Keras 1.2 Functional API 'GraphModel' object has no attribute 'get_layers'

I’m using the Keras 1.2 w/ TF (dev-tf branch) functional API & updated to use kc.convert_functional_model.

Here’s the error:

---------------------------------------------------------------------------
AttributeError                            Traceback (most recent call last)
<ipython-input-10-262bf3b052dd> in <module>()
      4 
      5 deeplift_model = revealcancel_model
----> 6 deeplift_prediction_func = compile_func([deeplift_model.get_layers()[0].get_activation_vars()],
      7                                        deeplift_model.get_layers()[-1].get_activation_vars())
      8 original_model_predictions = keras_model.predict(X_test.values, batch_size=200)

AttributeError: 'GraphModel' object has no attribute 'get_layers'

Here’s the code snippet I’m running…

...
revealcancel_model = kc.convert_functional_model(model=keras_model, nonlinear_mxts_mode=NonlinearMxtsMode.RevealCancel)

from  deeplift.util  import compile_func
import numpy as np
from keras import backend as K

deeplift_model = revealcancel_model
deeplift_prediction_func = compile_func([deeplift_model.get_layers()[0].get_activation_vars()],
                                       deeplift_model.get_layers()[-1].get_activation_vars())
original_model_predictions = keras_model.predict(X_test.values, batch_size=200)
converted_model_predictions = deeplift.util.run_function_in_batches(
                                input_data_list=[X_test.values],
                                func=deeplift_prediction_func,
                                batch_size=200,
                                progress_update=None)
print("difference in predictions:",np.max(np.array(converted_model_predictions)-np.array(original_model_predictions)))
assert np.max(np.array(converted_model_predictions)-np.array(original_model_predictions)) < 10**-5
predictions = converted_model_predictions

[donigian]

@AvantiShri It works now.

In summary in case anyone would like to use the Keras Functional API…

1. Run deeplift_model.get_name_to_blob().keys()
2. Update your compile_func to use layer name, example…

deeplift_prediction_func = compile_func([deeplift_model.get_name_to_blob()["input_1"].get_activation_vars()], deeplift_model.get_name_to_blob()["dense_4"].get_activation_vars())

1. Update get_target_contribs_func to use layer names as well, note you’ll need the pre-activation layer here…

revealcancel_func = revealcancel_model.get_target_contribs_func(find_scores_layer_name="input_1", pre_activation_target_layer_name="preact_dense_4")

[JanZrimec]

UPDATE: deeplift_model.get_name_to_layer().keys()