Currently EigenReporter supports streaming updates to its covariance statistics via the update method, but it does an unweighted average. For the training use-case, we'll want to use an exponential moving average since the representations of the model will change over time.
Currently
EigenReportersupports streaming updates to its covariance statistics via theupdatemethod, but it does an unweighted average. For the training use-case, we'll want to use an exponential moving average since the representations of the model will change over time.