Daniel Holler

LoRA Fine Tuning

36

This is just the first draft so we can start building this feature. - Added dataloader.py, which loads data for training - Added train.py, with the current training loop -...

Hey, most spectrogram generators are built for 80 bins and a hop size of 256. Has anybody succeeded in making WaveGrad work with these settings, making for better compatibility with...

License

1

Hey! Incredible work and results, and amazing due dilligence in the paper - really appreciated, and putting together RealEdit for evaluating results and fairly training and comparing to other SoTA...

Phoneme-level pronunciation control

2

Hey! I understand that the text tokens are currently encoded on the character-level and the model is trained with these tokens. What would be the process for getting phoneme level...

Hey, thank you for this wonderful work. I am wondering if it is possible to perform audio-driven inference on a video (or several frames), instead of just a singular frame...