test_results.json

{
    "eval_samples": 1233,
    "test_false_f1": 0.789407313997478,
    "test_loss": 0.616358757019043,
    "test_macro_f1": 0.6733562800163174,
    "test_micro_f1": 0.7923763179237632,
    "test_mixture_f1": 0.5283950617283951,
    "test_runtime": 37.1249,
    "test_samples_per_second": 33.212,
    "test_steps_per_second": 1.051,
    "test_true_f1": 0.9064869418702612,
    "test_unproven_f1": 0.46913580246913583
}