Datasets:
| from jiwer import wer | |
| def perform_computation(num_sentences): | |
| truth = ["this is a speed test" for _ in range(0, num_sentences)] | |
| hypo = ["this is not a speed test" for _ in range(0, num_sentences)] | |
| wer(truth, hypo) | |
| def test_speed_n1(benchmark): | |
| benchmark(perform_computation, 1) | |
| def test_speed_n10(benchmark): | |
| benchmark(perform_computation, 10) | |
| def test_speed_n100(benchmark): | |
| benchmark(perform_computation, 100) | |
| def test_speed_n1000(benchmark): | |
| benchmark(perform_computation, 1000) | |