Seems using a single, separate variable for (log?) standard deviation is more popular than making it part of the network, e.g. (Schulman et al., 2015). Should probably use this way instead of currently implemented, at least while comparing algorithms against baselines.
Can't use tf.get_variable() though, goes away in 2.0.
Seems using a single, separate variable for (log?) standard deviation is more popular than making it part of the network, e.g. (Schulman et al., 2015). Should probably use this way instead of currently implemented, at least while comparing algorithms against baselines.
Can't use
tf.get_variable()
though, goes away in 2.0.