急于执行时，Tensor对象不可迭代…使用Keras形状函数时

Question

I am trying to implement the following. I need help with the error, I get, during execution, and in general, suggestions, if there is a better way to implement the same idea, described below.

Network Model - https://i.imgur.com/LdKwXRP.png

I have a variable number of networks, eg in this fig, there are 3 illustrated. Each layer is executed sequentially.

The outputs of the other dense layers are used as an input for doing a dot product with the output of the dense layer that is currently being executed. The implementation for that is done by, storing the output of these intermediate layers, in a class variable list, called ActorNetwork.k_list.

During implementation, I get the error -

Traceback (most recent call last):
  File "training-code.py", line 80, in <module>
    main(args)
  File "training-code.py", line 36, in main
    ActorNetwork(sess, observation_dim[i], action_dim[i], float(args['actor_lr']), float(args['tau']), n))
  File "/home/rangwala/maddpg-attn/actorcriticv2.py", line 30, in __init__
    self.mainModel, self.mainModel_weights, self.mainModel_state = self._build_model()
  File "/home/rangwala/maddpg-attn/actorcriticv2.py", line 55, in _build_model
    keys = Input(shape=(K.shape(ActorNetwork.k_list,)))
  File "/home/rangwala/anaconda3/envs/comm-nav-cpu/lib/python3.6/site-packages/keras/engine/input_layer.py", line 171, in Input
    batch_shape = (None,) + tuple(shape)
  File "/home/rangwala/anaconda3/envs/comm-nav-cpu/lib/python3.6/site-packages/tensorflow/python/framework/ops.py", line 431, in __iter__
    "Tensor objects are not iterable when eager execution is not "
TypeError: Tensor objects are not iterable when eager execution is not enabled. To iterate over this tensor use tf.map_fn.

CODE -

class ActorNetwork(object):
"""
Implements actor network
"""

    k_list = [] #stores the values from the intermediate layers

# More code in between

    def _build_model(self):
        input_obs = Input(shape=(self.state_dim,))
        keys = Input(shape=(K.shape(ActorNetwork.k_list, ))) 
        #k_list is a class variable, of the ActorNetwork class.

        h = Dense(400)(input_obs)
        h = Activation('relu')(h)

        query = Dense(self.n_attn, name="keys_layer")(h)
        ky_list = ActorNetwork.k_list
        keys_list = ky_list.pop(self.n) #remove own entry from the list, for dot product
        concat_layer = Concatenate(axis=1)
        all_agents = concat_layer(keys_list)

        attn = tf.einsum('i, i->ij', [query, all_agents]) / self.temper #dot product

        attn = Activation('softmax')(attn)
        attn = Dropout(0.1)(attn)

        attn_out = tf.einsum('ik, k->i', [all_agents, attn])
        attn_add = Lambda(lambda x: x[0] + x[1])([query, attn_out]) #add own value to dot product value

        h = Dense(self.action_dim)(attn_add)
        pred = Activation('tanh')(h)
        pred = BatchNormalization()(pred)

        model = Model(inputs=[input_obs, keys], outputs=pred)
        model.compile(optimizer='Adam', loss='categorical_crossentropy')

        attn_layer_out = model.get_layer("keys_layer").output
        ActorNetwork.k_list[self.n] = attn_layer_out 

        return model, model.trainable_weights, input_obs

Answer 1

It looks like keras is trying to treat a shape tensor as an iterable - which it can only do in eager mode. Assuming you know the static shape of ActorNetwork.k_list , you can convert this to a list by using k_list.shape.as_list() .

I suspect its in the keys line, so try

keys = Input(shape=ActorNetwork.k_list.shape.as_list()))