Implemented Relaxed Distributions

[rachtsingh]

Not sure if this is the right name for this distribution (Concrete / GumbelSoftmax are other ideas), but this is what Tensorflow calls it. This PR uses the transforms machinery :)

I had to edit TransformedDistribution's log_prob method to take into account event_shape - this is probably not the right way to do it, but a quick first try that makes it work.

• [x] Implement RelaxedOneHotCategorical
• [x] Implement RelaxedBernoulli

cc @fritzo

[rachtsingh]

Ok, made changes based on comments, and added the RelaxedBernoulli distribution. No longer blocked, so ready for another review if necessary @fritzo.

[rachtsingh]

Added tests that:

1. Rounding the RelaxedBernoulli distribution gives the corresponding Bernoulli distribution
2. Taking the argmax of the RelaxedOneHotCategorical gives the corresponding Categorical distribution
3. As the temperature becomes very large, the first consistently gives 0.5 and the latter gives equal values for each index.