Development Plan

[ftshijt]

Refine current code

• remove abundant files

• modulization (e.g., mel layer, mfcc, fbank,, vocoder (world))

• normalization (norm, collect stats)

• make "egs", make separate data running scripts (\w different configuration and data_prep process)

• refine README.md

• autoregressive module

@ftshijt Hi, great work can you please add example audio and a google colab for inference?

[ftshijt]

@ftshijt Hi, great work can you please add example audio and a google colab for inference?

Thanks for your interest. We planned to have some samples available. Google colab would definitely be a good choice for that.

@ftshijt is this the repo for https://arxiv.org/abs/2010.12024

[ftshijt]

@ftshijt is this the repo for https://arxiv.org/abs/2010.12024

Yes, this is the repo for the paper. We are currently interested in making it a more comprehensive toolkit like nnsvs or espnet.