Closed zkokaja closed 2 years ago
we should add the rank of true word as part of process_extracted_logits functionality so we can easily classify correct/incorrect words based on arbitrary k.
k
duplicate https://github.com/hassonlab/247-pickling/issues/56
we should add the rank of true word as part of process_extracted_logits functionality so we can easily classify correct/incorrect words based on arbitrary
k
.