def eval_cnn(preds, placeholders, sess, graph, inputs, outputs, _):
with graph.as_default():
res = sess.run(tf.nn.softmax(preds["n1"]),
feed_dict={placeholders["i0"]: inputs["i0"]})
sess.close()
return evolution.accuracy_error(res, outputs["o0"]),
def eval_sequential(preds, placeholders, sess, graph, inputs, outputs, _):
"""
Here we compute the fitness of the model. It is used by the evolutionary internally and always is provided with the same parameters
:param preds: Dictionary created in the arranging and training function
:param placeholders: (Only) Input placeholders: ("i0", "i1", ..., "im").
:param sess: tf session to perform inference
:param graph: tf graph in which inference is performed
:param inputs: Data inputs for the model
:param outputs: Data outputs for the metric
:param _: hyperparameters, because we are evolving the optimizer selection and learning rate, they are unused when testing
:return: fitness of the model (as a tuple)
"""
with graph.as_default():
res = sess.run(tf.nn.softmax(preds["n1"]), feed_dict={placeholders["i0"]: inputs["i0"]})
sess.close()
return accuracy_error(res, outputs["o0"]),