FailedPreconditionError：尝试使用张量板时使用未初始化的值

Question

在我尝试实现tensorboard之前，代码工作正常。在我的model.fit中添加回调后，出现以下错误：

  

FailedPreconditionError：尝试使用未初始化的值   training_23 / Adam / Variable_7 [[节点：training_23 / Adam / Variable_7 / read   = IdentityT = DT_FLOAT，_class = [“ loc：@ training_23 / Adam / Assign_10”]，_ device =“ / job：localhost /副本：0 / task：0 / device：CPU：0”]]

我尝试添加到tf.global_variables_initializer()无济于事。我不知道缺少什么。下面是我的代码：

from keras.models import Sequential
from keras.layers import Dense, Dropout
import numpy
from tensorflow.keras.callbacks import TensorBoard
import tensorflow as tf

#random seed for reproducibility
numpy.random.seed(6)

NAME = "Test"
tensorboard = TensorBoard(log_dir="logs/{}".format(NAME))


# loading load prima indians diabetes dataset, past 5 years of medical 
history 
dataset = numpy.loadtxt("vmlb3.csv", delimiter=",")
test1 = numpy.loadtxt("mpredict.csv", delimiter=",")

# split into input (X) and output (Y) variables, splitting csv data
X = dataset[:,0:4]
Y = dataset[:,4]


# create model, add dense layers one by one specifying activation function
model = Sequential()
model.add(Dense(20, input_dim=4, activation='relu')) # input layer requires 
input_dim param
model.add(Dense(10, activation='relu'))
model.add(Dense(1, activation='sigmoid')) # sigmoid instead of relu for 
final probability between 0 and 1

# compile the model, adam gradient descent (optimized)
model.compile(loss="binary_crossentropy", optimizer="adam", metrics= 
['accuracy'])

# call the function to fit to the data (training the network)
model.fit(X, Y, epochs = 1, batch_size=20, validation_split = 0.3, 
**callbacks=[tensorboard]**)

# save the model
#model.save('weights.h5')