A fast Text-to-Speech (TTS) model. Work well for English, Mandarin/Chinese, Japanese, Korean, Russian and Tibetan (so far). 快速语音合成模型,适用于英语、普通话/中文、日语、韩语、俄语和藏语(当前已测试)。

Overview

简体中文 | English

并行语音合成

[TOC]

新进展

目录结构

.
|--- config/      # 配置文件
     |--- default.yaml
     |--- ...
|--- datasets/    # 数据处理
|--- encoder/     # 声纹编码器
     |--- voice_encoder.py
     |--- ...
|--- helpers/     # 一些辅助类
     |--- trainer.py
     |--- synthesizer.py
     |--- ...
|--- logdir/      # 训练过程保存目录
|--- losses/      # 一些损失函数
|--- models/      # 合成模型
     |--- layers.py
     |--- duration.py
     |--- parallel.py
|--- pretrained/  # 预训练模型(LJSpeech 数据集)
|--- samples/     # 合成样例
|--- utils/       # 一些通用方法
|--- vocoder/     # 声码器
     |--- melgan.py
     |--- ...
|--- wandb/       # Wandb 保存目录
|--- extract-duration.py
|--- extract-embedding.py
|--- LICENSE
|--- prepare-dataset.py  # 准备脚本
|--- README.md
|--- README_en.md
|--- requirements.txt    # 依赖文件
|--- synthesize.py       # 合成脚本
|--- train-duration.py   # 训练脚本
|--- train-parallel.py

合成样例

部分合成样例见这里

预训练

部分预训练模型见这里

快速开始

步骤(1):克隆仓库

$ git clone https://github.com/atomicoo/ParallelTTS.git

步骤(2):安装依赖

$ conda create -n ParallelTTS python=3.7.9
$ conda activate ParallelTTS
$ pip install -r requirements.txt

步骤(3):合成语音

$ python synthesize.py \
  --checkpoint ./pretrained/ljspeech-parallel-epoch0100.pth \
  --melgan_checkpoint ./pretrained/ljspeech-melgan-epoch3200.pth \
  --input_texts ./samples/english/synthesize.txt \
  --outputs_dir ./outputs/

如果要合成其他语种的语音,需要通过 --config 指定相应的配置文件。

如何训练

步骤(1):准备数据

$ python prepare-dataset.py

通过 --config 可以指定配置文件,默认的 default.yaml 针对 LJSpeech 数据集。

步骤(2):训练对齐模型

$ python train-duration.py

步骤(3):提取持续时间

$ python extract-duration.py

通过 --ground_truth 可以指定是否利用对齐模型生成 Ground-Truth 声谱图。

步骤(4):训练合成模型

$ python train-parallel.py

通过 --ground_truth 可以指定是否使用 Ground-Truth 声谱图进行模型训练。

训练日志

如果使用 TensorBoardX,则运行如下命令:

$ tensorboard --logdir logdir/[DIR]/

强烈推荐使用 Wandb(Weights & Biases),只需在上述训练命令中增加 --enable_wandb 选项。

数据集

  • LJSpeech:英语,女性,22050 Hz,约 24 小时
  • LibriSpeech:英语,多说话人(仅使用 train-clean-100 部分),16000 Hz,总计约 1000 小时
  • JSUT:日语,女性,48000 Hz,约 10 小时
  • BiaoBei:普通话,女性,48000 Hz,约 12 小时
  • KSS:韩语,女性,44100 Hz,约 12 小时
  • RuLS:俄语,多说话人(仅使用单一说话人音频),16000 Hz,总计约 98 小时
  • TWLSpeech(非公开,质量较差):藏语,女性(多说话人,音色相近),16000 Hz,约 23 小时

质量评估

TODO:待补充

速度指标

训练速度:对于 LJSpeech 数据集,设置批次尺寸为 64,可以在单张 8GB 显存的 GTX 1080 显卡上进行训练,训练 ~8h(~300 epochs)后即可合成质量较高的语音。

合成速度:以下测试在 CPU @ Intel Core i7-8550U / GPU @ NVIDIA GeForce MX150 下进行,每段合成音频在 8 秒左右(约 20 词)

批次尺寸 Spec
(GPU)
Audio
(GPU)
Spec
(CPU)
Audio
(CPU)
1 0.042 0.218 0.100 2.004
2 0.046 0.453 0.209 3.922
4 0.053 0.863 0.407 7.897
8 0.062 2.386 0.878 14.599

注意,没有进行多次测试取平均值,结果仅供参考。

一些问题

  • wavegan 分支中,vocoder 代码取自 ParallelWaveGAN,由于声学特征提取方式不兼容,需要进行转化,具体转化代码见这里
  • 普通话模型的文本输入选择拼音序列,因为 BiaoBei 的原始拼音序列不包含标点、以及对齐模型训练不完全,所以合成语音的节奏会有点问题。
  • 韩语模型没有专门训练对应的声码器,而是直接使用 LJSpeech(同为 22050 Hz)的声码器,可能稍微影响合成语音的质量。

参考资料

TODO

  • 合成语音质量评估(MOS)
  • 更多不同语种的测试
  • 语音风格迁移(音色)

欢迎交流

  • 微信号:Joee1995

  • 企鹅号:793071559

Owner
Atomicoo
Atomicoo
:mag: Transformers at scale for question answering & neural search. Using NLP via a modular Retriever-Reader-Pipeline. Supporting DPR, Elasticsearch, HuggingFace's Modelhub...

Haystack is an end-to-end framework that enables you to build powerful and production-ready pipelines for different search use cases. Whether you want

deepset 6.4k Jan 09, 2023
Stanford CoreNLP provides a set of natural language analysis tools written in Java

Stanford CoreNLP Stanford CoreNLP provides a set of natural language analysis tools written in Java. It can take raw human language text input and giv

Stanford NLP 8.8k Jan 07, 2023
[EMNLP 2021] Mirror-BERT: Converting Pretrained Language Models to universal text encoders without labels.

[EMNLP 2021] Mirror-BERT: Converting Pretrained Language Models to universal text encoders without labels.

Cambridge Language Technology Lab 61 Dec 10, 2022
Simple NLP based project without any use of AI

Simple NLP based project without any use of AI

Shripad Rao 1 Apr 26, 2022
Jarvis is a simple Chatbot with a GUI capable of chatting and retrieving information and daily news from the internet for it's user.

J.A.R.V.I.S Kindly consider starring this repository if you like the program :-) What/Who is J.A.R.V.I.S? J.A.R.V.I.S is an chatbot written that is bu

Epicalable 50 Dec 31, 2022
Create a machine learning model which will predict if the mortgage will be approved or not based on 5 variables

Mortgage-Application-Analysis Create a machine learning model which will predict if the mortgage will be approved or not based on 5 variables: age, in

1 Jan 29, 2022
Black for Python docstrings and reStructuredText (rst).

Style-Doc Style-Doc is Black for Python docstrings and reStructuredText (rst). It can be used to format docstrings (Google docstring format) in Python

Telekom Open Source Software 13 Oct 24, 2022
A Persian Image Captioning model based on Vision Encoder Decoder Models of the transformers🤗.

Persian-Image-Captioning We fine-tuning the Vision Encoder Decoder Model for the task of image captioning on the coco-flickr-farsi dataset. The implem

Hamtech-ai 15 Aug 25, 2022
Trains an OpenNMT PyTorch model and SentencePiece tokenizer.

Trains an OpenNMT PyTorch model and SentencePiece tokenizer. Designed for use with Argos Translate and LibreTranslate.

Argos Open Tech 61 Dec 13, 2022
Bpe algorithm can finetune tokenizer - Bpe algorithm can finetune tokenizer

"# bpe_algorithm_can_finetune_tokenizer" this is an implyment for https://github

张博 1 Feb 02, 2022
Uses Google's gTTS module to easily create robo text readin' on command.

Tool to convert text to speech, creating files for later use. TTRS uses Google's gTTS module to easily create robo text readin' on command.

0 Jun 20, 2021
Unsupervised text tokenizer focused on computational efficiency

YouTokenToMe YouTokenToMe is an unsupervised text tokenizer focused on computational efficiency. It currently implements fast Byte Pair Encoding (BPE)

VK.com 847 Dec 19, 2022
Implementation of TTS with combination of Tacotron2 and HiFi-GAN

Tacotron2-HiFiGAN-master Implementation of TTS with combination of Tacotron2 and HiFi-GAN for Mandarin TTS. Inference In order to inference, we need t

SunLu Z 7 Nov 11, 2022
Estimation of the CEFR complexity score of a given word, sentence or text.

NLP-Swedish … allows to estimate CEFR (Common European Framework of References) complexity score of a given word, sentence or text. CEFR scores come f

3 Apr 30, 2022
Prompt tuning toolkit for GPT-2 and GPT-Neo

mkultra mkultra is a prompt tuning toolkit for GPT-2 and GPT-Neo. Prompt tuning injects a string of 20-100 special tokens into the context in order to

61 Jan 01, 2023
[ICLR'19] Trellis Networks for Sequence Modeling

TrellisNet for Sequence Modeling This repository contains the experiments done in paper Trellis Networks for Sequence Modeling by Shaojie Bai, J. Zico

CMU Locus Lab 460 Oct 13, 2022
Header-only C++ HNSW implementation with python bindings

Hnswlib - fast approximate nearest neighbor search Header-only C++ HNSW implementation with python bindings. NEWS: version 0.6 Thanks to (@dyashuni) h

2.3k Jan 05, 2023
Princeton NLP's pre-training library based on fairseq with DeepSpeed kernel integration 🚃

This repository provides a library for efficient training of masked language models (MLM), built with fairseq. We fork fairseq to give researchers mor

Princeton Natural Language Processing 92 Dec 27, 2022
Kestrel Threat Hunting Language

Kestrel Threat Hunting Language What is Kestrel? Why we need it? How to hunt with XDR support? What is the science behind it? You can find all the ans

Open Cybersecurity Alliance 201 Dec 16, 2022
Applying "Load What You Need: Smaller Versions of Multilingual BERT" to LaBSE

smaller-LaBSE LaBSE(Language-agnostic BERT Sentence Embedding) is a very good method to get sentence embeddings across languages. But it is hard to fi

Jeong Ukjae 13 Sep 02, 2022