[Help]: Question of Data Preparation for TTA

open-mmlab / Amphion

Amphion (/æmˈfaɪən/) is a toolkit for Audio, Music, and Speech Generation. Its purpose is to support reproducible research and help junior researchers and engineers get started in the field of audio, music, and speech generation research and development.

https://openhlt.github.io/amphion/

MIT License

4.41k stars 373 forks source link

[Help]: Question of Data Preparation for TTA #176

Open jiusansan222 opened 4 months ago

jiusansan222 commented 4 months ago

Problem Overview

In the introduction to the TTA data preparation, there's no mention of the method to acquire the two specified JSON files. Additionally, these files were not included in the dataset you provided.

Screenshots

yuantuo666 commented 4 months ago

Hi, thanks for proposing this problem. It seems some Python script is missing or not mentioned in the RECIPE.md file, @HeCheng0625 could you please make a further explanation on this?

HeCheng0625 commented 4 months ago

Hi valid json: https://drive.google.com/file/d/18wXIJjO8RgLnaj5e3hcYkyOHGtBRtv8y/view?usp=drive_link train json: https://drive.google.com/file/d/1bquMJRyQ9F1In0w_seLEma0GzJFqfzNj/view?usp=drive_link

jiusansan222 commented 3 months ago

The file is inaccessible, please open the permissions, thank you！

hieuhthh commented 2 months ago

If I have raw WAV file data and a corresponding text prompt, how can I create my own dataset to train TTA? Are there any examples I can follow, such as how to create Mel, format the data, etc.? Thank you!