camoverride / language-classifier Goto Github PK

but describe how this might not be great in production

title

add diagram to show how app should work (independent servers)

• combine requirements file (?)
• missing version #'s

• add more details to notebook
• add info on naive bayes
• justification for train/test split

• spell check
• re-read / final check
• post to reddit (?)

the files in modeling should be moved to a jupyter notebook. Then create some bash scripts to move around the model artifacts

The below code reads from a dataset that we did not create in the tutorial,kindly advise on how to go about this:

TEST_DATA = []
for language in LANGUAGES:
    TEST_DATA.append(
        np.array([open('scraper/data/' + language + '/TEST_SET.db').read()])
    )