Multilabel Classification using random forest

Question

I have 5 labels and 499 datasets.

I tried using random forest classier. I have two inputs for the model: output of tfidf and review_length.

train_tfIdf