调用 Keras.backend.clear_session() 后无法导入预训练模型

Question

我正在尝试在 keras 循环中的每次迭代中使用新的数据样本训练模型(使用 tensorflow 后端(。由于一些迭代后的 GPU 内存错误，我附加了 K.clear_session((。但是，在一次迭代后，代码会引发错误：

'Cannot interpret feed_dict key as Tensor: ' + e.args[0])
TypeError: Cannot interpret feed_dict key as Tensor: Tensor Tensor("Placeholder:0", shape=(7, 7, 3, 64), dtype=float32) is not an element of this graph.

如果我在最后删除 K.clear_session((，则没有错误。有没有人可以解释为什么这个错误会在第二次迭代中出现？

我尝试了其他方法(对于 gpu 发布(，但没有一种有效，这是我的最后选择。但它会抛出错误。我粘贴了一个可以产生错误的示例代码。请注意，这不是实际代码，我只是做了一个例子来重现我在实际代码中面临的错误。

from __future__ import absolute_import, division, print_function, unicode_literals
import numpy as np
import tensorflow as tf
import random
seed_value= 0
import os
import keras
os.environ['PYTHONHASHSEED']=str(seed_value)
random.seed(0)
np.random.seed(0)
from keras import backend as K
from keras.datasets import cifar10
(x_train, y_train), (x_test, y_test) = cifar10.load_data()
for i in range(3):
    base_model = tf.keras.applications.resnet50.ResNet50(weights='imagenet', input_shape=(32, 32, 3),
                                               include_top=False)
    x = base_model.output
    x = tf.keras.layers.GlobalAveragePooling2D()(x)
    output = tf.keras.layers.Dense(10, activation='softmax',
                               kernel_initializer=tf.keras.initializers.RandomNormal(seed=4))(x)
    model = tf.keras.Model(inputs=base_model.input, outputs=output)
    y_train = keras.utils.to_categorical(y_train, 10)
    y_test = keras.utils.to_categorical(y_test, 10)
    for layer in base_model.layers:
        layer.trainable = False
    optimizer = tf.train.AdamOptimizer(learning_rate=0.0001)
    model.compile(optimizer=optimizer, loss='categorical_crossentropy',
              metrics=['accuracy'])
    model.fit(x_train,y_train,batch_size=1024,epochs=1,verbose=1)
    K.clear_session()

Traceback (most recent call last):
  File "C:UserssirshadAppDataLocalProgramsPythonPython36libsite-packagestensorflowpythonclientsession.py", line 1092, in _run
    subfeed, allow_tensor=True, allow_operation=False)
  File "C:UserssirshadAppDataLocalProgramsPythonPython36libsite-packagestensorflowpythonframeworkops.py", line 3490, in as_graph_element
    return self._as_graph_element_locked(obj, allow_tensor, allow_operation)
  File "C:UserssirshadAppDataLocalProgramsPythonPython36libsite-packagestensorflowpythonframeworkops.py", line 3569, in _as_graph_element_locked
    raise ValueError("Tensor %s is not an element of this graph." % obj)
ValueError: Tensor Tensor("Placeholder:0", shape=(7, 7, 3, 64), dtype=float32) is not an element of this graph.
During handling of the above exception, another exception occurred:
Traceback (most recent call last):
  File "E:/codes/experiments-AL/breakhis/40X-M-B/codes-AL/error_debug.py", line 22, in <module>
    include_top=False)
  File "C:UserssirshadAppDataLocalProgramsPythonPython36libsite-packagestensorflowpythonkerasapplications__init__.py", line 70, in wrapper
    return base_fun(*args, **kwargs)
  File "C:UserssirshadAppDataLocalProgramsPythonPython36libsite-packagestensorflowpythonkerasapplicationsresnet50.py", line 32, in ResNet50
    return resnet50.ResNet50(*args, **kwargs)
  File "C:UserssirshadAppDataLocalProgramsPythonPython36libsite-packageskeras_applicationsresnet50.py", line 291, in ResNet50
    model.load_weights(weights_path)
  File "C:UserssirshadAppDataLocalProgramsPythonPython36libsite-packagestensorflowpythonkerasenginenetwork.py", line 1544, in load_weights
    saving.load_weights_from_hdf5_group(f, self.layers)
  File "C:UserssirshadAppDataLocalProgramsPythonPython36libsite-packagestensorflowpythonkerasenginesaving.py", line 806, in load_weights_from_hdf5_group
    K.batch_set_value(weight_value_tuples)
  File "C:UserssirshadAppDataLocalProgramsPythonPython36libsite-packagestensorflowpythonkerasbackend.py", line 2784, in batch_set_value
    get_session().run(assign_ops, feed_dict=feed_dict)
  File "C:UserssirshadAppDataLocalProgramsPythonPython36libsite-packagestensorflowpythonclientsession.py", line 929, in run
    run_metadata_ptr)
  File "C:UserssirshadAppDataLocalProgramsPythonPython36libsite-packagestensorflowpythonclientsession.py", line 1095, in _run
    'Cannot interpret feed_dict key as Tensor: ' + e.args[0])
TypeError: Cannot interpret feed_dict key as Tensor: Tensor Tensor("Placeholder:0", shape=(7, 7, 3, 64), dtype=float32) is not an element of this graph.
Process finished with exit code 1

Answer 1

我能够通过将 imagenet 预训练模型保存到磁盘，然后在调用 tf.keras.backend.clear_session(( 后每次循环加载来克服这个问题。因此，将基本模型保存到文件然后加载即可。但我仍然很困惑为什么它以前不起作用

base_model = tf.keras.applications.resnet50.ResNet50