Ë  |hè'ãó”—dZddlZddlmcmZddlmZddlmZddlm Z ddl m Z ddl m Z ddl m Z dd l mZdd l mZdd l mZdd l mZdd l mZddl mZddl mZddlmZddlmZddlmZddlmZddlmZddlmZddlmZddlmZddlmZddlm Z ddlm!Z!ddlm"Z#ddlm$Z$ddlm%Z%ddl&m'Z'ddl&m(Z(dd l&m)Z)dd!l*m+Z+dd"l,m-Z-dd#l.m/Z0dd$l1m2Z2dd%l3m4Z4dd&l3m5Z6dd'l7m8Z8eee e e eeeeeeeee eeeeeeee!e#e$e%fZ9eee'e(e)fZ:ejv«ae8d+«d0d,„«Z?d-„Z@d1d.„ZAy)2z.Layer serialization/deserialization functions.éN)Ú base_layer)Ú input_layer)Ú input_spec)Ú activation)Ú attention)Ú convolutional)Úcore)Úlocally_connected)Úmerging)Úpooling)Úregularization)Ú reshaping)Úrnn)Úbatch_normalization)Úbatch_normalization_v1)Úgroup_normalization)Úlayer_normalization)Úunit_normalization)Úcategory_encoding)Údiscretization)Úhashed_crossing)Úhashing)Úimage_preprocessing)Úinteger_lookup)Ú normalization)Ú string_lookup)Útext_vectorization)Ú cell_wrappers)Úgru)Úlstm)Ú base_metric)Úserialization_lib)Ú serialization)Ú json_utils)Ú generic_utils)Ú tf_inspect)Ú keras_exportcó ‡—ttd«sit_dt_tjr:tjtj j j«k(ryit_tj j j«t_tjŠtjtjtˆfd„¬«tj j j«r-tjtjtˆfd„¬«tjtjd<t jtjd<dd lm}dd lm}dd lm}dd lm}t2j4tjd <t6j8tjd<|j:tjd<|j<tjd<|tjd<|j>tjd<|tjd<|tjd<tj j j«rddl m!}|tjd<nddl"m!}|tjd<tFjHtjd<tFjJtjd<tFjLtjd<tFjNtjd<tFjPtjd<tFjRtjd<tFjTtjd<tFjVtjd<y)z5Populates dict ALL_OBJECTS with every built-in layer.Ú ALL_OBJECTSNcóJ•—tj|«xr t|‰«S©N©ÚinspectÚisclassÚ issubclass©ÚxÚbase_clss €ú`/var/www/html/test/engine/venv/lib/python3.12/site-packages/tf_keras/src/layers/serialization.pyúz1populate_deserializable_objects..|sø€œWŸ_™_¨QÓ/ÒK´J¸qÀ(Ó4K€ó)Ú obj_filtercóJ•—tj|«xr t|‰«Sr+r,r0s €r3r4z1populate_deserializable_objects..„sø€¤§¡°Ó!3Ò!O¼ À1ÀhÓ8O€r5ÚBatchNormalizationV1ÚBatchNormalizationV2r)Úmodels)ÚSequenceFeatures)Ú LinearModel)Ú WideDeepModelÚInputÚ InputSpecÚ FunctionalÚModelr;Ú Sequentialr<r=)Ú DenseFeaturesrCÚaddÚsubtractÚmultiplyÚaverageÚmaximumÚminimumÚ concatenateÚdot),ÚhasattrÚLOCALr)ÚGENERATED_WITH_V2ÚtfÚ __internal__Útf2ÚenabledrÚLayerr%Ú!populate_dict_with_module_objectsÚ ALL_MODULESÚALL_V2_MODULESrÚBatchNormalizationrÚ tf_keras.srcr:Ú3tf_keras.src.feature_column.sequence_feature_columnr;Ú"tf_keras.src.premade_models.linearr<Ú%tf_keras.src.premade_models.wide_deepr=rr>rr?r@rArBÚ-tf_keras.src.feature_column.dense_features_v2rCÚ*tf_keras.src.feature_column.dense_featuresr rDrErFrGrHrIrJrK)r:r;r<r=rCr2s @r3Úpopulate_deserializable_objectsr^fs«ø€ô ”5˜-Ô (ØŒÔØ"&ŒÔô ×ÒÜ × #Ñ #¤r§¡×':Ñ':×'BÑ'BÓ'DÒ Dð à„EÔÜ Ÿo™o×1Ñ1×9Ñ9Ó;„EÔä×Ñ€HÜ×3Ñ3Ü ×ÑÜÛKõô ‡×Ñ×"Ñ"Ô$Ü×7Ñ7Ü × Ñ Ü ÛOõ ô ×1Ñ1ô ×ÑØñô ×.Ñ.ô ×ÑØñõ $õõõô"-×!2Ñ!2„E×ÑgÑÜ%/×%9Ñ%9„E×ÑkÑ"Ø&,×&7Ñ&7„E×ÑlÑ#Ø!'§¡„E×ÑgÑØ,<„E×ÑÐ(Ñ)Ø&,×&7Ñ&7„E×ÑlÑ#Ø'2„E×ÑmÑ$Ø)6„E×ÑoÑ&ä ‡×Ñ×"Ñ"Ô$õ ð.;Œ×ј/Ò*õ ð.;Œ×ј/Ñ*ô 'Ÿ{™{„E×ÑeÑÜ$+×$4Ñ$4„E×ÑjÑ!Ü$+×$4Ñ$4„E×ÑjÑ!Ü#*§?¡?„E×ÑiÑ Ü#*§?¡?„E×ÑiÑ Ü#*§?¡?„E×ÑiÑ Ü'.×':Ñ':„E×ÑmÑ$Ü&Ÿ{™{„E×ÑeÒr5zkeras.layers.serializecó¬—t|tj«rtd|›d«‚|rt j |«St j |«S)aSerializes a `Layer` object into a JSON-compatible representation. Args: layer: The `Layer` object to serialize. Returns: A JSON-serializable dict representing the object's config. Example: ```python from pprint import pprint model = tf.keras.models.Sequential() model.add(tf.keras.Input(shape=(16,))) model.add(tf.keras.layers.Dense(32, activation='relu')) pprint(tf.keras.layers.serialize(model)) # prints the configuration of the model, as a dict. zCannot serialize zŽ since it is a metric. Please use the `keras.metrics.serialize()` and `keras.metrics.deserialize()` APIs to serialize and deserialize metrics.)Ú isinstancer!ÚMetricÚ ValueErrorÚlegacy_serializationÚserialize_keras_objectr")ÚlayerÚuse_legacy_formats r3Ú serializerg¿sY€ô*%œ×+Ñ+Ô,ÜØ ˜wð''ð 'ó ð ñ Ü#×:Ñ:¸5ÓAÐAä × 3Ñ 3°EÓ :Ð:r5zkeras.layers.deserializecóÖ—t«|std|›«‚|r'tj|tj |d¬«St j|tj |d¬«S)a8Instantiates a layer from a config dictionary. Args: config: dict of the form {'class_name': str, 'config': dict} custom_objects: dict mapping class names (or function names) of custom (non-Keras) objects to class/functions Returns: Layer instance (may be Model, Sequential, Network, Layer...) Example: ```python # Configuration of Dense(32, activation='relu') config = { 'class_name': 'Dense', 'config': { 'activation': 'relu', 'activity_regularizer': None, 'bias_constraint': None, 'bias_initializer': {'class_name': 'Zeros', 'config': {}}, 'bias_regularizer': None, 'dtype': 'float32', 'kernel_constraint': None, 'kernel_initializer': {'class_name': 'GlorotUniform', 'config': {'seed': None}}, 'kernel_regularizer': None, 'name': 'dense', 'trainable': True, 'units': 32, 'use_bias': True } } dense_layer = tf.keras.layers.deserialize(config) ``` z2Cannot deserialize empty config. Received: config=re)Úmodule_objectsÚcustom_objectsÚprintable_module_name)r^rbrcÚdeserialize_keras_objectrMr)r")Úconfigrjrfs r3Ú deserializernásv€ôL$Ô%Ù ÜØ@ÀÀÐ Ió ð ñÜ#×<Ñ<Ø Ü ×,Ñ,Ø)Ø")ô  ð ô × 5Ñ 5ØÜ×(Ñ(Ø%Ø%ô  ðr5cót—ttd«s t«tjj |«S)z?Returns class if `class_name` is registered, else returns None.r))rLrMr^r)Úget)Ú class_names r3Úget_builtin_layerrrs)€ä ”5˜-Ô (Ü'Ô)Ü × Ñ × Ñ  Ó ,Ð,r5cóz—t«tj|tj|¬«}t ||«S)z(Instantiates a layer from a JSON string.)rirj)r^r$Údecode_and_deserializerMr)rn)Ú json_stringrjrms r3Údeserialize_from_jsonrv#s6€ä#Ô%Ü × .Ñ .ØÜ×(Ñ(Ø%ô€Fô v˜~Ó .Ð.r5)F)NFr+)BÚ__doc__Ú threadingÚtensorflow.compat.v2ÚcompatÚv2rOÚtf_keras.src.enginerrrÚtf_keras.src.layersrrrr r r r r rrÚ!tf_keras.src.layers.normalizationrrrrrÚ!tf_keras.src.layers.preprocessingrrrrrrrÚpreprocessing_normalizationrrÚtf_keras.src.layers.rnnrrr Útf_keras.src.metricsr!Útf_keras.src.savingr"Útf_keras.src.saving.legacyr#rcÚ&tf_keras.src.saving.legacy.saved_modelr$Útf_keras.src.utilsr%r&r-Ú tensorflow.python.util.tf_exportr'rUrVÚlocalrMr^rgrnrrrv©r5r3úrŠsKðñ5ãç!Ð!å*Ý+Ý*Ý*Ý)Ý-Ý$Ý1Ý'Ý'Ý.Ý)Ý#ÝAÝDÝAÝAÝ@Ý?Ý<Ý=Ý5ÝAÝ<õõ<Ý@Ý1Ý'Ý(Ý,Ý1ÝLÝ=Ý,Ý4õ:ðØØØ ØØØØ ØØØØØ ØØØ ØØ ØØØØØØØð3€ ð8ØØØØð €ð ˆ ‰Ó€òV+ñrÐ&Ó'ò;ó(ð;ñBÐ(Ó)ò7ó*ð7òt-ô/r5