yeyupiaoling/MASR

Issues

• masr界面识别显存不释放

  #75 opened 5 months ago by upupbo
  2

• 这个是不是不能用于cuda为11.4的

  #74 opened 4 months ago by king-hnu
  4

• windows系统多卡训练失败

  #76 opened 5 months ago by yjyz1011
  2

• Aishell-1测试集问题

  #73 opened 6 months ago by CSLJingyu
  1

• infer出现问题

  #72 opened 6 months ago by iakRulan
  1

• train time

  #71 opened 7 months ago by iakRulan
  3

• 下载了squeezeformer的best model之后导出报错

  #70 opened 10 months ago by DannyWang920
  1

• 运行train.py报错

  #69 opened 10 months ago by happywch
  1

• 运行train.py时报错

  #66 opened a year ago by Vickywhy
  4

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

  #62 opened 2 years ago by dianxin556
  2

• 在train.py运行开始时发生如下报错

  #65 opened a year ago by Isaiah-pq
  5

• RuntimeError: `lengths` array must be sorted in decreasing order when `enforce_sorted` is True. You can pass `enforce_sorted=False` to pack_padded_sequence and/or pack_sequence to sidestep this requirement if you do not need ONNX exportability

  #64 opened 2 years ago by TszSimLaw
  3

• Is it possible to build a WeChat group for better communication?

  #63 opened 2 years ago by Alex-Songs
  1

• 在数据库WenetSpeech上预训练的模型conformere免费下载

  #60 opened 2 years ago by eeric
  1

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

  #57 opened 2 years ago by eeric
  4

• 缺失vocabulary文件

  #61 opened 2 years ago by DrewdropLife
  0

• No such file or directory: 'dataset/manifest.test'

  #59 opened 2 years ago by eeric
  2

• AIShell (179小时) 的预训练模型，是哪个版本torch训练的

  #58 opened 2 years ago by eeric
  3

• online 和offline自己炼的话 相同条件下是不是offline效果好点？

  #55 opened 2 years ago by 2651084156
  67

• no module named "torch.inference"

  #56 opened 2 years ago by hxhcreate
  1

• 运行tune.py报错，提示没有mean_std.npz文件

  #54 opened 2 years ago by jiangduwang
  2

• 训练loss不收敛

  #25 opened 3 years ago by 827379852
  12

• 模型训练GPU利用率

  #30 opened 3 years ago by jcl-gx
  4

• 实时语音识别，报hidden层size不匹配

  #42 opened 2 years ago by moneypi
  7

• 数据准备部分怎么生成数据列表文件，存在dataset/annotation/目录下

  #53 opened 2 years ago by jiangduwang
  2

• 语音识别的拼音输出

  #50 opened 2 years ago by tgarm
  8

• 选择音频处理方式前向计算维度错误

  #49 opened 2 years ago by jackjieliu
  1

• 没有python_speech_features包

  #48 opened 2 years ago by jackjieliu
  1

• About dataset

  #47 opened 2 years ago by qzhsdu
  2

• 数据集标注信息

  #41 opened 2 years ago by leishenzhupi
  2

• 默认类型应该改为float

  #44 opened 2 years ago by youshengithub
  1

• 是spectrogram还是mfcc特征呢？

  #40 opened 2 years ago by lywhz
  7

• 效果好像很差？

  #39 opened 2 years ago by hroken
  1

• 缺少文件

  #38 opened 2 years ago by journeyzx
  2

• 使用预训练数聚集进行finetune 大概需要多少数据呢，训练多少轮次可以不降低本身精度呢

  #37 opened 3 years ago by 827379852
  1

• predict.py中import paddle应该注释掉，否则会报错

  #35 opened 3 years ago by leishenzhupi
  2

• 直接录制语音和上传语音文件，识别效果的差异

  #36 opened 3 years ago by bird7code
  6

• 用自己的数据集报错，格式检查没问题

  #34 opened 3 years ago by XWT999
  1

• 加入自己的数据训练，预测同样的数据得分却不高

  #31 opened 3 years ago by bird7code
  5

• WenetSpeech数据集的训练参数

  #33 opened 3 years ago by bird7code
  4

• 流式语音识别的功能貌似没有看到，请问现今实现了么？

  #29 opened 3 years ago by RabbitBoss
  5

• 能否支持在线识别，边说便识别

  #26 opened 3 years ago by swpucwf
  1

• WenetSpeech数据集下载出现404错误

  #32 opened 3 years ago by bird7code
  0

• 模型训练问题

  #28 opened 3 years ago by Chenwe111
  10

• 预训练模型加载错误

  #27 opened 3 years ago by Crescentz
  4

• Android预计什么时候支持

  #24 opened 3 years ago by wy676579037
  1

• infer_path.py 消耗时间：3432ms 是不是数据集的关系

  #23 opened 3 years ago by cgisky
  2

• 大佬，流式的是怎么个思路啊

  #22 opened 3 years ago by wyt1234
  1

• 大佬，cpu能跑嘛

  #20 opened 3 years ago by hy1079503225
  5

• 大佬，cpu能跑嘛

  #21 opened 3 years ago by hy1079503225
  1