Chung-Ming Chien
Chung-Ming Chien
@rspiewak47 It is for the case you want to restart the training from an available checkpoint.
@xDuck @KinamSalad I think the code should be modified to enable the use of ``torch.jit``. It's in my future plan for the next major update.
@xDuck Yeah I think the length regulator may be a major problem of scripting the whole model. Looking forward to your result!
@xDuck Great job!!!! Thanks for your work! I will try it several days later!
@xukai98 No I just train the model from scratch. It's just a simple demo.
@xukai98 Yes, they are trained together. There is a speaker embedding table containing the speaker representations of all the speakers. And this embedding table, is end-to-end trained, together with the...
@SamuelLarkin It seems that the length of your pitch sequence is longer than the length of the phoneme sequence. Did you set ``preprocessing.pitch.feature`` in ``preprocess.yaml`` to ``"frame-level"`` while preprocessing the...
@cuongnguyengit Are the MFA boundaries accurate? And how do you think about the results, for example, do you think the pitch or prosody of the synthesized samples is strange, or...
@cuongnguyengit I am just guessing that maybe you forget to normalize the pitch and energy features so the pitch and energy losses are so large. Just turn on ``preprocessing.pitch.normalization`` and...
@EuphoriaCelestial Could you please print out the values of these tensors, or give more information?