Tele-AI/TeleSpeech-ASR

Issues

• 请问预训练数据都包含哪些方言

  #49 opened 3 months ago by aoji0606
  0

• 在微调阶段，fairseq格式的 dict.${label}.txt是在微调的哪一步骤用到

  #47 opened 3 months ago by uncleunclelu
  1

• 关于微调参数设置

  #48 opened 3 months ago by GaoZhipengCN
  0

• 请问后续会提供base模型在8种方言的finetune结果吗？

  #41 opened 6 months ago by mikeswf
  4

• 识别的时候能输出置信度吗

  #46 opened 4 months ago by kuikui111222
  0

• 该预训练模型能否支持 类似 语音质量判别的下游任务

  #45 opened 4 months ago by xuetaolue
  0

• 请问推理的转录结果只有result中的infer.log吗，有没有utt和转录内容一一对应的的文件

  #44 opened 4 months ago by xingjiahao
  1

• 可以提供transformer直接使用finetune model的教程吗

  #43 opened 5 months ago by CuiMingyu
  1

• decode.sh运行报错

  #3 opened 7 months ago by lq0104
  32

• 如何生成wav.scp文件呢？

  #42 opened 5 months ago by ShengHee
  0

• 执行prepare_kaldi_feats.sh后，如何生成 .tsv 结尾的文件？

  #27 opened 7 months ago
  11

• 提供一个不依赖fairseq和kaldi的runtime

  #19 opened 7 months ago by lovemefan
  18

• 模型

  #40 opened 6 months ago by mikeswf
  2

• 有商用 API 可以调用吗？

  #36 opened 6 months ago by hlthu
  1

• 关于fairseq和wenet中模型结构的区别

  #38 opened 6 months ago by llh666521
  4

• 请问现在公布的aishell1的结果是在fairseq还是wenet上微调的？

  #39 opened 6 months ago by llh666521
  2

• 训练参数设置问题

  #37 opened 6 months ago by rainskyfyy
  5

• 关于8k流式模型

  #34 opened 6 months ago by srdfjy
  2

• 方言识别质量非常差

  #33 opened 6 months ago by dfengpo
  4

• 训练资源要求

  #35 opened 6 months ago by rainskyfyy
  4

• 您好，请问可以具体列一下目前支持的30种方言吗，后续会开源少量的数据

  #20 opened 7 months ago by csf123123
  1

• 粤语或者闽南语直接可用 asr 模型

  #26 opened 7 months ago by pengzhendong
  3

• 希望能支持方言检测

  #32 opened 7 months ago by dfengpo
  1

• 预训练代码

  #29 opened 7 months ago by wowerqq
  1

• decode.sh运行报错

  #30 opened 7 months ago by GaoZhipengCN
  0

• 如何不使用CUDA运行

  #31 opened 7 months ago by GaoZhipengCN
  2

• 关于finetune 速度

  #28 opened 7 months ago by wowerqq
  1

• Mark: 技术报告啥时候出呀？

  #23 opened 7 months ago by MingLunHan
  0

• Great work, so when will 1B speech model be released?

  #2 opened 7 months ago by marvin-nj
  1

• 数据集问题

  #22 opened 7 months ago by llleohk
  4

• 如何获取tsv文件

  #24 opened 7 months ago by XiaoYu2022
  1

• 微调版的方言ASR模型计划开源吗？

  #6 opened 7 months ago by lovemefan
  5

• 特征为啥用mfcc不用fbank呢？

  #25 opened 7 months ago by brainbpe
  0

• 请问后续会提供小模型以方便在手机等终端上直接使用嘛

  #14 opened 7 months ago by hasayakey
  2

• 推理数据准备

  #16 opened 7 months ago by ShiyuRaymond
  5

• 执行decode.sh报错

  #21 opened 7 months ago by PeaceAndJoyAaron
  2

• mfcc特征处理

  #18 opened 7 months ago by oslijunw
  5

• 采样频率

  #17 opened 7 months ago by liugz18
  3

• 是否可以借鉴一下funasr，提供docker部署的使用方式

  #13 opened 7 months ago by Hunter6324
  1

• 可以考虑提供一个非huggingface的模型下载路径，不然不好体验效果

  #15 opened 7 months ago by locasxe
  1

• 请问支持哪三十种语言?包括维吾尔语和蒙古语吗？

  #1 opened 8 months ago by cj-wade
  5

• 推理脚本

  #8 opened 7 months ago by zui-jiang
  5

• 怎么用呢？

  #9 opened 7 months ago by luck1020
  2

• 请问finetune_large_kespeech.pt这模型怎么才能跑起来

  #11 opened 7 months ago by fanxun271
  1

• cannot import name 'checkpoint_utils' from 'fairseq'

  #10 opened 7 months ago by Min808
  1

• infer.py脚本有问题

  #12 opened 7 months ago by lostsollar
  1

• 模型可以直接检测方言类型吗？

  #5 opened 7 months ago by pengzhendong
  2

• 数据开源吗？

  #4 opened 7 months ago by Mddct
  1

• 技术报告有计划公开吗

  #7 opened 7 months ago by uloveqian2021
  1