KeyError: 'batch_input_shape' while using deeplift on keras

[AmirHoseinSafari]

Hi,

I am using Keras (V 2.2.4) and my model is LRCN (a combination of CNN + LSTM). You can see my code here: https://github.com/AmirHoseinSafari/LRCN-drug-resistance/blob/master/models/Bayesian_optimizer.py I am trying to run it this way:

   X_train2 = np.array(X_train).astype(np.float)
   X_test2 = np.array(X_test).astype(np.float)
   X_val2 = np.array(X_val).astype(np.float)

   y_train2 = np.array(y_train).astype(np.float)
   y_test2 = np.array(y_test).astype(np.float)
   y_val2 = np.array(y_val).astype(np.float)

  model.compile(
       loss=masked_loss_function,
       optimizer='Adam',
       metrics=[masked_accuracy]
   )
   history = model.fit(
       X_train2,
       y_train2,
       epochs=epochs,
       batch_size=128,
       verbose=2,
       validation_data=(X_val2, y_val2)
   )

  model.save('model_save.h5')

import deeplift
from deeplift.conversion import kerasapi_conversion as kc

deeplift_model = \
    kc.convert_model_from_saved_files(
        'model_save.h5',
        nonlinear_mxts_mode=deeplift.layers.NonlinearMxtsMode.DeepLIFT_GenomicsDefault)

find_scores_layer_idx = 0

deeplift_contribs_func = deeplift_model.get_target_contribs_func(
    find_scores_layer_idx=find_scores_layer_idx,
    target_layer_idx=-1)

scores = np.array(deeplift_contribs_func(task_idx=0,
                                         input_data_list=[X],
                                         batch_size=10,
                                         progress_update=1000))`

My model is something like:

model = Sequential()
model.add(Dropout(dropout2_rate))
model.add(Conv1D(filters=filterCNN1, kernel_size=kernelCNN1, activation='relu', padding='same'))
model.add(MaxPooling1D(pool_size=poolCNN1, padding='same'))
model.add(LSTM(LSTM1, return_sequences=False, recurrent_dropout=0.3))
model.add(Dropout(dropout2_rate))
model.add(Dense(dense_1_neurons))
model.add(Dropout(dropout2_rate))
model.add(Dense(12, activation='sigmoid'))

I am getting this error (at convert_model_from_saved_files function) and I couldn't find any way to resolve it. Could you please help me with this?

Here is the error:

Traceback (most recent call last): File "/Users/amir/PycharmProjects/Lab/LSTM-DR/LSTM-drug-resistance/pipe_line_gene.py", line 31, in train() File "/Users/amir/PycharmProjects/Lab/LSTM-DR/LSTM-drug-resistance/pipe_line_gene.py", line 14, in train model_gene_based.run_bayesian(df_train, labels) File "/Users/amir/PycharmProjects/Lab/LSTM-DR/LSTM-drug-resistance/models/model_gene_based.py", line 578, in run_bayesian Bayesian_optimizer.BO(X_train, X_test, X_val, y_train, y_test, y_val, limited, portion) File "/Users/amir/PycharmProjects/Lab/LSTM-DR/LSTM-drug-resistance/models/Bayesian_optimizer.py", line 409, in BO fit_with_partial(dropout2_rate=0.2, dense_1_neurons_x128=1, dense_2_neurons_x128=1, dense_3_neurons_x128=1, File "/Users/amir/PycharmProjects/Lab/LSTM-DR/LSTM-drug-resistance/models/Bayesian_optimizer.py", line 191, in fit_with return run_one_fold(model) File "/Users/amir/PycharmProjects/Lab/LSTM-DR/LSTM-drug-resistance/models/Bayesian_optimizer.py", line 304, in run_one_fold kc.convert_model_from_saved_files( File "/Users/amir/venv/p3/lib/python3.8/site-packages/deeplift/conversion/kerasapi_conversion.py", line 414, in convert_model_from_saved_files return model_conversion_function(model_config=model_config, **kwargs) File "/Users/amir/venv/p3/lib/python3.8/site-packages/deeplift/conversion/kerasapi_conversion.py", line 459, in convert_sequential_model batch_input_shape = model_config[0]['config'][KerasKeys.batch_input_shape] KeyError: 'batch_input_shape'