Merge pull request #1 from MartinDelophy/codex/vocal-worker-discord
Move vocal separation to a worker and add Discord community links
This commit is contained in:
MartinDelophy
@@ -6,6 +6,7 @@ Language: **English** | [中文](README.zh-CN.md)
|
||||
[](https://github.com/MartinDelophy/ai-video-editor/stargazers)
|
||||
[](https://github.com/MartinDelophy/ai-video-editor/forks)
|
||||
[](LICENSE)
|
||||
[](https://discord.gg/uq2uvUTBr)
|
||||
|
||||
A local-first, browser-based AI video editor with ONNX voiceover generation, Whisper automatic captions, JoyVASA + LivePortrait talking avatars, multi-track timeline editing, and MP4/WebM export. Timeline Studio runs its core AI workflows directly in the browser and is inspired by modern editors such as CapCut.
|
||||
|
||||
@@ -218,6 +219,10 @@ Released under the [MIT License](LICENSE). You may use, modify, and distribute t
|
||||
|
||||
## Community
|
||||
|
||||
Join the Discord community for questions, feedback, and project updates:
|
||||
|
||||
[Join the Timeline Studio Discord](https://discord.gg/uq2uvUTBr)
|
||||
|
||||
The current star count is updated live in the GitHub Stars badge at the top of this README.
|
||||
|
||||
[View current stargazers](https://github.com/MartinDelophy/ai-video-editor/stargazers) · [Star Timeline Studio](https://github.com/MartinDelophy/ai-video-editor)
|
||||
|
||||
@@ -6,6 +6,7 @@
|
||||
[](https://github.com/MartinDelophy/ai-video-editor/stargazers)
|
||||
[](https://github.com/MartinDelophy/ai-video-editor/forks)
|
||||
[](LICENSE)
|
||||
[](https://discord.gg/uq2uvUTBr)
|
||||
|
||||
一个本地优先的浏览器 AI 视频编辑器，支持 ONNX AI 配音、Whisper 自动字幕、JoyVASA + LivePortrait 数字人、多轨时间线以及 MP4/WebM 导出。核心 AI 流程直接在浏览器运行，交互参考 CapCut/剪映等现代时间线编辑器。
|
||||
|
||||
@@ -203,6 +204,10 @@ npx netlify-cli deploy --prod --dir=dist
|
||||
|
||||
## 社区
|
||||
|
||||
欢迎加入 Discord 社区，交流使用问题、产品建议和项目进展：
|
||||
|
||||
[加入 Timeline Studio Discord 社区](https://discord.gg/uq2uvUTBr)
|
||||
|
||||
README 顶部的 GitHub Stars 徽章会实时显示当前 Star 数量。
|
||||
|
||||
[查看当前 Stargazers](https://github.com/MartinDelophy/ai-video-editor/stargazers) · [为 Timeline Studio 点 Star](https://github.com/MartinDelophy/ai-video-editor)
|
||||
|
||||
@@ -3,6 +3,7 @@ import { useEffect } from "react";
|
||||
import { RATIO_OPTIONS } from "../config/editor.js";
|
||||
import { decodeWaveform } from "../lib/media.js";
|
||||
import { disposeVisionWorker } from "../lib/vision.js";
|
||||
import { disposeVocalSeparationWorker } from "../lib/vocalSeparation.js";
|
||||
import {
|
||||
getNearestRatioIdForSize,
|
||||
revokeVisionObjectUrls,
|
||||
@@ -158,6 +159,7 @@ export function useEditorLifecycle(d) {
|
||||
d.visionObjectUrlsRef.current.forEach((urls) => revokeVisionObjectUrls(urls));
|
||||
d.visionObjectUrlsRef.current.clear();
|
||||
disposeVisionWorker();
|
||||
disposeVocalSeparationWorker();
|
||||
d.voiceRecorderStreamRef.current?.getTracks().forEach((track) => track.stop());
|
||||
window.clearInterval(d.voiceRecorderTimerRef.current);
|
||||
}, []);
|
||||
|
||||
+60
-133
@@ -1,144 +1,71 @@
|
||||
const MODEL_URL = "https://huggingface.co/haixin/timeline-studio-vocal-remover/resolve/main/model.json";
|
||||
const CHUNK_SIZE = 31744;
|
||||
const PADDING = 3072;
|
||||
const FFT_SIZE = 6144;
|
||||
const HOP_SIZE = 1024;
|
||||
let worker;
|
||||
let nextRequestId = 0;
|
||||
const requests = new Map();
|
||||
|
||||
let modelPromise;
|
||||
let runtimePromise;
|
||||
let activeBackend = "webgl";
|
||||
function rejectRequests(error) {
|
||||
requests.forEach(({ reject }) => reject(error));
|
||||
requests.clear();
|
||||
}
|
||||
|
||||
function loadScript(src) {
|
||||
function resetWorker(error) {
|
||||
worker?.terminate();
|
||||
worker = null;
|
||||
if (error) rejectRequests(error);
|
||||
}
|
||||
|
||||
function getWorker() {
|
||||
if (typeof Worker === "undefined") throw new Error("VOCAL_WORKER_UNAVAILABLE");
|
||||
if (worker) return worker;
|
||||
|
||||
worker = new Worker(new URL("../workers/vocal-separation.worker.js", import.meta.url));
|
||||
worker.addEventListener("message", ({ data }) => {
|
||||
const request = requests.get(data?.requestId);
|
||||
if (!request) return;
|
||||
if (data.type === "progress") {
|
||||
request.onProgress(data.progress, data.phase);
|
||||
return;
|
||||
}
|
||||
requests.delete(data.requestId);
|
||||
if (data.type === "result") {
|
||||
request.resolve({
|
||||
vocals: new Blob([data.vocalsBuffer], { type: "audio/wav" }),
|
||||
accompaniment: new Blob([data.accompanimentBuffer], { type: "audio/wav" }),
|
||||
backend: data.backend,
|
||||
});
|
||||
return;
|
||||
}
|
||||
request.reject(new Error(data.error || "VOCAL_MODEL_FAILED"));
|
||||
});
|
||||
worker.addEventListener("error", (event) => {
|
||||
resetWorker(new Error(event.message || "VOCAL_WORKER_FAILED"));
|
||||
});
|
||||
return worker;
|
||||
}
|
||||
|
||||
function runWorker(left, right, sampleRate, onProgress) {
|
||||
const requestId = `vocal-${++nextRequestId}`;
|
||||
const activeWorker = getWorker();
|
||||
return new Promise((resolve, reject) => {
|
||||
const existing = document.querySelector(`script[src="${src}"]`);
|
||||
if (existing) { if (window.tf) resolve(); else existing.addEventListener("load", resolve, { once: true }); return; }
|
||||
const script = document.createElement("script"); script.src = src; script.crossOrigin = "anonymous";
|
||||
script.onload = resolve; script.onerror = () => reject(new Error("VOCAL_RUNTIME_DOWNLOAD_FAILED")); document.head.appendChild(script);
|
||||
requests.set(requestId, { resolve, reject, onProgress });
|
||||
activeWorker.postMessage(
|
||||
{ type: "separate", requestId, leftBuffer: left.buffer, rightBuffer: right.buffer, sampleRate },
|
||||
[left.buffer, right.buffer],
|
||||
);
|
||||
});
|
||||
}
|
||||
|
||||
async function getRuntime() {
|
||||
if (!runtimePromise) runtimePromise = (async () => {
|
||||
await loadScript("/vendor/vocal-remover/tf.min.js?v=4.22.0");
|
||||
if (navigator.gpu) await loadScript("/vendor/vocal-remover/tf-backend-webgpu.js?v=4.22.0");
|
||||
return window.tf;
|
||||
})();
|
||||
return runtimePromise;
|
||||
}
|
||||
|
||||
async function getModel(tf) {
|
||||
if (!modelPromise) {
|
||||
modelPromise = (async () => {
|
||||
const requestedBackend = navigator.gpu && tf.findBackend("webgpu") ? "webgpu" : "webgl";
|
||||
let ready = false;
|
||||
try { ready = await tf.setBackend(requestedBackend); } catch { ready = false; }
|
||||
if (!ready && requestedBackend === "webgpu") ready = await tf.setBackend("webgl");
|
||||
if (!ready) throw new Error("VOCAL_BACKEND_UNAVAILABLE");
|
||||
activeBackend = tf.getBackend();
|
||||
await tf.ready();
|
||||
return tf.loadGraphModel(MODEL_URL);
|
||||
})();
|
||||
}
|
||||
return modelPromise;
|
||||
}
|
||||
|
||||
function stft(tf, input) {
|
||||
return tf.tidy(() => {
|
||||
const spectrum = tf.signal.stft(input, FFT_SIZE, HOP_SIZE, FFT_SIZE, (length) => tf.signal.hannWindow(length));
|
||||
const real = tf.real(spectrum).slice([0, 0], [32, 3072]).transpose();
|
||||
const imag = tf.imag(spectrum).slice([0, 0], [32, 3072]).transpose();
|
||||
const output = tf.stack([real, imag], 0);
|
||||
return tf.where(tf.isNaN(output), tf.zerosLike(output), output);
|
||||
});
|
||||
}
|
||||
|
||||
async function inverseChannel(tf, spectrogram) {
|
||||
const frames = spectrogram.shape[1];
|
||||
const output = new Float32Array(FFT_SIZE + HOP_SIZE * (frames - 1));
|
||||
const window = await tf.signal.hannWindow(FFT_SIZE).data();
|
||||
for (let index = 0; index < frames; index += 1) {
|
||||
const frame = tf.tidy(() => {
|
||||
const values = spectrogram.slice([0, index, 0], [-1, 1, 2]).squeeze([1]);
|
||||
return tf.spectral.irfft(tf.complex(values.slice([0, 0], [-1, 1]).squeeze(), values.slice([0, 1], [-1, 1]).squeeze()));
|
||||
});
|
||||
const samples = await frame.data();
|
||||
frame.dispose();
|
||||
const offset = index * HOP_SIZE;
|
||||
for (let cursor = 0; cursor < samples.length; cursor += 1) output[offset + cursor] += samples[cursor] * window[cursor];
|
||||
}
|
||||
return output;
|
||||
}
|
||||
|
||||
async function istft(tf, tensor) {
|
||||
const padded = tf.pad(tensor, [[0, 0], [0, 0], [0, 1], [0, 0]]);
|
||||
const shaped = padded.reshape([2, 2, 3073, 32]).transpose([0, 2, 3, 1]);
|
||||
const left = shaped.slice([0, 0, 0, 0], [1, -1, -1, -1]).squeeze([0]);
|
||||
const right = shaped.slice([1, 0, 0, 0], [1, -1, -1, -1]).squeeze([0]);
|
||||
const result = await Promise.all([inverseChannel(tf, left), inverseChannel(tf, right)]);
|
||||
tf.dispose([padded, shaped, left, right]);
|
||||
return result;
|
||||
}
|
||||
|
||||
function wavBlob(channels, sampleRate) {
|
||||
const frames = channels[0].length;
|
||||
const buffer = new ArrayBuffer(44 + frames * 4);
|
||||
const view = new DataView(buffer);
|
||||
const text = (offset, value) => [...value].forEach((char, index) => view.setUint8(offset + index, char.charCodeAt(0)));
|
||||
text(0, "RIFF"); view.setUint32(4, buffer.byteLength - 8, true); text(8, "WAVE"); text(12, "fmt ");
|
||||
view.setUint32(16, 16, true); view.setUint16(20, 1, true); view.setUint16(22, 2, true);
|
||||
view.setUint32(24, sampleRate, true); view.setUint32(28, sampleRate * 4, true); view.setUint16(32, 4, true); view.setUint16(34, 16, true);
|
||||
text(36, "data"); view.setUint32(40, frames * 4, true);
|
||||
let offset = 44;
|
||||
for (let frame = 0; frame < frames; frame += 1) for (let channel = 0; channel < 2; channel += 1) {
|
||||
const sample = Math.max(-1, Math.min(1, channels[channel][frame] || 0));
|
||||
view.setInt16(offset, sample < 0 ? sample * 0x8000 : sample * 0x7fff, true); offset += 2;
|
||||
}
|
||||
return new Blob([buffer], { type: "audio/wav" });
|
||||
}
|
||||
|
||||
export async function separateVocals(blob, onProgress = () => {}) {
|
||||
const tf = await getRuntime();
|
||||
onProgress(4, { key: "vocalSeparationLoadingModel" });
|
||||
const model = await getModel(tf);
|
||||
onProgress(7, {
|
||||
key: activeBackend === "webgpu" ? "vocalSeparationUsingWebGpu" : "vocalSeparationUsingWebGl",
|
||||
});
|
||||
const context = new AudioContext({ sampleRate: 44100 });
|
||||
try {
|
||||
const decoded = await context.decodeAudioData((await blob.arrayBuffer()).slice(0));
|
||||
const left = decoded.getChannelData(0);
|
||||
const right = decoded.numberOfChannels > 1 ? decoded.getChannelData(1) : left;
|
||||
const accompaniment = [[], []];
|
||||
const vocals = [[], []];
|
||||
const chunks = Math.ceil(left.length / CHUNK_SIZE);
|
||||
for (let index = 0; index < chunks; index += 1) {
|
||||
const start = index * CHUNK_SIZE;
|
||||
const valid = Math.min(CHUNK_SIZE, left.length - start);
|
||||
const paddedLeft = new Float32Array(CHUNK_SIZE + PADDING * 2); paddedLeft.set(left.subarray(start, start + valid), PADDING);
|
||||
const paddedRight = new Float32Array(CHUNK_SIZE + PADDING * 2); paddedRight.set(right.subarray(start, start + valid), PADDING);
|
||||
const input = tf.tidy(() => {
|
||||
const l = stft(tf, tf.tensor1d(paddedLeft)); const r = stft(tf, tf.tensor1d(paddedRight));
|
||||
return tf.stack([l, r], 3).transpose([0, 3, 1, 2]).reshape([1, 4, 3072, 32]);
|
||||
});
|
||||
const musicTensor = model.predict(input);
|
||||
const vocalTensor = tf.sub(input, musicTensor);
|
||||
const [music, voice] = await Promise.all([istft(tf, musicTensor), istft(tf, vocalTensor)]);
|
||||
for (let channel = 0; channel < 2; channel += 1) {
|
||||
accompaniment[channel].push(...music[channel].slice(PADDING, PADDING + valid));
|
||||
vocals[channel].push(...voice[channel].slice(PADDING, PADDING + valid));
|
||||
}
|
||||
tf.dispose([input, musicTensor, vocalTensor]);
|
||||
onProgress(8 + Math.round(((index + 1) / chunks) * 88), {
|
||||
key: "vocalSeparationProcessingChunk",
|
||||
current: index + 1,
|
||||
total: chunks,
|
||||
});
|
||||
await tf.nextFrame();
|
||||
}
|
||||
return {
|
||||
vocals: wavBlob(vocals, decoded.sampleRate),
|
||||
accompaniment: wavBlob(accompaniment, decoded.sampleRate),
|
||||
backend: activeBackend,
|
||||
};
|
||||
} finally { await context.close().catch(() => {}); }
|
||||
const left = decoded.getChannelData(0).slice();
|
||||
const right = decoded.numberOfChannels > 1 ? decoded.getChannelData(1).slice() : left.slice();
|
||||
return await runWorker(left, right, decoded.sampleRate, onProgress);
|
||||
} finally {
|
||||
await context.close().catch(() => {});
|
||||
}
|
||||
}
|
||||
|
||||
export function disposeVocalSeparationWorker() {
|
||||
resetWorker(new Error("VOCAL_WORKER_DISPOSED"));
|
||||
}
|
||||
|
||||
@@ -0,0 +1,161 @@
|
||||
const MODEL_URL = "https://huggingface.co/haixin/timeline-studio-vocal-remover/resolve/main/model.json";
|
||||
const CHUNK_SIZE = 31744;
|
||||
const PADDING = 3072;
|
||||
const FFT_SIZE = 6144;
|
||||
const HOP_SIZE = 1024;
|
||||
|
||||
let modelPromise;
|
||||
let runtimePromise;
|
||||
let activeBackend = "webgl";
|
||||
|
||||
function postProgress(requestId, progress, phase) {
|
||||
self.postMessage({ type: "progress", requestId, progress, phase });
|
||||
}
|
||||
|
||||
async function getRuntime() {
|
||||
runtimePromise ??= (async () => {
|
||||
try {
|
||||
importScripts("/vendor/vocal-remover/tf.min.js?v=4.22.0");
|
||||
if (self.navigator.gpu) importScripts("/vendor/vocal-remover/tf-backend-webgpu.js?v=4.22.0");
|
||||
} catch {
|
||||
throw new Error("VOCAL_RUNTIME_DOWNLOAD_FAILED");
|
||||
}
|
||||
return self.tf;
|
||||
})();
|
||||
return runtimePromise;
|
||||
}
|
||||
|
||||
async function getModel(runtime) {
|
||||
modelPromise ??= (async () => {
|
||||
const requestedBackend = self.navigator.gpu && runtime.findBackend("webgpu") ? "webgpu" : "webgl";
|
||||
let ready;
|
||||
try { ready = await runtime.setBackend(requestedBackend); } catch { ready = false; }
|
||||
if (!ready && requestedBackend === "webgpu") ready = await runtime.setBackend("webgl");
|
||||
if (!ready) throw new Error("VOCAL_BACKEND_UNAVAILABLE");
|
||||
activeBackend = runtime.getBackend();
|
||||
await runtime.ready();
|
||||
return runtime.loadGraphModel(MODEL_URL);
|
||||
})();
|
||||
return modelPromise;
|
||||
}
|
||||
|
||||
function stft(runtime, input) {
|
||||
return runtime.tidy(() => {
|
||||
const spectrum = runtime.signal.stft(input, FFT_SIZE, HOP_SIZE, FFT_SIZE, (length) => runtime.signal.hannWindow(length));
|
||||
const real = runtime.real(spectrum).slice([0, 0], [32, 3072]).transpose();
|
||||
const imag = runtime.imag(spectrum).slice([0, 0], [32, 3072]).transpose();
|
||||
const output = runtime.stack([real, imag], 0);
|
||||
return runtime.where(runtime.isNaN(output), runtime.zerosLike(output), output);
|
||||
});
|
||||
}
|
||||
|
||||
function createHannWindow() {
|
||||
const window = new Float32Array(FFT_SIZE);
|
||||
for (let index = 0; index < FFT_SIZE; index += 1) {
|
||||
window[index] = 0.5 - 0.5 * Math.cos((2 * Math.PI * index) / FFT_SIZE);
|
||||
}
|
||||
return window;
|
||||
}
|
||||
|
||||
const hannWindow = createHannWindow();
|
||||
|
||||
async function inverseChannel(runtime, spectrogram) {
|
||||
const frames = spectrogram.shape[1];
|
||||
const timeFrames = runtime.tidy(() => {
|
||||
const frameMajor = spectrogram.transpose([1, 0, 2]);
|
||||
const real = frameMajor.slice([0, 0, 0], [-1, -1, 1]).squeeze([2]);
|
||||
const imag = frameMajor.slice([0, 0, 1], [-1, -1, 1]).squeeze([2]);
|
||||
return runtime.spectral.irfft(runtime.complex(real, imag));
|
||||
});
|
||||
const samples = await timeFrames.data();
|
||||
timeFrames.dispose();
|
||||
|
||||
const output = new Float32Array(FFT_SIZE + HOP_SIZE * (frames - 1));
|
||||
for (let frame = 0; frame < frames; frame += 1) {
|
||||
const sourceOffset = frame * FFT_SIZE;
|
||||
const outputOffset = frame * HOP_SIZE;
|
||||
for (let cursor = 0; cursor < FFT_SIZE; cursor += 1) {
|
||||
output[outputOffset + cursor] += samples[sourceOffset + cursor] * hannWindow[cursor];
|
||||
}
|
||||
}
|
||||
return output;
|
||||
}
|
||||
|
||||
async function istft(runtime, tensor) {
|
||||
const padded = runtime.pad(tensor, [[0, 0], [0, 0], [0, 1], [0, 0]]);
|
||||
const shaped = padded.reshape([2, 2, 3073, 32]).transpose([0, 2, 3, 1]);
|
||||
const left = shaped.slice([0, 0, 0, 0], [1, -1, -1, -1]).squeeze([0]);
|
||||
const right = shaped.slice([1, 0, 0, 0], [1, -1, -1, -1]).squeeze([0]);
|
||||
const result = await Promise.all([inverseChannel(runtime, left), inverseChannel(runtime, right)]);
|
||||
runtime.dispose([padded, shaped, left, right]);
|
||||
return result;
|
||||
}
|
||||
|
||||
function wavBuffer(channels, sampleRate) {
|
||||
const frames = channels[0].length;
|
||||
const buffer = new ArrayBuffer(44 + frames * 4);
|
||||
const view = new DataView(buffer);
|
||||
const text = (offset, value) => [...value].forEach((char, index) => view.setUint8(offset + index, char.charCodeAt(0)));
|
||||
text(0, "RIFF"); view.setUint32(4, buffer.byteLength - 8, true); text(8, "WAVE"); text(12, "fmt ");
|
||||
view.setUint32(16, 16, true); view.setUint16(20, 1, true); view.setUint16(22, 2, true);
|
||||
view.setUint32(24, sampleRate, true); view.setUint32(28, sampleRate * 4, true); view.setUint16(32, 4, true); view.setUint16(34, 16, true);
|
||||
text(36, "data"); view.setUint32(40, frames * 4, true);
|
||||
let offset = 44;
|
||||
for (let frame = 0; frame < frames; frame += 1) for (let channel = 0; channel < 2; channel += 1) {
|
||||
const sample = Math.max(-1, Math.min(1, channels[channel][frame] || 0));
|
||||
view.setInt16(offset, sample < 0 ? sample * 0x8000 : sample * 0x7fff, true); offset += 2;
|
||||
}
|
||||
return buffer;
|
||||
}
|
||||
|
||||
async function separate(requestId, left, right, sampleRate) {
|
||||
const runtime = await getRuntime();
|
||||
postProgress(requestId, 4, { key: "vocalSeparationLoadingModel" });
|
||||
const model = await getModel(runtime);
|
||||
postProgress(requestId, 7, { key: activeBackend === "webgpu" ? "vocalSeparationUsingWebGpu" : "vocalSeparationUsingWebGl" });
|
||||
const accompaniment = [[], []];
|
||||
const vocals = [[], []];
|
||||
const chunks = Math.ceil(left.length / CHUNK_SIZE);
|
||||
for (let index = 0; index < chunks; index += 1) {
|
||||
const start = index * CHUNK_SIZE;
|
||||
const valid = Math.min(CHUNK_SIZE, left.length - start);
|
||||
const paddedLeft = new Float32Array(CHUNK_SIZE + PADDING * 2); paddedLeft.set(left.subarray(start, start + valid), PADDING);
|
||||
const paddedRight = new Float32Array(CHUNK_SIZE + PADDING * 2); paddedRight.set(right.subarray(start, start + valid), PADDING);
|
||||
const input = runtime.tidy(() => {
|
||||
const l = stft(runtime, runtime.tensor1d(paddedLeft));
|
||||
const r = stft(runtime, runtime.tensor1d(paddedRight));
|
||||
return runtime.stack([l, r], 3).transpose([0, 3, 1, 2]).reshape([1, 4, 3072, 32]);
|
||||
});
|
||||
const musicTensor = model.predict(input);
|
||||
const vocalTensor = runtime.sub(input, musicTensor);
|
||||
const [music, voice] = await Promise.all([istft(runtime, musicTensor), istft(runtime, vocalTensor)]);
|
||||
for (let channel = 0; channel < 2; channel += 1) {
|
||||
accompaniment[channel].push(...music[channel].slice(PADDING, PADDING + valid));
|
||||
vocals[channel].push(...voice[channel].slice(PADDING, PADDING + valid));
|
||||
}
|
||||
runtime.dispose([input, musicTensor, vocalTensor]);
|
||||
postProgress(requestId, 8 + Math.round(((index + 1) / chunks) * 88), {
|
||||
key: "vocalSeparationProcessingChunk", current: index + 1, total: chunks,
|
||||
});
|
||||
}
|
||||
const vocalsBuffer = wavBuffer(vocals, sampleRate);
|
||||
const accompanimentBuffer = wavBuffer(accompaniment, sampleRate);
|
||||
self.postMessage(
|
||||
{ type: "result", requestId, vocalsBuffer, accompanimentBuffer, backend: activeBackend },
|
||||
[vocalsBuffer, accompanimentBuffer],
|
||||
);
|
||||
}
|
||||
|
||||
self.addEventListener("message", async ({ data }) => {
|
||||
if (data?.type !== "separate") return;
|
||||
try {
|
||||
await separate(
|
||||
data.requestId,
|
||||
new Float32Array(data.leftBuffer),
|
||||
new Float32Array(data.rightBuffer),
|
||||
data.sampleRate,
|
||||
);
|
||||
} catch (error) {
|
||||
self.postMessage({ type: "error", requestId: data.requestId, error: error?.message || "VOCAL_MODEL_FAILED" });
|
||||
}
|
||||
});
|
||||
@@ -14,4 +14,10 @@ export default defineConfig({
|
||||
},
|
||||
},
|
||||
plugins: [react()],
|
||||
test: {
|
||||
coverage: {
|
||||
include: ["src/**/*.{js,jsx,ts,tsx}"],
|
||||
exclude: ["src/**/*.test.*", "src/**/__fixtures__/**"],
|
||||
},
|
||||
},
|
||||
});
|
||||
|
||||
Reference in New Issue
Block a user