ㅤㅤ👇 Applio's official links below 👇ㅤㅤ
🌐 Website • 📚 Documentation • ☎️ Discord
Goal of this project is to have a much more flexible base than mainline rvc.
ㅤ
Features that are, at the time of writing, already added in:
-
New loss functions: Multi-scale mel spectrogram loss ( optimized ), Envelope loss
-
Configurable learning-rate warmup.
( Provides an ability to give your training a lil warmup, potentially yielding better results. ) -
New logging mechanism for losses: Average loss per epoch logged as the standard loss,
and rolling average loss over 5 epochs to evaluate general trends and the model's performance over time.Both work for each metric individually. -
Features a different optimizer: RAdam ( Rectified Adam )
( More init. stability and compared to AdamW, doesn't require using / configuring warmup. )
Most likely better convergence / generalization on average, compared to plain AdamW without a warmup. -
Support for following vocoders: HiFi-GAN, MRF-HiFi-gan and Refine-GAN.
-
Mel spectrogram % similarity metric.
-
SoX resampler in "VHQ" mode used by default in place of "soxr_medium" or "kaiser_best" ( fallback. ).
-
Checkpointing and in-place support for both Gen. and Disc.
( Decreases the vram consumption on cost of the computation / training speed. ) -
Customization for preprocessing workflow ( Including 'mute' files usage and various slicing methods. )
⚠️ 1: HiFi-gan is the stock rvc/applio vocoder, hence it's what you use for og pretrains and customs ( for now ).
⚠️ 2: MRF-HiFi-GAN and Refine-GAN require new pretrained models. They can't be used with original rvc's G/D pretrains.
ㅤ ✨ to-do list ✨
- Ability to switch back to AdamW optimizer
( In case someone wants to try it with a warmup. )- Adjustable hop length for RMVPE.
- Custom initial learning rate per Generator and Discriminator.
( Not 100% sure about that just yet. )- Ability to delay / headstart the Generator or Discriminator.
- and more...
Run the installation script based on your operating system:
- Windows: Double-click
run-install.bat. - Linux/macOS: Execute
run-install.sh.
Start Applio using:
- Windows: Double-click
run-fork.bat. - Linux/macOS: Run
run-fork.sh.
This launches the Gradio interface in your default browser.
To monitor training or visualize data:
- Windows: Run
run-tensorboard.bat. - Linux/macOS: Run
run-tensorboard.sh.
For more detailed instructions, visit the documentation.
The creators of the original Applio repository, Applio's contributors, and the maintainer of this fork (Codename;0), built upon Applio, are not responsible for any legal issues, damages, or consequences arising from the use of this repository or the content generated from it. By using this fork, you acknowledge that:
- The use of this fork is at your own risk.
- This repository is intended solely for educational, and experimental purposes.
- Any misuse, including but not limited to illegal activities or violation of third-party rights,
is not the responsibility of the original creators, contributors, or this fork’s maintainer. - You willingly agree to comply with this repository's Terms of Use