unravel_index
    • unstack
    • variable_creator_scope
    • vectorized_map
    • where
    • while_loop
    • zeros
    • zeros_initializer
    • zeros_like
  • tf.audio
    • Overview
    • decode_wav
    • encode_wav
  • tf.autodiff
    • Overview
    • ForwardAccumulator
  • tf.autograph
    • Overview
    • set_verbosity
    • to_code
    • to_graph
    • trace
    • experimental
      • Overview
      • Feature
      • do_not_convert
      • set_loop_options
  • tf.bitwise
    • Overview
    • bitwise_and
    • bitwise_or
    • bitwise_xor
    • invert
    • left_shift
    • right_shift
  • tf.compat
    • Overview
    • as_bytes
    • as_str
    • as_str_any
    • as_text
    • dimension_at_index
    • dimension_value
    • forward_compatibility_horizon
    • forward_compatible
    • path_to_str
    • v1
      • Overview
      • AttrValue
      • AttrValue.ListValue
      • ConditionalAccumulator
      • ConditionalAccumulatorBase
      • ConfigProto
      • ConfigProto.DeviceCountEntry
      • ConfigProto.Experimental
      • DeviceSpec
      • Dimension
      • Event
      • FixedLengthRecordReader
      • GPUOptions
      • GPUOptions.Experimental
      • GPUOptions.Experimental.VirtualDevices
      • GraphDef
      • GraphKeys
      • GraphOptions
      • HistogramProto
      • IdentityReader
      • InteractiveSession
      • LMDBReader
      • LogMessage
      • MetaGraphDef
      • MetaGraphDef.CollectionDefEntry
      • MetaGraphDef.MetaInfoDef
      • MetaGraphDef.MetaInfoDef.FunctionAliasesEntry
      • MetaGraphDef.SignatureDefEntry
      • NameAttrList
      • NameAttrList.AttrEntry
      • NodeDef
      • NodeDef.AttrEntry
      • NodeDef.ExperimentalDebugInfo
      • OptimizerOptions
      • Print
      • ReaderBase
      • RunMetadata
      • RunMetadata.FunctionGraphs
      • RunOptions
      • RunOptions.Experimental
      • RunOptions.Experimental.RunHandlerPoolOptions
      • Session
      • SessionLog
      • SparseConditionalAccumulator
      • SparseTensorValue
      • Summary
      • Summary.Audio
      • Summary.Image
      • Summary.Value
      • SummaryMetadata
      • SummaryMetadata.PluginData
      • TFRecordReader
      • TensorInfo
      • TensorInfo.CompositeTensor
      • TensorInfo.CooSparse
      • TextLineReader
      • Variable
      • VariableAggregation
      • VariableScope
      • WholeFileReader
      • add_check_numerics_ops
      • add_to_collection
      • add_to_collections
      • all_variables
      • arg_max
      • arg_min
      • argmax
      • argmin
      • assert_equal
      • assert_greater
      • assert_greater_equal
      • assert_integer
      • assert_less
      • assert_less_equal
      • assert_near
      • assert_negative
      • assert_non_negative
      • assert_non_positive
      • assert_none_equal
      • assert_positive
      • assert_rank
      • assert_rank_at_least
      • assert_rank_in
      • assert_scalar
      • assert_type
      • assert_variables_initialized
      • assign
      • assign_add
      • assign_sub
      • batch_gather
      • batch_scatter_update
      • batch_to_space
      • batch_to_space_nd
      • bincount
      • boolean_mask
      • case
      • clip_by_average_norm
      • colocate_with
      • cond
      • confusion_matrix
      • constant
      • container
      • control_flow_v2_enabled
      • convert_to_tensor
      • convert_to_tensor_or_indexed_slices
      • convert_to_tensor_or_sparse_tensor
      • count_nonzero
      • count_up_to
      • create_partitioned_variables
      • decode_csv
      • decode_raw
      • delete_session_tensor
      • depth_to_space
      • device
      • disable_control_flow_v2
      • disable_eager_execution
      • disable_resource_variables
      • disable_tensor_equality
      • disable_v2_behavior
      • disable_v2_tensorshape
      • div
      • enable_control_flow_v2
      • enable_eager_execution
      • enable_resource_variables
      • enable_tensor_equality
      • enable_v2_behavior
      • enable_v2_tensorshape
      • executing_eagerly
      • executing_eagerly_outside_functions
      • expand_dims
      • extract_image_patches
      • fixed_size_partitioner
      • floor_div
      • foldl
      • foldr
      • gather
      • gather_nd
      • get_collection
      • get_collection_ref
      • get_default_graph
      • get_default_session
      • get_local_variable
      • get_seed
      • get_session_handle
      • get_session_tensor
      • get_variable
      • get_variable_scope
      • global_variables
      • global_variables_initializer
      • gradients
      • hessians
      • initialize_all_tables
      • initialize_all_variables
      • initialize_local_variables
      • initialize_variables
      • is_variable_initialized
      • load_file_system_library
      • local_variables
      • local_variables_initializer
      • make_template
      • map_fn
      • min_max_variable_partitioner
      • model_variables
      • moving_average_variables
      • multinomial
      • no_regularizer
      • norm
      • ones_like
      • op_scope
      • pad
      • parse_example
      • parse_single_example
      • placeholder
      • placeholder_with_default
      • py_func
      • quantize_v2
      • random_normal_initializer
      • random_poisson
      • random_uniform_initializer
      • reduce_all
      • reduce_any
      • reduce_join
      • reduce_logsumexp
      • reduce_max
      • reduce_mean
      • reduce_min
      • reduce_prod
      • reduce_sum
      • report_uninitialized_variables
      • reset_default_graph
      • resource_variables_enabled
      • reverse_sequence
      • scalar_mul
      • scan
      • scatter_add
      • scatter_div
      • scatter_max
      • scatter_min
      • scatter_mul
      • scatter_nd_add
      • scatter_nd_sub
      • scatter_nd_update
      • scatter_sub
      • scatter_update
      • serialize_many_sparse
      • serialize_sparse
      • set_random_seed
      • setdiff1d
      • shape
      • size
      • space_to_batch
      • space_to_depth
      • sparse_add
      • sparse_concat
      • sparse_matmul
      • sparse_merge
      • sparse_placeholder
      • sparse_reduce_max
      • sparse_reduce_max_sparse
      • sparse_reduce_sum
      • sparse_reduce_sum_sparse
      • sparse_segment_mean
      • sparse_segment_sqrt_n
      • sparse_segment_sum
      • sparse_split
      • sparse_to_dense
      • squeeze
      • string_split
      • string_to_hash_bucket
      • string_to_number
      • substr
      • tables_initializer
      • to_bfloat16
      • to_complex128
      • to_complex64
      • to_double
      • to_float
      • to_int32
      • to_int64
      • trainable_variables
      • transpose
      • truncated_normal_initializer
      • tuple
      • uniform_unit_scaling_initializer
      • variable_axis_size_partitioner
      • variable_creator_scope
      • variable_op_scope
      • variable_scope
      • variables_initializer
      • verify_tensor_all_finite
      • where
      • while_loop
      • wrap_function
      • zeros_like
      • app
        • Overview
        • run
      • audio
        • Overview
      • autograph
        • Overview
        • to_code
        • to_graph
        • experimental
          • Overview
      • bitwise
        • Overview
      • compat
        • Overview
      • config
        • Overview
        • experimental
          • Overview
        • optimizer
          • Overview
        • threading
          • Overview
      • data
        • Overview
        • Dataset
        • FixedLengthRecordDataset
        • Iterator
        • TFRecordDataset
        • TextLineDataset
        • get_output_classes
        • get_output_shapes
        • get_output_types
        • make_initializable_iterator
        • make_one_shot_iterator
        • experimental
          • Overview
          • Counter
          • CsvDataset
          • RaggedTensorStructure
          • RandomDataset
          • SparseTensorStructure
          • SqlDataset
          • TensorArrayStructure
          • TensorStructure
          • choose_from_datasets
          • make_batched_features_dataset
          • make_csv_dataset
          • map_and_batch_with_legacy_function
          • sample_from_datasets
          • service
            • Overview
      • debugging
        • Overview
        • assert_shapes
        • experimental
          • Overview
      • distribute
        • Overview
        • MirroredStrategy
        • OneDeviceStrategy
        • ReplicaContext
        • Strategy
        • StrategyExtended
        • get_loss_reduction
        • cluster_resolver
          • Overview
        • experimental
          • Overview
          • CentralStorageStrategy
          • MultiWorkerMirroredStrategy
          • ParameterServerStrategy
          • TPUStrategy
      • distributions
        • Overview
        • Bernoulli
        • Beta
        • Categorical
        • Dirichlet
        • DirichletMultinomial
        • Distribution
        • Exponential
        • Gamma
        • Laplace
        • Multinomial
        • Normal
        • RegisterKL
        • ReparameterizationType
        • StudentT
        • Uniform
        • kl_divergence
      • dtypes
        • Overview
      • errors
        • Overview
        • error_code_from_exception_type
        • exception_type_from_error_code
        • raise_exception_on_not_ok_status
      • estimator
        • Overview
        • BaselineClassifier
        • BaselineEstimator
        • BaselineRegressor
        • DNNClassifier
        • DNNEstimator
        • DNNLinearCombinedClassifier
        • DNNLinearCombinedEstimator
        • DNNLinearCombinedRegressor
        • DNNRegressor
        • Estimator
        • LinearClassifier
        • LinearEstimator
        • LinearRegressor
        • classifier_parse_example_spec
        • regressor_parse_example_spec
        • experimental
          • Overview
          • KMeans
          • dnn_logit_fn_builder
          • linear_logit_fn_builder
        • export
          • Overview
        • inputs
          • Overview
          • numpy_input_fn
          • pandas_input_fn
        • tpu
          • Overview
          • InputPipelineConfig
          • RunConfig
          • TPUConfig
          • TPUEstimator
          • TPUEstimatorSpec
          • experimental
            • Overview
            • EmbeddingConfigSpec
      • experimental
        • Overview
        • output_all_intermediates
      • feature_column
        • Overview
        • categorical_column_with_vocabulary_file
        • input_layer
        • linear_model
        • make_parse_example_spec
        • shared_embedding_columns
      • flags
        • Overview
        • ArgumentParser
        • ArgumentSerializer
        • BaseListParser
        • BooleanFlag
        • BooleanParser
        • CantOpenFlagFileError
        • CsvListSerializer
        • DEFINE
        • DEFINE_alias
        • DEFINE_bool
        • DEFINE_enum
        • DEFINE_enum_class
        • DEFINE_flag
        • DEFINE_float
        • DEFINE_integer
        • DEFINE_list
        • DEFINE_multi
        • DEFINE_multi_enum
        • DEFINE_multi_enum_class
        • DEFINE_multi_float
        • DEFINE_multi_integer
        • DEFINE_multi_string
        • DEFINE_spaceseplist
        • DEFINE_string
        • DuplicateFlagError
        • EnumClassFlag
        • EnumClassListSerializer
        • EnumClassParser
        • EnumClassSerializer
        • EnumFlag
        • EnumParser
        • Error
        • FLAGS
        • Flag
        • FlagHolder
        • FlagNameConflictsWithMethodError
        • FlagValues
        • FloatParser
        • IllegalFlagValueError
        • IntegerParser
        • ListParser
        • ListSerializer
        • MultiEnumClassFlag
        • MultiFlag
        • UnparsedFlagAccessError
        • UnrecognizedFlagError
        • ValidationError
        • WhitespaceSeparatedListParser
        • adopt_module_key_flags
        • declare_key_flag
        • disclaim_key_flags
        • doc_to_help
        • flag_dict_to_args
        • get_help_width
        • mark_bool_flags_as_mutual_exclusive
        • mark_flag_as_required
        • mark_flags_as_mutual_exclusive
        • mark_flags_as_required
        • multi_flags_validator
        • register_multi_flags_validator
        • register_validator
        • set_default
        • text_wrap
        • validator
        • tf_decorator
          • Overview
          • TFDecorator
          • make_decorator
          • rewrap
          • unwrap
      • gfile
        • Overview
        • Copy
        • DeleteRecursively
        • Exists
        • FastGFile
        • Glob
        • IsDirectory
        • ListDirectory
        • MakeDirs
        • MkDir
        • Remove
        • Rename
        • Stat
        • Walk
      • graph_util
        • Overview
        • convert_variables_to_constants
        • extract_sub_graph
        • must_run_on_cpu
        • remove_training_nodes
        • tensor_shape_from_node_def_name
      • image
        • Overview
        • ResizeMethod
        • crop_and_resize
        • draw_bounding_boxes
        • extract_glimpse
        • resize
        • resize_area
        • resize_bicubic
        • resize_bilinear
        • resize_image_with_pad
        • resize_nearest_neighbor
        • sample_distorted_bounding_box
      • initializers
        • Overview
        • he_normal
        • he_uniform
        • lecun_normal
        • lecun_uniform
      • io
        • Overview
        • TFRecordCompressionType
        • tf_record_iterator
        • gfile
          • Overview
      • keras
        • Overview
        • activations
          • Overview
        • applications
          • Overview
          • densenet
            • Overview
          • efficientnet
            • Overview
          • efficientnet_v2
            • Overview
          • imagenet_utils
            • Overview
          • inception_resnet_v2
            • Overview
          • inception_v3
            • Overview
          • mobilenet
            • Overview
          • mobilenet_v2
            • Overview
          • mobilenet_v3
            • Overview
          • nasnet
            • Overview
          • resnet
            • Overview
          • resnet50
            • Overview
          • resnet_v2
            • Overview
          • vgg16
            • Overview
          • vgg19
            • Overview
          • xception
            • Overview
        • backend
          • Overview
          • get_session
          • name_scope
          • set_session
        • callbacks
          • Overview
          • TensorBoard
        • constraints
          • Overview
        • datasets
          • Overview
          • boston_housing
            • Overview
          • cifar10
            • Overview
          • cifar100
            • Overview
          • fashion_mnist
            • Overview
          • imdb
            • Overview
          • mnist
            • Overview
          • reuters
            • Overview
        • estimator
          • Overview
          • model_to_estimator
        • experimental
          • Overview
          • export_saved_model
          • load_from_saved_model
        • initializers
          • Overview
          • Constant
          • Identity
          • Ones
          • Orthogonal
          • RandomNormal
          • RandomUniform
          • TruncatedNormal
          • VarianceScaling
          • Zeros
          • glorot_normal
          • glorot_uniform
          • he_normal
          • he_uniform
          • lecun_normal
          • lecun_uniform
        • layers
          • Overview
          • BatchNormalization
          • CuDNNGRU
          • CuDNNLSTM
          • DenseFeatures
          • GRU
          • GRUCell
          • LSTM
          • LSTMCell
          • disable_v2_dtype_behavior
          • enable_v2_dtype_behavior
          • experimental
            • Overview
            • preprocessing
              • Overview
        • losses
          • Overview
        • metrics
          • Overview
        • mixed_precision
          • Overview
          • experimental
            • Overview
        • models
          • Overview
        • optimizers
          • Overview
          • schedules
            • Overview
        • preprocessing
          • Overview
          • image
            • Overview
          • sequence
            • Overview
          • text
            • Overview
        • regularizers
          • Overview
        • utils
          • Overview
          • DeterministicRandomTestTool
          • get_or_create_layer
          • track_tf1_style_variables
        • wrappers
          • Overview
          • scikit_learn
            • Overview
      • layers
        • Overview
        • AveragePooling1D
        • AveragePooling2D
        • AveragePooling3D
        • BatchNormalization
        • Conv1D
        • Conv2D
        • Conv2DTranspose
        • Conv3D
        • Conv3DTranspose
        • Dense
        • Dropout
        • Flatten
        • Layer
        • MaxPooling1D
        • MaxPooling2D
        • MaxPooling3D
        • SeparableConv1D
        • SeparableConv2D
        • average_pooling1d
        • average_pooling2d
        • average_pooling3d
        • batch_normalization
        • conv1d
        • conv2d
        • conv2d_transpose
        • conv3d
        • conv3d_transpose
        • dense
        • dropout
        • flatten
        • max_pooling1d
        • max_pooling2d
        • max_pooling3d
        • separable_conv1d
        • separable_conv2d
        • experimental
          • Overview
          • keras_style_scope
          • set_keras_style
      • linalg
        • Overview
        • experimental
          • Overview
      • lite
        • Overview
        • OpHint
        • OpHint.OpHintArgumentTracker
        • TFLiteConverter
        • TocoConverter
        • toco_convert
        • constants
          • Overview
        • experimental
          • Overview
          • convert_op_hints_to_stubs
          • authoring
            • Overview
      • logging
        • Overview
        • TaskLevelStatusMessage
        • debug
        • error
        • fatal
        • flush
        • get_verbosity
        • info
        • log
        • log_every_n
        • log_first_n
        • log_if
        • set_verbosity
        • vlog
        • warn
        • warning
      • lookup
        • Overview
        • StaticHashTable
        • StaticVocabularyTable
        • experimental
          • Overview
      • losses
        • Overview
        • Reduction
        • absolute_difference
        • add_loss
        • compute_weighted_loss
        • cosine_distance
        • get_losses
        • get_regularization_loss
        • get_regularization_losses
        • get_total_loss
        • hinge_loss
        • huber_loss
        • log_loss
        • mean_pairwise_squared_error
        • mean_squared_error
        • sigmoid_cross_entropy
        • softmax_cross_entropy
        • sparse_softmax_cross_entropy
      • manip
        • Overview
      • math
        • Overview
        • in_top_k
        • log_softmax
        • softmax
        • special
          • Overview
      • metrics
        • Overview
        • accuracy
        • auc
        • average_precision_at_k
        • false_negatives
        • false_negatives_at_thresholds
        • false_positives
        • false_positives_at_thresholds
        • mean
        • mean_absolute_error
        • mean_cosine_distance
        • mean_iou
        • mean_per_class_accuracy
        • mean_relative_error
        • mean_squared_error
        • mean_tensor
        • percentage_below
        • precision
        • precision_at_k
        • precision_at_thresholds
        • precision_at_top_k
        • recall
        • recall_at_k
        • recall_at_thresholds
        • recall_at_top_k
        • root_mean_squared_error
        • sensitivity_at_specificity
        • sparse_average_precision_at_k
        • sparse_precision_at_k
        • specificity_at_sensitivity
        • true_negatives
        • true_negatives_at_thresholds
        • true_positives
        • true_positives_at_thresholds
      • mixed_precision
        • Overview
        • MixedPrecisionLossScaleOptimizer
        • disable_mixed_precision_graph_rewrite
        • enable_mixed_precision_graph_rewrite
        • experimental
          • Overview
      • mlir
        • Overview
        • experimental
          • Overview
      • nest
        • Overview
      • nn
        • Overview
        • avg_pool
        • batch_norm_with_global_normalization
        • bidirectional_dynamic_rnn
        • conv1d
        • conv2d
        • conv2d_backprop_filter
        • conv2d_backprop_input
        • conv2d_transpose
        • conv3d
        • conv3d_backprop_filter
        • conv3d_transpose
        • convolution
        • crelu
        • ctc_beam_search_decoder
        • ctc_loss
        • ctc_loss_v2
        • depthwise_conv2d
        • depthwise_conv2d_native
        • dilation2d
        • dropout
        • dynamic_rnn
        • embedding_lookup
        • embedding_lookup_sparse
        • erosion2d
        • fractional_avg_pool
        • fractional_max_pool
        • fused_batch_norm
        • max_pool
        • max_pool_with_argmax
        • moments
        • nce_loss
        • pool
        • quantized_avg_pool
        • quantized_conv2d
        • quantized_max_pool
        • quantized_relu_x
        • raw_rnn
        • relu_layer
        • safe_embedding_lookup_sparse
        • sampled_softmax_loss
        • separable_conv2d
        • sigmoid_cross_entropy_with_logits
        • softmax_cross_entropy_with_logits
        • softmax_cross_entropy_with_logits_v2
        • sparse_softmax_cross_entropy_with_logits
        • static_bidirectional_rnn
        • static_rnn
        • static_state_saving_rnn
        • sufficient_statistics
        • weighted_cross_entropy_with_logits
        • weighted_moments
        • xw_plus_b
        • experimental
          • Overview
        • rnn_cell
          • Overview
          • BasicLSTMCell
          • BasicRNNCell
          • DeviceWrapper
          • DropoutWrapper
          • GRUCell
          • LSTMCell
          • LSTMStateTuple
          • MultiRNNCell
          • RNNCell
          • ResidualWrapper
      • profiler
        • Overview
        • AdviceProto
        • AdviceProto.Checker
        • AdviceProto.CheckersEntry
        • GraphNodeProto
        • GraphNodeProto.InputShapesEntry
        • MultiGraphNodeProto
        • OpLogProto
        • OpLogProto.IdToStringEntry
        • ProfileOptionBuilder
        • Profiler
        • advise
        • profile
        • write_op_log
      • python_io
        • Overview
      • quantization
        • Overview
      • queue
        • Overview
      • ragged
        • Overview
        • RaggedTensorValue
        • constant_value
        • placeholder
      • random
        • Overview
        • stateless_multinomial
        • experimental
          • Overview
      • raw_ops
        • Overview
      • resource_loader
        • Overview
        • get_data_files_path
        • get_path_to_datafile
        • get_root_dir_with_all_resources
        • load_resource
        • readahead_file_path
      • saved_model
        • Overview
        • Builder
        • build_signature_def
        • build_tensor_info
        • classification_signature_def
        • contains_saved_model
        • get_tensor_from_tensor_info
        • is_valid_signature
        • load
        • main_op_with_restore
        • predict_signature_def
        • regression_signature_def
        • simple_save
        • builder
          • Overview
        • constants
          • Overview
        • experimental
          • Overview
        • loader
          • Overview
        • main_op
          • Overview
          • main_op
        • signature_constants
          • Overview
        • signature_def_utils
          • Overview
          • MethodNameUpdater
        • tag_constants
          • Overview
        • utils
          • Overview
      • sets
        • Overview
      • signal
        • Overview
      • sparse
        • Overview
      • spectral
        • Overview
      • strings
        • Overview
        • length
        • split
        • substr
      • summary
        • Overview
        • FileWriter
        • FileWriterCache
        • SummaryDescription
        • TaggedRunMetadata
        • all_v2_summary_ops
        • audio
        • get_summary_description
        • histogram
        • image
        • initialize
        • merge
        • merge_all
        • scalar
        • tensor_summary
        • text
      • sysconfig
        • Overview
      • test
        • Overview
        • StubOutForTesting
        • assert_equal_graph_def
        • compute_gradient
        • compute_gradient_error
        • get_temp_dir
        • test_src_dir_path
      • tpu
        • Overview
        • CrossShardOptimizer
        • PaddingSpec
        • batch_parallel
        • bfloat16_scope
        • core
        • cross_replica_sum
        • initialize_system
        • outside_compilation
        • replicate
        • rewrite
        • shard
        • shutdown_system
        • experimental
          • Overview
          • AdagradParameters
          • AdamParameters
          • FtrlParameters
          • StochasticGradientDescentParameters
          • embedding_column
          • shared_embedding_columns
          • embedding
            • Overview
      • train
        • Overview
        • AdadeltaOptimizer
        • AdagradDAOptimizer
        • AdagradOptimizer
        • AdamOptimizer
        • Checkpoint
        • ChiefSessionCreator
        • FtrlOptimizer
        • GradientDescentOptimizer
        • LooperThread
        • MomentumOptimizer
        • MonitoredSession
        • MonitoredSession.StepContext
        • MonitoredTrainingSession
        • NewCheckpointReader
        • Optimizer
        • ProximalAdagradOptimizer
        • ProximalGradientDescentOptimizer
        • QueueRunner
        • RMSPropOptimizer
        • Saver
        • SaverDef
        • Scaffold
        • SessionCreator
        • SessionManager
        • SingularMonitoredSession
        • Supervisor
        • SyncReplicasOptimizer
        • WorkerSessionCreator
        • add_queue_runner
        • assert_global_step
        • basic_train_loop
        • batch
        • batch_join
        • checkpoint_exists
        • cosine_decay
        • cosine_decay_restarts
        • create_global_step
        • do_quantize_training_on_graphdef
        • exponential_decay
        • export_meta_graph
        • generate_checkpoint_state_proto
        • get_checkpoint_mtimes
        • get_global_step
        • get_or_create_global_step
        • global_step
        • import_meta_graph
        • init_from_checkpoint
        • input_producer
        • inverse_time_decay
        • limit_epochs
        • linear_cosine_decay
        • maybe_batch
        • maybe_batch_join
        • maybe_shuffle_batch
        • maybe_shuffle_batch_join
        • natural_exp_decay
        • noisy_linear_cosine_decay
        • piecewise_constant
        • polynomial_decay
        • range_input_producer
        • remove_checkpoint
        • replica_device_setter
        • sdca_fprint
        • sdca_optimizer
        • sdca_shrink_l1
        • shuffle_batch
        • shuffle_batch_join
        • slice_input_producer
        • start_queue_runners
        • string_input_producer
        • summary_iterator
        • update_checkpoint_state
        • warm_start
        • experimental
          • Overview
        • queue_runner
          • Overview
      • types
        • Overview
        • experimental
          • Overview
      • user_ops
        • Overview
        • my_fact
      • version
        • Overview
      • xla
        • Overview
        • experimental
          • Overview
  • tf.config
    • Overview
    • LogicalDevice
    • LogicalDeviceConfiguration
    • PhysicalDevice
    • experimental_connect_to_cluster
    • experimental_connect_to_host
    • experimental_functions_run_eagerly
    • experimental_run_functions_eagerly
    • functions_run_eagerly
    • get_logical_device_configuration
    • get_soft_device_placement
    • get_visible_devices
    • list_logical_devices
    • list_physical_devices
    • run_functions_eagerly
    • set_logical_device_configuration
    • set_soft_device_placement
    • set_visible_devices
    • experimental
      • Overview
      • ClusterDeviceFilters
      • disable_mlir_bridge
      • disable_mlir_graph_optimization
      • enable_mlir_bridge
      • enable_mlir_graph_optimization
      • enable_op_determinism
      • enable_tensor_float_32_execution
      • get_device_details
      • get_device_policy
      • get_memory_growth
      • get_memory_info
      • get_memory_usage
      • get_synchronous_execution
      • reset_memory_stats
      • set_device_policy
      • set_memory_growth
      • set_synchronous_execution
      • tensor_float_32_execution_enabled
    • optimizer
      • Overview
      • get_experimental_options
      • get_jit
      • set_experimental_options
      • set_jit
    • threading
      • Overview
      • get_inter_op_parallelism_threads
      • get_intra_op_parallelism_threads
      • set_inter_op_parallelism_threads
      • set_intra_op_parallelism_threads
  • tf.data
    • Overview
    • Dataset
    • DatasetSpec
    • FixedLengthRecordDataset
    • Iterator
    • IteratorSpec
    • Options
    • TFRecordDataset
    • TextLineDataset
    • ThreadingOptions
    • experimental
      • Overview
      • AutoShardPolicy
      • AutotuneAlgorithm
      • AutotuneOptions
      • CheckpointInputPipelineHook
      • Counter
      • CsvDataset
      • DatasetInitializer
      • DistributeOptions
      • ExternalStatePolicy
      • OptimizationOptions
      • RandomDataset
      • Reducer
      • SqlDataset
      • TFRecordWriter
      • assert_cardinality
      • bucket_by_sequence_length
      • cardinality
      • choose_from_datasets
      • copy_to_device
      • dense_to_ragged_batch
      • dense_to_sparse_batch
      • enable_debug_mode
      • enumerate_dataset
      • from_variant
      • get_next_as_optional
      • get_single_element
      • get_structure
      • group_by_reducer
      • group_by_window
      • ignore_errors
      • index_table_from_dataset
      • load
      • make_batched_features_dataset
      • make_csv_dataset
      • make_saveable_from_iterator
      • map_and_batch
      • parallel_interleave
      • parse_example_dataset
      • prefetch_to_device
      • rejection_resample
      • sample_from_datasets
      • save
      • scan
      • shuffle_and_repeat
      • snapshot
      • table_from_dataset
      • take_while
      • to_variant
      • unbatch
      • unique
      • service
        • Overview
        • DispatchServer
        • DispatcherConfig
        • ShardingPolicy
        • WorkerConfig
        • WorkerServer
        • distribute
        • from_dataset_id
        • register_dataset
  • tf.debugging
    • Overview
    • Assert
    • assert_all_finite
    • assert_equal
    • assert_greater
    • assert_greater_equal
    • assert_integer
    • assert_less
    • assert_less_equal
    • assert_near
    • assert_negative
    • assert_non_negative
    • assert_non_positive
    • assert_none_equal
    • assert_positive
    • assert_proper_iterable
    • assert_rank
    • assert_rank_at_least
    • assert_rank_in
    • assert_same_float_dtype
    • assert_scalar
    • assert_shapes
    • assert_type
    • check_numerics
    • disable_check_numerics
    • disable_traceback_filtering
    • enable_check_numerics
    • enable_traceback_filtering
    • get_log_device_placement
    • is_numeric_tensor
    • is_traceback_filtering_enabled
    • set_log_device_placement
    • experimental
      • Overview
      • disable_dump_debug_info
      • enable_dump_debug_info
  • tf.distribute
    • Overview
    • CrossDeviceOps
    • DistributedDataset
    • DistributedIterator
    • DistributedValues
    • HierarchicalCopyAllReduce
    • InputContext
    • InputOptions
    • InputReplicationMode
    • MirroredStrategy
    • MultiWorkerMirroredStrategy
    • NcclAllReduce
    • OneDeviceStrategy
    • ReduceOp
    • ReductionToOneDevice
    • ReplicaContext
    • RunOptions
    • Server
    • Strategy
    • StrategyExtended
    • TPUStrategy
    • experimental_set_strategy
    • get_replica_context
    • get_strategy
    • has_strategy
    • in_cross_replica_context
    • cluster_resolver
      • Overview
      • ClusterResolver
      • GCEClusterResolver
      • KubernetesClusterResolver
      • SimpleClusterResolver
      • SlurmClusterResolver
      • TFConfigClusterResolver
      • TPUClusterResolver
      • UnionResolver
    • coordinator
      • Overview
    • experimental
      • Overview
      • CentralStorageStrategy
      • CollectiveHints
      • CommunicationImplementation
      • CommunicationOptions
      • MultiWorkerMirroredStrategy
      • ParameterServerStrategy
      • TPUStrategy
      • ValueContext
      • coordinator
        • Overview
        • ClusterCoordinator
        • PerWorkerValues
        • RemoteValue
      • partitioners
        • Overview
        • FixedShardsPartitioner
        • MaxSizePartitioner
        • MinSizePartitioner
        • Partitioner
      • rpc
        • Overview
        • Client
        • Server
  • tf.dtypes
    • Overview
    • DType
    • as_dtype
    • complex
    • saturate_cast
  • tf.errors
    • Overview
    • AbortedError
    • AlreadyExistsError
    • CancelledError
    • DataLossError
    • DeadlineExceededError
    • FailedPreconditionError
    • InternalError
    • InvalidArgumentError
    • NotFoundError
    • OpError
    • OperatorNotAllowedInGraphError
    • OutOfRangeError
    • PermissionDeniedError
    • ResourceExhaustedError
    • UnauthenticatedError
    • UnavailableError
    • UnimplementedError
    • UnknownError
  • tf.estimator
    • Overview
    • BaselineClassifier
    • BaselineEstimator
    • BaselineRegressor
    • BestExporter
    • BinaryClassHead
    • BoostedTreesClassifier
    • BoostedTreesEstimator
    • BoostedTreesRegressor
    • CheckpointSaverHook
    • CheckpointSaverListener
    • DNNClassifier
    • DNNEstimator
    • DNNLinearCombinedClassifier
    • DNNLinearCombinedEstimator
    • DNNLinearCombinedRegressor
    • DNNRegressor
    • Estimator
    • EstimatorSpec
    • EvalSpec
    • Exporter
    • FeedFnHook
    • FinalExporter
    • FinalOpsHook
    • GlobalStepWaiterHook
    • Head
    • LatestExporter
    • LinearClassifier
    • LinearEstimator
    • LinearRegressor
    • LoggingTensorHook
    • LogisticRegressionHead
    • ModeKeys
    • MultiClassHead
    • MultiHead
    • MultiLabelHead
    • NanLossDuringTrainingError
    • NanTensorHook
    • PoissonRegressionHead
    • ProfilerHook
    • RegressionHead
    • RunConfig
    • SecondOrStepTimer
    • SessionRunArgs
    • SessionRunContext
    • SessionRunHook
    • SessionRunValues
    • StepCounterHook
    • StopAtStepHook
    • SummarySaverHook
    • TrainSpec
    • VocabInfo
    • WarmStartSettings
    • add_metrics
    • classifier_parse_example_spec
    • regressor_parse_example_spec
    • train_and_evaluate
    • experimental
      • Overview
      • InMemoryEvaluatorHook
      • LinearSDCA
      • RNNClassifier
      • RNNEstimator
      • build_raw_supervised_input_receiver_fn
      • call_logit_fn
      • make_early_stopping_hook
      • make_stop_at_checkpoint_step_hook
      • stop_if_higher_hook
      • stop_if_lower_hook
      • stop_if_no_decrease_hook
      • stop_if_no_increase_hook
    • export
      • Overview
      • ClassificationOutput
      • EvalOutput
      • ExportOutput
      • PredictOutput
      • RegressionOutput
      • ServingInputReceiver
      • TensorServingInputReceiver
      • build_parsing_serving_input_receiver_fn
      • build_raw_serving_input_receiver_fn
  • tf.experimental
    • Overview
    • BatchableExtensionType
    • ExtensionType
    • ExtensionTypeBatchEncoder
    • Optional
    • async_clear_error
    • async_scope
    • dispatch_for_api
    • dispatch_for_binary_elementwise_apis
    • dispatch_for_unary_elementwise_apis
    • function_executor_type
    • register_filesystem_plugin
    • unregister_dispatch_for
    • dlpack
      • Overview
      • from_dlpack
      • to_dlpack
    • numpy
      • Overview
      • abs
      • absolute
      • add
      • all
      • allclose
      • amax
      • amin
      • angle
      • any
      • append
      • arange
      • arccos
      • arccosh
      • arcsin
      • arcsinh
      • arctan
      • arctan2
      • arctanh
      • argmax
      • argmin
      • argsort
      • around
      • array
      • array_equal
      • asanyarray
      • asarray
      • ascontiguousarray
      • atleast_1d
      • atleast_2d
      • atleast_3d
      • average
      • bitwise_and
      • bitwise_not
      • bitwise_or
      • bitwise_xor
      • bool_
      • broadcast_arrays
      • broadcast_to
      • cbrt
      • ceil
      • clip
      • complex128
      • complex64
      • compress
      • concatenate
      • conj
      • conjugate
      • copy
      • cos
      • cosh
      • count_nonzero
      • cross
      • cumprod
      • cumsum
      • deg2rad
      • diag
      • diag_indices
      • diagflat
      • diagonal
      • diff
      • divide
      • divmod
      • dot
      • dsplit
      • dstack
      • einsum
      • empty
      • empty_like
      • equal
      • exp
      • exp2
      • expand_dims
      • experimental_enable_numpy_behavior
      • expm1
      • eye
      • fabs
      • finfo
      • fix
      • flip
      • fliplr
      • flipud
      • float16
      • float32
      • float64
      • float_power
      • floor
      • floor_divide
      • full
      • full_like
      • gcd
      • geomspace
      • greater
      • greater_equal
      • heaviside
      • hsplit
      • hstack
      • hypot
      • identity
      • iinfo
      • imag
      • inexact
      • inner
      • int16
      • int32
      • int64
      • int8
      • isclose
      • iscomplex
      • iscomplexobj
      • isfinite
      • isinf
      • isnan
      • isneginf
      • isposinf
      • isreal
      • isrealobj
      • isscalar
      • issubdtype
      • ix_
      • kron
      • lcm
      • less
      • less_equal
      • linspace
      • log
      • log10
      • log1p
      • log2
      • logaddexp
      • logaddexp2
      • logical_and
      • logical_not
      • logical_or
      • logical_xor
      • logspace
      • matmul
      • max
      • maximum
      • mean
      • meshgrid
      • min
      • minimum
      • mod
      • moveaxis
      • multiply
      • nanmean
      • nanprod
      • nansum
      • ndim
      • negative
      • nextafter
      • nonzero
      • not_equal
      • object_
      • ones
      • ones_like
      • outer
      • pad
      • polyval
      • positive
      • power
      • prod
      • promote_types
      • ptp
      • rad2deg
      • ravel
      • real
      • reciprocal
      • remainder
      • repeat
      • reshape
      • result_type
      • roll
      • rot90
      • round
      • select
      • shape
      • sign
      • signbit
      • s