R: stm() package Prevalence regression failing to converge within iteration limit

I'm trying to estimate a manyTopics stm model but I get this error:

Error in vb.variational.reg(Y = lambda[, i], X = covar, Xcorr = Xcorr, : Prevalence regression failing to converge within iteration limit. May want to try gamma.prior='L1'. You >can change max iterations using control. See stm documentation

My code:

temp <- textProcessor(documents=sample$policies,
                      metadata=sample,
                      lowercase=FALSE, 
                      removestopwords=FALSE, 
                      removenumbers=FALSE, 
                      removepunctuation=FALSE, 
                      stem=FALSE,
                      sparselevel=.90)
meta  <- temp$meta
vocab <- temp$vocab
docs  <- temp$documents
out   <- prepDocuments(docs, vocab, meta)
docs  <- out$documents
vocab <- out$vocab
length(vocab)
meta  <- out$meta

storage <- manyTopics(docs, 
                      vocab, 
                      K=c(12),
                      prevalence=~Organization+discuss,
                      max.em.its=500, 
                      data=meta, 
                      runs=50)

I only have 135 observations. Could someone please explain what the exact problem is and how could solve it.