Is there any way of getting the topic_word distribution from LightLDA?

[cbravo93]

I am interested in using LightLDA; however I am unable to find how to extract the topic word distribution after training the model (with fit_transform only the doc_topic distribution is returned).

Is there any solution for this? If yes, could it be included in the examples? If no, is there any chance of implementing it?

Similarly is it possible to get the log likelihood values?

Thanks!