How to train better？

Open 1zxLi opened this issue 5 years ago • 3 comments

I did not modify any parameters. After 400K epochs, the loss oscillated around 5.3 and could not continue to converge. Is there a more suitable parameter setting scheme to continue to optimize the loss? （ljspeech data）

Jun 10 '20 02:06 1zxLi

I met the same problem...

Sep 25 '20 15:09 xuexidi

I met the same problem too, I do not modify anything, but the loss is around 5.3 after 600k steps. The samples it generated is terrible under this situtation.

Dec 30 '20 10:12 harryfyodor

I did not modify any parameters. After 400K epochs, the loss oscillated around 5.3 and could not continue to converge. Is there a more suitable parameter setting scheme to continue to optimize the loss? （ljspeech data）

Hi, i met the same problem too, have you solved it?

Dec 30 '20 10:12 harryfyodor