Evaluates the functional correctness of generated samples, and writes results to f"{sample_file}_results.jsonl.gz"
(
sample_file: str,
k: str = "1,10,100",
n_workers: int = 4,
timeout: float = 3.0,
)
| 75 | return model |
| 76 | |
| 77 | def entry_point( |
| 78 | sample_file: str, |
| 79 | k: str = "1,10,100", |
| 80 | n_workers: int = 4, |
| 81 | timeout: float = 3.0, |
| 82 | ): |
| 83 | """ |
| 84 | Evaluates the functional correctness of generated samples, and writes |
| 85 | results to f"{sample_file}_results.jsonl.gz" |
| 86 | """ |
| 87 | k = list(map(int, k.split(","))) |
| 88 | results = evaluate_functional_correctness( |
| 89 | sample_file, k, n_workers, timeout |
| 90 | ) |
| 91 | |
| 92 | return results |
| 93 | |
| 94 | def filter_code(completion: str) -> str: |
| 95 | # The program tends to overwrite, we only take the first function |
no test coverage detected