FastSpeech-Pytorch
The Implementation of FastSpeech Based on Pytorch.
Update (2020/07/20)
- Optimize the training process.
- Optimize the implementation of length regulator.
- Use the same hyper parameter as FastSpeech2.
- The measures of the 1, 2 and 3 make the training process 3 times faster than before.
- Better speech quality.
Model
My Blog
Prepare Dataset
- Download and extract LJSpeech dataset.
- Put LJSpeech dataset in
data
.
- Unzip
alignments.zip
.
- Put Nvidia pretrained waveglow model in the
waveglow/pretrained_model
and rename as waveglow_256channels.pt
;
- Run
python3 preprocess.py
.
Training
Run python3 train.py
.
Evaluation
Run python3 eval.py
.
Notes
- In the paper of FastSpeech, authors use pre-trained Transformer-TTS model to provide the target of alignment. I didn't have a well-trained Transformer-TTS model so I use Tacotron2 instead.
- I use the same hyper-parameter as FastSpeech2.
- The examples of audio are in
sample
.
- pretrained model.
Reference
Repository
Paper