Zhenwen Dai

> @meissnereric how close is Gibbs sampling to getting merged? Thanks for you interest. This implementation only contains Metropolis Hasting and does not support Gibbs sampling, because a proper Gibbs...

@anirudhacharya Implementing a LDA model with stochastic variational inference is hard. See this issue about implementing LDA in Edward ([https://github.com/blei-lab/edward/issues/463](https://github.com/blei-lab/edward/issues/463)). Gibbs sampling implementation has very different comparing the inference methods...

Why it is unsafe? I think it is the python way to call ```self.model.inducing_inputs_str(layer)``` dynamically.