Slow Inference

[luantunez]

Hello and thank you for sharing your work!
I am encountering the problem of the inference time for your model being quite slow in the colab you shared.
Could you explain why? Is there any workarround?
I have seen you recommended ParallelWaveGan on issue #208 but I would really like to use your repository models and results.
Thank you very much in advance!

[luantunez]

Also, what are the npy files in "tacotron_output/eval/", like "tacotron_output/eval/speech-mel-00001.npy" ?
How can I create these to perform new inferences?
Thank you again!

[KopJacob]

Hello and thank you for sharing your work! I am encountering the problem of the inference time for your model being quite slow in the colab you shared. Could you explain why? Is there any workarround? I have seen you recommended ParallelWaveGan on issue #208 but I would really like to use your repository models and results. Thank you very much in advance!

Hello. Could you please tell me roughly how long your inference time is and how much audio you are generating? I have also encountered a similar issue