Matthias Rosenkranz
Matthias Rosenkranz
Likely fixed with PR #824. Could you try running your code on current master?
Is it possible to post the full example including your inference? Your code works for me on master if I just use HMC with an `Empirical` posterior for `Eps` and...
It's not in iteration or any of the other cells (uncollapsed qx not shown because it didn't fit on screen): ![edward_tf_session_issue_uncollapsed](https://user-images.githubusercontent.com/2349370/36645991-a7f104ec-1a71-11e8-9100-4279e3cb1119.png)
That did it. Thanks. Browsing through the code I guess my issue was that the default graph is written in `Inference.initialize()`, which is called from `VariationalInference.initialize()` before any optimizer or...
Hi, please have a look at the mixture density networks tutorial. Basically, you need to update the inference stepwise and save the loss at each step.
That's my understanding as well. Happy to be corrected if there is an easier way.