I am trying to fit a GMM in sklearn and i see that the model converges at around epoch 3 but i cannot seems to access the log-likelihood score computed at each epoch.