Can the DynamicLDA do inference for new corpus?

We want to apply DTM (dynamic topic model) to see the topic evaluation via time, and our data will be fresh daily. 
However, I applied DTM provided by Gensim before but it only can training batch dataset. Hence it is not suitable for our scenario. The speed it bad and we cannot put all of the historical data to this model. 

Want to know if the DynamicLDA can do this? Thanks.