B ӻdV@s4dZddlZddlZddlZddlZddlZddlZddlZddlZ ddl m Z ddl m Z ddlmZddlmZddlmZddlmZdd lmZdd lmZdd lmZdd lmZdd lmZddlmZddlm Z ddlm!Z!ddlm"Z"ddlm#Z#ddl$m%Z%ddl$m&Z&ddl$m'Z'ddl$m(Z(ddl)m*Z*ddl)m+Z+ddl)m,Z,ddl)m-Z.ddl/m0Z0ddl/m1Z1ddl/m2Z2ddl3m4Z4ddl5m6Z6dd l5m7Z7dd!l5m8Z8dd"l5m9Z9dd#l5m:Z:dd$l5m;Z;dd%lm?Z?dd'l@mAZAdd(lBmCZCdd)lBmDZDdd*lBmEZFdd+lGmHZHdd,lImJZJdd-lKmLZLdd.lMmNZNdd/lMmOZPdd0lMmQZQdd1lRmSZSdd2lRmTZTdd3lUmVZVdd4lUmWZWdd5lXmYZYe6Zd6e[d7Z\d8Z]e j^e!j_eJj`fZaeWd9Gd:d;d;eAjbe;jcZdGdd?d?edZfGd@dAdAedZgdBdCZhdDdEZie+jjZjdS)Fz=Contains the base Layer class, from which all layers inherit.N) json_format) node_def_pb2)tf2)ag_ctx)api)distribution_strategy_context)backprop)context) def_function) constant_op)dtypes) func_graph)ops) sparse_tensor) tensor_spec) tensor_util)backend) constraints) initializers) regularizers)base_layer_utils) input_spec) keras_tensor)node)autocast_variable)loss_scale_optimizer)policy)layer_serialization) generic_utils) layer_utils)object_identity) tf_inspect)tf_utils) version_utils) to_snake_case)is_tensor_or_tensor_list)module) array_ops)math_ops) variables) np_arrays) ragged_tensor) tf_logging) autotrackable)base)data_structures)compat)nest)get_canonical_name_for_symbol) keras_export) doc_controls metrics_modztensorflow.python.keras.metricsZ tf_op_layer_zkeras.layers.Layerc s@eZdZdZeedejj Z dZ ddZ ddZ e jdd d Ze jejd d ZejddZejddZejd d d d d d d d ejjejjf ddZejddZeddZddZ ejddZ!ddZ"ddZ#ejdd d!Z$d"d#Z%d$d%Z&d&d'Z'd(d)Z(e)d*d+Z*e)d,d-Z+e)d.d/Z,e,j-d0d/Z,e)d1d2Z.e)ej/d3d4Z0e0j-d5d4Z0e)d6d7Z1e1j-d8d7Z1e)d9d:Z2e2j-d;d:Z2e)dd=Z3e)d?d@Z4e)dAdBZ5e)dCdDZ6e)ej7dEdFZ8e)dGdHZ9dIdJZ:dKdLZ;e)dMdNZdSdTZ?dUdVZ@ej7dWdXZAej7dYdZZBej7d[d\ZCej/d]d^ZDej/d_d`ZEe)ej/dadbZFe)ej/dcddZGej/dedfZHej/dgdhZIej/didjZJej/dkdlZKe)dmdnZLe)dodpZMe)ej/dqdrZNdsdtZOe)ej/dudvZPe)ej/dwdxZQe)ej/dydzZRej/d{d|ZSej/d}d~ZTe)ej7ddZUe)ej7ddZVe)ej7ddZWe)ddZXeXj-e jddZXe)ddZYeYj-e jddZYddZZe)ddZ[e)ddZ\e)ddZ]e)ddZ^dddZ_ddZ`ddZae)ddZbebj-ddZbddZcdddZddddZeddZfddZgddZhddZiddZjdddZkdddZldddZmddZnddZoddZpddZqddZrddZse)ddZte jddÄZufddńZvfddDŽZwddɄZxddd˄Zyddd̈́ZzddτZ{dddфZ|e)e}j~ddӄZe)e}j~ddՄZe)e}j~ddׄZe)e}j~ddلZe)e}j~ddۄZe)dd݄Zej-dd݄ZddZddZe jddZdddZe)ddZe)ddZe)ddZdfdd Ze)ddZddZddZZS(Layera'This is the class from which all layers inherit. A layer is a callable object that takes as input one or more tensors and that outputs one or more tensors. It involves *computation*, defined in the `call()` method, and a *state* (weight variables), defined either in the constructor `__init__()` or in the `build()` method. Users will just instantiate a layer and then treat it as a callable. Args: trainable: Boolean, whether the layer's variables should be trainable. name: String name of the layer. dtype: The dtype of the layer's computations and weights. Can also be a `tf.keras.mixed_precision.Policy`, which allows the computation and weight dtype to differ. Default of `None` means to use `tf.keras.mixed_precision.global_policy()`, which is a float32 policy unless set to different value. dynamic: Set this to `True` if your layer should only be run eagerly, and should not be used to generate a static computation graph. This would be the case for a Tree-RNN or a recursive network, for example, or generally for any layer that manipulates tensors using Python control flow. If `False`, we assume that the layer can safely be used to generate a static computation graph. Attributes: name: The name of the layer (string). dtype: The dtype of the layer's weights. variable_dtype: Alias of `dtype`. compute_dtype: The dtype of the layer's computations. Layers automatically cast inputs to this dtype which causes the computations and output to also be in this dtype. When mixed precision is used with a `tf.keras.mixed_precision.Policy`, this will be different than `variable_dtype`. dtype_policy: The layer's dtype policy. See the `tf.keras.mixed_precision.Policy` documentation for details. trainable_weights: List of variables to be included in backprop. non_trainable_weights: List of variables that should not be included in backprop. weights: The concatenation of the lists trainable_weights and non_trainable_weights (in this order). trainable: Whether the layer should be trained (boolean), i.e. whether its potentially-trainable weights should be returned as part of `layer.trainable_weights`. input_spec: Optional (list of) `InputSpec` object(s) specifying the constraints on inputs that can be accepted by the layer. We recommend that descendants of `Layer` implement the following methods: * `__init__()`: Defines custom layer attributes, and creates layer state variables that do not depend on input shapes, using `add_weight()`. * `build(self, input_shape)`: This method can be used to create weights that depend on the shape(s) of the input(s), using `add_weight()`. `__call__()` will automatically build the layer (if it has not been built yet) by calling `build()`. * `call(self, inputs, *args, **kwargs)`: Called in `__call__` after making sure `build()` has been called. `call()` performs the logic of applying the layer to the input tensors (which should be passed in as argument). Two reserved keyword arguments you can optionally use in `call()` are: - `training` (boolean, whether the call is in inference mode or training mode). See more details in [the layer/model subclassing guide]( https://www.tensorflow.org/guide/keras/custom_layers_and_models#privileged_training_argument_in_the_call_method) - `mask` (boolean tensor encoding masked timesteps in the input, used in RNN layers). See more details in [the layer/model subclassing guide]( https://www.tensorflow.org/guide/keras/custom_layers_and_models#privileged_mask_argument_in_the_call_method) A typical signature for this method is `call(self, inputs)`, and user could optionally add `training` and `mask` if the layer need them. `*args` and `**kwargs` is only useful for future extension when more input parameters are planned to be added. * `get_config(self)`: Returns a dictionary containing the configuration used to initialize this layer. If the keys differ from the arguments in `__init__`, then override `from_config(self)` as well. This method is used when saving the layer or a model that contains this layer. Examples: Here's a basic example: a layer with two variables, `w` and `b`, that returns `y = w . x + b`. It shows how to implement `build()` and `call()`. Variables set as attributes of a layer are tracked as weights of the layers (in `layer.weights`). ```python class SimpleDense(Layer): def __init__(self, units=32): super(SimpleDense, self).__init__() self.units = units def build(self, input_shape): # Create the state of the layer (weights) w_init = tf.random_normal_initializer() self.w = tf.Variable( initial_value=w_init(shape=(input_shape[-1], self.units), dtype='float32'), trainable=True) b_init = tf.zeros_initializer() self.b = tf.Variable( initial_value=b_init(shape=(self.units,), dtype='float32'), trainable=True) def call(self, inputs): # Defines the computation from inputs to outputs return tf.matmul(inputs, self.w) + self.b # Instantiates the layer. linear_layer = SimpleDense(4) # This will also call `build(input_shape)` and create the weights. y = linear_layer(tf.ones((2, 2))) assert len(linear_layer.weights) == 2 # These weights are trainable, so they're listed in `trainable_weights`: assert len(linear_layer.trainable_weights) == 2 ``` Note that the method `add_weight()` offers a shortcut to create weights: ```python class SimpleDense(Layer): def __init__(self, units=32): super(SimpleDense, self).__init__() self.units = units def build(self, input_shape): self.w = self.add_weight(shape=(input_shape[-1], self.units), initializer='random_normal', trainable=True) self.b = self.add_weight(shape=(self.units,), initializer='random_normal', trainable=True) def call(self, inputs): return tf.matmul(inputs, self.w) + self.b ``` Besides trainable weights, updated via backpropagation during training, layers can also have non-trainable weights. These weights are meant to be updated manually during `call()`. Here's a example layer that computes the running sum of its inputs: ```python class ComputeSum(Layer): def __init__(self, input_dim): super(ComputeSum, self).__init__() # Create a non-trainable weight. self.total = tf.Variable(initial_value=tf.zeros((input_dim,)), trainable=False) def call(self, inputs): self.total.assign_add(tf.reduce_sum(inputs, axis=0)) return self.total my_sum = ComputeSum(2) x = tf.ones((2, 2)) y = my_sum(x) print(y.numpy()) # [2. 2.] y = my_sum(x) print(y.numpy()) # [4. 4.] assert my_sum.weights == [my_sum.total] assert my_sum.non_trainable_weights == [my_sum.total] assert my_sum.trainable_weights == [] ``` For more information about creating layers, see the guide [Making new Layers and Models via subclassing]( https://www.tensorflow.org/guide/keras/custom_layers_and_models) )_obj_reference_counts_dictFcCs6t|jddd}|dk r"d|S|jjd|jjS)NkerasT)api_nameZadd_prefix_to_v1_namesztf.{}.)r2 __class__format __module____name__)selfcanonical_namerA[/var/www/html/venv/lib/python3.7/site-packages/tensorflow/python/keras/engine/base_layer.py_get_cell_name$s  zLayer._get_cell_namecCsBd|_d|_d|_t|dds>d|_t|ddr8d|_nd|_dS)NFZ_disable_keras_instrumentationTZ_is_model_for_instrumentation)Z_instrumented_keras_apiZ_instrumented_keras_layer_classZ_instrumented_keras_model_classgetattr)r?rArArB_instrument_layer_creation+s  z Layer._instrument_layer_creationTNc Ks|ddddddddh}t||||_d |_d |_d|_d|_d|_t |j  |_ | |t |dd|_|d g|d gg|_t|_g|_g|_g|_t|_|||dt|_|d gg|_g|_ |!||_"d|kr"d|kr"|df|d<d|ks6d|krd|krNt#|d}n4d|krd|krl|d}nd}|ft#|d}||_$|dd|_%d |_&d |_'dS)NZ input_dim input_shapebatch_input_shape batch_sizeweightsactivity_regularizerautocastimplementationF_trainable_weights_non_trainable_weights_self_tracked_trackablesT)(rErZvalidate_kwargs _trainable _statefulbuilt _input_spec_build_input_shape_saved_model_inputs_spec is_default compute_mask_supports_masking_init_set_namergetpop_activity_regularizer_maybe_create_attributeZ_updates threadinglocal _thread_local_callable_losses_losses_metricsLock _metrics_lock_set_dtype_policyrZv2_dtype_behavior_enabled _autocast_inbound_nodes_value_outbound_nodes_value_init_call_fn_args_dynamictuple_batch_input_shape_initial_weights_auto_track_sub_layersZ#_preserve_input_structure_in_config) r? trainablenamedtypedynamickwargsZallowed_kwargsrGrHrArArB__init__6sd             zLayer.__init__cCst|jds||_d|_dS)aCreates the variables of the layer (optional, for subclass implementers). This is a method that implementers of subclasses of `Layer` or `Model` can override if they need a state-creation step in-between layer instantiation and layer call. This is typically used to create the weights of `Layer` subclasses. Args: input_shape: Instance of `TensorShape`, or list of instances of `TensorShape` if the layer expects a list of inputs (one instance per input). _is_defaultTN)hasattrbuildrTrR)r?rFrArArBrxs z Layer.buildcOs|S)aThis is where the layer's logic lives. Note here that `call()` method in `tf.keras` is little bit different from `keras` API. In `keras` API, you can pass support masking for layers as additional arguments. Whereas `tf.keras` has `compute_mask()` method to support masking. Args: inputs: Input tensor, or dict/list/tuple of input tensors. The first positional `inputs` argument is subject to special rules: - `inputs` must be explicitly passed. A layer cannot have zero arguments, and `inputs` cannot be provided via the default value of a keyword argument. - NumPy array or Python scalar values in `inputs` get cast as tensors. - Keras mask metadata is only collected from `inputs`. - Layers are built (`build(input_shape)` method) using shape info from `inputs` only. - `input_spec` compatibility is only checked against `inputs`. - Mixed precision input casting is only applied to `inputs`. If a layer has tensor arguments in `*args` or `**kwargs`, their casting behavior in mixed precision should be handled manually. - The SavedModel input specification is generated using `inputs` only. - Integration with various ecosystem packages like TFMOT, TFLite, TF.js, etc is only supported for `inputs` and not for tensors in positional and keyword arguments. *args: Additional positional arguments. May contain tensors, although this is not recommended, for the reasons above. **kwargs: Additional keyword arguments. May contain tensors, although this is not recommended, for the reasons above. The following optional keyword arguments are reserved: - `training`: Boolean scalar tensor of Python boolean indicating whether the `call` is meant for training or inference. - `mask`: Boolean input mask. If the layer's `call()` method takes a `mask` argument, its default value will be set to the mask generated for `inputs` by the previous layer (if `input` did come from a layer that generated a corresponding mask, i.e. if it came from a Keras layer with masking support). Returns: A tensor or list/tuple of tensors. rA)r?inputsargsrtrArArBcalls+z Layer.callcCs>t|tjr|}n t|}|r.|j|n |j||S)aAdds a Trackable object to this layer's state. Args: trackable_object: The tf.tracking.Trackable object to add. trainable: Boolean, whether the variable should be part of the layer's "trainable_variables" (e.g. variables, biases) or "non_trainable_variables" (e.g. BatchNorm mean and variance). Returns: The TrackableWeightHandler used to track this object. ) isinstancerTrackableWeightHandlerrMappendrN)r?Ztrackable_objectrphandlerrArArB_add_trackables   zLayer._add_trackablec  sf|dkr d}| ddx| D]} | dkrtd| qW| dd} | dd}| d d}|dkrr|jppt}t|}|jjdkr| t |j j t|}t|}t|}| tjjkr|rtd qd }n |dkrd}|dkrH|jrtd }nD|js|js|jr(td }n d| krHtd||j |j f| dtj}|r|jj|jjkr|jr|fdd}|dk rtdd}|j|||d|||||| | | |d }|dk r|j d|j d}|!|||t"|r8xd|D]0}t#||r$|j$%|n |j&%|qWn*t#||rV|j$%|n |j&%||S)agAdds a new variable to the layer. Args: name: Variable name. shape: Variable shape. Defaults to scalar if unspecified. dtype: The type of the variable. Defaults to `self.dtype`. initializer: Initializer instance (callable). regularizer: Regularizer instance (callable). trainable: Boolean, whether the variable should be part of the layer's "trainable_variables" (e.g. variables, biases) or "non_trainable_variables" (e.g. BatchNorm mean and variance). Note that `trainable` cannot be `True` if `synchronization` is set to `ON_READ`. constraint: Constraint instance (callable). use_resource: Whether to use `ResourceVariable`. synchronization: Indicates when a distributed a variable will be aggregated. Accepted values are constants defined in the class `tf.VariableSynchronization`. By default the synchronization is set to `AUTO` and the current `DistributionStrategy` chooses when to synchronize. If `synchronization` is set to `ON_READ`, `trainable` must not be set to `True`. aggregation: Indicates how a distributed variable will be aggregated. Accepted values are constants defined in the class `tf.VariableAggregation`. **kwargs: Additional keyword arguments. Accepted values are `getter`, `collections`, `experimental_autocast` and `caching_device`. Returns: The variable created. Raises: ValueError: When giving unsupported dtype and no initializer or when trainable has been set to True with synchronization set as `ON_READ`. NrAZ partitioner) collectionsexperimental_autocastcaching_devicegetterzUnknown keyword argument:rrTrzSynchronization value can be set to VariableSynchronization.ON_READ only for non-trainable variables. You have specified trainable=True and synchronization=VariableSynchronization.ON_READ.FZglorot_uniformZzerosrzBAn initializer for variable %s of type %s is required for layer %scs||}t|S)N)rZcreate_autocast_variable)rzrtvariable) old_getterrArBrs z Layer.add_weight..getterzb`caching_device` does not work with mixed precision API. Ignoring user specified `caching_device`.) rqshaper overwrite initializerrr constraintrp use_resourcersynchronization aggregationr:)'r[ TypeErrorrrrfloatxr as_dtype _dtype_policyvariable_dtyperfrPolicy base_dtyperqrrZrr tf_variablesVariableSynchronizationZON_READ ValueError is_floating is_integerZ is_unsignedZis_boolrZ make_variable compute_dtyper,warningZ _add_variable_with_custom_getterfind_handle_weight_regularizationis_split_variabletrack_variablerMr~rN)r?rqrrrr regularizerrprrrrrtkwargZcollections_argrKrrrZ name_in_scopevrA)rrB add_weights/                       zLayer.add_weightcst|jj}|j|jd}t|dr0|j|d<t |j |d<t|drn|j r\|j |d<nd|krn| d| fdd|D}t|dkrt|jd rtd |jj|S) aReturns the config of the layer. A layer config is a Python dictionary (serializable) containing the configuration of a layer. The same layer can be reinstantiated later (without its trained weights) from this configuration. The config of a layer does not include connectivity information, nor the layer class name. These are handled by `Network` (one layer of abstraction above). Note that `get_config()` does not guarantee to return a fresh copy of dict every time it is called. The callers should make a copy of the returned dict if they want to modify it. Returns: Python dictionary. )rqrprmrGrrrscsg|]}|kr|qSrArA).0arg) expected_argsrArB sz$Layer.get_config..rvzNLayer %s has arguments in `__init__` and therefore must override `get_config`.)r!getfullargspecrurzrqrprwrmr serializerrsremovekeyslen get_configNotImplementedErrorr;r>)r?all_argsconfig extra_argsrA)rrBrs"       zLayer.get_configcCs |f|S)aCreates a layer from its config. This method is the reverse of `get_config`, capable of instantiating the same layer from the config dictionary. It does not handle layer connectivity (handled by Network), nor weights (handled by `set_weights`). Args: config: A Python dictionary, typically the output of get_config. Returns: A layer instance. rA)clsrrArArB from_configszLayer.from_configc str|ttjdrtj |dd}fdd}t ||}y|dd}Wn4t k r}zt djj|Wdd}~XYnXWdQRXt d d |St d jjdS) aComputes the output shape of the layer. If the layer has not been built, this method will call `build` on the layer. This assumes that the layer will later be used with inputs that match the input shape provided here. Args: input_shape: Shape tuple (tuple of integers) or list of shape tuples (one per output tensor of the layer). Shape tuples can include None for free dimensions, instead of an integer. Returns: An input shape tuple. _scratch_graphF) to_tuplescstj|jd}d|_|S)N)rrr)r placeholderrr _keras_mask)rph)r?rArB_make_placeholder_likesz:Layer.compute_output_shape.._make_placeholder_like)trainingzWe could not automatically infer the static shape of the layer's output. Please implement the `compute_output_shape` method on your layer (%s).NcSs|jS)N)r)trArArBz,Layer.compute_output_shape..z[Please run in eager mode or implement the `compute_output_shape` method on your layer (%s).)r executing_eagerly _maybe_buildr FuncGraphstrrq as_defaultr"convert_shapesr1 map_structurerrr;r>)r?rFrryoutputserA)r?rBcompute_output_shapes    *zLayer.compute_output_shapecs\dd}t||}||}|jdkrHddt|D}|dtfdd|S) a^Compute the output tensor signature of the layer based on the inputs. Unlike a TensorShape object, a TensorSpec object contains both shape and dtype information for a tensor. This method allows layers to provide output dtype information if it is different from the input dtype. For any layer that doesn't implement this function, the framework will fall back to use `compute_output_shape`, and will assume that the output dtype matches the input dtype. Args: input_signature: Single TensorSpec or nested structure of TensorSpec objects, describing a candidate input for the layer. Returns: Single TensorSpec or nested structure of TensorSpec objects, describing how the layer would transform the provided input. Raises: TypeError: If input_signature contains a non-TensorSpec object. cSs t|tjstd||jS)NzKOnly TensorSpec signature types are supported, but saw signature entry: {}.)r|r TensorSpecrr<r)srArArBcheck_type_return_shape7s z?Layer.compute_output_signature..check_type_return_shapeNcSsg|] }|jqSrA)rr)rrrArArBr@sz2Layer.compute_output_signature..rcstj|dS)N)rrr)rr)r)rrrArBrErz0Layer.compute_output_signature..)r1rr_compute_dtypeflatten)r?input_signaturerrF output_shapeZ input_dtypesrA)rrrBcompute_output_signature!s   zLayer.compute_output_signaturecCsB|jr.tdd|}||}ttj|S|||||SdS)NcSstj|j|jdS)N)rrr)rrrrr)xrArArBrQrz3Layer._keras_tensor_symbolic_call..)rsr1rrr KerasTensor_infer_output_signature)r?ry input_masksrzrtrZoutput_signaturerArArB_keras_tensor_symbolic_callHs z!Layer._keras_tensor_symbolic_callc Cs4|j}t|r,t|s,t|jt}t t |j d}| t tj|}t tj|}t tj|}t tj|}t|Nt|j*||||}||f||}WdQRX|||WdQRX|j|||ddt tj|}WdQRXt|dr.|js.|||~|S)zTODO(kaftan): Docstring.rNF) build_graph _set_inputs)r{r is_subclassedfrom_saved_model autograph tf_convertrcontrol_status_ctxr rrrqrr1rrZkeras_tensor_to_placeholderr name_scope _name_scoperenable_auto_cast_variables_compute_dtype_objectr_maybe_cast_inputs_handle_activity_regularization_set_mask_metadataZkeras_tensor_from_tensorrwryr)r?ryrzrtrcall_fnZ scratch_graphrrArArBrXs:            zLayer._infer_output_signaturecCs@|jssz%Layer.compute_mask..zLayer z9 does not support masking, but was passed an input_mask: N)rXanyr1rrrqr)r?rymaskrArArBrWs zLayer.compute_maskc Ost|dstd|||\}}}t|}t|||||rN|||||St}t dd|Dr~t t |}t|}| ||||\}}|j r|r||d<||||\}}}|js|t} |j||| |dt|j||j| r |j} |j} n|} |} t| |js8|||jrL| ||}t!"|j#| |f||} WdQRX|j$r|%|| |j&r|'|| || |j(dkr|)|| SQRXWdQRXdS)a2Wraps `call`, applying pre- and post-processing steps. Args: *args: Positional arguments to be passed to `self.call`. **kwargs: Keyword arguments to be passed to `self.call`. Returns: Output tensor(s). Note: - The following optional keyword arguments are reserved for specific uses: * `training`: Boolean scalar tensor of Python boolean indicating whether the `call` is meant for training or inference. * `mask`: Boolean input mask. - If the layer's `call` method takes a `mask` argument (as some Keras layers do), its default value will be set to the mask generated for `inputs` by the previous layer (if `input` did come from a layer that generated a corresponding mask, i.e. if it came from a Keras layer with masking support. - If the layer is not built, the method will call `build`. Raises: ValueError: if the layer's `call` method returns None (an invalid value). RuntimeError: if `super().__init__()` was not called in the constructor. r`z.r)layerryrrN)*rw RuntimeError_split_out_first_argr1r _in_functional_construction_mode_functional_construction_callr call_contextrr_convert_numpy_or_python_types_get_input_masks_expects_mask_arg_set_training_modein_call _clear_lossesr renterrassert_input_compatibilityrqr{_namer_autographed_callrZ name_scope_v2rRrrgrrrrr\rrXrrU_set_save_spec) r?rzrtry input_listrrmask_is_implicit training_modeeagerrrrrArArB__call__s`                zLayer.__call__c Cst}tdd|Dr8dd}t||}t|}d}|||||\}} |jrf| rf||d<d}d} d} |d||r| d||} |j s| d| dkr|j dk r|j } n:t rt } t| rt| tj} qt| } n|j} |j r|d| ||\}}d} |j||d| d t|||||} | dkrPtd |jd | rn|jdd||dd \}}|r~| d||f||| } | SQRXdS) Ncss$|]}t|tjtjttfVqdS)N)r|r*rrrr)rrrArArBr#sz6Layer._functional_construction_call..cSs$t|tjtjttfr t|S|S)N)r|r*rrrrr"convert_to_tensor_v2_with_dispatch)rrArArB_convert_non_tensor&s z@Layer._functional_construction_call.._convert_non_tensorFrTr)rryrrzVA layer's `call` method should return a Tensor or a list of Tensors, not None (layer: z).)pop_kwarg_if_none)rrrr1rrrr_call_arg_was_passed_get_call_arg_value_expects_training_argr[rrglobal_learning_phase_is_setlearning_phaser is_tf_typer(castr bool_default_training_arg_set_call_arg_valuerrrrq_set_connectivity_metadata) r?ryrzrtrrrZmask_arg_passed_by_frameworkrrZtraining_valueZ training_arg_passed_by_frameworkrrArArBrs\             z#Layer._functional_construction_callcCsd}|jr|d||r&|d||}|dkr|j}|dk rB|}nFtrt}t|tr^qt |rxt |t j}qt|}n|j}|d|||\}}nd|kr|d}n|j}|||fS)Nr)r rrrrr r r|rrr r(r r rrr[)r?rzrtrrZcall_ctx_trainingrArArBrrs,      zLayer._set_training_modecCs0t|r&t|s&t|jtS|jSdS)N)rrrrrr{rr)r?rArArBrs  zLayer._autographed_callcCs|jjS)zThe dtype of the layer weights. This is equivalent to `Layer.dtype_policy.variable_dtype`. Unless mixed precision is used, this is the same as `Layer.compute_dtype`, the dtype of the layer's computations. )rr)r?rArArBrrsz Layer.dtypecCs|jS)z3Name of the layer (string), set in the constructor.)r)r?rArArBrqsz Layer.namecCs|jS)zBWhether this layer supports computing a mask using `compute_mask`.)rX)r?rArArBsupports_maskingszLayer.supports_maskingcCs ||_dS)N)rX)r?valuerArArBrscCstdd|DS)zBWhether the layer is dynamic (eager-only); set in the constructor.css|] }|jVqdS)N)rk)rrrArArBrsz Layer.dynamic..)r_flatten_layers)r?rArArBrssz Layer.dynamiccCstdd|DS)Ncss|] }|jVqdS)N)rQ)rrrArArBrsz!Layer.stateful..)rr)r?rArArBstatefulszLayer.statefulcCs ||_dS)N)rQ)r?rrArArBrscCs|jS)N)rP)r?rArArBrpszLayer.trainablecCsx|D] }||_q WdS)N)rrP)r?rrrArArBrpscCs|jS)z;Optional regularizer function for the output of this layer.)r\)r?rArArBrJszLayer.activity_regularizercCs ||_dS)z;Optional regularizer function for the output of this layer.N)r\)r?rrArArBrJscCs|jS)a `InputSpec` instance(s) describing the input format for this layer. When you create a layer subclass, you can set `self.input_spec` to enable the layer to run input compatibility checks when it is called. Consider a `Conv2D` layer: it can only be called on a single input tensor of rank 4. As such, you can set, in `__init__()`: ```python self.input_spec = tf.keras.layers.InputSpec(ndim=4) ``` Now, if you try to call the layer on an input that isn't rank 4 (for instance, an input of shape `(2,)`, it will raise a nicely-formatted error: ``` ValueError: Input 0 of layer conv2d is incompatible with the layer: expected ndim=4, found ndim=1. Full shape received: [2] ``` Input checks that can be specified via `input_spec` include: - Structure (e.g. a single input, a list of 2 inputs, etc) - Shape - Rank (ndim) - Dtype For more information, see `tf.keras.layers.InputSpec`. Returns: A `tf.keras.layers.InputSpec` instance, or nested structure thereof. )rS)r?rArArBrs!zLayer.input_speccCs>x2t|D]$}|dk r t|ts td|q W||_dS)Nz:Layer input_spec must be an instance of InputSpec. Got: {})r1rr| InputSpecrr<rS)r?rrrArArBrs  cCs(|jr |d}||j|SgSdS)zList of all trainable weights tracked by this layer. Trainable weights are updated via gradient descent during training. Returns: A list of trainable variables. trainable_variablesN)rp_gather_children_attribute_dedup_weightsrM)r?children_weightsrArArBtrainable_weightss  zLayer.trainable_weightscCs@|jr|d}|j|}n|d}|j|j|}||S)zList of all non-trainable weights tracked by this layer. Non-trainable weights are *not* updated during training. They are expected to be updated manually in `call()`. Returns: A list of non-trainable variables. non_trainable_variablesr))rprrNrMr)r?rnon_trainable_weightsrArArBr s    zLayer.non_trainable_weightscCs |j|jS)z^Returns the list of all layer variables/weights. Returns: A list of variables. )rr)r?rArArBrI5sz Layer.weightscCstdgS)Nz`layer.updates` will be removed in a future version. This property should not be used in TensorFlow 2.0, as `updates` are applied automatically.)warningswarn)r?rArArBupdates>s z Layer.updatescCsrg}xh|D]\}|jr6|jdtjk rB||jn ||jx&|jD]}|}|dk rJ||qJWqW|S)a List of losses added using the `add_loss()` API. Variable regularization tensors are created when this property is accessed, so it is eager safe: accessing `losses` under a `tf.GradientTape` will propagate gradients back to the corresponding variables. Examples: >>> class MyLayer(tf.keras.layers.Layer): ... def call(self, inputs): ... self.add_loss(tf.abs(tf.reduce_mean(inputs))) ... return inputs >>> l = MyLayer() >>> l(np.ones((10, 1))) >>> l.losses [1.0] >>> inputs = tf.keras.Input(shape=(10,)) >>> x = tf.keras.layers.Dense(10)(inputs) >>> outputs = tf.keras.layers.Dense(1)(x) >>> model = tf.keras.Model(inputs, outputs) >>> # Activity regularization. >>> len(model.losses) 0 >>> model.add_loss(tf.abs(tf.reduce_mean(x))) >>> len(model.losses) 1 >>> inputs = tf.keras.Input(shape=(10,)) >>> d = tf.keras.layers.Dense(10, kernel_initializer='ones') >>> x = d(inputs) >>> outputs = tf.keras.layers.Dense(1)(x) >>> model = tf.keras.Model(inputs, outputs) >>> # Weight regularization. >>> model.add_loss(lambda: tf.reduce_mean(d.kernel)) >>> model.losses [] Returns: A list of tensors. rN)r _eager_lossesrZREVIVED_LOSS_PLACEHOLDERextendrbrar~)r?Zcollected_lossesrrZ loss_tensorrArArBlossesFs+  z Layer.lossesc KsJ|dd|r"td|fdd}t|}g}g}g}x|D]}t|rf|t||qF|dkrpqFt |st |t j stj|td}t|st |t j rts||qFt |rF||qFW|j|tj}|r|std|j|x4|D],} t|dd r4|| n |j| qWdS) aAdd loss tensor(s), potentially dependent on layer inputs. Some losses (for instance, activity regularization losses) may be dependent on the inputs passed when calling a layer. Hence, when reusing the same layer on different inputs `a` and `b`, some entries in `layer.losses` may be dependent on `a` and some on `b`. This method automatically keeps track of dependencies. This method can be used inside a subclassed layer or model's `call` function, in which case `losses` should be a Tensor or list of Tensors. Example: ```python class MyLayer(tf.keras.layers.Layer): def call(self, inputs): self.add_loss(tf.abs(tf.reduce_mean(inputs))) return inputs ``` This method can also be called directly on a Functional Model during construction. In this case, any loss Tensors passed to this Model must be symbolic and be able to be traced back to the model's `Input`s. These losses become part of the model's topology and are tracked in `get_config`. Example: ```python inputs = tf.keras.Input(shape=(10,)) x = tf.keras.layers.Dense(10)(inputs) outputs = tf.keras.layers.Dense(1)(x) model = tf.keras.Model(inputs, outputs) # Activity regularization. model.add_loss(tf.abs(tf.reduce_mean(x))) ``` If this is not the case for your loss (if, for example, your loss references a `Variable` of one of the model's layers), you can wrap your loss in a zero-argument lambda. These losses are not tracked as part of the model's topology since they can't be serialized. Example: ```python inputs = tf.keras.Input(shape=(10,)) d = tf.keras.layers.Dense(10) x = d(inputs) outputs = tf.keras.layers.Dense(1)(x) model = tf.keras.Model(inputs, outputs) # Weight regularization. model.add_loss(lambda: tf.reduce_mean(d.kernel)) ``` Args: losses: Loss tensor, or list/tuple of tensors. Rather than tensors, losses may also be zero-argument callables which create a loss tensor. **kwargs: Additional keyword arguments for backward compatibility. Accepted values: inputs - Deprecated, will be automatically inferred. ryNzUnknown keyword arguments: %sc SsVt|r$td |}WdQRX|dkr0dSt|sLtj|td}d|_ |S)z3Tags callable loss tensor as `_unconditional_loss`.N)rrT) callablerrrr rrrrZ_unconditional_loss)lossrArArB _tag_callables  z%Layer.add_loss.._tag_callable)rrz|Expected a symbolic Tensors or a callable for the loss value. Please wrap your loss computation in a zero argument `lambda`._is_graph_networkF)r[rrr1rr$r~ functoolspartialrr r|rrrrrrr"Zis_symbolic_tensorrZis_in_tf_functionrar"rrrr!rDZ_graph_network_add_lossrb) r?r#rtr&Zcallable_lossesZ eager_lossesZsymbolic_lossesr%in_call_contextZ symbolic_lossrArArBadd_losssD=               zLayer.add_losscCs4t|ddsg|j_nx|D] }g|j_q WdS)z)Used every step in eager to reset losses.rON)rDr`r!r)r?rrArArBrs  zLayer._clear_lossesc Cs8g}x.|D]"}|j||jWdQRXqW|S)aList of metrics added using the `add_metric()` API. Example: >>> input = tf.keras.layers.Input(shape=(3,)) >>> d = tf.keras.layers.Dense(2) >>> output = d(input) >>> d.add_metric(tf.reduce_max(output), name='max') >>> d.add_metric(tf.reduce_min(output), name='min') >>> [m.name for m in d.metrics] ['max', 'min'] Returns: A list of `Metric` objects. N)rrer"rc)r?Zcollected_metricsrrArArBmetricss z Layer.metricsc Ksft|}t|dks0t|dkrB|ddkrBtdt|t|d}t|tj}t j }|dkrx|sxt dn |r|j j}|s|st dt||st|d d s8t|dd}| } |r|jn|}|jN||} | r| }n4|r|j|n"tj|t|d dd }|j|WdQRX| rb||n*|rFt d |rPdnd} ||| |dS)aAdds metric tensor to the layer. This method can be used inside the `call()` method of a subclassed layer or model. ```python class MyMetricLayer(tf.keras.layers.Layer): def __init__(self): super(MyMetricLayer, self).__init__(name='my_metric_layer') self.mean = tf.keras.metrics.Mean(name='metric_1') def call(self, inputs): self.add_metric(self.mean(inputs)) self.add_metric(tf.reduce_sum(inputs), name='metric_2') return inputs ``` This method can also be called directly on a Functional Model during construction. In this case, any tensor passed to this Model must be symbolic and be able to be traced back to the model's `Input`s. These metrics become part of the model's topology and are tracked when you save the model via `save()`. ```python inputs = tf.keras.Input(shape=(10,)) x = tf.keras.layers.Dense(10)(inputs) outputs = tf.keras.layers.Dense(1)(x) model = tf.keras.Model(inputs, outputs) model.add_metric(math_ops.reduce_sum(x), name='metric_1') ``` Note: Calling `add_metric()` with the result of a metric object on a Functional Model, as shown in the example below, is not supported. This is because we cannot trace the metric result tensor back to the model's inputs. ```python inputs = tf.keras.Input(shape=(10,)) x = tf.keras.layers.Dense(10)(inputs) outputs = tf.keras.layers.Dense(1)(x) model = tf.keras.Model(inputs, outputs) model.add_metric(tf.keras.metrics.Mean()(x), name='metric_1') ``` Args: value: Metric tensor. name: String metric name. **kwargs: Additional keyword arguments for backward compatibility. Accepted values: `aggregation` - When the `value` tensor provided is not the result of calling a `keras.Metric` instance, it will be aggregated by default using a `keras.Metric.Mean`. rrrzUnknown keyword arguments: _metric_objNzkPlease provide a name for your metric like `self.add_metric(tf.reduce_sum(inputs), name='mean_activation')`z;Expected a symbolic Tensor for the metric value, received: r'Frr)rqrrzUsing the result of calling a `Metric` object when calling `add_metric` on a Functional Model is not supported. Please pass the Tensor to monitor directly.Zmean)listrrrrrwr|rrrrrrr-rqrDre_get_existing_metricrcr~r5ZMeanZ_graph_network_add_metric) r?rrqrtZ kwargs_keysZfrom_metric_objZ is_symbolicr*Z metric_objZshould_update_statematchrrArArB add_metricsB5          zLayer.add_metriccCsP|dk rtdt}|jr$dS|jsLx t|D]}t|r6|q6WdS)aAdd update op(s), potentially dependent on layer inputs. Weight updates (for instance, the updates of the moving mean and variance in a BatchNormalization layer) may be dependent on the inputs passed when calling a layer. Hence, when reusing the same layer on different inputs `a` and `b`, some entries in `layer.updates` may be dependent on `a` and some on `b`. This method automatically keeps track of dependencies. This call is ignored when eager execution is enabled (in that case, variable updates are run on the fly and thus do not need to be tracked for later execution). Args: updates: Update op, or list/tuple of update ops, or zero-arg callable that returns an update op. A zero-arg callable should be passed in order to disable running the updates by setting `trainable=False` on this Layer, when executing in Eager mode. inputs: Deprecated, will be automatically inferred. Nz`add_update` `inputs` kwarg has been deprecated. You no longer need to pass a value to `inputs` as it is being automatically inferred.) r,rrrZin_keras_graphfrozenr1rr$)r?r ryrupdaterArArB add_updateszLayer.add_updatec Cs2|j}d}x,|D]$}t|tjr,||j7}q|d7}qW|t|krjtd|jt||t|ddfd}g}x|D]}t|tjr|j}||||}| |||7}qx||} t | dr| j nd} |j } | | std| | f| || f|d7}qxWt|x|D]} | qWdS) atSets the weights of the layer, from NumPy arrays. The weights of a layer represent the state of the layer. This function sets the weight values from numpy arrays. The weight values should be passed in the order they are created by the layer. Note that the layer's weights must be instantiated before calling this function, by calling the layer. For example, a `Dense` layer returns a list of two values: the kernel matrix and the bias vector. These can be used to set the weights of another `Dense` layer: >>> layer_a = tf.keras.layers.Dense(1, ... kernel_initializer=tf.constant_initializer(1.)) >>> a_out = layer_a(tf.convert_to_tensor([[1., 2., 3.]])) >>> layer_a.get_weights() [array([[1.], [1.], [1.]], dtype=float32), array([0.], dtype=float32)] >>> layer_b = tf.keras.layers.Dense(1, ... kernel_initializer=tf.constant_initializer(2.)) >>> b_out = layer_b(tf.convert_to_tensor([[10., 20., 30.]])) >>> layer_b.get_weights() [array([[2.], [2.], [2.]], dtype=float32), array([0.], dtype=float32)] >>> layer_b.set_weights(layer_a.get_weights()) >>> layer_b.get_weights() [array([[1.], [1.], [1.]], dtype=float32), array([0.], dtype=float32)] Args: weights: a list of NumPy arrays. The number of arrays and their shape must match number of the dimensions of the weights of the layer (i.e. it should match the output of `get_weights`). Raises: ValueError: If the provided weights list does not match the layer's specifications. rrzYou called `set_weights(weights)` on layer "%s" with a weight list of length %s, but the layer was expecting %s weights. Provided weights: %s...N2rrAzBLayer weight shape %s not compatible with provided weight shape %s)rIr|rr} num_tensorsrrrqr set_weightsrwrZis_compatible_withr~rZbatch_set_valuerfinalize_state) r?rIparamsZexpected_num_weightsparamZ weight_indexZweight_value_tuplesr6ZtensorsweightZ weight_shapeZ ref_shaperrArArBr7s<,     "        zLayer.set_weightscCsH|j}g}x2|D]*}t|tjr0||q||qWt|S)a9Returns the current weights of the layer, as NumPy arrays. The weights of a layer represent the state of the layer. This function returns both trainable and non-trainable weight values associated with this layer as a list of NumPy arrays, which can in turn be used to load state into similarly parameterized layers. For example, a `Dense` layer returns a list of two values: the kernel matrix and the bias vector. These can be used to set the weights of another `Dense` layer: >>> layer_a = tf.keras.layers.Dense(1, ... kernel_initializer=tf.constant_initializer(1.)) >>> a_out = layer_a(tf.convert_to_tensor([[1., 2., 3.]])) >>> layer_a.get_weights() [array([[1.], [1.], [1.]], dtype=float32), array([0.], dtype=float32)] >>> layer_b = tf.keras.layers.Dense(1, ... kernel_initializer=tf.constant_initializer(2.)) >>> b_out = layer_b(tf.convert_to_tensor([[10., 20., 30.]])) >>> layer_b.get_weights() [array([[2.], [2.], [2.]], dtype=float32), array([0.], dtype=float32)] >>> layer_b.set_weights(layer_a.get_weights()) >>> layer_b.get_weights() [array([[1.], [1.], [1.]], dtype=float32), array([0.], dtype=float32)] Returns: Weights values as a list of NumPy arrays. ) rIr|rr}r"Z get_tensorsr~rZbatch_get_value)r?rIZoutput_weightsr;rArArB get_weightss#  zLayer.get_weightscCsdS)zFinalizes the layers state after updating layer weights. This function can be subclassed in a layer and will be called after updating a layer weights. It can be overridden to finalize any additional layer state after a weight update. NrA)r?rArArBr8<szLayer.finalize_statecCstd|jS)zDeprecated, do NOT use! Retrieves updates relevant to a specific set of inputs. Args: inputs: Input tensor or list/tuple of input tensors. Returns: List of update ops of the layer that depend on `inputs`. zy`layer.get_updates_for` is deprecated and will be removed in a future version. Please use `layer.updates` method instead.)rrr )r?ryrArArBget_updates_forFs zLayer.get_updates_forcCstd|jS)zDeprecated, do NOT use! Retrieves losses relevant to a specific set of inputs. Args: inputs: Input tensor or list/tuple of input tensors. Returns: List of loss tensors of the layer that depend on `inputs`. zp`layer.get_losses_for` is deprecated and will be removed in a future version. Please use `layer.losses` instead.)rrr#)r?ryrArArBget_losses_forWs zLayer.get_losses_forcCs2||}t|tr"dd|DSt|ddSdS)avRetrieves the input mask tensor(s) of a layer at a given node. Args: node_index: Integer, index of the node from which to retrieve the attribute. E.g. `node_index=0` will correspond to the first time the layer was called. Returns: A mask tensor (or list of tensors if the layer has multiple inputs). cSsg|]}t|ddqS)rN)rD)rrrArArBrxsz+Layer.get_input_mask_at..rN) get_input_atr|r.rD)r? node_indexryrArArBget_input_mask_aths  zLayer.get_input_mask_atcCs2||}t|tr"dd|DSt|ddSdS)axRetrieves the output mask tensor(s) of a layer at a given node. Args: node_index: Integer, index of the node from which to retrieve the attribute. E.g. `node_index=0` will correspond to the first time the layer was called. Returns: A mask tensor (or list of tensors if the layer has multiple outputs). cSsg|]}t|ddqS)rN)rD)rrrArArBrsz,Layer.get_output_mask_at..rN) get_output_atr|r.rD)r?r@outputrArArBget_output_mask_at|s  zLayer.get_output_mask_atcCs.|j}t|trdd|DSt|ddSdS)aqRetrieves the input mask tensor(s) of a layer. Only applicable if the layer has exactly one inbound node, i.e. if it is connected to one incoming layer. Returns: Input mask tensor (potentially None) or list of input mask tensors. Raises: AttributeError: if the layer is connected to more than one incoming layers. cSsg|]}t|ddqS)rN)rD)rrrArArBrsz$Layer.input_mask..rN)inputr|r.rD)r?ryrArArB input_masks zLayer.input_maskcCs.|j}t|trdd|DSt|ddSdS)atRetrieves the output mask tensor(s) of a layer. Only applicable if the layer has exactly one inbound node, i.e. if it is connected to one incoming layer. Returns: Output mask tensor (potentially None) or list of output mask tensors. Raises: AttributeError: if the layer is connected to more than one incoming layers. cSsg|]}t|ddqS)rN)rD)rrrArArBrsz%Layer.output_mask..rN)rCr|r.rD)r?rCrArArB output_masks zLayer.output_maskcCs||ddS)aRetrieves the input shape(s) of a layer at a given node. Args: node_index: Integer, index of the node from which to retrieve the attribute. E.g. `node_index=0` will correspond to the first time the layer was called. Returns: A shape tuple (or list of shape tuples if the layer has multiple inputs). Raises: RuntimeError: If called in Eager mode. input_shapesz input shape)_get_node_attribute_at_index)r?r@rArArBget_input_shape_atszLayer.get_input_shape_atcCs||ddS)aRetrieves the output shape(s) of a layer at a given node. Args: node_index: Integer, index of the node from which to retrieve the attribute. E.g. `node_index=0` will correspond to the first time the layer was called. Returns: A shape tuple (or list of shape tuples if the layer has multiple outputs). Raises: RuntimeError: If called in Eager mode. output_shapesz output shape)rI)r?r@rArArBget_output_shape_atszLayer.get_output_shape_atcCs||ddS)aRetrieves the input tensor(s) of a layer at a given node. Args: node_index: Integer, index of the node from which to retrieve the attribute. E.g. `node_index=0` will correspond to the first input node of the layer. Returns: A tensor (or list of tensors if the layer has multiple inputs). Raises: RuntimeError: If called in Eager mode. input_tensorsrE)rI)r?r@rArArBr?szLayer.get_input_atcCs||ddS)aRetrieves the output tensor(s) of a layer at a given node. Args: node_index: Integer, index of the node from which to retrieve the attribute. E.g. `node_index=0` will correspond to the first output node of the layer. Returns: A tensor (or list of tensors if the layer has multiple outputs). Raises: RuntimeError: If called in Eager mode. output_tensorsrC)rI)r?r@rArArBrBszLayer.get_output_atcCs&|jstd|jd|dddS)aFRetrieves the input tensor(s) of a layer. Only applicable if the layer has exactly one input, i.e. if it is connected to one incoming layer. Returns: Input tensor or list of input tensors. Raises: RuntimeError: If called in Eager mode. AttributeError: If no inbound nodes are found. zLayer z& is not connected, no input to return.rrMrE)_inbound_nodesAttributeErrorrqrI)r?rArArBrE s z Layer.inputcCs&|jstd|jd|dddS)amRetrieves the output tensor(s) of a layer. Only applicable if the layer has exactly one output, i.e. if it is connected to one incoming layer. Returns: Output tensor or list of output tensors. Raises: AttributeError: if the layer is connected to more than one incoming layers. RuntimeError: if called in Eager mode. zLayer z has no inbound nodes.rrNrC)rOrPrqrI)r?rArArBrCsz Layer.outputcCsT|jstdtdd|jD}t|dkr:|jdjStdt|jddS) aRetrieves the input shape(s) of a layer. Only applicable if the layer has exactly one input, i.e. if it is connected to one incoming layer, or if all inputs have the same shape. Returns: Input shape, as an integer shape tuple (or list of shape tuples, one tuple per input tensor). Raises: AttributeError: if the layer has no defined input_shape. RuntimeError: if called in Eager mode. zDThe layer has never been called and thus has no defined input shape.cSsg|]}t|jqSrA)rrH)rrrArArBrEsz%Layer.input_shape..rrz The layer "z has multiple inbound nodes, with different input shapes. Hence the notion of "input shape" is ill-defined for the layer. Use `get_input_shape_at(node_index)` instead.N)rOrPsetrrHrrq)r?Zall_input_shapesrArArBrF0s  zLayer.input_shapec Cs^|jsRt|ddr6t|||jWdQRXntd|jd|jdt |j S)zCount the total number of scalars composing the weights. Returns: An integer count. Raises: ValueError: if the layer isn't yet built (in which case its weights aren't yet defined). r'FNz$You tried to call `count_params` on z=, but the layer isn't built. You can build it manually via: `z.build(batch_input_shape)`.) rRrDr"maybe_init_scoperryrrqr count_paramsrI)r?rArArBrSQs   zLayer.count_paramscCsL|jstdtdd|jD}t|dkr:|jdjStd|jdS)aRetrieves the output shape(s) of a layer. Only applicable if the layer has one output, or if all outputs have the same shape. Returns: Output shape, as an integer shape tuple (or list of shape tuples, one tuple per output tensor). Raises: AttributeError: if the layer has no defined output shape. RuntimeError: if called in Eager mode. zEThe layer has never been called and thus has no defined output shape.cSsg|]}t|jqSrA)rrK)rrrArArBrzsz&Layer.output_shape..rrzThe layer "%s" has multiple inbound nodes, with different output shapes. Hence the notion of "output shape" is ill-defined for the layer. Use `get_output_shape_at(node_index)` instead.N)rOrPrQrrKrq)r?Zall_output_shapesrArArBrfs  zLayer.output_shapecCs|jS)zCDeprecated, do NOT use! Only for compatibility with external Keras.)rO)r?rArArB inbound_nodesszLayer.inbound_nodescCs|jS)zCDeprecated, do NOT use! Only for compatibility with external Keras.)_outbound_nodes)r?rArArBoutbound_nodesszLayer.outbound_nodescOstd|j|f||S)a*Deprecated, do NOT use! This is an alias of `self.__call__`. Args: inputs: Input tensor(s). *args: additional positional arguments to be passed to `self.call`. **kwargs: additional keyword arguments to be passed to `self.call`. Returns: Output tensor(s). zp`layer.apply` is deprecated and will be removed in a future version. Please use `layer.__call__` method instead.)rrr)r?ryrzrtrArArBapplys z Layer.applycOstd|j||S)z/Deprecated, do NOT use! Alias for `add_weight`.zy`layer.add_variable` is deprecated and will be removed in a future version. Please use `layer.add_weight` method instead.)rrr)r?rzrtrArArB add_variables zLayer.add_variablecCs|jS)zReturns the list of all layer variables/weights. Alias of `self.weights`. Note: This will not track the weights of nested `tf.Modules` that are not themselves Keras layers. Returns: A list of variables. )rI)r?rArArBr)s zLayer.variablescCs|jS)N)r)r?rArArBrszLayer.trainable_variablescCs|jS)N)r)r?rArArBrszLayer.non_trainable_variablescCs|jS)N)rh)r?rArArBrOszLayer._inbound_nodescCs ||_dS)N)rh)r?rrArArBrOscCs|jS)N)ri)r?rArArBrUszLayer._outbound_nodescCs ||_dS)N)ri)r?rrArArBrUscCst|tjr||_n\t|tr,t||_nDt|trL|dkrLt||_n$|rftt|j |_n t |_|jj dkrt st }td|jj|jj f|jjrt|jj|_nd|_dS)zSets self._dtype_policy.) mixed_float16Zmixed_bfloat16rYzMixed precision is not supported with the tf.distribute.Strategy: %s. Either stop using mixed precision by removing the use of the "%s" policy or use a different Strategy, e.g. a MirroredStrategy.N)r|rrrdictZ deserializerr rrqZ global_policyrZstrategy_supports_loss_scaling ds_contextZ get_strategyrr;r>rr)r?rrZstrategyrArArBrfs$    zLayer._set_dtype_policycCs|jS)zrThe dtype policy associated with this layer. This is an instance of a `tf.keras.mixed_precision.Policy`. )r)r?rArArB dtype_policy szLayer.dtype_policycCs|jjS)aThe dtype of the layer's computations. This is equivalent to `Layer.dtype_policy.compute_dtype`. Unless mixed precision is used, this is the same as `Layer.dtype`, the dtype of the weights. Layers automatically cast their inputs to the compute dtype, which causes computations and the output to be in the compute dtype as well. This is done by the base Layer class in `Layer.__call__`, so you do not have to insert these casts if implementing your own layer. Layers often perform certain internal computations in higher precision when `compute_dtype` is float16 or bfloat16 for numeric stability. The output will still typically be float16 or bfloat16 in such cases. Returns: The layer's compute dtype. )rr)r?rArArBr szLayer.compute_dtypecCs|jjS)z$Deprecated alias of `compute_dtype`.)rr)r?rArArBr# szLayer._compute_dtypecCs|jS)z1Alias of `Layer.dtype`, the dtype of the weights.)rr)r?rArArBr( szLayer.variable_dtypecCsN|st|}|j}|jo"|o"|j}|rFtt|j|rFt|j |S|SdS)aMaybe casts the inputs to the compute dtype. If self._compute_dtype is floating-point, and self_autocast is True, floating-point inputs are casted to self._compute_dtype. Args: inputs: Input tensor, or structure of input tensors. input_list: Flat list of input tensors. Returns: `inputs`, but tensors may have been casted to self._compute_dtype N) r1rrrgrrmap_should_cast_single_inputr_cast_single_input)r?ryrZcompute_dtype_objectZshould_autocastrArArBr- s   zLayer._maybe_cast_inputscCs(t|tr$|jo"|j|jko"|jjSdS)NF)r|_AUTOCAST_TYPESrrrr)r?rrArArBr^I s zLayer._should_cast_single_inputcCs ||rt||jS|SdS)z8Cast a single Tensor or TensorSpec to the compute dtype.N)r^r(r r)r?rrArArBr_O s zLayer._cast_single_inputcCs|jjS)N)rr)r?rArArB_dtypeY sz Layer._dtypecCs t|j}|t|dS)N)r rrqrfrr)r?rrArArBra` s cCs<ts|jS|j}t}|r,|d|}|r8|d7}|S)N/)renabledrqrZget_name_scope)r?rZcurrent_name_scoperArArBre s zLayer._name_scopecCs4|s tjt|jj|d|_nt|||_dS)N) zero_based)rZunique_object_namerr$r;r>rZobserve_object_name)r?rqrdrArArBrYs s    zLayer._init_set_namecsDfdd|jD}|sdSt|dkr.rzfPlease provide different names for the metrics you have added. We found {} metrics with the name: "{}"r)rcrrr<)r?rqr0rA)rqrBr/| s zLayer._get_existing_metriccsPfdd}t|r:x2|D]}|t||qWn|t||dS)z3Create lambdas which compute regularization losses.c s&td|}WdQRX|S)z8Creates a regularization loss `Tensor` for variable `v`.z /RegularizerN)rr)rZregularization)rqrrArB_loss_for_variable sz?Layer._handle_weight_regularization.._loss_for_variableN)rrr+r(r))r?rqrrrerrA)rqrrBr s   z#Layer._handle_weight_regularizationc Csl|jrht|}tdHx@|D]8}||}tt|d|j }||}| |q"WWdQRXdS)NZActivityRegularizerr) r\r1rrrr(r r'rrrr+)r?ryrZ output_listrCZ activity_lossrHZmean_activity_lossrArArBr s    z%Layer._handle_activity_regularizationc Cs|js dSt|}t|ddp0tdd|D}|rH|rD||dS|||}|dkr`dSt|}x6t||D](\} } y | | _Wqvt k rYqvXqvW|r||dS)NZ _compute_output_and_mask_jointlyFcss|]}t|dddk VqdS)rN)rD)rrrArArBr sz+Layer._set_mask_metadata..) rXr1rrDall_set_mask_keras_history_checkedrWziprrP) r?ryrZ previous_maskr flat_outputsZmask_already_computedZ output_masksZ flat_maskstensorrrArArBr s(       zLayer._set_mask_metadatacCs*x$|D]}t|dddk rd|j_qWdS)NrT)rDrZ_keras_history_checked)r?rirCrArArBrg s z%Layer._set_mask_keras_history_checkedcCsz|js|jsd}d}n\|d||r8|d||}d}n:dd|D}tdd|Drbd}d}nt||}d}||fS)NFrcSsg|]}t|ddqS)rN)rD)rrrArArBr sz*Layer._get_input_masks..css|]}|dkVqdS)NrA)rrrArArBr sz)Layer._get_input_masks..T)rXrrrrfr1pack_sequence_as)r?ryrrzrtrZ implicit_maskrArArBr s  zLayer._get_input_maskscCs@|s |s dS||krdS|j}|s.|dd}|tt||kS)NFTr) _call_fn_argsrZrh)r?arg_namerzrtinputs_in_args call_fn_argsrArArBr s zLayer._call_arg_was_passedcCs<||kr||S|j}|s&|dd}tt||}||S)Nr)rlrZrh)r?rmrzrtrnroZ args_dictrArArBr s zLayer._get_call_arg_valuecCst|j|d}|dk rJ|s"|d}t||krJt|}|||<t||fS|dkrd|rd||dn|||<||fS)Nr)_call_fn_arg_positionsrZrr.rlr[)r?rm new_valuerzrtrnrZarg_posrArArBr s   zLayer._set_call_arg_valuec Cst|}t||f}dd|D}g}xD|D]<}t||krbt|jt|}WdQRX||q0Wt ||}t j ||||d|S)NcSsh|] }t|qSrA)id)rirArArB sz3Layer._set_connectivity_metadata..) call_argsZ call_kwargsr) r1rrrrrrqr'identityr~rk node_moduleNode) r?rzrtrriZ flat_inputsZ input_ids_setZ outputs_copyrrArArBr s    z Layer._set_connectivity_metadatacCs|jstd|dt|j|ksRtd|dt|dtt|jdt|j||}t|trt|dkr|dS|Sd S) aPrivate utility to retrieves an attribute (e.g. inputs) from a node. This is used to implement the methods: - get_input_shape_at - get_output_shape_at - get_input_at etc... Args: node_index: Integer index of the node from which to retrieve the attribute. attr: Exact node attribute name. attr_name: Human-readable attribute name, for error messages. Returns: The layer's attribute `attr` at the node of index `node_index`. Raises: RuntimeError: If the layer has no inbound nodes, or if called in Eager mode. ValueError: If the index provided does not match any node. z8The layer has never been called and thus has no defined r:z Asked to get z at node z, but the layer has only z inbound nodes.rrN)rOrrrrrDr|r.)r?r@attr attr_namevaluesrArArBrI s,z"Layer._get_node_attribute_at_indexc Cs"|jst|j||jt|}|rj|jjdkrjy|djj j}Wnt k rXYnX| t |d}tdd|Drt|}n(ytj|dd}Wntk rYnXt|jdst|||WdQRXt|||jdk rt||jWdQRXd|_dS)Nrcss|]}t|dVqdS)rN)rw)rrrArArBrP sz%Layer._maybe_build..F)rrv)rRrrrqr1rrrrrrrPrfrrrfr"Z get_shapesrrrwrxrRr6rnrZ init_scoper7)r?ryrrrrHrArArBrA s2       zLayer._maybe_buildcs@tdd|}|}tj|dd}fdd}t||S)NcSs|jS)N)r)rrArArBrm rz&Layer._symbolic_call..F)rcstj|jd}d|_|S)N)rrr)rrrrr)rr)r?rArBrs sz4Layer._symbolic_call.._make_placeholder_like)r1rrr"r)r?ryrHrKrrA)r?rB_symbolic_calll s   zLayer._symbolic_callcCs(t}x|D]}|j||<qW|S)z}Get the `trainable` state of each sublayer. Returns: A dict mapping all sublayers to their `trainable` value. )weakrefWeakKeyDictionaryrrp)r?trainable_staterrArArB_get_trainable_statey szLayer._get_trainable_statecCs(x"|D]}||kr |||_q WdS)z(Set `trainable` state for each sublayer.N)rrp)r?rrrArArB_set_trainable_state szLayer._set_trainable_statecCs|dt|jS)zEA dictionary counting the number of attributes referencing an object.r7)r]r ZObjectIdentityDictionaryr7)r?rArArB_obj_reference_counts s zLayer._obj_reference_countscCst||s|||dS)aCreate the attribute with the default value if it hasn't been created. This is useful for fields that is used for tracking purpose, _trainable_weights, or _layers. Note that user could create a layer subclass and assign an internal field before invoking the Layer.__init__(), the __setattr__() need to create the tracking fields and __init__() need to not override them. Args: name: String, the name of the attribute. default_value: Object, the default value of the attribute. N)rw __setattr__)r?rq default_valuerArArBr] s zLayer._maybe_create_attributecs t||d|j}|kr0ttj||dS|}|dkrb|d|<ttj||dS|=ttj||ttst rttj| dfdd|j Dtt j rttj| dfdd|jDttj| dfdd|jDdS) NrrOcsg|]}|k r|qSrArA)rl)existing_valuerArBr sz%Layer.__delattr__..rMcsg|]}|k r|qSrArA)rw)rrArBr srNcsg|]}|k r|qSrArA)rr)rrArBr s)rDrsuperr- AutoTrackable __delattr__r|r6r has_weightsrrOrVariablerMrN)r?rqreference_countsZreference_count)r;)rrBr s0       zLayer.__delattr__cs|dks t|ddr t|j|r`yttj||Wn"tk rZtd|YnXdSt j ||d|j }| dd|<y| |Wntk rYnXx4tD]&ttjrt|dr|jqWt|ddrPttjs trP|d gtfd d |jDsP|jtd rPd_xtjdd D]ttjsvq`|dg|dgjrtfdd |j Drq`|j n*tfdd |j!Drq`|j!t"#q`Wttj||dS)NZ_self_setattr_trackingTzCan't set the attribute "{}", likely because it conflicts with an existing read-only @property of the object. Please choose a different name.) trackablerrqrrrcrorOc3s|]}|kVqdS)NrA)rr)rrArBr sz$Layer.__setattr__.._use_resource_variables)Zexpand_compositesrMrNc3s|]}|kVqdS)NrA)rr)valrArBr sc3s|]}|kVqdS)NrA)rr)rrArBr s)$rDrwr;rr-rrrPr<r/Zsticky_attribute_assignmentrrZrr1rr|r5Metricrcr~r&Modulerrr]rrOrrrrprMrNrr)r?rqrr)r;)rrrBr sV          zLayer.__setattr__csFdks tt|drB|jddd}ttjfdd|DSgS)N>rrr)rOF) include_self recursivec3s|]}t|VqdS)N)rD)rr) attributerArBr sz3Layer._gather_children_attribute..)AssertionErrorrw_flatten_modulesr. itertoolschain from_iterable)r?rZ nested_layersrA)rrBr s  z Layer._gather_children_attributeccs,x&|j||dD]}t|tr|VqWdS)N)rr)rr|r6)r?rrrrArArBr" s zLayer._flatten_layersc cs|r |Vt|dd}|rt}t|}x|r|}t|}||krJq,||t|tj rt|t j s|V|rt|dd}|r| t |q,t|tjr,|j} | r,| t | q,WdS)aFlattens `tf.Module` instances (excluding `Metrics`). Args: recursive: Whether to recursively flatten through submodules. include_self: Whether to include this `Layer` instance. Yields: `tf.Module` instance tracked by this `Layer`. rON)rDrQrdequepopleftrraddr|r&rr5r extendleftreversedr/ZTrackableDataStructureZ_values) r?rrZ trackablesZseen_object_idsrZ trackable_objZ trackable_idZ subtrackablesZtracked_valuesrArArBr( s0       zLayer._flatten_modulescCsdS)NTrA)r?rArArB _is_layerT szLayer._is_layercCs|jjjj|d|jjjj|d|jjjj|d|j}||jjpNg7}|dkrld|kpf|j|_n||_|j }| |jj pi| d|_d|kp|j|_dS)Nrr)r;_call_full_argspecfgetcacher[rl_call_accepts_kwargs kwonlyargsr _call_fn_arg_defaultscopyr3kwonlydefaultsrZrr)r?Zexpects_training_argroZcall_fn_arg_defaultsrArArBrjW s   zLayer._init_call_fn_argscCs t|jS)N)r!rr{)r?rArArBrn szLayer._call_full_argspeccCs(|jj}|r$|ddkr$|ddS|S)Nrr?r)rrz)r?rrArArBrlu s zLayer._call_fn_argscCsH|j}|jjpg}t}x*tdt|dD]}|||||<q,W|S)Nr)rlrdefaultsrZranger)r?roZcall_fn_defaultsrrsrArArBr~ s  zLayer._call_fn_arg_defaultscCs*t}xt|jD]\}}|||<qW|S)N)rZ enumeraterl)r?Zcall_fn_arg_positionsposrrArArBrp s zLayer._call_fn_arg_positionscCs |jjdk S)N)rvarkw)r?rArArBr szLayer._call_accepts_kwargscCst|jdsg|j_|jjS)Nr!)rwr`r!)r?rArArBr! s zLayer._eager_lossescCs ||j_dS)N)r`r!)r?r#rArArBr! scCsBgt}}x0|D](}t||kr|||t|qW|S)z;Dedupe weights while maintaining order as much as possible.)rQrrr~r)r?rIrCZseen_idsrrArArBr s     zLayer._dedup_weightscCsV|r|d}|dd}n2|jd|krDt|}||jd}ntd|||fS)Nrrz9The first argument to `Layer.call` must always be passed.)rlrr[r)r?rzrtryrArArBr s zLayer._split_out_first_argcCs"|jdk rdSttj||_dS)N)rUr1rr"get_tensor_spec)r?ryrArArBr s zLayer._set_save_speccs$|jdkrdStfdd|jS)Ncstj|dS)N) dynamic_batch)r"r)r)rrArBr rz&Layer._get_save_spec..)rUr1r)r?rrA)rrB_get_save_spec s   zLayer._get_save_speccCs t|S)N)rZLayerSavedModelSaver)r?rArArB_trackable_saved_model_saver sz"Layer._trackable_saved_model_savercCs|jjS)N)rZobject_identifier)r?rArArB_object_identifier szLayer._object_identifiercCs|jjS)z6Info about this layer to be saved into the SavedModel.)rZtracking_metadata)r?rArArB_tracking_metadata szLayer._tracking_metadata checkpointc s<|dkr|d}|j|}ni}|tj|f||S)NZ savedmodelr)rZtrackable_childrenr3r_trackable_children)r? save_typertrchildren)r;rArBr s zLayer._trackable_childrencCstt|dddk S)Nr8)r9)r2type)r?rArArB!_use_input_spec_as_call_signature sz'Layer._use_input_spec_as_call_signaturecCs&|j}|dd|dd|S)Nr`re)__dict__rr[)r?staterArArB __getstate__ s   zLayer.__getstate__cCs*t|d<t|d<t|d|dS)Nr`rer)r^r_rdobjectr)r?rrArArB __setstate__ s  zLayer.__setstate__)TNNF)N)N)N)N)T)N)F)F)FF)TT)TT)N)T)r)r>r= __qualname____doc__ frozensetrrr&rZ_TF_MODULE_IGNORED_PROPERTIES_must_restore_from_configrCrEr no_automatic_dependency_trackingrurdefaultrxr4Zfor_subclass_implementersr{rrrZAUTOZVariableAggregationNONErr classmethodrrrrrrWrrrrpropertyrrrqrsetterrsZdo_not_doc_inheritablerrprJrrrrIZdo_not_generate_docsr r#r+rr,r1r4r7r<r8r=r>rArDrFrGrJrLr?rBrErCrFrSrrTrVrWrXr)rrrOrUrfr\rrrrr^r_rarrYr/rrrrgrrrrrrIrr|rrrr]rrrrrrrjrZcached_per_instancerrlrrprr!rrrrrrrrrrr __classcell__rArA)r;rBr6asP-    -- -'6 {S+      #    >x  y $U,       %         $+    , I  ,          r6csZeZdZdZejdfdd ZddZdd Zd d Z e j d d Z fddZ ZS)TensorFlowOpLayeraWraps a TensorFlow Operation in a Layer. This class is used internally by the Functional API. When a user uses a raw TensorFlow Operation on symbolic tensors originating from an `Input` Layer, the resultant operation will be wrapped with this Layer object in order to make the operation compatible with the Keras API. This Layer will create a new, identical operation (except for inputs and outputs) every time it is called. If `run_eagerly` is `True`, the op creation and calculation will happen inside an Eager function. Instances of this Layer are created when `autolambda` is called, which is whenever a Layer's `__call__` encounters symbolic inputs that do not have Keras metadata, or when a Network's `__init__` encounters outputs that do not have Keras metadata. Attributes: node_def: String, the serialized NodeDef of the Op this layer will wrap. name: String, the name of the Layer. constants: Dict of NumPy arrays, the values of any Tensors needed for this Operation that do not originate from a Keras `Input` Layer. Since all placeholders must come from Keras `Input` Layers, these Tensors must be treated as constant in the Functional API. trainable: Bool, whether this Layer is trainable. Currently Variables are not supported, and so this parameter has no effect. dtype: The default dtype of this Layer. Inherited from `Layer` and has no effect on this class, however is used in `get_config`. NTcstt|jt|||ddt|tr:t|t |_ n"t|t sN| d}tj ||_ |dk rvdd|Dni|_d|_d|_dS)NF)rqrprrrKzutf-8cSsi|]\}}|t|qSrA)r)rindexconstantrArArB 9 sz.TensorFlowOpLayer.__init__..T)rrru_TF_OP_LAYER_NAME_PREFIXr|rZrZ ParseDictrNodeDefnode_defbytesencodeZ FromStringitems constantsrRr)r?rrqrrprr)r;rArBru& s      zTensorFlowOpLayer.__init__cCstr||S||S)N)r r _defun_call_make_op)r?ryrArArBr{C s zTensorFlowOpLayer.callcCs2t}||jd|jd_||j|_|S)NTZ_cloned)rrZCopyFromrrybZ unique_namerq)r?graphrrArArB_make_node_defH s   z TensorFlowOpLayer._make_node_defc Cs(t|}|dj}||}|xH|jD]:\}}t|}|dk rbt j ||j |d}| ||q4Wt j|||gd}||}|t|jj} dd|jjD} g} x&| D]} | | | || qWt| } t| |j| |jt|jdkr|jdS|jSQRXdS)Nr)rq)Zcontrol_inputscSsg|]}t|jqSrA)r0as_strrq)rryrArArBre sz.TensorFlowOpLayer._make_op..r)r1rrrrrrrZconstant_valuer rrEinsertrZ _create_c_opZ_create_op_from_tf_operationZ_control_flow_post_processingr0rZop_defrqryr~Zget_attrrlrZrecord_gradientryrr) r?ryrrrrrZc_opopZop_typeZ attr_namesattrsrzrArArBrQ s.         zTensorFlowOpLayer._make_opcCs ||S)zDWraps the op creation method in an Eager function for `run_eagerly`.)r)r?ryrArArBrq szTensorFlowOpLayer._defun_callcsLtt|}||dttdt|jdd|j Dd|S)NrqcSsi|]\}}t||qSrA)r get_value)rrscrArArBr| sz0TensorFlowOpLayer.get_config..)rqrr) rrrr3rrrZ MessageToDictrrr)r?r)r;rArBrv s zTensorFlowOpLayer.get_config)NTN)r>r=rrrrrur{rrr functionrrrrArA)r;rBr s  rcs4eZdZdZfddZddZfddZZS)AddLossz|Adds its inputs as a loss. Attributes: unconditional: Whether or not the loss should be conditioned on the inputs. c s$d|d<tt|jf|||_dS)NFrK)rrru unconditional)r?rrt)r;rArBru szAddLoss.__init__cCs|j||j d|S)N)ry)r+r)r?ryrArArBr{ sz AddLoss.callcs"tt|}|d|ji|S)Nr)rrrr3r)r?r)r;rArBr szAddLoss.get_config)r>r=rrrur{rrrArA)r;rBr s rcs6eZdZdZd fdd ZddZfddZZS) AddMetriczAdds its inputs as a metric. Attributes: aggregation: 'mean' or None. How the inputs should be aggregated. metric_name: The name to use for this metric. Nc s"tt|jf|||_||_dS)N)rrrur metric_name)r?rrrt)r;rArBru szAddMetric.__init__cCs|j||j|jd|S)N)rrq)r1rr)r?ryrArArBr{ szAddMetric.callcs&tt|}||j|jd|S)N)rr)rrrr3rr)r?r)r;rArBr s  zAddMetric.get_config)NN)r>r=rrrur{rrrArA)r;rBr srcCstddt|||gDS)zECheck the arguments to see if we are constructing a functional model.css|]}t|tjVqdS)N)r|rr)rrjrArArBr sz3_in_functional_construction_mode..)rr1r)rryrzrtrrArArBr srcCs$t|tjtjttfr t|S|S)N)r|r*rrrrrr)rrArArBr s r)krrrr(rr^rr}numpyrZgoogle.protobufrZtensorflow.core.frameworkrZtensorflow.pythonrZ tensorflow.python.autograph.corerZ tensorflow.python.autograph.implrrZtensorflow.python.distributerr[Ztensorflow.python.eagerrr r Ztensorflow.python.frameworkr r r rrrrZtensorflow.python.kerasrrrrZtensorflow.python.keras.enginerrrrrwZ'tensorflow.python.keras.mixed_precisionrrrZ*tensorflow.python.keras.saving.saved_modelrZtensorflow.python.keras.utilsrrr r!r"r#Z+tensorflow.python.keras.utils.generic_utilsr$Z&tensorflow.python.keras.utils.tf_utilsr%Ztensorflow.python.moduler&Ztensorflow.python.opsr'r(r)rZtensorflow.python.ops.numpy_opsr*Ztensorflow.python.ops.raggedr+Ztensorflow.python.platformr,Ztensorflow.python.trackabler-r.rr/Ztensorflow.python.utilr0r1Z tensorflow.python.util.tf_exportr2r3Ztensorflow.tools.docsr4 LazyLoaderglobalsr5rZTensorZ SparseTensorZ RaggedTensorr`rZLayerVersionSelectorr6rrrrrrrArArArBs                                                   <|