Hi, I am trying to use ADVI to train a model on the GPU. I then want to use the fit results and generate the trace on the CPU for memory reasons. I have followed How to save fitted ADVI Result? which doesn’t appear to work. with model: approx=pm.ADVI() approx.fit() saveparam = [(param.name…

You are right… seems the ordering mapping in logp_dlogp_function is not exactly the same as the one used in opvi. Try this: ... # save inference bij = inference.approx.groups[0].bij saveparam = {param.name: bij.rmap(param.eval()) for param in inference.approx.params} # new model model2 = constr…

Saving ADVI results and reloading

Questions

junpenglao April 13, 2018, 5:19am 2

I checked on a simple model and it still work, are you sure model and model2 are identical?

Topic		Replies	Views
How to save fitted ADVI Result? Questions	3	2129	August 6, 2018
Setting ADVI start state Questions	3	739	October 14, 2019
ADVI start with initialization Questions	7	2090	September 23, 2017
Minibatch for a large dataset ADVI Questions	2	1223	September 7, 2018
Pickling trace object Questions	6	2320	September 16, 2018

Saving ADVI results and reloading

Related topics