chazo1994

I have same problems.

@KevinMIN95 help me please.

> to match with other baselines. I haven't tried with 22050 but I think it may be helpful to increase the capacity of the model. Thanks, I will try. Maybe...

@sanjeevani279 I have same problem with 22050 hz, while 1600hz is ok. Did you resolve this problem ?

> @chazo1994 Are you using 1600Hz and the batchsize is 16? How is the synthesis effect? @Summerxu86 I train two model 22050khz and 16khz, both use batchsize 48. Model 16k...

> @chazo1994 how do we use the pretrained models, I'm a little confused with that? Just use load_checkpoint function in utils.py for both generator and discriminator.

I have some results of magphase vocoder with lossless and lowdim demo. https://drive.google.com/file/d/1LONGI3xb4hYV8fMqdS8CDDNCbdFxbtkY/view?usp=drivesdk The voice quality of lowdim demo still Very bad. I don't know the version which intergrated into...

Pytorch vesion 1.6.0 stable not support autograd for 'bucketize'. So you should change to lastest version, even so I think the output tensor does not require gradients. https://github.com/pytorch/pytorch/issues/45119#issuecomment-696971735

I added cqs question into my question file and it's work.

I have same problems