B d2@s@dZddlmmZddlmZddlmZddlm Z ddl m Z ddl m Z e dGd d d Ze d d Gd ddeZe ddGdddeZe ddGdddeZe ddGdddeZe ddGdddeZeZeZeZeZeZeZeZeZe dd d!Ze d"d(d#d$Ze d%d&d'ZdS))z@Constraints: functions that impose constraints on weight values.N)backend)deserialize_keras_object)serialize_keras_object) keras_export) doc_controlszkeras.constraints.Constraintc@s,eZdZdZddZddZeddZdS) ConstraintaoBase class for weight constraints. A `Constraint` instance works like a stateless function. Users who subclass this class should override the `__call__` method, which takes a single weight parameter and return a projected version of that parameter (e.g. normalized or clipped). Constraints can be used with various Keras layers via the `kernel_constraint` or `bias_constraint` arguments. Here's a simple example of a non-negative weight constraint: >>> class NonNegative(tf.keras.constraints.Constraint): ... ... def __call__(self, w): ... return w * tf.cast(tf.math.greater_equal(w, 0.), w.dtype) >>> weight = tf.constant((-1.0, 1.0)) >>> NonNegative()(weight) >>> tf.keras.layers.Dense(4, kernel_constraint=NonNegative()) cCs|S)agApplies the constraint to the input weight variable. By default, the inputs weight variable is not modified. Users should override this method to implement their own projection function. Args: w: Input weight variable. Returns: Projected variable (by default, returns unmodified inputs). )selfwrrC/var/www/html/venv/lib/python3.7/site-packages/keras/constraints.py__call__8s zConstraint.__call__cCsiS)aReturns a Python dict of the object config. A constraint config is a Python dictionary (JSON-serializable) that can be used to reinstantiate the same object. Returns: Python dict containing the configuration of the constraint object. r)r rrr get_configGs zConstraint.get_configcCs |f|S)aInstantiates a weight constraint from a configuration dictionary. Example: ```python constraint = UnitNorm() config = constraint.get_config() constraint = UnitNorm.from_config(config) ``` Args: config: A Python dictionary, the output of `get_config`. Returns: A `tf.keras.constraints.Constraint` instance. r)clsconfigrrr from_configRszConstraint.from_configN)__name__ __module__ __qualname____doc__r r classmethodrrrrr rs rzkeras.constraints.MaxNormzkeras.constraints.max_normc@s6eZdZdZd ddZejddZejdd Zd S) MaxNormaQMaxNorm weight constraint. Constrains the weights incident to each hidden unit to have a norm less than or equal to a desired value. Also available via the shortcut function `tf.keras.constraints.max_norm`. Args: max_value: the maximum norm value for the incoming weights. axis: integer, axis along which to calculate weight norms. For instance, in a `Dense` layer the weight matrix has shape `(input_dim, output_dim)`, set `axis` to `0` to constrain each weight vector of length `(input_dim,)`. In a `Conv2D` layer with `data_format="channels_last"`, the weight tensor has shape `(rows, cols, input_depth, output_depth)`, set `axis` to `[0, 1, 2]` to constrain the weights of each filter tensor of size `(rows, cols, input_depth)`. rcCs||_||_dS)N) max_valueaxis)r rrrrr __init__szMaxNorm.__init__cCsBttjt||jdd}t|d|j}||t|S)NT)rkeepdimsr) rsqrttf reduce_sumsquarercliprepsilon)r r normsdesiredrrr r szMaxNorm.__call__cCs|j|jdS)N)rr)rr)r rrr r szMaxNorm.get_configN)rr) rrrrrrdo_not_generate_docsr r rrrr rgs rzkeras.constraints.NonNegzkeras.constraints.non_negc@seZdZdZddZdS)NonNegz}Constrains the weights to be non-negative. Also available via the shortcut function `tf.keras.constraints.non_neg`. cCs|tt|dtS)Ng)rcastZ greater_equalrZfloatx)r r rrr r szNonNeg.__call__N)rrrrr rrrr r%sr%zkeras.constraints.UnitNormzkeras.constraints.unit_normc@s6eZdZdZd ddZejddZejddZd S) UnitNormaConstrains the weights incident to each hidden unit to have unit norm. Also available via the shortcut function `tf.keras.constraints.unit_norm`. Args: axis: integer, axis along which to calculate weight norms. For instance, in a `Dense` layer the weight matrix has shape `(input_dim, output_dim)`, set `axis` to `0` to constrain each weight vector of length `(input_dim,)`. In a `Conv2D` layer with `data_format="channels_last"`, the weight tensor has shape `(rows, cols, input_depth, output_depth)`, set `axis` to `[0, 1, 2]` to constrain the weights of each filter tensor of size `(rows, cols, input_depth)`. rcCs ||_dS)N)r)r rrrr rszUnitNorm.__init__c Cs*|tttjt||jddS)NT)rr)rr!rrrrr)r r rrr r szUnitNorm.__call__cCs d|jiS)Nr)r)r rrr r szUnitNorm.get_configN)r) rrrrrrr$r r rrrr r's  r'zkeras.constraints.MinMaxNormzkeras.constraints.min_max_normc@s6eZdZdZd ddZejddZejd d Zd S) MinMaxNorma(MinMaxNorm weight constraint. Constrains the weights incident to each hidden unit to have the norm between a lower bound and an upper bound. Also available via the shortcut function `tf.keras.constraints.min_max_norm`. Args: min_value: the minimum norm for the incoming weights. max_value: the maximum norm for the incoming weights. rate: rate for enforcing the constraint: weights will be rescaled to yield `(1 - rate) * norm + rate * norm.clip(min_value, max_value)`. Effectively, this means that rate=1.0 stands for strict enforcement of the constraint, while rate<1.0 means that weights will be rescaled at each step to slowly move towards a value inside the desired interval. axis: integer, axis along which to calculate weight norms. For instance, in a `Dense` layer the weight matrix has shape `(input_dim, output_dim)`, set `axis` to `0` to constrain each weight vector of length `(input_dim,)`. In a `Conv2D` layer with `data_format="channels_last"`, the weight tensor has shape `(rows, cols, input_depth, output_depth)`, set `axis` to `[0, 1, 2]` to constrain the weights of each filter tensor of size `(rows, cols, input_depth)`. ?rcCs||_||_||_||_dS)N) min_valuerrater)r r+rr,rrrr rszMinMaxNorm.__init__cCsXttjt||jdd}|jt||j|j d|j|}||t |S)NT)rr) rrrrrrr,r r+rr!)r r r"r#rrr r s zMinMaxNorm.__call__cCs|j|j|j|jdS)N)r+rr,r)r+rr,r)r rrr r szMinMaxNorm.get_configN)r)r*r*r) rrrrrrr$r r rrrr r(s  r(z"keras.constraints.RadialConstraintz#keras.constraints.radial_constraintc@s&eZdZdZejddZddZdS)RadialConstraintaConstrains `Conv2D` kernel weights to be the same for each radius. Also available via the shortcut function `tf.keras.constraints.radial_constraint`. For example, the desired output for the following 4-by-4 kernel: ``` kernel = [[v_00, v_01, v_02, v_03], [v_10, v_11, v_12, v_13], [v_20, v_21, v_22, v_23], [v_30, v_31, v_32, v_33]] ``` is this:: ``` kernel = [[v_11, v_11, v_11, v_11], [v_11, v_33, v_33, v_11], [v_11, v_33, v_33, v_11], [v_11, v_11, v_11, v_11]] ``` This constraint can be applied to any `Conv2D` layer version, including `Conv2DTranspose` and `SeparableConv2D`, and with either `"channels_last"` or `"channels_first"` data format. The method assumes the weight tensor is of shape `(rows, cols, input_depth, output_depth)`. cCs|j}|jdks|jdkr(td||\}}}}t|||||f}t|jtjtj |dddd}ttjtj |dddd||||fS)NzGThe weight tensor must have rank 4. Received weight tensor with shape: )rr) shapeZrank ValueErrorrZreshapeZmap_fn_kernel_constraintstackrZunstack)r r Zw_shapeheightwidthZchannelsZkernelsrrr r s  zRadialConstraint.__call__c stjddgddggddtd}t|ddtttj|ddfddfd d}tttj|ddd dd d}fd d}fd d}tjj j ||||g| t ddggd\}}|S)zKRadially constraints a kernel with shape (height, width, channels).r-int32)dtyperrboolcsddfS)Nr-rr)kernelstartrr >z5RadialConstraint._kernel_constraint..cs,ddftjdjdS)Nr-)rr)r8)rZzerosr8r)r:r;rr r<?scSstjdddS)Nrr7)r8)rconstantrrrr r<Dr=cSstjdddS)Nr-r7)r8)rr>rrrr r<Er=cs t|S)N)rless)indexargs)r;rr r<Gr=cs(|dtj|||fdfS)Nr-)Zconstant_values)rpad)iarray)r:paddingr;rr body_fnIs z4RadialConstraint._kernel_constraint..body_fnN)Zshape_invariants) rr>r1r&switchrmathZfloormodcompatZv1Z while_loop get_shapeZ TensorShape)r r:Z kernel_shapeZ kernel_newr@Zwhile_conditionrF_r)r:rEr;r r34s&   z#RadialConstraint._kernel_constraintN)rrrrrr$r r3rrrr r.sr.zkeras.constraints.serializecCst|S)N)r) constraintrrr serializeesrMzkeras.constraints.deserializecCst|t|ddS)NrL)Zmodule_objectscustom_objectsZprintable_module_name)rglobals)rrNrrr deserializejs rPzkeras.constraints.getcCs\|dkr dSt|trt|St|tr>t|id}t|St|rJ|Std|dS)z&Retrieves a Keras constraint function.N) class_namerz4Could not interpret constraint function identifier: ) isinstancedictrPstrcallabler2) identifierrrrr getts  rW)N)rZtensorflow.compat.v2rIZv2rZkerasrZ!keras.saving.legacy.serializationrrZ tensorflow.python.util.tf_exportrZtensorflow.tools.docsrrrr%r'r(r.Zmax_normZnon_negZ unit_normZ min_max_normZradial_constraintZmaxnormZnonnegZunitnormrMrPrWrrrr s<     I) $;Y