[Python] Learn words' sentiment based on labels

[Artificious]

I have a large database with the following setup for feature and label:

list_of_list = [[["Many", "words", "forming", "sentences"], "POSITIVE"], ["Some, "are", "bad"], "NEGATIVE"]]

I want to build a model to that learns the sentiment of the words, given the labels. Can someone point me in the right direction?

First time poster, please tell me if I made a mistake in the process of posting here.

[Larz60+]

you may want to investigate the NLTK sentiment package here: https://www.nltk.org/api/nltk.sentiment.html