From e491663a7648f982883ead3191b13ed4b79d9207 Mon Sep 17 00:00:00 2001 From: jinzr Date: Thu, 29 Feb 2024 11:11:12 +0800 Subject: [PATCH] Update README.md --- egs/vctk/TTS/README.md | 24 +++++++++++++++++++++++- 1 file changed, 23 insertions(+), 1 deletion(-) diff --git a/egs/vctk/TTS/README.md b/egs/vctk/TTS/README.md index 724bbfc2b..c07516b77 100644 --- a/egs/vctk/TTS/README.md +++ b/egs/vctk/TTS/README.md @@ -12,4 +12,26 @@ This recipe provides a VITS model trained on the VCTK dataset. Pretrained model can be found [here](https://huggingface.co/zrjin/icefall-tts-vctk-vits-2023-12-05), note that this model was pretrained on the Edinburgh DataShare VCTK dataset. -For tutorial and more details, please refer to the [VITS documentation](https://k2-fsa.github.io/icefall/recipes/TTS/vctk/vits.html). \ No newline at end of file +For tutorial and more details, please refer to the [VITS documentation](https://k2-fsa.github.io/icefall/recipes/TTS/vctk/vits.html). + +The training command is given below: +``` +export CUDA_VISIBLE_DEVICES="0,1,2,3" +./vits/train.py \ + --world-size 4 \ + --num-epochs 1000 \ + --start-epoch 1 \ + --use-fp16 1 \ + --exp-dir vits/exp \ + --tokens data/tokens.txt + --max-duration 350 +``` + +To inference, use: +``` +./vits/infer.py \ + --epoch 1000 \ + --exp-dir vits/exp \ + --tokens data/tokens.txt \ + --max-duration 500 +``` \ No newline at end of file