SillyTavern/public/scripts/extensions/tts/vits.js
Risenafis bb39e852b8
TTS: Add support for VITS/W2V2-VITS/Bert-VITS2 (#2439)
* add VITSTtsProvider

* add options

* fix params, drop reference_audio_path

* post with body

* improve preview lang

* add space in label
2024-06-29 14:15:37 +03:00

405 lines
16 KiB
JavaScript

import { getPreviewString, saveTtsProviderSettings } from './index.js';
export { VITSTtsProvider };
class VITSTtsProvider {
//########//
// Config //
//########//
settings;
ready = false;
voices = [];
separator = '. ';
audioElement = document.createElement('audio');
/**
* Perform any text processing before passing to TTS engine.
* @param {string} text Input text
* @returns {string} Processed text
*/
processText(text) {
return text;
}
audioFormats = ['wav', 'ogg', 'silk', 'mp3', 'flac'];
languageLabels = {
'Auto': 'auto',
'Chinese': 'zh',
'English': 'en',
'Japanese': 'ja',
'Korean': 'ko',
};
langKey2LangCode = {
'zh': 'zh-CN',
'en': 'en-US',
'ja': 'ja-JP',
'ko': 'ko-KR',
};
modelTypes = {
VITS: 'VITS',
W2V2_VITS: 'W2V2-VITS',
BERT_VITS2: 'BERT-VITS2',
};
defaultSettings = {
provider_endpoint: 'http://localhost:23456',
format: 'wav',
lang: 'auto',
length: 1.0,
noise: 0.33,
noisew: 0.4,
segment_size: 50,
streaming: false,
dim_emotion: 0,
sdp_ratio: 0.2,
emotion: 0,
text_prompt: '',
style_text: '',
style_weight: 1,
};
get settingsHtml() {
let html = `
<label for="vits_lang">Text Language</label>
<select id="vits_lang">`;
for (let language in this.languageLabels) {
if (this.languageLabels[language] == this.settings?.lang) {
html += `<option value="${this.languageLabels[language]}" selected="selected">${language}</option>`;
continue;
}
html += `<option value="${this.languageLabels[language]}">${language}</option>`;
}
html += `
</select>
<label>VITS / W2V2-VITS / Bert-VITS2 Settings:</label><br/>
<label for="vits_endpoint">Provider Endpoint:</label>
<input id="vits_endpoint" type="text" class="text_pole" maxlength="250" value="${this.defaultSettings.provider_endpoint}"/>
<span>Use <a target="_blank" href="https://github.com/Artrajz/vits-simple-api">vits-simple-api</a>.</span><br/>
<label for="vits_format">Audio format:</label>
<select id="vits_format">`;
for (let format of this.audioFormats) {
if (format == this.settings?.format) {
html += `<option value="${format}" selected="selected">${format}</option>`;
continue;
}
html += `<option value="${format}">${format}</option>`;
}
html += `
</select>
<label for="vits_length">Audio length: <span id="vits_length_output">${this.defaultSettings.length}</span></label>
<input id="vits_length" type="range" value="${this.defaultSettings.length}" min="0.0" max="5" step="0.01" />
<label for="vits_noise">Noise: <span id="vits_noise_output">${this.defaultSettings.noise}</span></label>
<input id="vits_noise" type="range" value="${this.defaultSettings.noise}" min="0.1" max="2" step="0.01" />
<label for="vits_noisew">SDP noise: <span id="vits_noisew_output">${this.defaultSettings.noisew}</span></label>
<input id="vits_noisew" type="range" value="${this.defaultSettings.noisew}" min="0.1" max="2" step="0.01" />
<label for="vits_segment_size">Segment Size: <span id="vits_segment_size_output">${this.defaultSettings.segment_size}</span></label>
<input id="vits_segment_size" type="range" value="${this.defaultSettings.segment_size}" min="0" max="1000" step="1" />
<label for="vits_streaming" class="checkbox_label">
<input id="vits_streaming" type="checkbox" />
<span>Streaming</span>
</label>
<label>W2V2-VITS Settings:</label><br/>
<label for="vits_dim_emotion">Dimensional emotion:</label>
<input id="vits_dim_emotion" type="number" class="text_pole" min="0" max="5457" step="1" value="${this.defaultSettings.dim_emotion}"/>
<label>BERT-VITS2 Settings:</label><br/>
<label for="vits_sdp_ratio">sdp_ratio: <span id="vits_sdp_ratio_output">${this.defaultSettings.sdp_ratio}</span></label>
<input id="vits_sdp_ratio" type="range" value="${this.defaultSettings.sdp_ratio}" min="0.0" max="1" step="0.01" />
<label for="vits_emotion">emotion: <span id="vits_emotion_output">${this.defaultSettings.emotion}</span></label>
<input id="vits_emotion" type="range" value="${this.defaultSettings.emotion}" min="0" max="9" step="1" />
<label for="vits_text_prompt">Text Prompt:</label>
<input id="vits_text_prompt" type="text" class="text_pole" maxlength="512" value="${this.defaultSettings.text_prompt}"/>
<label for="vits_style_text">Style text:</label>
<input id="vits_style_text" type="text" class="text_pole" maxlength="512" value="${this.defaultSettings.style_text}"/>
<label for="vits_style_weight">Style weight <span id="vits_style_weight_output">${this.defaultSettings.style_weight}</span></label>
<input id="vits_style_weight" type="range" value="${this.defaultSettings.style_weight}" min="0" max="1" step="0.01" />
`;
return html;
}
onSettingsChange() {
// Used when provider settings are updated from UI
this.settings.provider_endpoint = $('#vits_endpoint').val();
this.settings.lang = $('#vits_lang').val();
this.settings.format = $('#vits_format').val();
this.settings.dim_emotion = $('#vits_dim_emotion').val();
this.settings.text_prompt = $('#vits_text_prompt').val();
this.settings.style_text = $('#vits_style_text').val();
// Update the default TTS settings based on input fields
this.settings.length = $('#vits_length').val();
this.settings.noise = $('#vits_noise').val();
this.settings.noisew = $('#vits_noisew').val();
this.settings.segment_size = $('#vits_segment_size').val();
this.settings.streaming = $('#vits_streaming').is(':checked');
this.settings.sdp_ratio = $('#vits_sdp_ratio').val();
this.settings.emotion = $('#vits_emotion').val();
this.settings.style_weight = $('#vits_style_weight').val();
// Update the UI to reflect changes
$('#vits_length_output').text(this.settings.length);
$('#vits_noise_output').text(this.settings.noise);
$('#vits_noisew_output').text(this.settings.noisew);
$('#vits_segment_size_output').text(this.settings.segment_size);
$('#vits_sdp_ratio_output').text(this.settings.sdp_ratio);
$('#vits_emotion_output').text(this.settings.emotion);
$('#vits_style_weight_output').text(this.settings.style_weight);
saveTtsProviderSettings();
this.changeTTSSettings();
}
async loadSettings(settings) {
// Pupulate Provider UI given input settings
if (Object.keys(settings).length == 0) {
console.info('Using default TTS Provider settings');
}
// Only accept keys defined in defaultSettings
this.settings = this.defaultSettings;
for (const key in settings) {
if (key in this.settings) {
this.settings[key] = settings[key];
} else {
console.debug(`Ignoring non-user-configurable setting: ${key}`);
}
}
// Set initial values from the settings
$('#vits_endpoint').val(this.settings.provider_endpoint);
$('#vits_lang').val(this.settings.lang);
$('#vits_format').val(this.settings.format);
$('#vits_length').val(this.settings.length);
$('#vits_noise').val(this.settings.noise);
$('#vits_noisew').val(this.settings.noisew);
$('#vits_segment_size').val(this.settings.segment_size);
$('#vits_streaming').prop('checked', this.settings.streaming);
$('#vits_dim_emotion').val(this.settings.dim_emotion);
$('#vits_sdp_ratio').val(this.settings.sdp_ratio);
$('#vits_emotion').val(this.settings.emotion);
$('#vits_text_prompt').val(this.settings.text_prompt);
$('#vits_style_text').val(this.settings.style_text);
$('#vits_style_weight').val(this.settings.style_weight);
// Update the UI to reflect changes
$('#vits_length_output').text(this.settings.length);
$('#vits_noise_output').text(this.settings.noise);
$('#vits_noisew_output').text(this.settings.noisew);
$('#vits_segment_size_output').text(this.settings.segment_size);
$('#vits_sdp_ratio_output').text(this.settings.sdp_ratio);
$('#vits_emotion_output').text(this.settings.emotion);
$('#vits_style_weight_output').text(this.settings.style_weight);
// Register input/change event listeners to update settings on user interaction
$('#vits_endpoint').on('input', () => { this.onSettingsChange(); });
$('#vits_lang').on('change', () => { this.onSettingsChange(); });
$('#vits_format').on('change', () => { this.onSettingsChange(); });
$('#vits_length').on('change', () => { this.onSettingsChange(); });
$('#vits_noise').on('change', () => { this.onSettingsChange(); });
$('#vits_noisew').on('change', () => { this.onSettingsChange(); });
$('#vits_segment_size').on('change', () => { this.onSettingsChange(); });
$('#vits_streaming').on('change', () => { this.onSettingsChange(); });
$('#vits_dim_emotion').on('change', () => { this.onSettingsChange(); });
$('#vits_sdp_ratio').on('change', () => { this.onSettingsChange(); });
$('#vits_emotion').on('change', () => { this.onSettingsChange(); });
$('#vits_text_prompt').on('change', () => { this.onSettingsChange(); });
$('#vits_style_text').on('change', () => { this.onSettingsChange(); });
$('#vits_style_weight').on('change', () => { this.onSettingsChange(); });
await this.checkReady();
console.info('VITS: Settings loaded');
}
// Perform a simple readiness check by trying to fetch voiceIds
async checkReady() {
await Promise.allSettled([this.fetchTtsVoiceObjects(), this.changeTTSSettings()]);
}
async onRefreshClick() {
return;
}
//#################//
// TTS Interfaces //
//#################//
async getVoice(voiceName) {
if (this.voices.length == 0) {
this.voices = await this.fetchTtsVoiceObjects();
}
const match = this.voices.filter(
v => v.name == voiceName,
)[0];
if (!match) {
throw `TTS Voice name ${voiceName} not found`;
}
return match;
}
async getVoiceById(voiceId) {
if (this.voices.length == 0) {
this.voices = await this.fetchTtsVoiceObjects();
}
const match = this.voices.filter(
v => v.voice_id == voiceId,
)[0];
if (!match) {
throw `TTS Voice id ${voiceId} not found`;
}
return match;
}
async generateTts(text, voiceId) {
const response = await this.fetchTtsGeneration(text, voiceId);
return response;
}
//###########//
// API CALLS //
//###########//
async fetchTtsVoiceObjects() {
const response = await fetch(`${this.settings.provider_endpoint}/voice/speakers`);
if (!response.ok) {
throw new Error(`HTTP ${response.status}: ${await response.json()}`);
}
const jsonData = await response.json();
const voices = [];
const addVoices = (modelType) => {
jsonData[modelType].forEach(voice => {
voices.push({
name: `[${modelType}] ${voice.name} (${voice.lang})`,
voice_id: `${modelType}&${voice.id}`,
preview_url: false,
lang: voice.lang,
});
});
};
for (const key in this.modelTypes) {
addVoices(this.modelTypes[key]);
}
this.voices = voices; // Assign to the class property
return voices; // Also return this list
}
// Each time a parameter is changed, we change the configuration
async changeTTSSettings() {
}
/**
* Fetch TTS generation from the API.
* @param {string} inputText Text to generate TTS for
* @param {string} voiceId Voice ID to use (model_type&speaker_id))
* @returns {Promise<Response|string>} Fetch response
*/
async fetchTtsGeneration(inputText, voiceId, lang = null, forceNoStreaming = false) {
console.info(`Generating new TTS for voice_id ${voiceId}`);
const streaming = !forceNoStreaming && this.settings.streaming;
const [model_type, speaker_id] = voiceId.split('&');
const params = new URLSearchParams();
params.append('text', inputText);
params.append('id', speaker_id);
if (streaming) {
params.append('streaming', streaming);
// Streaming response only supports MP3
}
else {
params.append('format', this.settings.format);
}
params.append('lang', lang ?? this.settings.lang);
params.append('length', this.settings.length);
params.append('noise', this.settings.noise);
params.append('noisew', this.settings.noisew);
params.append('segment_size', this.settings.segment_size);
if (model_type == this.modelTypes.W2V2_VITS) {
params.append('emotion', this.settings.dim_emotion);
}
else if (model_type == this.modelTypes.BERT_VITS2) {
params.append('sdp_ratio', this.settings.sdp_ratio);
params.append('emotion', this.settings.emotion);
if (this.settings.text_prompt) {
params.append('text_prompt', this.settings.text_prompt);
}
if (this.settings.style_text) {
params.append('style_text', this.settings.style_text);
params.append('style_weight', this.settings.style_weight);
}
}
const url = `${this.settings.provider_endpoint}/voice/${model_type.toLowerCase()}`;
if (streaming) {
return url + `?${params.toString()}`;
}
const response = await fetch(
url,
{
method: 'POST',
headers: {
'Content-Type': 'application/x-www-form-urlencoded',
},
body: params,
},
);
if (!response.ok) {
toastr.error(response.statusText, 'TTS Generation Failed');
throw new Error(`HTTP ${response.status}: ${await response.text()}`);
}
return response;
}
/**
* Preview TTS for a given voice ID.
* @param {string} id Voice ID
*/
async previewTtsVoice(id) {
this.audioElement.pause();
this.audioElement.currentTime = 0;
const voice = await this.getVoiceById(id);
const lang = voice.lang.includes(this.settings.lang) ? this.settings.lang : voice.lang[0];
let lang_code = this.langKey2LangCode[lang];
const text = getPreviewString(lang_code);
const response = await this.fetchTtsGeneration(text, id, lang, true);
if (typeof response != 'string') {
if (!response.ok) {
throw new Error(`HTTP ${response.status}: ${await response.text()}`);
}
const audio = await response.blob();
const url = URL.createObjectURL(audio);
this.audioElement.src = url;
this.audioElement.play();
}
}
// Interface not used
async fetchTtsFromHistory(history_item_id) {
return Promise.resolve(history_item_id);
}
}