Using float("-inf")

[jacobjinkelly]

ptr_decoder.py uses negative infinity (i.e. logarithm of 0). It is unclear if this is best practice or if there is a better way to model the log probability of probability 0.