Chinese-speech-to-text

Chinese-speech-to-text copied to clipboard

I want to learn to use your model, but due to my current computer performance limitations, I am not able to train a model, so ask me if you can send me your trained model.

8

Thanks

Dylan-Dyb

1124146862

- [ ] 请问这个对wav文件的时长有限制嘛？ - [ ]不是中文普通话，翻译效果咋样？ @liangstein

gaowenxin95

when i train the model，it raise an error,please help me!

2

20/10000 [..............................] - ETA: 1:00:01 - loss: 2282.6917 40/10000 [..............................] - ETA: 31:47 - loss: 1886.1443 60/10000 [..............................] - ETA: 22:25 - loss: 1565.0140 80/10000 [..............................] - ETA: 17:43 -...

lzx168

I am a new one to speech machine learning. How to start the your code? I mean from download training data, and how to read train set to train and...

tianke0711

你好,请问您能提供您训练好的模型供我测试您的数据吗?之前您给别人提供的模型链接失效了

zhouhao-learning

No 'mfcc_vec.npy'

3

Could you please add the 'mfcc_vec.npy' in this repo? Thank you~

ShengleiH

您好，我用您一模一样的代码运行，无法收敛，即使跑完124个epochs，loss还是两百多，请问是什么问题呢

pinweihelai

输出总是只有一个字

1

您好，我进行了模型的训练，但是在尝试用1.wav,2.wav进行测试时，输出结果总是'一‘,用每一步epoch训练得到的模型进行测试时发现结果向着只输出一个’一‘收敛，请问这是什么原因呢？ 另外，我理解ctc_loss的标准值应该是零，您的代码当中将model.fit中的y设为1是否会对结果有影响呢？（不管是否改变y的值输出都只有一个字）

wanzzc

This is my test results with 1.wav, which use your upload model. does their any method that can i enhance results ?

7

袅创纵艇婆表底禁兹禁坡跷唐那发同皮盆齐盆坏老桃奎袅疫诱映佻往库提柏库亿异养秧护难赏岗克快库畴肠冈汉二换唠恐卡鱼疫一打二孔琴异语很样环忘矿坡毫爱语滩畅汉网口槐李永伟以同逃它卡网火时场传盆话房旷集框王彭梅逃扰剧阳彭冈屠画语闹末宝毫忘啊唐逃一忘冈逃硬谊弥票意依印莺重奥样乒同很二库重表二氧运逃爬察潮笔倒订口卡已映氧养影同爬投重样事上话笔澡重忘影重逃唐同样意考笔堂重杭二王老让奥忘逃后很烧谈商秧忘盘弥涛样茶察拿毫号逃尧往雅样奥误网了已厂逃拧逃跷爬爬逃重创让号穿痛团同库意席奥伊以以同网爬啊同鱼味布顶迎图疫余基第逃坡涛魄图团他同他同他坡逃脏扭迎偶二逃抵叠姆急弹拿她急怕普它第桃禁普几去集第婴集第集尽第现集朝集二去集朝去二集一集您一去二您二去二去急去集急去集其集一集一二老二婴集二急二婴靠哇把婴婴印厅普

Alvin2580du