refactor training - separate staggered training; make differences as small as possible

models/pauls_networks.py

@@ -3,7 +3,6 @@ from collections import namedtuple
 import keras
 from keras.engine import Input, Model as KerasModel
 from keras.layers import Activation, Conv1D, Dense, Dropout, Embedding, GlobalMaxPooling1D, TimeDistributed
-from keras.regularizers import l2
 
 import dataset
 
@@ -58,7 +57,7 @@ def get_model(cnnDropout, flow_features, domain_features, window_size, domain_le
     # remove temporal dimension by global max pooling
     y = GlobalMaxPooling1D()(y)
     y = Dropout(cnnDropout)(y)
-    y = Dense(dense_dim, kernel_regularizer=l2(0.1), activation='relu')(y)
+    y = Dense(dense_dim, activation='relu')(y)
     out_client = Dense(1, activation='sigmoid', name="client")(y)
     out_server = Dense(1, activation='sigmoid', name="server")(y)