Results of training on my own dataset

Question

Results of training on my own dataset

dongdongdashen opened this issue 3 years ago · comments

Hi,When I train with my own dataset, the result is as shown in the following picture.Can you tell me where is the problem? May be the length of data is too short? (my dataset is about 20 minutes ) Looking forward to your reply ：

）

Domhnall-Liopa · Answer 1 · Wed Mar 02 2022 18:28:18 GMT+0800 (China Standard Time)

Hi, is it possible you've got teacher forcing enabled? Going by the attention matrix, the encoder doesn't seem to have an impact on the decoder

dongdong · Answer 2 · Wed Mar 02 2022 19:51:08 GMT+0800 (China Standard Time)

Hi, is it possible you've got teacher forcing enabled? Going by the attention matrix, the encoder doesn't seem to have an impact on the decoder

Hi,I check the hparams.py and train.py in synthesizer folder,how can I stop teacher forcing when training? is it ok?

Domhnall-Liopa · Answer 3 · Wed Mar 02 2022 19:59:13 GMT+0800 (China Standard Time)

I think if you use the following in hparams.py:

tacotron_teacher_forcing_mode="constant",
tacotron_teacher_forcing_ratio=0.,

It will disable teacher forcing in training

dongdong · Answer 4 · Wed Mar 02 2022 20:02:52 GMT+0800 (China Standard Time)

I think if you use the following in hparams.py:
tacotron_teacher_forcing_mode="constant",
tacotron_teacher_forcing_ratio=0.,
It will disable teacher forcing in training

thanks!

comvee · Answer 5 · Thu Mar 10 2022 14:13:44 GMT+0800 (China Standard Time)

Could you share the attention plot after disabling the teacher-forcing? Did you solve the problem?