hub / github.com/Xflick/EEND_PyTorch / calc_diarization_error

Function calc_diarization_error

eend/pytorch_backend/loss.py:70–102 · view source on GitHub ↗

Calculates diarization error stats for reporting. Args: pred (torch.FloatTensor): (T,C)-shaped pre-activation values label (torch.FloatTensor): (T,C)-shaped labels in {0,1} label_delay: if label_delay == 5: pred: 0 1 2 3 4 | 5 6 ... 99 100 | label: x

(pred, label, label_delay=0)

Source from the content-addressed store, hash-verified

source not stored for this graph (policy: none)

Callers 1

report_diarization_errorFunction · 0.85

Calls

no outgoing calls

Tested by

no test coverage detected