Can we save the lda model with old data and use trained model for new data?

271 Views Asked by At

I have 100 sentences in a excel sheet.. I need to generate topics from this sentences. For that im using LDA model.

I have trained the model, fit transform the model and got the optimum number of topics from 100 sentences.

Now wrt optimum number of topics, i have save(dump) the model in a pickle file.

The reason for saving the model is when the new sentences arrived, i dont need to retrain the model again.

Example: First when the train the lda model, i got optimum_topic=5. I have save the model wrt optimum_topic.

Now when i load this model, and try to use the model for new sentence(assume i have 50 new sentences), then i'm unable to do so.. I'm getting below error

IndexError: index 5 is out of bounds for axis 0 with size 5

Is there any other way via which i can train the model with old data and when new data arrives, i can use the old trained model?

0

There are 0 best solutions below