How to train for new speaker and minimum required dataset required? about dc_tts

[iampartho]

Did you resolve the issue , can you teach me how you resolved it? Thanks.

[vinamramunot-tech]

@navinthenapster @iampartho Can you provide steps on how to create your own custom dataset and then how to train that dataset? Not sure how to run the training as not sure about the format of the dataset.

[iampartho]

Well it has been pretty long since I did the project so my memory is not that sharp about what stepped I followed for my project. But here it goes:

For the creation of dataset I think I used an app named lyrebird ai , it basically records like some minutes of voice data (which sort of use all sort of pronunciation and usually used words) , so I created my dataset using that app. I was not good and comfortable in tensorflow and codes were not well documented here so I shifted from here and used this project for my work as it was well documented and easy to implement. But honestly my results were not great but anyway I hope this information helps you.

[vinamramunot-tech]

@iampartho thank you.