For each `customer_id` in the test dataset, you must predict the target label for `churn`. The submission file should contain only 2 columns - `customer_id` and `churn` as shown below:

| customer_id| churn |
| 165034 | 0 |
| 165035 | 1  |