k2-fsa/icefall

Issues

• Nan outputs from encoder

  #1731 opened a month ago by Manjunath-mlp
  12

• zipformer and SSL feature adaptation

  #1799 opened a month ago by alanshaoTT
  8

• Whisper微调是否支持不定长度的模型解码?

  #1836 opened a month ago by teamtee
  1

• 'scaling.py' throws torch error: 'function definitions aren't supported' when executed by 'train.py' and other scripts

  #1834 opened a month ago by ChrystianKacki
  7

• RuntimeError: cuDNN error: CUDNN_STATUS_INTERNAL_ERROR

  #1833 opened a month ago by lightandblue
  4

• Zipformer missing bias parameter

  #1832 opened a month ago by fmac2000
  3

• How to configure zipformer-large model parameters

  #1831 opened a month ago by mayiachen
  1

• 请问目前有在k2框架训练ngram lm的脚本或者文档嘛

  #1824 opened a month ago by alanshaoTT
  3

• Getting Nans while training

  #1806 opened 2 months ago by Manjunath-mlp
  1

• multi gpus for decoding

  #1813 opened a month ago by Manjunath-mlp
  1

• VITS : stack expects each tensor to be equal size, but got [445852] at entry 0 and [445851] at entry 17

  #1826 opened a month ago by possum-code
  1

• 在从kaldi脚本生成cut的时候遇到json问题

  #1811 opened a month ago by 728826058
  1

• checkpoint.py:113: FutureWarning: You are using `torch.load` with `weights_only=False`

  #1822 opened a month ago by mablue
  1

• export 'icefall-asr-multi-zh-hans-zipformer-large' to onnx successfully but deploy failed

  #1823 opened a month ago by goddamnVincent
  5

• [Bug] [ljspeech/TTS/./vits/train.py:387] UnicodeDecodeError: 'ascii' codec can't decode byte 0xc9 in position 1236: ordinal not in range(128)

  #1817 opened a month ago by mablue
  0

• Zipformer Training Issues:Gradient too small and cuda out of memory issue

  #1751 opened 2 months ago by bsshruthi22
  15

• I used to compile it in windows but still an error presenting

  #1809 opened 2 months ago by mablue
  3

• RuntimeError: Too many grads were not finite

  #1810 opened 2 months ago by SSwethaSel0609
  0

• Fine-grained Text Normalization for Libriheavy

  #1803 opened 2 months ago by yfyeung
  0

• An error occurs in Stage 5 (lhotse) in the audio tagging egs

  #1804 opened 2 months ago by humblemh
  0

• Dimension mismatch Error

  #1801 opened 2 months ago by SSwethaSel0609
  2

• How to do data preparation for a custom dataset for training a streaming zipformer transducer model?

  #1807 opened 2 months ago by mukherjeesougata
  0

• multi GPU training issue

  #1798 opened 2 months ago by sanjuktasr
  10

• Prepare for PromptASR model training

  #1796 opened 2 months ago by ye0njinkim
  0

• location of CR-CTC code

  #1795 opened 2 months ago by Zilai-WANG
  1

• devision by zero

  #1737 opened 2 months ago by Airgods
  2

• 设置了start_bacth似乎是无效的

  #1783 opened 3 months ago by alanshaoTT
  2

• Lost end of audio with jit_pretrained_streaming.py

  #1779 opened 3 months ago by kfmn
  2

• Mismatch of pretrained weights size of SURT model downloaded from huggingface

  #1784 opened 2 months ago by nightwatcher-of-abyss
  15

• Run onnx_check.py RuntimeError: PytorchStreamReader failed locating file constants.pkl: file not found

  #1760 opened 2 months ago by kellkwang
  2

• Decoding through .pt model using Audio id

  #1765 opened 2 months ago by bhaswa
  2

• How to train zipformer multi-language ASR model,such as include English, German, French

  #1761 opened 3 months ago by TszSimLaw
  1

• NANs in CTC loss: solution

  #1777 opened 3 months ago by kfmn
  2

• Issue with LODR for RNN Transducer.

  #1771 opened 3 months ago by sangeet2020
  1

• Illegal memory access during zipformer training

  #1764 opened 3 months ago by ngoel17
  26

• Unused words.txt Generated in prepare.sh for GigaSpeech Recipe

  #1756 opened 4 months ago by yfyeung
  0

• Adaptation using K2

  #1762 opened 3 months ago by AlexandderGorodetski
  1

• Kmeans in SSL

  #1759 opened 3 months ago by SSwethaSel0609
  1

• Multi GPU Training

  #1758 opened 4 months ago by bhaswa
  1

• Assertion error,batch size mismatch in Training zipformer

  #1733 opened 4 months ago by bsshruthi22
  14

• data/manifests/musan_cuts.jsonl.gz can not found when running alimeeting train script ,plz check script issue

  #1742 opened 4 months ago by duhumail
  1

• bug report : alimeeting prepare script: prepare_alimeeting_gss.sh some bug found. plz fix

  #1738 opened 4 months ago by duhumail
  1

• bug report : alimeeting prepare script: code comment not correct on stage 2 Prepare musan manifest

  #1741 opened 4 months ago by duhumail
  1

• multi-zh-hans/ASR/zipformer的训练失败

  #1740 opened 4 months ago by 1716775457damn
  1

• modified_beam_search has higher WER than greedy_search

  #1739 opened 4 months ago by Ryuk17
  0

• pad_length in streaming decode

  #1736 opened 4 months ago by 1215thebqtic
  0

• No duration or confidence info for alignments

  #1735 opened 4 months ago by AhmedSalah98
  2

• execute prepare.sh and encountered an error in gigaspeech

  #1728 opened 4 months ago by lmee
  2

• LLM post processing

  #1729 opened 5 months ago by AlexandderGorodetski
  0

• [Transducer Loss] Why not normalize transducer loss

  #1726 opened 5 months ago by ncakhoa
  0