NATURAL TTS SYNTHESIS BY CONDITIONING WAVENET ON MEL SPECTROGRAM PREDICTIONS https://arxiv.org/pdf/1712.05884.pdf
WaveNet: A Generative Model for Raw Audio
- Simple LJ Speech DataLoader
- Mel Spectrogram Prediction network (text to Spectrogram)
- [TODO] WaveNet Vocoder (Spectrogram to raw audio)
https://arxiv.org/abs/1609.03499
- install pytorch and torchvision:
conda install pytorch -c pytorch
- install tensorflow and tensorboardX for logging.
pip install tensorboard
pip install tensorboardX
train Spectrogram Prediction Network
python train.py
view logs in Tensorboard
tensorboard --logdir runs
