Hello, I am trying to make fully reproducible ADVI runs, meaning that the approx.hist path should be exactly the same between runs. The current specification of seed at pm.ADVI seems to work. with pm.Model() as model: .... with model : advi = pm.ADVI(random_seed=74) approx = advi.fit(n…

ADVI reproducibility

ricardoV94 July 7, 2022, 10:32am 6

What was out of scope? If seeding is not possible we should open an issue on GitHub

Topic		Replies	Views
Saving ADVI results and reloading Questions	6	2029	April 15, 2018
Poor Accuracy of ADVI for Linear Regression Questions	12	3457	April 18, 2018
ADVI start with initialization Questions	7	2090	September 23, 2017
How to save fitted ADVI Result? Questions	3	2129	August 6, 2018
What's the difference between pm.fit(..., method = 'ADVI' ) and pm.ADVI() Questions	4	1567	November 16, 2018