This helps us convert the predictions into a more interpretable format. Let's read the test dataframe and insert the predictions in the correct columns:

test_df = pd.read_csv("data/test.csv")
for i, col in enumerate(["toxic", "severe_toxic", "obscene", "threat", "insult", "identity_hate"]):
    test_df[col] = test_preds[:, i]

Now, we can preview a few of the rows of the DataFrame:

test_df.head(3)

We get the following output: