The scoring of an Approximate Solution question is automatic which is done based on the custom checker. If you do not write a custom checker, then the approximate solution questions are scored in the same way as coding questions; with the test cases determining the final score. For coding questions, individual test cases consisting of input and output pairs are set up by the question setter. Each test case is assigned a score, and the total possible score for the question is the sum of all the scores of the test cases.
For an approximate solution where you have written a custom checker, the custom checker is run once for each test case. When a candidate runs their code to check their solution during a test (and when their final solution is evaluated after the test completion), their code is successively supplied to each test case input to STDIN. For each test case, we then run the custom checker to get the score for each test case. The candidate's score for an approximate solution question is the sum of the scores returned by the custom checker for each test case.
Refer to Creating a Custom Checker to understand the custom checker.