Issues

• wer_summary reports WER=0.0 when every scored reference is empty, discarding the accumulated edit count

  speechbrain/speechbrain#3090 · Filter by author shaurya416 opened on Sep 30, 2026

• Verify evals on Papers with Code

  speechbrain/speechbrain#3076 · Filter by author NielsRogge opened on Aug 28, 2026

• FutureWarning: torch.nn.utils.weight_norm is deprecated in favor of torch.nn.utils.parametrizations.weight_norm.

  bug
  speechbrain/speechbrain#3074 · Filter by author Decodetalkers opened on Aug 26, 2026

• Training efficiency regression since SpeechBrain 1.1.0

  bug
  speechbrain/speechbrain#3071 · Filter by author Craya opened on Jul 29, 2026
  4

• make_pe() in class RelPosEncXL

  bug
  speechbrain/speechbrain#3070 · Filter by author coollip opened on Jul 27, 2026
  1

• Feature request: Add FunASR/SenseVoice model integration

  speechbrain/speechbrain#3058 · Filter by author LauraGPT opened on May 31, 2026
  1

• HyperConformer and Conformer Achieve Higher WER on test-other When Trained on 100h LibriSpeech

  bug
  speechbrain/speechbrain#3041 · Filter by author zzm196 opened on Mar 13, 2026
  4

• I can not download RIR.zip and noises.zip for voxceleb train speaker rec

  bug
  speechbrain/speechbrain#3014 · Filter by author haha010508 opened on Dec 11, 2025

• AttributeError: module 'torchaudio' has no attribute 'list_audio_backends'

  bug
  speechbrain/speechbrain#3012 · Filter by author matrn opened on Dec 3, 2025
  11

• LazyModule not playing well with %autoreload

  bug
  speechbrain/speechbrain#2995 · Filter by author egaznep opened on Nov 12, 2025
  2

• Performance dismatch on xvector

  bug
  speechbrain/speechbrain#2979 · Filter by author wxlsummer opened on Oct 8, 2025

• VAD fails with cuDNN error when using GPU

  bug
  speechbrain/speechbrain#2970 · Filter by author peterATIn2Dialog opened on Sep 5, 2025

• ASR training from scratch

  bug
  speechbrain/speechbrain#2968 · Filter by author nata-kostina opened on Aug 22, 2025
  2

• CUFFT_INTERNAL_ERROR

  bug
  speechbrain/speechbrain#2956 · Filter by author programindz opened on Jul 25, 2025

• fetching double path from classifier

  bug
  speechbrain/speechbrain#2941 · Filter by author ayan013 opened on Jul 2, 2025

• 训练两种语言

  bug
  speechbrain/speechbrain#2936 · Filter by author fan732 opened on Jun 12, 2025

• 没有办法调整阈值

  bug
  speechbrain/speechbrain#2917 · Filter by author SMK1311 opened on May 15, 2025

• Brain hparams can only be accessed with dot notation, not by name

  bug
  speechbrain/speechbrain#2898 · Filter by author tomaz-suller opened on May 8, 2025
  3

• sb.utils.quirks.apply_quirks() causes jit problems when interacting with other libraries with cuda

  bug
  speechbrain/speechbrain#2896 · Filter by author JamesDeAntonis opened on May 5, 2025
  1

• SepformerSeparation can not load audio.

  bug
  speechbrain/speechbrain#2891 · Filter by author imxtx opened on May 1, 2025

• Recipe CommonVoice/ASR/transformer fails with error 'WhisperTokenizerFast' object has no attribute 'normalize'

  bug
  speechbrain/speechbrain#2884 · Filter by author Craya opened on Apr 11, 2025
  2

• Add Beam Search tutorial

  enhancement
  speechbrain/speechbrain#2879 · Filter by author Adel-Moumen opened on Apr 9, 2025
  

• Fluent Speech Commands WER wrong since eos is counted as token

  bug
  speechbrain/speechbrain#2863 · Filter by author tomaz-suller opened on Mar 14, 2025
  2

  

• Weights only load failed. In PyTorch 2.6, we changed the default value of the weights_only argument in torch.load from False to True.

  bug
  speechbrain/speechbrain#2856 · Filter by author lixl4433 opened on Mar 10, 2025
  1

• Add basic FSDP support for sharded training

  enhancement
  Feature

  speechbrain/speechbrain#2848 · Filter by author pplantinga opened on Mar 4, 2025
  3