B «»ˆd2ã@sždZddlmmZddlmZddlmZddlmZddlm Z ddlm Z ddl m Z dd l mZdd lmZdd lmZed ƒGd d„deƒƒZdS)zContains the Dense layer.éN)Ú activations)Úbackend)Ú constraints)Ú initializers)Ú regularizers)Úutils)ÚLayer)Ú InputSpec)Ú keras_exportzkeras.layers.Densec sLeZdZdZejd‡fdd„ ƒZdd „Zd d „Zd d „Z ‡fdd„Z ‡Z S)ÚDensea_ Just your regular densely-connected NN layer. `Dense` implements the operation: `output = activation(dot(input, kernel) + bias)` where `activation` is the element-wise activation function passed as the `activation` argument, `kernel` is a weights matrix created by the layer, and `bias` is a bias vector created by the layer (only applicable if `use_bias` is `True`). These are all attributes of `Dense`. Note: If the input to the layer has a rank greater than 2, then `Dense` computes the dot product between the `inputs` and the `kernel` along the last axis of the `inputs` and axis 0 of the `kernel` (using `tf.tensordot`). For example, if input has dimensions `(batch_size, d0, d1)`, then we create a `kernel` with shape `(d1, units)`, and the `kernel` operates along axis 2 of the `input`, on every sub-tensor of shape `(1, 1, d1)` (there are `batch_size * d0` such sub-tensors). The output in this case will have shape `(batch_size, d0, units)`. Besides, layer attributes cannot be modified after the layer has been called once (except the `trainable` attribute). When a popular kwarg `input_shape` is passed, then keras will create an input layer to insert before the current layer. This can be treated equivalent to explicitly defining an `InputLayer`. Example: >>> # Create a `Sequential` model and add a Dense layer as the first layer. >>> model = tf.keras.models.Sequential() >>> model.add(tf.keras.Input(shape=(16,))) >>> model.add(tf.keras.layers.Dense(32, activation='relu')) >>> # Now the model will take as input arrays of shape (None, 16) >>> # and output arrays of shape (None, 32). >>> # Note that after the first layer, you don't need to specify >>> # the size of the input anymore: >>> model.add(tf.keras.layers.Dense(32)) >>> model.output_shape (None, 32) Args: units: Positive integer, dimensionality of the output space. activation: Activation function to use. If you don't specify anything, no activation is applied (ie. "linear" activation: `a(x) = x`). use_bias: Boolean, whether the layer uses a bias vector. kernel_initializer: Initializer for the `kernel` weights matrix. bias_initializer: Initializer for the bias vector. kernel_regularizer: Regularizer function applied to the `kernel` weights matrix. bias_regularizer: Regularizer function applied to the bias vector. activity_regularizer: Regularizer function applied to the output of the layer (its "activation"). kernel_constraint: Constraint function applied to the `kernel` weights matrix. bias_constraint: Constraint function applied to the bias vector. Input shape: N-D tensor with shape: `(batch_size, ..., input_dim)`. The most common situation would be a 2D input with shape `(batch_size, input_dim)`. Output shape: N-D tensor with shape: `(batch_size, ..., units)`. For instance, for a 2D input with shape `(batch_size, input_dim)`, the output would have shape `(batch_size, units)`. NTÚglorot_uniformÚzerosc  s¶tƒjfd|i| —Žt|tƒs(t|ƒn||_|jdkrFtd|›ƒ‚t |¡|_||_ t  |¡|_ t  |¡|_ t  |¡|_t  |¡|_t | ¡|_t | ¡|_tdd|_d|_dS)NÚactivity_regularizerrzTReceived an invalid value for `units`, expected a positive integer. Received: units=é)Úmin_ndimT)ÚsuperÚ__init__Ú isinstanceÚintÚunitsÚ ValueErrorrÚgetÚ activationÚuse_biasrÚkernel_initializerÚbias_initializerrÚkernel_regularizerÚbias_regularizerrÚkernel_constraintÚbias_constraintr Ú input_specZsupports_masking) ÚselfrrrrrrrrrrÚkwargs)Ú __class__©úI/var/www/html/venv/lib/python3.7/site-packages/keras/layers/core/dense.pyrfs          zDense.__init__c CsÖt |jpt ¡¡}|js.|js.td|›ƒ‚t |¡}tj   |d¡}|dkr^t d|›ƒ‚t dd|id|_ |jd||jg|j|j|j|jdd|_|jrÆ|jd |jg|j|j|j|jdd|_nd|_d|_dS) NzMA Dense layer can only be built with a floating-point dtype. Received: dtype=éÿÿÿÿzlThe last dimension of the inputs to a Dense layer should be defined. Found None. Full input shape received: r)rZaxesÚkernelT)ÚshapeZ initializerZ regularizerÚ constraintÚdtypeZ trainableÚbias)ÚtfZas_dtyper*rZfloatxZ is_floatingZ is_complexÚ TypeErrorÚ TensorShapeÚcompatÚdimension_valuerr r Z add_weightrrrrr'rrrrr+Zbuilt)r!Ú input_shaper*Zlast_dimr$r$r%Úbuild‰s:      z Dense.buildc Csê|jj|jjkr tj||jd}t|tjƒ}|rÊtj |j d¡dkrXt d|j ›dƒ‚|}|j j j dkrr|j }nX|j j dkrŒ|  ¡}d}n>xt|jdƒD] }|j}qœW|  ¡}tj ||jdd…¡}|j j }|dksä|dkrTt|tjƒrBtj |d¡\}}tj|j|jdd…df|jd }|}tjj|j||d d }ntj||jd }nTt ||j|dgdgg¡}t ¡s¨|j  ¡} | dd…|jj dg} | | ¡|j rÀtj !||j"¡}|j#dk rÖ| #|¡}|ræ| $|¡}|S) N)r*r&zkDense layer only supports RaggedTensors when the innermost dimension is non-ragged. Received: inputs.shape=Ú.érFr)ÚindicesÚvaluesÚ dense_shapeÚsum)Zcombiner)ÚaÚb)%r*Z base_dtypeZ_compute_dtype_objectr,ÚcastrZ RaggedTensorr/r0r(rZ flat_valuesÚrankZ to_tensorÚrangeZ ragged_rankr6Zfrom_nested_row_splitsZnested_row_splitsZ SparseTensorÚsparseZfill_empty_rowsr5r7ÚnnZembedding_lookup_sparser'ÚmatmulZ tensordotZexecuting_eagerlyÚas_listÚ set_shaperZbias_addr+rZwith_flat_values) r!ÚinputsZ is_raggedZoriginal_inputsÚ_r<ZidsÚweightsÚoutputsr(Z output_shaper$r$r%Úcall³sT           z Dense.callcCsJt |¡}| d¡}tj |d¡dkr6td|›ƒ‚|dd… |j¡S)Nrr&zlThe last dimension of the input shape of a Dense layer should be defined. Found None. Received: input_shape=)r,r.Zwith_rank_at_leastr/r0rZ concatenater)r!r1r$r$r%Úcompute_output_shapes    zDense.compute_output_shapecsrtƒ ¡}| |jt |j¡|jt |j ¡t |j ¡t  |j ¡t  |j ¡t  |j¡t |j¡t |j¡dœ ¡|S)N) rrrrrrrrrr)rÚ get_configÚupdaterrÚ serializerrrrrrrrrrrr)r!Úconfig)r#r$r%rIs&  zDense.get_config) NTr r NNNNN) Ú__name__Ú __module__Ú __qualname__Ú__doc__rZallow_initializer_layoutrr2rGrHrIÚ __classcell__r$r$)r#r%r !sC*S r )rPZtensorflow.compat.v2r/Zv2r,ZkerasrrrrrZ keras.dtensorrZkeras.engine.base_layerrZkeras.engine.input_specr Z tensorflow.python.util.tf_exportr r r$r$r$r%Ús