search

TensorSpeech / TensorFlowTTS

:stuck_out_tongue_closed_eyes: TensorFlowTTS: Real-Time State-of-the-art Speech Synthesis for Tensorflow 2 (supported including English, French, Korean, Chinese, German and Easy to adapt for other languages)
https://tensorspeech.github.io/TensorFlowTTS/
Apache License 2.0
3.8k stars 810 forks source link

NaN output while infering/decoding #717

Closed supreet21 closed 2 years ago

supreet21 commented 2 years ago

I trained a Tactoron2 model using custom data, while the training is going fine as the losses are converging also the predictions from validation set while training also shows, the mel spectograms. but when I use the same model in the decoding script the mels in .npy files have NaN value. Can someone explain why this could be the case?

dathudeptrai commented 2 years ago

Can you check the inference and decoding on validation data as well ?

stale[bot] commented 2 years ago

This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs.