Training 7000 documents is taking more than 5 hours

[fabriziox1234]

As stated, I'm making an instance of the LDA model using multiple cores yet it still taking over 5 hours for the code to finish. As seen in the code, I'm trying to gather the coherence score for 5 different k. Any suggestions? I feel like this is awfully slow for what I'm trying to do

coherence = []

for k in range(5, 10):
print('Round: '+str(k))
lda_model = gensim.models.LdaMulticore(corpus=corpus,
id2word=id2word,
num_topics=k,
workers=3,
passes = 2,
iterations = 2,
chunksize = 1000,
eval_every=None
)

cm = gensim.models.coherencemodel.CoherenceModel(model=lda_model, texts=corpus,
                                               dictionary=id2word, coherence='c_v')
coherence.append((k,cm.get_coherence()))

[shadeMe]

This appears to be a specific usage question about the gensim library. Please use their support forum for such questions.