Issues

• md_eval_22 / dscore drop reference sessions that are missing from the hypothesis before running md-eval-22.pl, so the corpus DER omits them (md-eval-22.pl itself scores them as 100% missed)

  fgnt/meeteval#132 · Filter by author shaurya416 opened on Sep 30, 2026

• error when use seglst cal wer

  fgnt/meeteval#129 · Filter by author Wang-Zeyu opened on May 21, 2026
  3

• [enhancement] multi-talker Word Error Rate (mtWER) as a metric?

  fgnt/meeteval#127 · Filter by author namgyaaal opened on Apr 8, 2026
  2

• Add optional character-level metrics (CER / cpCER / tcpCER) for languages with ambiguous word segmentation

  fgnt/meeteval#126 · Filter by author hwanython opened on Mar 16, 2026
  2

• Alignment from meeteval.wer.wer.time_constrained.align

  fgnt/meeteval#113 · Filter by author m-wiesner opened on Jun 13, 2025
  3

• Does mdeval computes DER taking overlapped speech into account?

  fgnt/meeteval#62 · Filter by author AntoineBlanot opened on Feb 6, 2024
  9

• Computing time-constrained WER

  fgnt/meeteval#17 · Filter by author desh2608 opened on May 25, 2023
  7

• ImportError: ...cy_levenshtein.cpython-38-x86_64-linux-gnu.so: undefined symbol: _ZSt28__throw_bad_array_new_lengthv

  fgnt/meeteval#15 · Filter by author boeddeker opened on May 15, 2023
  1

• Getting ORC-WER breakdown

  fgnt/meeteval#6 · Filter by author desh2608 opened on Feb 28, 2023
  4