Content overview 
 
 
 
 
 
 
 
 Ukucinga Iimoduli zeTensorFlow Module Ukwakhiwa Ukukhangela ukuvelisa i-variables Ukuphucula ubunzima Ukucaciswa Functions Ukwenza i-SavedModel Ukwenza machine learning kwi-TensorFlow, kufuneka uqhagamshelane, ukugcina kunye nokuguqula imodeli. Umzekelo umzekelo, abstractly: 
 
 
 A function that calculates something on tensors (a forward pass) Zonke iimvavanyo ezidlulileyo ezidlulileyo ngokuvumelana nokufunda Kule nqakraza, uya kufutshane phantsi kwindawo ye-Keras ukuze ubone njani iimodeli ze-TensorFlow. Oku kuthetha njani iimodeli ze-TensorFlow zihlanganisa kunye neemodeli, kunye nokusetyenziswa kunye nokuguqulwa. Ukucinga import tensorflow as tf
import keras
from datetime import datetime

%load_ext tensorboard
 2023-10-18 01:21:05.536666: E tensorflow/compiler/xla/stream_executor/cuda/cuda_dnn.cc:9342] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
2023-10-18 01:21:05.536712: E tensorflow/compiler/xla/stream_executor/cuda/cuda_fft.cc:609] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered
2023-10-18 01:21:05.536766: E tensorflow/compiler/xla/stream_executor/cuda/cuda_blas.cc:1518] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered
 Iimoduli zeTensorFlow Iimodeli ezininzi zihlanganisa izihlangu. Iimodeli zihlanganisa izakhiwo ezaziwayo ezisemathemathemikhali ezidlulileyo kwaye zihlanganisa iimodeli ezidlulileyo. Kwi-TensorFlow, iimodeli ezininzi zihlanganisa izihlangu kunye neemodeli, ezifana neKeras okanye , ziye ziye ziye ziye ziye ziye ziye ziye ziye ziye ziye ziye ziye ziye ziye ziye ziye zibe: . Ukucinga tf.Module Module Ukwakhiwa Nazi isibonelo esiyingqayizivele Ukusebenza kwi-tensor ye-scalar: tf.Module class SimpleModule(tf.Module):
  def __init__(self, name=None):
    super().__init__(name=name)
    self.a_variable = tf.Variable(5.0, name="train_me")
    self.non_trainable_variable = tf.Variable(5.0, trainable=False, name="do_not_train_me")
  def __call__(self, x):
    return self.a_variable * x + self.non_trainable_variable

simple_module = SimpleModule(name="simple")

simple_module(tf.constant(5.0))
 2023-10-18 01:21:08.181350: W tensorflow/core/common_runtime/gpu/gpu_device.cc:2211] Cannot dlopen some GPU libraries. Please make sure the missing libraries mentioned above are installed properly if you would like to use GPU. Follow the guide at https://www.tensorflow.org/install/gpu for how to download and setup the required libraries for your platform.
Skipping registering GPU devices...
<tf.Tensor: shape=(), dtype=float32, numpy=30.0>
 Iimoduli kunye, ngokubanzi, iihlabathi ziquka i-deep-learning terminology ye"i-object": zinezimo zangaphakathi, kunye nezindlela ezisetyenziswayo. Akukho nto elide ukuba ngaphandle kokusebenza njenge a ; unako ukufumana iimodeli zakho kunye neengxaki zakho. __call__ Iphasiwedi yePython Uyakwazi ukucwangcisa i-trainability yeempawu kwi-akhawunti ye-akhawunti ye-akhawunti ye-akhawunti ye-akhawunti ye-akhawunti ye-akhawunti ye-akhawunti ye-akhawunti ye-akhawunti. 
 
 Qaphela: i-tf.Module yi-class base ye-tf.keras.layers.Layer kunye ne-tf.keras.Model, ngoko yonke into efanelekileyo kwi-Keras. Kwiimeko ze-compatibility ye-historical, i-Keras layers ayifumana amaxabiso ezivela kwi-modules, ngoko iimodeli akho kufuneka usebenzise i-modules kuphela okanye i-Keras layers kuphela. Nangona kunjalo, iindlela ezilandelayo zokuhlolwa amaxabiso ziyafumaneka kwimeko ezininzi.   I-class ye-base ye-both iimveliso , ngoko yonke into ufumane apha kusetyenziswa Keras. Kuba izizathu zokuxhomekeka zomlando, i-Keras layers ayifumana ama-variables evela kumoduli, ngoko iimodeli bakho kufuneka usebenzise i-modules kuphela okanye i-Keras layers kuphela. Nangona kunjalo, izindlela ezilandelayo zokufunda ama-variables ziyafumaneka kwimeko ezimbini. Note: tf.Module tf.keras.layers.Layer tf.keras.Model Ngokuba subclassing Yintoni okanye iiyunithi ezisekelwe izakhiwo le-object zithunyelwe ngokuzenzakalelayo. Oku kukuvumela ukugcina kunye nokukhuthaza i-variables, kwaye ukwakha i-collections of iimveliso tf.Module tf.Variable tf.Module tf.Module # All trainable variables
print("trainable variables:", simple_module.trainable_variables)
# Every variable
print("all variables:", simple_module.variables)
 trainable variables: (<tf.Variable 'train_me:0' shape=() dtype=float32, numpy=5.0>,)
all variables: (<tf.Variable 'train_me:0' shape=() dtype=float32, numpy=5.0>, <tf.Variable 'do_not_train_me:0' shape=() dtype=float32, numpy=5.0>)
 Kuyinto umzekelo we-two-layer linear layer model eyenziwe ngama-modules. Okokuqala i-densy (linear) layer: class Dense(tf.Module):
  def __init__(self, in_features, out_features, name=None):
    super().__init__(name=name)
    self.w = tf.Variable(
      tf.random.normal([in_features, out_features]), name='w')
    self.b = tf.Variable(tf.zeros([out_features]), name='b')
  def __call__(self, x):
    y = tf.matmul(x, self.w) + self.b
    return tf.nn.relu(y)
 Emva koko, isampula epheleleyo, leyo yenza iindidi ezimbini kunye nokuvelisa: class SequentialModule(tf.Module):
  def __init__(self, name=None):
    super().__init__(name=name)

    self.dense_1 = Dense(in_features=3, out_features=3)
    self.dense_2 = Dense(in_features=3, out_features=2)

  def __call__(self, x):
    x = self.dense_1(x)
    return self.dense_2(x)

# You have made a model!
my_model = SequentialModule(name="the_model")

# Call it, with random results
print("Model results:", my_model(tf.constant([[2.0, 2.0, 2.0]])))
 Model results: tf.Tensor([[0.       3.415034]], shape=(1, 2), dtype=float32)
 iindidi ziquka ngokuzenzekelayo, ngokugqithisileyo, nayiphi na okanye iindidi ezisekelweyo. Oku kukuvumela ukulawula iincwadi ze s kunye ne-model instance eyodwa, kwaye ukugcina kunye nokulanda iimodeli ezininzi. tf.Module tf.Variable tf.Module tf.Module print("Submodules:", my_model.submodules)
 Submodules: (<__main__.Dense object at 0x7f7931aea250>, <__main__.Dense object at 0x7f77ed5b8a00>)
 for var in my_model.variables:
  print(var, "\n")
 <tf.Variable 'b:0' shape=(3,) dtype=float32, numpy=array([0., 0., 0.], dtype=float32)> 

<tf.Variable 'w:0' shape=(3, 3) dtype=float32, numpy=
array([[-2.8161757, -2.6065955,  1.9061812],
       [-0.9430401, -0.4624743, -0.4531979],
       [-1.3428234,  0.7062293,  0.7874674]], dtype=float32)> 

<tf.Variable 'b:0' shape=(2,) dtype=float32, numpy=array([0., 0.], dtype=float32)> 

<tf.Variable 'w:0' shape=(3, 2) dtype=float32, numpy=
array([[ 1.0474309 , -0.6920227 ],
       [ 1.2405277 ,  0.36411622],
       [-1.6990206 ,  0.762131  ]], dtype=float32)>
 Ukukhangela ukuvelisa i-variables Ingaba uye uyakufuneka apha ukuba kufuneka uqhagamshelane kunye neengxaki ze-input kunye neengxaki ze-output kwi-layer. I-variable ine-shape enokuthi ifumaneka kwaye ingathintela. w Ukukhishwa kwe-variable yokufunda kwi-module yokufumana okokuqala nge-input ye-shape ye-specific, akufuneka ukucacisa ubungakanani ye-input kwi-front. class FlexibleDenseModule(tf.Module):
  # Note: No need for `in_features`
  def __init__(self, out_features, name=None):
    super().__init__(name=name)
    self.is_built = False
    self.out_features = out_features

  def __call__(self, x):
    # Create variables on first call.
    if not self.is_built:
      self.w = tf.Variable(
        tf.random.normal([x.shape[-1], self.out_features]), name='w')
      self.b = tf.Variable(tf.zeros([self.out_features]), name='b')
      self.is_built = True

    y = tf.matmul(x, self.w) + self.b
    return tf.nn.relu(y)
 # Used in a module
class MySequentialModule(tf.Module):
  def __init__(self, name=None):
    super().__init__(name=name)

    self.dense_1 = FlexibleDenseModule(out_features=3)
    self.dense_2 = FlexibleDenseModule(out_features=2)

  def __call__(self, x):
    x = self.dense_1(x)
    return self.dense_2(x)

my_model = MySequentialModule(name="the_model")
print("Model results:", my_model(tf.constant([[2.0, 2.0, 2.0]])))
 Model results: tf.Tensor([[0. 0.]], shape=(1, 2), dtype=float32)
 Ukucaciswa oku kuthetha ukuba izihlangu ze-TensorFlow ziquka kuxhomekeke ukucaciswa kwizithuba zayo, njenge , kunokuba ubungakanani ingxelo kunye nokuveliswa. tf.keras.layers.Dense Ukuphucula ubunzima Ngaba unako ukufumana a Njengoko i-A i-A . tf.Module Qhagamshelana Ukucaciswa I-Checkpoints yi-weights kuphela (yintoni, i-values ye-set ye-variables kwi-module kunye ne-submodules yayo): chkp_path = "my_checkpoint"
checkpoint = tf.train.Checkpoint(model=my_model)
checkpoint.write(chkp_path)
 'my_checkpoint'
 Iifayile ye-Checkpoint ibandakanya iifayile ezimbini: idatha ngokwayo kunye neifayile ye-indice ye-metadata. Iifayile ye-indice ibandakanya into efanelekileyo ezihlabathi kunye ne-numbering yeefayile ze-checkpoint, kwaye i-data ye-checkpoint ibandakanya iimpawu ze-variable kunye neengxaki zayo ze-attribute. ls my_checkpoint*
 my_checkpoint.data-00000-of-00001  my_checkpoint.index
 Uyakwazi ukuyifaka kwindawo yokucaciswa ukuze uqinisekisa ukuba inkxaso lwezinto zihlanganisa, zihlanganiswa ngama-Python i-object ebonakalayo. tf.train.list_variables(chkp_path)
 [('_CHECKPOINTABLE_OBJECT_GRAPH', []),
 ('model/dense_1/b/.ATTRIBUTES/VARIABLE_VALUE', [3]),
 ('model/dense_1/w/.ATTRIBUTES/VARIABLE_VALUE', [3, 3]),
 ('model/dense_2/b/.ATTRIBUTES/VARIABLE_VALUE', [2]),
 ('model/dense_2/w/.ATTRIBUTES/VARIABLE_VALUE', [3, 2])]
 Ngexesha lokuzalwa (i-multi-machine) zokusebenza ze-sharding, yaye ngoko zithunyelwe (isib. '00000-of-00001'). Kule mzekelo, kunjalo, kukho kuphela shard. Xa uqhagamshelane iimodeli, uqhagamshelane iimpawu kwi-Python yakho. new_model = MySequentialModule()
new_checkpoint = tf.train.Checkpoint(model=new_model)
new_checkpoint.restore("my_checkpoint")

# Should be the same result as above
new_model(tf.constant([[2.0, 2.0, 2.0]]))
 <tf.Tensor: shape=(1, 2), dtype=float32, numpy=array([[0., 0.]], dtype=float32)>
 
 
 Qaphela: Njengoko i-checkpoints ziquka iinkqubo zokusebenza ezininzi zokusebenza, bheka i-checkpoint.CheckpointManager iyisiqingatha yokusebenza le-checkpoint. Qhagamshelana nenkqubo ye-checkpoint yokusebenza kwizifundo ezininzi. Qaphela: Njengoko i-checkpoints ziquka iinkqubo zokusebenza ezininzi zokusebenza, bheka i-checkpoint.CheckpointManager iyisiqingatha yokusebenza le-checkpoint. Qhagamshelana nenkqubo ye-checkpoint yokusebenza kwizifundo ezininzi. Ukuqeqeshwa Checkpoints Guide Ukucaciswa Functions I-TensorFlow inokufumana iimodeli ngaphandle kwe-Original Python Objects, njengoko kuboniswa iimveliso , ngelixa xa ufake isampuli esebenzayo ukusuka TensorFlow Serving TensorFlow Lite TensorFlow Hub. TensorFlow kufuneka ukwazi ukwenza izicwangciso ezaziwayo ePython, kodwa Ukwenza oku, unako ukwenza a Yintoni iingcebiso ezidlulileyo . without the original code graph Ukuqhagamshelwano kwi-Graphs and Functions Guide Oku kubandakanya imisebenzi, okanye Yintoni ukuvelisa umsebenzi. iimveliso Uyakwazi ukucacisa i-graph kwi-model apha ngokongeza i-graph inkcazelo ukuba le ikhowudi kufuneka ifumaneke njenge-graph. @tf.function class MySequentialModule(tf.Module):
  def __init__(self, name=None):
    super().__init__(name=name)

    self.dense_1 = Dense(in_features=3, out_features=3)
    self.dense_2 = Dense(in_features=3, out_features=2)

  @tf.function
  def __call__(self, x):
    x = self.dense_1(x)
    return self.dense_2(x)

# You have made a model with a graph!
my_model = MySequentialModule(name="the_model")
 I-module owenziwe kusebenza ngexesha elifanayo. Yonke i-signature eyenziwe kwi-function ikhiqiza i-diagram eyahlukileyo. Qhagamshelana Kuba imibuzo. Ukuqhagamshelwano kwi-Graphs and Functions Guide print(my_model([[2.0, 2.0, 2.0]]))
print(my_model([[[2.0, 2.0, 2.0], [2.0, 2.0, 2.0]]]))
 tf.Tensor([[0.31593648 0.        ]], shape=(1, 2), dtype=float32)
tf.Tensor(
[[[0.31593648 0.        ]
  [0.31593648 0.        ]]], shape=(1, 2, 2), dtype=float32)
 Uyakwazi ukubonisa i-diagram ngokucacisa ngaphakathi kwe-TensorBoard summary. # Set up logging.
stamp = datetime.now().strftime("%Y%m%d-%H%M%S")
logdir = "logs/func/%s" % stamp
writer = tf.summary.create_file_writer(logdir)

# Create a new model to get a fresh trace
# Otherwise the summary will not see the graph.
new_model = MySequentialModule()

# Bracket the function call with
# tf.summary.trace_on() and tf.summary.trace_export().
tf.summary.trace_on(graph=True)
tf.profiler.experimental.start(logdir)
# Call only one tf.function when tracing.
z = print(new_model(tf.constant([[2.0, 2.0, 2.0]])))
with writer.as_default():
  tf.summary.trace_export(
      name="my_func_trace",
      step=0,
      profiler_outdir=logdir)
 tf.Tensor([[0. 0.]], shape=(1, 2), dtype=float32)
 Yenza i-TensorBoard ukuze ubone i-trace efanelekileyo: #docs_infra: no_execute
%tensorboard --logdir logs/func
 Ukwakhiwa a SavedModel Umgangatho olusetyenziselwa ukusuka kwimodeli epheleleyo isebenzisa Ngathi kuquka kunye nenkundla yeentlobo kunye nenkundla yeentlobo. SavedModel SavedModel Uyakwazi ukugcina umzila owayephiweyo ngexesha elandelayo: tf.saved_model.save(my_model, "the_saved_model")
 INFO:tensorflow:Assets written to: the_saved_model/assets
 # Inspect the SavedModel in the directory
ls -l the_saved_model
 total 32
drwxr-sr-x 2 kbuilder kokoro  4096 Oct 18 01:21 assets
-rw-rw-r-- 1 kbuilder kokoro    58 Oct 18 01:21 fingerprint.pb
-rw-rw-r-- 1 kbuilder kokoro 17704 Oct 18 01:21 saved_model.pb
drwxr-sr-x 2 kbuilder kokoro  4096 Oct 18 01:21 variables
 # The variables/ directory contains a checkpoint of the variables
ls -l the_saved_model/variables
 total 8
-rw-rw-r-- 1 kbuilder kokoro 490 Oct 18 01:21 variables.data-00000-of-00001
-rw-rw-r-- 1 kbuilder kokoro 356 Oct 18 01:21 variables.index
 Yintoni ifayile A Ukubonisa i-functional . saved_model.pb inkqubo buffer tf.Graph Iimodeli kunye neengxaki ziyafumaneka ukusuka kule umzekelo ngaphandle kokwenza umzekelo weqela owenziwe. Oku kufuneka kwiimeko apho unayo (okanye ufuna) umxholo wePython, njenge-serving kwi-scale okanye kwi-edge device, okanye kwiimeko apho ikhowudi yePython yokuqala ayikho okanye efanelekileyo ukusetyenziswa. Uyakwazi ukulanda i-model njenge-object entsha: new_model = tf.saved_model.load("the_saved_model")
 , esekelwe ukusetyenziswa kwimodeli ebhalisiweyo, i-internal TensorFlow user object ngaphandle kweenkcukacha ze-class. It is not of type . new_model SequentialModule isinstance(new_model, SequentialModule)
 False
 Le model entsha isebenza kwi-input signatures ezidlulileyo. Unako ukongeza i-signatures ezininzi kwi-model eyenziwe ngexesha elandelayo. print(my_model([[2.0, 2.0, 2.0]]))
print(my_model([[[2.0, 2.0, 2.0], [2.0, 2.0, 2.0]]]))
 tf.Tensor([[0.31593648 0.        ]], shape=(1, 2), dtype=float32)
tf.Tensor(
[[[0.31593648 0.        ]
  [0.31593648 0.        ]]], shape=(1, 2, 2), dtype=float32)
 Ukusetyenziswa , unako ukugcina i-TensorFlow ama-weights kunye ne-graphs usebenzisa Yaye ngoko kwakhona. SavedModel tf.Module 
 
 Okokuqala ifumaneka kwi-website yeTensorFlow, le nqaku ifumaneka apha phantsi kwegama elitsha kwaye ifumaneka phantsi kweCC BY 4.0. Iimveliso ze-code zithunyelwe phantsi kwe-Apache 2.0 License. Okokuqala ifumaneka kwi kwisayithi, le nqaku linokufumaneka apha phantsi kwi-headline esitsha kwaye ilawulwa phantsi kwe-CC BY 4.0. Isampula ye-code ilawulwa phantsi kwe-Apache 2.0 License. TensorFlow TensorFlow

Le audio iveliswe ngolwimi lokuqala lwebali!

Inde kakhulu; Ukufunda

Make resilience your competitive advantage

Indlela yokufaka, ukugcina, kunye nokuguqula iimodeli kwi-TensorFlow

About Author

IZIMVO

ZIJONGE IIMPAWU

ELI NQAKU LINIKEZELWE KU

Related Stories

Light-Mode

Classic

Newspaper

Dark-Mode

Neon Noir

Minty

HN StartUps