From 5d925eaa4dd83bb2696b4fb399bfbed2d0eb36f9 Mon Sep 17 00:00:00 2001 From: Edresson Casanova Date: Thu, 8 Dec 2022 14:01:04 -0300 Subject: [PATCH] Fix lint --- recipes/vctk/yourtts/train_yourtts.py | 15 ++++++++++++--- 1 file changed, 12 insertions(+), 3 deletions(-) diff --git a/recipes/vctk/yourtts/train_yourtts.py b/recipes/vctk/yourtts/train_yourtts.py index 0d3ec12c..32a46024 100644 --- a/recipes/vctk/yourtts/train_yourtts.py +++ b/recipes/vctk/yourtts/train_yourtts.py @@ -1,14 +1,23 @@ -import torch import os + +import torch from trainer import Trainer, TrainerArgs from TTS.config.shared_configs import BaseDatasetConfig -from TTS.tts.datasets import load_tts_samples -from TTS.tts.models.vits import VitsArgs, VitsAudioConfig, Vits from TTS.tts.configs.vits_config import VitsConfig +from TTS.tts.datasets import load_tts_samples +from TTS.tts.models.vits import Vits, VitsArgs, VitsAudioConfig torch.set_num_threads(24) +# pylint: disable=W0105 +""" + This recipe replicates the first experiment proposed in the YourTTS paper (https://arxiv.org/abs/2112.02418). + YourTTS model is based on the VITS model however it uses external speaker embeddings extracted from a pre-trained speaker encoder and has small architecture changes. + In addition, YourTTS can be trained in multilingual data, however, this recipe replicates the single language training using the VCTK dataset. + The VitsArgs instance has commented parameters used to enable the multilingual training. +""" + # Name of the run for the Trainer RUN_NAME = "YourTTS-EN-VCTK"