Saving gsdmm model

[Jedevold]

Once the model is trained, how do I save it for future classification?

[dtemkin]

It hasn't been merged yet, but you can try my branch. I added a save() and load() function.

[jonathan-chin]

@dtemkin I think I'm using these new functions correctly but when trying to load a saved model, I get the following error:

TypeError: MovieGroupProcess._from_data() missing 1 required positional argument: 'vocab_size'

any ideas? (also, thanks for adding these functions!)

[jonathan-chin]

@dtemkin I figured it out. your save() function is missing a parameter. it should read:

payload = {"K": self.K, "alpha":self.alpha, 'beta': self.beta, "D": self.number_docs, 'vocab_size': self.vocab_size, 'cluster_doc_count': self.cluster_doc_count,
                   "cluster_word_count": self.cluster_word_count,
                   "cluster_word_distribution": self.cluster_word_distribution,
                   "n_iters": self.n_iters}

notice the addition of vocab_size after D and before cluster_doc_count