rob-base-superqa1 / eval_results.json
nbroad's picture
nbroad HF staff
results
ee7653e
{
"epoch": 3.0,
"eval_HasAns_exact": 69.65206535309125,
"eval_HasAns_f1": 76.1771308342428,
"eval_HasAns_total": 18423,
"eval_NoAns_exact": 77.6604520706154,
"eval_NoAns_f1": 77.6604520706154,
"eval_NoAns_total": 6061,
"eval_best_exact": 71.63453684038556,
"eval_best_exact_thresh": 0.0,
"eval_best_f1": 76.5443261460238,
"eval_best_f1_thresh": 0.0,
"eval_exact": 71.63453684038556,
"eval_f1": 76.54432614602389,
"eval_samples": 44895,
"eval_total": 24484
}