Add easy access to the hmmlearn ConvergenceMonitor
eonu opened this issue · 1 comments
eonu commented
Currently when fitting a GMMHMM, there is no way to track the convergence of the Baum–Welch algorithm. This can be done with the ConvergenceMonitor class in hmmlearn, which is instantiated when the HMM is created, and can be accessed by hmm.monitor_.
In our fit() method, we would have to save the ConvergenceMonitor as an instance variable before we call the fit() method from hmmlearn.