From 7eb045206c59e09063c5a015911a07af0a165b36 Mon Sep 17 00:00:00 2001 From: rafaelvalle Date: Thu, 14 Jun 2018 11:25:42 -0700 Subject: [PATCH] README.md: updating readme to explicitly mention that mel representation of WaveNet and Tacotron2 must be the same --- README.md | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/README.md b/README.md index e1fe810..ff32b14 100644 --- a/README.md +++ b/README.md @@ -38,9 +38,13 @@ wavenet](https://github.com/r9y9/wavenet_vocoder/) 1. `python -m multiproc train.py --output_directory=outdir --log_directory=logdir --hparams=distributed_run=True,fp16_run=True` ## Inference +When performing Mel-Spectrogram to Audio synthesis with a WaveNet model, make sure Tacotron 2 and WaveNet were trained on the same mel-spectrogram representation. Follow these steps to use a a simple inference pipeline using griffin-lim: + 1. `jupyter notebook --ip=127.0.0.1 --port=31337` 2. load inference.ipynb + + ## Related repos [nv-wavenet](https://github.com/NVIDIA/nv-wavenet/): Faster than real-time wavenet inference