&Vf)ddlZddlZddlmZddlmZddlm Z ddl m Z ddl m Z edGdd Zd Zd Zd ZdS) N)backend) keras_export) global_state) jax_utils) auto_namezkeras.random.SeedGeneratorc>eZdZdZddZd dZdZedZdS) SeedGeneratoraAGenerates variable seeds upon each call to a RNG-using function. In Keras, all RNG-using methods (such as `keras.random.normal()`) are stateless, meaning that if you pass an integer seed to them (such as `seed=42`), they will return the same values at each call. In order to get different values at each call, you must use a `SeedGenerator` instead as the seed argument. The `SeedGenerator` object is stateful. Example: ```python seed_gen = keras.random.SeedGenerator(seed=42) values = keras.random.normal(shape=(2, 3), seed=seed_gen) new_values = keras.random.normal(shape=(2, 3), seed=seed_gen) ``` Usage in a layer: ```python class Dropout(keras.Layer): def __init__(self, **kwargs): super().__init__(**kwargs) self.seed_generator = keras.random.SeedGenerator(1337) def call(self, x, training=False): if training: return keras.random.dropout( x, rate=0.5, seed=self.seed_generator ) return x ``` Nc |tjj}|_|dd}|rt d|||_n t __tttst dfd}t j j5j |dddd _ ddddS#1swxYwYdS) Nrz Unrecognized keyword arguments: z3Argument `seed` must be an integer. Received: seed=cl|dd}jdg|S)Ndtyperr )getrconvert_to_tensor)argskwargsr seedselfs \/var/www/html/software/conda/lib/python3.11/site-packages/keras/src/random/seed_generator.pyseed_initializerz0SeedGenerator.__init__..seed_initializerFs5JJw--E<114)51II I)caller)uint32Fseed_generator_state)shaper trainablename)r __class____name__rpop ValueErrorr _initial_seedmake_default_seed isinstanceint name_scopeVariablestate)rrrrcustom_backendrs`` r__init__zSeedGenerator.__init__0s <T^455D It44  JHHHII I  %)DLL"DL! <$&&D$$$ OOO  J J J J J J  $ 7 7 7  .. + /DJ                  s%C66C:=C:Tc|j}|jdz}|rN|jt jddgd}|j||zn#|j|dzdzdz|S)Nrrr i iO=)r(valuerrnparrayassign)rordered seed_statenew_seed_value increments rnextzSeedGenerator.nextSsZ #)A-  @ 66!Q  7I J  j94 5 5 5 5 J  zA~5> ? ? ?rcd|jiS)Nr)r")rs r get_configzSeedGenerator.get_configbs*++rc|di|S)Nr9)clsconfigs r from_configzSeedGenerator.from_configess}}V}}r)NN)T) r __module__ __qualname____doc__r*r5r7 classmethodr<r9rrr r sr  D!!!!F    ,,,[rr ctjrtdtjd}|#t }tjd||S)Na[JAX RNG] When tracing a JAX function, you should only use seeded random ops, e.g. you should create a `SeedGenerator` instance, attach it to your layer/model, and pass the instance as the `seed` argument when calling random ops. Unseeded random ops would get incorrectly traced by JAX and would become constant after tracing. Example: ``` # Make sure to set the seed generator as a layer attribute self.seed_generator = keras.random.SeedGenerator(seed=1337) ... out = keras.random.normal(shape=(1,), seed=self.seed_generator) ```global_seed_generator)ris_in_jax_tracing_scoper!rget_global_attributer set_global_attribute)gens rrBrBjsd(**       +,C D DC {oo)*A3GGG JrcFtjdtdS)Nr,geA) python_randomrandintr%r9rrr#r#s  CHH - --rc6ddlm}t|tr|St|t r||dgdS|"t dStd|dt|d ) Nr)rrr F)r1z\Argument `seed` must be either an integer or an instance of `SeedGenerator`. Received: seed=z (of type )) keras.src.backendrr$r r5r%rBr!type)rrs r draw_seedrNs333333$ &&;yy{{ D#  ;  $(;;;; $&&++E+:::  8 8 8*.t** 8 8 8  r)randomrHnumpyr. keras.srcrkeras.src.api_exportrkeras.src.backend.commonrkeras.src.utilsrkeras.src.utils.namingrr rBr#rNr9rrrVs------111111%%%%%%,,,,,,*++ZZZZZZZ,+Zz0...     r