bramblexu / aspect-term-extraction Goto Github PK

Hello,
I am trying to implement bi-lstm-crf-embedding.ipynb on my dataset using google colab. However, I have the following problem by compiling the model:

`---------------------------------------------------------------------------
AttributeError Traceback (most recent call last)
in ()
10 model = TimeDistributed(Dense(50, activation="tanh"))(model) # a dense layer as suggested by neuralNer
11 crf = CRF(n_tags) # CRF layer
---> 12 out = crf(model) # output
13
14 model = Model(input, out)

4 frames
/usr/local/lib/python3.6/dist-packages/tensorflow/python/autograph/impl/api.py in wrapper(*args, **kwargs)
668 except Exception as e: # pylint:disable=broad-except
669 if hasattr(e, 'ag_error_metadata'):
--> 670 raise e.ag_error_metadata.to_exception(e)
671 else:
672 raise

AttributeError: in user code:

/usr/local/lib/python3.6/dist-packages/keras_contrib/layers/crf.py:292 call  *
    test_output = self.viterbi_decoding(X, mask)
/usr/local/lib/python3.6/dist-packages/keras_contrib/layers/crf.py:564 viterbi_decoding  *
    argmin_tables = self.recursion(input_energy, mask, return_logZ=False)
/usr/local/lib/python3.6/dist-packages/keras_contrib/layers/crf.py:521 _step  *
    return self.step(input_energy_i, states, return_logZ)
/usr/local/lib/python3.6/dist-packages/keras_contrib/layers/crf.py:463 step  *
    m = K.slice(states[3], [0, t], [-1, 2])

AttributeError: module 'keras.backend' has no attribute 'slice'`

Could you help me please?

Best regards!

Why does this model work so badly? Tenorfolw implements a model that works so well.