直方图频率激活时的 Keras 张量板错误

Question

当尝试使用Tesorboard监控训练时，我正在使用后端Tensorflow的Keras运行一个简单的神经网络。

我的模型如下：

import keras
from keras.layers.core import Dense, Activation, Dropout
from keras.models import Sequential
model = Sequential()
model.add(Dense(32, input_dim=500))
model.add(Activation('relu'))
model.add(Dropout(0.2))
model.add(Dense(2, activation='softmax'))
model.compile(optimizer='rmsprop',
loss='binary_crossentropy',
metrics=['accuracy'])

我尝试使用以下配置运行它：

tensorboard = keras.callbacks.TensorBoard(log_dir="../TFlogs",
histogram_freq=2, batch_size=300, 
write_grads=True, write_images=True)
one_hot_labels = keras.utils.to_categorical(y_train, num_classes=2)
history = model.fit(X_train, one_hot_labels, epochs=10,
validation_split=0.1, batch_size=300, callbacks=[tensorboard])

但是由于某种原因，我收到以下错误：

File "C:/Users/user/PycharmProjects/SpamFilter/Filter/NeuralModel.py", line 108, in fit
history = model.fit(X_train, one_hot_labels, epochs=10, validation_split=0.1, batch_size=300, callbacks=[tensorboard])
File "C:UsersuserPycharmProjectsSpamFiltervenvKeraslibsite-packageskerasmodels.py", line 1002, in fit
validation_steps=validation_steps)
File "C:UsersuserPycharmProjectsSpamFiltervenvKeraslibsite-packageskerasenginetraining.py", line 1705, in fit
validation_steps=validation_steps)
File "C:UsersuserPycharmProjectsSpamFiltervenvKeraslibsite-packageskerasenginetraining.py", line 1256, in _fit_loop
callbacks.on_epoch_end(epoch, epoch_logs)
File "C:UsersuserPycharmProjectsSpamFiltervenvKeraslibsite-packageskerascallbacks.py", line 77, in on_epoch_end
callback.on_epoch_end(epoch, logs)
File "C:UsersuserPycharmProjectsSpamFiltervenvKeraslibsite-packageskerascallbacks.py", line 855, in on_epoch_end
result = self.sess.run([self.merged], feed_dict=feed_dict)
File "C:UsersuserPycharmProjectsSpamFiltervenvKeraslibsite-packagestensorflowpythonclientsession.py", line 905, in run
run_metadata_ptr)
File "C:UsersuserPycharmProjectsSpamFiltervenvKeraslibsite-packagestensorflowpythonclientsession.py", line 1109, in _run
np_val = np.asarray(subfeed_val, dtype=subfeed_dtype)
File "C:UsersuserPycharmProjectsSpamFiltervenvKeraslibsite-packagesnumpycorenumeric.py", line 492, in asarray
return array(a, dtype, copy=False, order=order)
ValueError: setting an array element with a sequence.

当我设置TensorBoard选项histogram_freq=0时，我没有收到错误，但我想获取直方图信息，有人可以帮助我找到原因吗？

Answer 1

在创建模型之前尝试运行K.clear_session()。

有关此错误的参考，请参阅此 GitHub 线程。

Answer 2

这可能与K.learing_phase()有关。 特别是如果你以前做过K.set_learning_phase(1)。

要诊断： 运行print(K.learning_phase())，如果它返回一个 int，那么这个问题几乎肯定与这个问题有关。 尝试删除所有与K.set_learing_phase(1)相关的句子，看看是否有所作为。