primeqa.mrc.metrics.nq_f1.nq_eval.compute_final_f1#

primeqa.mrc.metrics.nq_f1.nq_eval.compute_final_f1(long_answer_stats, short_answer_stats)#

Computes overall F1 given long and short answers, ignoring scores.

Note: this assumes that the answers have been thresholded.

Parameters

  • long_answer_stats – List of long answer scores.

  • short_answer_stats – List of short answer scores.

Returns

Dictionary of name (string) -> score.