1Vh3xdZddlmZddlmZddlmZejZdZ dZ dZ dZ d Z d Zd Zy ) aJUtilities for handling Keras model in graph plugin. Two canonical types of Keras model are Functional and Sequential. A model can be serialized as JSON and deserialized to reconstruct a model. This utility helps with dealing with the serialized Keras model. They have distinct structures to the configurations in shapes below: Functional: config name: Name of the model. If not specified, it is 'model' with an optional suffix if there are more than one instance. input_layers: Keras.layers.Inputs in the model. output_layers: Layer names that are outputs of the model. layers: list of layer configurations. layer: [*] inbound_nodes: inputs to this layer. Sequential: config name: Name of the model. If not specified, it is 'sequential' with an optional suffix if there are more than one instance. layers: list of layer configurations. layer: [*] [*]: Note that a model can be a layer. Please refer to https://github.com/tensorflow/tfjs-layers/blob/master/src/keras_format/model_serialization.ts for more complete definition. )GraphDef)dtypes) tb_loggingc#,Kd|f|jdjdrj|jdjd}|jdjdD]&}t|D]\}}|r|d|n|}||f(yyw)aYWalks the nested keras layer configuration in preorder. Args: keras_layer: Keras configuration from model.to_json. Yields: A tuple of (name_scope, layer_config). name_scope: a string representing a scope name, similar to that of tf.name_scope. layer_config: a dict representing a Keras layer configuration. configlayersname/N)get _walk_layers) keras_layer name_scopelayersub_name_scopesublayers T/home/dcms/DCMS/lib/python3.12/site-packages/tensorboard/plugins/graph/keras_util.pyr r 4s { x $$X. __X.226:  __X.228< 1E,8,? 1(& *>:# &x00  1 1/sBBc|r|d|S|S)a/Returns scoped name for a node as a string in the form '/'. Args: name_scope: a string representing a scope name, similar to that of tf.name_scope. node_name: a string representing the current node name. Returns A string representing a scoped name. r )r node_names r _scoped_namerLs$i00 cF|jdjdduS)zReturns True if layer is a model. Args: layer: a dict representing a Keras model configuration. Returns: bool: True if layer is a model. rr N)r )rs r _is_modelr\s# 99X  " "8 ,D 88rc4t|dtfr|S|gS)aNormalizes to a list of layers. Args: maybe_layers: A list of data[1] or a list of list of data. Returns: List of list of data. [1]: A Functional model has fields 'inbound_nodes' and 'output_layers' which can look like below: - ['in_layer_name', 0, 0] - [['in_layer_is_model', 1, 0], ['in_layer_is_model', 1, 1]] The data inside the list seems to describe [name, size, index]. r) isinstancelist) maybe_layerss r_norm_to_list_of_layersrhs% #<?TG< CO.rchg}|jd|jdgD]}t|ts|jdgD]a}t|ts|g}|D]G}|jdijdg}t |dkr4|j |ddIc|S)zOReturns a list of [name, size, index] for all inbound nodes of the given layer. inbound_nodesNargsr keras_history)r rdictrlenappend)rr!maybe_inbound_nodeinbound_node_argsarghistorys r_get_inbound_nodesr,|sM yy!-"'))OR"@ 6 0$7%7%;%;FB%G 6!""3T:):(;%,6C!ggh377LG7|a' !((!5 6 6  6 rc<|jd}|jds tdt||jd}|jd}|jd}t|} t |xr|} t | } | r7| r5t || D]%\} } t|| }t|| d}|||<'nT| r| st|| dd}n=| s;|r9| r7t |dk(sJd t  z|d} t|| }|||<| r/|r-t|}|Dcgc]}t||d}}|||<nF|jdd }|jdjd}t||}|g||<|||fScc}w) aUpdates input_to_in_layer, model_name_to_output, and prev_node_name based on the model_layer. Args: name_scope: a string representing a scope name, similar to that of tf.name_scope. model_layer: a dict representing a Keras model configuration. input_to_in_layer: a dict mapping Keras.layers.Input to inbound layer. model_name_to_output: a dict mapping Keras Model name to output layer of the model. prev_node_name: a string representing a previous, in sequential model layout, node name. Returns: A tuple of (input_to_in_layer, model_name_to_output, prev_node_name). input_to_in_layer: a dict mapping Keras.layers.Input to inbound layer. model_name_to_output: a dict mapping Keras Model name to output layer of the model. prev_node_name: a string representing a previous, in sequential model layout, node name. rr zlayer is not a model.r input_layers output_layersrzhCannot have multi-input Functional model when parent model is not Functional. Number of input layers: %d)r ValueErrorrr,boolzipr&r)r model_layerinput_to_in_layermodel_name_to_outputprev_node_name layer_configrr.r/r!is_functional_modelis_parent_functional_model input_layer inbound_nodeinput_layer_nameinbound_node_namer r layer_names last_layerlast_layer_name output_nodes r _update_dictsrDs2??8,L   H %011Z)9)9&)ABI##N3L $$_5M&{3M|= >"&m!4!&9),\=)I D %K+I{C  ,Za I 2C . / D $,?&j-2B12EF &  < A%  <>A+>N O %#1o ' ;?.<*+}(7FLMU|IuQx8M M*5Y'!%%h/3 $..266v>"9o> +6-Y' 3^ DDNs0Fci}i}t}d}t|D]'\}}t|rt|||||\}}}&|j d}t ||j d}|j j} || _|j d4|j djd} | | jd_ |j d} d} d} t| tr%d| vr!| j dj d} n| | } | 5 tj| }|j | jd_n| d } | rt&j)d | |j d gt+|D]X\}}}t ||}|j ||g}|t-|kr||n|d }| j.j1|Zn|| j.j1|||vr*| j.j1|j || j}*|S#t$$rd } YwxYw) aReturns a GraphDef representation of the Keras model in a dict form. Note that it only supports models that implemented to_json(). Args: keras_layer: A dict from Keras model.to_json(). Returns: A GraphDef representation of the layers in the model. Nrr class_nameascii keras_classdtypeFTz8Unsupported dtype value in graph model config (json): %sr!r1)rr rrDr rnodeaddr encodeattrsrr%ras_dtypeas_datatype_enumtype TypeErrorloggerwarningr,r&inputr')rinput_to_layerr7gr8rrr9rnode_defkeras_cls_namedtype_or_policyrIhas_unsupported_valuetf_dtyper sizeindex inbound_nameinbound_node_names input_names rkeras_model_to_graph_defrbsN AN)+6W' E U  $   $ yy*  \-=-=f-EF 66::<! 99\ " ."YY|4;;GDN-;HMM- ( *&**73 % ot ,_1L#''155f=E  (#E   -!??51.6.G.G g&+ ($( ! NNK  99_ % 1%7%> 2!dE+J= &:%=%= <.&"s#566'u-+B/ %%j11 22 ' NN ! !. 1  & NN ! !."4"4Y"? @!oW'r Ha -(,% -s%3I  IIN)__doc__"tensorboard.compat.proto.graph_pb2r"tensorboard.compat.tensorflow_stubrtensorboard.utilr get_loggerrSr rrrr,rDrbrrrrhsN 885'    10  9(.FERl r