Philipp Brendel

Thanks for the confirmation! My idea was to use apply_output_transform in order to use a metric on a reference solution during training, but I think I get your point. I...

Dear @lululxvi, thank you for the clarification, I will check other options! I have a slightly related question that you might be able to answer quickly: Can PDEOperatorCartesianProd work with...

* found the solution via [this stackoverflow](https://stackoverflow.com/questions/47765595/tensorflow-attempting-to-use-uninitialized-value-beta1-power) ```py ... self.model.compile("adam", lr=1e-3, decay=None, loss_weights=[1, 1, 1, 1]) self.model.sess.run(tf.global_variables_initializer()) ``` Sorry for posting this a little early (in hindsight), but maybe this...