I have calculated a zero-truncated negative-binomial model using glmmadmb, and also vglm, in R. I need to report some measure of goodness of fit for this model. I have already done cross validation (training-test sets), but would like to report another goodness of fit measure too. For a logistic regression I would report % deviance explained, or a pseudo r-squared. Are there similar tests for a zero-truncated negative-binomial model? The R functions do not report deviance, and I am not sure how to calculate it.