after training epoch=150， the generaed wav is only noisy

HappyBall / tacotron

tacotron for research on Chinese speech synthesis and Taiwanese speech synthesis from Chinese input text sequence with different granularities

Apache License 2.0

24 stars 13 forks source link

Open peter374530731 opened 6 years ago

frozen-finger commented 5 years ago

I faced with the same problem. Did you solve

No description provided.

I faced with the same problem. Did u solve it by keeping training?( my model over 23k steps)

yuye-img commented 5 years ago

请问你们的资源在那里下载呢？有其他的下载地址吗？github上面的打不开了，而且下载了之后也不完整。

ccuiyuhan commented 4 years ago

你好，中文数据集你下载成功了吗？可以分享一下吗937903992@qq.com