issues
search
TencentGameMate
/
chinese_speech_pretrain
chinese speech pretrained models
965
stars
83
forks
source link
issues
Newest
Newest
Most commented
Recently updated
Oldest
Least commented
Least recently updated
如何将预训练的权重转换成huggingface格式？
#54
CodeMrSheep
opened
2 months ago
1
采样率是多少啊？
#53
sunjian2015
opened
2 months ago
0
音频fps如何调整为25
#52
tailangjun
opened
3 months ago
1
哪个模型最好？
#51
piwawa
opened
3 months ago
0
关于该项目的bibtex格式引用
#50
mixxs
opened
4 months ago
1
如何提取音频特征
#49
tailangjun
opened
4 months ago
1
Error
#48
ChengsongLu
closed
5 months ago
3
如何获得1024维特征的离散id
#47
wcr369
opened
5 months ago
4
请问我们在espnet/egs2/aishell/asr1/下使用，报TypeError: wav2vec2_custom() missing 1 required positional argument: 'ckpt'错误，怎么解决，非常感谢！！！
#46
MELABIPCAS
opened
6 months ago
1
fairseq和huggingface输出结果不同
#45
hao-qiang
opened
6 months ago
1
.
#44
Bingtai1015
opened
6 months ago
0
可以提取采样率为22050的音频的特征吗？
#43
Bingtai1015
opened
6 months ago
2
Add WavLM
#42
Blakey-Gavin
closed
10 months ago
1
k-means参数的读取
#41
jidanhuang
opened
11 months ago
0
请问该预训练模型们的语音的采样率是多少呢？
#40
ywh-my
opened
11 months ago
1
用CTC直接微调效果非常差
#39
zyh3826
opened
1 year ago
4
这个可以用于speaker-diarization任务吗
#38
luomingjun2023
closed
1 year ago
1
能否使用预训练模型同时更改参数？
#37
LwLiu-2012
opened
1 year ago
1
可以同时提取中英文语音的特征吗
#36
milely
closed
1 year ago
1
你好请问large的特征聚类的时候使用了百分之多少的特征?10%的话需要内存多大的机器？
#35
manmushanhe
opened
1 year ago
0
如何获得最后的unit?
#34
mikesun4096
opened
1 year ago
0
hubert特征，用的是哪层的特征啊，还是哪些层的特征进行了加权和？比例是多少
#33
yangsuxia
closed
1 year ago
0
开源出来的hubert large 模型，有对应的kmean模型么？还是base和large使用同一个kmeans就可以？
#32
joan126
opened
1 year ago
2
求一个能够输出最终文字的代码案例
#31
moresun
opened
1 year ago
1
Problem about time shape
#30
huutuongtu
closed
1 year ago
0
请问hubert模型训练时的batch_size大小是多少
#29
dancinghui
opened
1 year ago
0
请问如何使用huggingface代码finetune
#28
Yonnie1331
opened
1 year ago
1
请问最长能处理多长的语音？
#27
ddlBoJack
closed
1 year ago
0
最终输出是768维还是1024维呢？
#26
ZiqiaoPeng
opened
1 year ago
5
请问如何用 fairseq 训练 wenetspeech
#25
panpan-wu
opened
1 year ago
1
How many days did the pre-training phase take on large model?
#24
Qoboty
opened
1 year ago
0
采用预训练模型提取语音特征，怎么处理长语音，直接切割或滑窗处理？
#23
Owen1234560
opened
1 year ago
2
Fine-tune with my own dataset, wer is 1
#22
abcdbosh
opened
1 year ago
0
您好，改怎么进行微调呢？
#21
SinLT
closed
1 year ago
0
你好，有WavLM的中文预训练模型吗？
#20
dengcunqin
opened
1 year ago
0
请问预训练好模型之后提取音频特征时加权求和的具体做法是什么？
#19
zdaaaaa
opened
1 year ago
2
关于模型中没有task_cfg、model_cfg、model_weight、dictionaries_symbols这一问题，求大佬解答
#18
646312715
opened
1 year ago
11
能期待下vq-wav2vec的自监督backbone吗？
#17
splinter21
closed
1 year ago
0
预训练超参mask_prob设置
#16
212wzt5A
opened
1 year ago
0
如何测试？
#15
qiuyuzhao
opened
1 year ago
2
请问wenet speech中用于训练的100小时数据选取有技巧吗？还是任意选取都可以？
#14
user-ZJ
closed
1 year ago
1
模型小型化
#13
xuwenshen
closed
1 year ago
1
可以用作特征的是哪个字段里面的值
#12
kejom-ou
closed
2 years ago
3
ASR finetune收敛速度问题
#11
qinyuenlp
closed
1 year ago
15
Failed to load pretrained model from huggingface
#10
teinhonglo
closed
2 years ago
12
与原始版本预训练模型对比
#9
zhangxueyangjuxie
closed
2 years ago
4
请问还传ESPnet的训练代码吗？
#8
qixing-ai
closed
1 year ago
18
HuBERT模型对应的kmeans模型
#7
ziyichen-paii
closed
2 years ago
3
About fairseq checkpoint link
#6
godiclee
closed
2 years ago
3
有没有更详细的教程
#5
hello2013
closed
2 years ago
3
Next