Keras优化器的TensorFlow和Keras中期问题

如何解决Keras优化器的TensorFlow和Keras中期问题

目前，我正在学习聊天机器人编程的基础知识，并且几乎没有TensorFlow和Keras的经验。在对程序进行编码时，我遇到一条错误消息： AttributeError：模块'keras.optimizers'没有属性'TFOptimizer' 版 ：Tensorflow 2.1.0 ：keras 2.3.1 ：Python 3.7

import nltk
from nltk.stem import WordNetLemmatizer
lemmatizer = WordNetLemmatizer()
import json
import pickle
import tensorflow 
from tensorflow import keras
from tensorflow.keras import layers

from tensorflow.keras import optimizers

from tensorflow.python.keras.optimizers import TFOptimizer



import numpy as np
np.array(object,dtype=object,copy=True,order='K',subok=False,ndmin=0)
from keras.models import Sequential,Model
from keras.layers import Dense,Activation,Dropout,Lambda

import tensorflow as tf
physical_devices = tf.config.list_physical_devices('GPU') 
tf.config.experimental.set_memory_growth(physical_devices[0],True)
import random
words=[]
classes = []
documents = []
ignore_words = ['?','!']
data_file = open('intents.json',encoding='utf-8').read()
intents = json.loads(data_file)

问题：

model = Sequential()
model.add(Dense(128,input_shape=(len(train_x[0]),),activation='relu'))
model.add(Dropout(0.5))
model.add(Dense(64,activation='relu'))
model.add(Dropout(0.5))
model.add(Dense(len(train_y[0]),activation='softmax'))
sgd = keras.optimizers.Adam(lr=0.01,decay=1e-6)
model.compile(loss='categorical_crossentropy',optimizer=sgd,metrics=['accuracy']) 
hist = model.fit(np.array(train_x),np.array(train_y),epochs=200,batch_size=5,verbose=1)
model.save('chatbot_model.h5',hist)
print("model created")

错误消息：

AttributeError                            Traceback (most recent call last)
<ipython-input-25-54920be00d53> in <module>
     12 #fitting and saving the model
     13 hist = model.fit(np.array(train_x),verbose=1)
---> 14 model.save('chatbot_model.h5',hist)
     15 print("model created")

~\anaconda3\envs\tensorflow\lib\site-packages\keras\engine\network.py in save(self,filepath,overwrite,include_optimizer)
   1150             raise NotImplementedError
   1151         from ..models import save_model
-> 1152         save_model(self,include_optimizer)
   1153 
   1154     @saving.allow_write_to_gcs

~\anaconda3\envs\tensorflow\lib\site-packages\keras\engine\saving.py in save_wrapper(obj,*args,**kwargs)
    447                 os.remove(tmp_filepath)
    448         else:
--> 449             save_function(obj,**kwargs)
    450 
    451     return save_wrapper

~\anaconda3\envs\tensorflow\lib\site-packages\keras\engine\saving.py in save_model(model,include_optimizer)
    539                 return
    540         with H5Dict(filepath,mode='w') as h5dict:
--> 541             _serialize_model(model,h5dict,include_optimizer)
    542     elif hasattr(filepath,'write') and callable(filepath.write):
    543         # write as binary stream

~\anaconda3\envs\tensorflow\lib\site-packages\keras\engine\saving.py in _serialize_model(model,include_optimizer)
    161             layer_group[name] = val
    162     if include_optimizer and model.optimizer:
--> 163         if isinstance(model.optimizer,optimizers.TFOptimizer):
    164             warnings.warn(
    165                 'TensorFlow optimizers do not '

AttributeError: module 'keras.optimizers' has no attribute 'TFOptimizer'

解决方法

keras和tensorflow.keras是Keras API的两种不同实现，因此不应混用。 According to the creator of the Keras API，用户应优先选择tensorflow.keras实施。

新版本的多后端Keras：2.3.0

https://github.com/keras-team/keras/releases/tag/2.3.0

• 首个全面支持TF 2的多后端Keras版本
• 继续支持Theano / CNTK
• 将是多后端Keras的最后一个主要版本

我们建议您将Keras代码切换为tf.keras。

有关更多信息，请参见https://stackoverflow.com/a/63377877/5666087。