Shaowu Pan
Shaowu Pan
**Is your feature request related to a problem? Please describe.** The whole purpose of TFR is to make hundreds of GB level training data possible by dividing them into little...
I use Tensorflow default save model API, it works. It can save and read model. But simply for a two layer neural network with 8 hidden units costs me 68MB...
In the API [page](http://edwardlib.org/api/inference) about model parameter inference, i.e., simply a tensorflow functionality. The code given in the API page is not working with error: ``` Traceback (most recent call...
I know I can pass tensorflow optimizer into edward. But what is the default optimization in KLqp? Is the one written in the 2016 JMLR-ADVI paper ?
edward PPCA example is given here: http://edwardlib.org/tutorials/probabilistic-pca In edward implementation, I would say it is a **VI** to minimize the KL divergence between a parameterized Gaussian for **W** and the...
This example `ode_demo.py` under `Example` is considering a NN with input cubic nonlinearity. Since the original dynamics is `y**3` so it seems a bit weird. This implies that as long...
I am using it to compare memory consumption for my code framework with others. how to cite?
It seems that no one really talk about the effect of activation functions. Personally, I found **Relu** gives what the most human-favorable uncertainty result while **tanh** has too much confidence...
I want to cite pele since I used your code to generate data using AMR. Which reference should I go for?