CODEWITHSUNDEEP
pythontensorflowkerastheano
InternetRelayCat
InternetRelayCat

Reputation: 13

Theano, TypeError: Unknown parameter type: <class 'tensorflow.python.framework.ops.Tensor'>

When I run the following code (which is from this IPython notebook), I get an error:

import theano

def get_submodel(model, start, end):
    return theano.function([model.layers[start].input],
                          model.layers[end].get_output(train=False),
                          allow_input_downcast=True)

def get_encoder(ae):
    return get_submodel(ae, 0, (len(ae.layers) // 2) - 1)

ae_encoder = get_encoder(ae)

This is the error message:

Traceback (most recent call last):
  File "Parametric t-SNE (Keras).py", line 432, in <module>
    ae_encoder = get_encoder(ae)
  File "Parametric t-SNE (Keras).py", line 424, in get_encoder
    return get_submodel(ae, 0, (len(ae.layers) // 2) - 1)
  File "Parametric t-SNE (Keras).py", line 422, in get_submodel
    allow_input_downcast=True)
  File "/usr/lib64/python3.4/site-packages/theano/compile/function.py", line 326, in function
    output_keys=output_keys)
  File "/usr/lib64/python3.4/site-packages/theano/compile/pfunc.py", line 397, in pfunc
    for p in params]
  File "/usr/lib64/python3.4/site-packages/theano/compile/pfunc.py", line 397, in <listcomp>
    for p in params]
  File "/usr/lib64/python3.4/site-packages/theano/compile/pfunc.py", line 496, in _pfunc_param_to_in
    raise TypeError('Unknown parameter type: %s' % type(param))
TypeError: Unknown parameter type: <class 'tensorflow.python.framework.ops.Tensor'>

For reference, here is where ae is defined:

n = X_train.shape[1]
ae = Sequential()
ae.add(Dense(500, activation='relu', weights=encoder.layers[0].get_weights(), input_shape=(n,)))
ae.add(Dense(500, activation='relu', weights=encoder.layers[1].get_weights()))
ae.add(Dense(2000, activation='relu', weights=encoder.layers[2].get_weights()))
ae.add(Dense(2, weights=encoder.layers[3].get_weights()))
ae.add(Dense(2000, activation='relu', weights=decoder.layers[0].get_weights()))
ae.add(Dense(500, activation='relu', weights=decoder.layers[1].get_weights()))
ae.add(Dense(500, activation='relu', weights=decoder.layers[2].get_weights()))
ae.add(Dense(n, weights=decoder.layers[3].get_weights()))
ae.compile(loss='mse', optimizer='rmsprop')
ae.fit(X_train, X_train, nb_epoch=100, verbose=2, batch_size=32)

Based on the answer to a similar question, I suspect that get_submodel may need to be modified to use a symbolic variable instead of a tensor/matrix. However, I am not sure how to do this, and why it would even give an error since the IPython notebook on GitHub did not seem to contain any error messages. I have not been able to find more specific advice about the tensorflow.python.framework.ops.Tensor error message.

Upvotes: 1

Views: 2414

Answers (1)

Marcin Możejko
Marcin Możejko

Reputation: 40516

It seems that you have a tensorflow backend - not a theano one. That's why using theano.function generates an error. Try using keras.backend.function.

Upvotes: 1

Related Questions