Implement WebGPU detection and integrate Kokoro TTS worker; update TypeScript config and add accordion component

This commit is contained in:
Jack Merrill 2025-04-08 13:22:04 -04:00
parent 5c303b594b
commit 73a0fba45e
Signed by: jack
GPG Key ID: F6BFCA1B80EA6AF7
11 changed files with 741 additions and 373 deletions

View File

@ -1,4 +1,5 @@
import { AppSidebar } from "@/components/app-sidebar";
import KokoroReader from "@/components/KokoroReader";
import { NavActions } from "@/components/nav-actions";
import {
Breadcrumb,
@ -6,6 +7,12 @@ import {
BreadcrumbList,
BreadcrumbPage,
} from "@/components/ui/breadcrumb";
import { Button } from "@/components/ui/button";
import {
Popover,
PopoverContent,
PopoverTrigger,
} from "@/components/ui/popover";
import { Separator } from "@/components/ui/separator";
import {
SidebarInset,
@ -13,6 +20,7 @@ import {
SidebarTrigger,
} from "@/components/ui/sidebar";
import { createClient } from "@/utils/supabase/server";
import { Speech } from "lucide-react";
import { redirect } from "next/navigation";
import { remark } from "remark";
import remarkHtml from "remark-html";
@ -96,7 +104,7 @@ export default async function DocumentPage({
</Breadcrumb>
</div>
<div className="ml-auto px-3">
<NavActions />
<NavActions pages={pages} />
</div>
</header>
<div
@ -104,6 +112,7 @@ export default async function DocumentPage({
text-white
prose-h1:font-semibold prose-h1:text-2xl prose-h1:mb-4 prose-h1:text-white
prose-h2:font-medium prose-h2:text-xl prose-h2:mb-3 prose-h2:text-white
prose-h4:font-medium prose-h4:text-lg prose-h4:mb-2 prose-h4:text-gray-300
prose-a:text-blue-400 hover:prose-a:underline
prose-p:leading-7 prose-p:text-gray-200
prose-blockquote:italic prose-blockquote:border-l-4 prose-blockquote:pl-4 prose-blockquote:border-gray-600 prose-blockquote:text-gray-300

View File

@ -26,9 +26,33 @@ export default async function Page() {
return redirect("/login");
}
const { data: documents, error } = await supabase
.from("documents")
.select("*")
.eq("owner", user.id)
.order("created_at", { ascending: false });
if (error) {
console.error("Failed to fetch documents:", error);
// Optionally handle the error, e.g., show a message to the user
return (
<div className="p-4">
<p className="text-red-600">Failed to load documents.</p>
</div>
);
}
return (
<SidebarProvider>
<AppSidebar />
<AppSidebar
documents={documents.map((d) => {
return {
name: d.file_name,
url: `/dashboard/documents/${d.id}`,
emoji: "📄",
};
})}
/>
<SidebarInset>
<header className="flex h-14 shrink-0 items-center gap-2">
<div className="flex flex-1 items-center gap-2 px-3">
@ -40,8 +64,8 @@ export default async function Page() {
<Breadcrumb>
<BreadcrumbList>
<BreadcrumbItem>
<BreadcrumbPage className="line-clamp-1">
Project Management & Task Tracking
<BreadcrumbPage className="line-clamp-1 text-muted-foreground">
Select a document...
</BreadcrumbPage>
</BreadcrumbItem>
</BreadcrumbList>

BIN
bun.lockb

Binary file not shown.

175
components/KokoroReader.tsx Normal file
View File

@ -0,0 +1,175 @@
"use client";
import { useRef, useState, useEffect } from "react";
import { Button } from "./ui/button";
import { Play } from "lucide-react";
import {
Accordion,
AccordionContent,
AccordionItem,
AccordionTrigger,
} from "./ui/accordion";
import { Label } from "./ui/label";
export default function KokoroReader({ pages }: { pages: any[] }) {
// Create a reference to the worker object.
const worker = useRef<Worker>(null);
const [inputText, setInputText] = useState(
"Life is like a box of chocolates. You never know what you're gonna get."
);
const [selectedSpeaker, setSelectedSpeaker] = useState("af_heart");
const [voices, setVoices] = useState<any[]>([]);
const [status, setStatus] = useState<"ready" | "running" | null>(null);
const [error, setError] = useState(null);
const [loadingMessage, setLoadingMessage] = useState("Loading...");
const [results, setResults] = useState<{ text: string; src: string }[]>([]);
// We use the `useEffect` hook to setup the worker as soon as the `App` component is mounted.
useEffect(() => {
// Create the worker if it does not yet exist.
console.log("Initializing worker...");
worker.current ??= new Worker("/workers/kokoro-worker.js", {
type: "module",
});
console.log("Worker initialized");
// Create a callback function for messages from the worker thread.
const onMessageReceived = (e: any) => {
switch (e.data.status) {
case "device":
setLoadingMessage(`Loading model (device="${e.data.device}")`);
break;
case "ready":
setStatus("ready");
setVoices(e.data.voices);
break;
case "error":
setError(e.data.data);
break;
case "complete":
const { audio, text } = e.data;
// Generation complete: re-enable the "Generate" button
setResults((prev) => [{ text, src: audio }, ...prev]);
setStatus("ready");
break;
}
};
console.log("onmessagereceived");
const onErrorReceived = (e: any) => {
console.error("Worker error:", e);
setError(e.message);
};
console.log("Attaching event listeners to worker");
// Attach the callback function as an event listener.
worker.current.addEventListener("message", onMessageReceived);
worker.current.addEventListener("error", onErrorReceived);
console.log(worker.current);
// Define a cleanup function for when the component is unmounted.
return () => {
worker.current!.removeEventListener("message", onMessageReceived);
worker.current!.removeEventListener("error", onErrorReceived);
};
}, []);
const handleSubmit = (e: any) => {
e.preventDefault();
setStatus("running");
worker.current!.postMessage({
type: "generate",
text: inputText.trim(),
voice: selectedSpeaker,
});
};
return (
<div className="flex flex-col items-center justify-center pt-4 relative overflow-hidden font-sans">
<div className="max-w-3xl w-full relative z-[2]">
<div className="items-center justify-center text-center">
<Button variant="ghost" size="icon" className="h-10 w-10">
<Play />
</Button>
</div>
<Accordion type="single" collapsible>
<AccordionItem value="item-1">
<AccordionTrigger className="text-white pb-2">
Settings
</AccordionTrigger>
<AccordionContent className="pb-2">
<Label>Voice</Label>
<select
value={selectedSpeaker}
onChange={(e) => setSelectedSpeaker(e.target.value)}
className="w-full bg-gray-700/50 backdrop-blur-sm border-2 border-gray-600 rounded-md text-gray-100 px-3 py-2 focus:outline-none focus:ring-2 focus:ring-blue-500 focus:border-transparent"
>
{Object.entries(voices).map(([id, voice]) => (
<option key={id} value={id}>
{voice.name} (
{voice.language === "en-us" ? "American" : "British"}{" "}
{voice.gender})
</option>
))}
</select>
</AccordionContent>
</AccordionItem>
</Accordion>
</div>
{/* <div className="bg-gray-800/50 backdrop-blur-sm border border-gray-700 rounded-lg p-6">
<form onSubmit={handleSubmit} className="space-y-4">
<textarea
placeholder="Enter text..."
value={inputText}
onChange={(e) => setInputText(e.target.value)}
className="w-full min-h-[100px] max-h-[300px] bg-gray-700/50 backdrop-blur-sm border-2 border-gray-600 rounded-xl resize-y text-gray-100 placeholder-gray-400 px-3 py-2 focus:outline-none focus:ring-2 focus:ring-blue-500 focus:border-transparent"
rows={Math.min(8, inputText.split("\n").length)}
/>
<div className="flex flex-col items-center space-y-4">
<select
value={selectedSpeaker}
onChange={(e) => setSelectedSpeaker(e.target.value)}
className="w-full bg-gray-700/50 backdrop-blur-sm border-2 border-gray-600 rounded-xl text-gray-100 px-3 py-2 focus:outline-none focus:ring-2 focus:ring-blue-500 focus:border-transparent"
>
{Object.entries(voices).map(([id, voice]) => (
<option key={id} value={id}>
{voice.name} (
{voice.language === "en-us" ? "American" : "British"}{" "}
{voice.gender})
</option>
))}
</select>
<button
type="submit"
className="inline-flex justify-center items-center px-6 py-2 text-lg font-semibold bg-gradient-to-t from-blue-600 to-purple-600 hover:from-blue-700 hover:to-purple-700 transition-colors duration-300 rounded-xl text-white disabled:opacity-50"
disabled={status === "running" || inputText.trim() === ""}
>
{status === "running" ? "Generating..." : "Generate"}
</button>
</div>
</form>
</div> */}
{results.map((result, i) => (
<div key={i}>
<div className="text-white bg-gray-800/70 backdrop-blur-sm border border-gray-700 rounded-lg p-4 z-10">
<span className="absolute right-5 font-bold">
#{results.length - i}
</span>
<p className="mb-3 max-w-[95%]">{result.text}</p>
<audio controls src={result.src} className="w-full">
Your browser does not support the audio element.
</audio>
</div>
</div>
))}
</div>
);
}

View File

@ -1,6 +1,6 @@
"use client"
"use client";
import * as React from "react"
import * as React from "react";
import {
ArrowDown,
ArrowUp,
@ -14,17 +14,18 @@ import {
Link,
MoreHorizontal,
Settings2,
Speech,
Star,
Trash,
Trash2,
} from "lucide-react"
} from "lucide-react";
import { Button } from "@/components/ui/button"
import { Button } from "@/components/ui/button";
import {
Popover,
PopoverContent,
PopoverTrigger,
} from "@/components/ui/popover"
} from "@/components/ui/popover";
import {
Sidebar,
SidebarContent,
@ -33,7 +34,8 @@ import {
SidebarMenu,
SidebarMenuButton,
SidebarMenuItem,
} from "@/components/ui/sidebar"
} from "@/components/ui/sidebar";
import KokoroReader from "./KokoroReader";
const data = [
[
@ -96,20 +98,30 @@ const data = [
icon: ArrowDown,
},
],
]
];
export function NavActions() {
const [isOpen, setIsOpen] = React.useState(false)
export function NavActions({ pages }: { pages: any[] }) {
const [isOpen, setIsOpen] = React.useState(false);
React.useEffect(() => {
setIsOpen(true)
}, [])
setIsOpen(true);
}, []);
return (
<div className="flex items-center gap-2 text-sm">
<div className="text-muted-foreground hidden font-medium md:inline-block">
{/* <div className="text-muted-foreground hidden font-medium md:inline-block">
Edit Oct 08
</div>
</div> */}
<Popover>
<PopoverTrigger asChild>
<Button variant="ghost" size="icon" className="h-7 w-7">
<Speech />
</Button>
</PopoverTrigger>
<PopoverContent>
<KokoroReader pages={pages} />
</PopoverContent>
</Popover>
<Button variant="ghost" size="icon" className="h-7 w-7">
<Star />
</Button>
@ -149,5 +161,5 @@ export function NavActions() {
</PopoverContent>
</Popover>
</div>
)
);
}

View File

@ -0,0 +1,66 @@
"use client"
import * as React from "react"
import * as AccordionPrimitive from "@radix-ui/react-accordion"
import { ChevronDownIcon } from "lucide-react"
import { cn } from "@/lib/utils"
function Accordion({
...props
}: React.ComponentProps<typeof AccordionPrimitive.Root>) {
return <AccordionPrimitive.Root data-slot="accordion" {...props} />
}
function AccordionItem({
className,
...props
}: React.ComponentProps<typeof AccordionPrimitive.Item>) {
return (
<AccordionPrimitive.Item
data-slot="accordion-item"
className={cn("border-b last:border-b-0", className)}
{...props}
/>
)
}
function AccordionTrigger({
className,
children,
...props
}: React.ComponentProps<typeof AccordionPrimitive.Trigger>) {
return (
<AccordionPrimitive.Header className="flex">
<AccordionPrimitive.Trigger
data-slot="accordion-trigger"
className={cn(
"focus-visible:border-ring focus-visible:ring-ring/50 flex flex-1 items-start justify-between gap-4 rounded-md py-4 text-left text-sm font-medium transition-all outline-none hover:underline focus-visible:ring-[3px] disabled:pointer-events-none disabled:opacity-50 [&[data-state=open]>svg]:rotate-180",
className
)}
{...props}
>
{children}
<ChevronDownIcon className="text-muted-foreground pointer-events-none size-4 shrink-0 translate-y-0.5 transition-transform duration-200" />
</AccordionPrimitive.Trigger>
</AccordionPrimitive.Header>
)
}
function AccordionContent({
className,
children,
...props
}: React.ComponentProps<typeof AccordionPrimitive.Content>) {
return (
<AccordionPrimitive.Content
data-slot="accordion-content"
className="data-[state=closed]:animate-accordion-up data-[state=open]:animate-accordion-down overflow-hidden text-sm"
{...props}
>
<div className={cn("pt-0 pb-4", className)}>{children}</div>
</AccordionPrimitive.Content>
)
}
export { Accordion, AccordionItem, AccordionTrigger, AccordionContent }

View File

@ -4,3 +4,12 @@ import { twMerge } from "tailwind-merge";
export function cn(...inputs: ClassValue[]) {
return twMerge(clsx(inputs));
}
export async function detectWebGPU() {
try {
const adapter = await navigator.gpu.requestAdapter();
return !!adapter;
} catch (e) {
return false;
}
}

View File

@ -2,6 +2,20 @@ import type { NextConfig } from "next";
const nextConfig: NextConfig = {
/* config options here */
webpack: (config, { isServer }) => {
if (!isServer) {
config.module.rules.push({
test: /kokoro-worker\.js$/,
use: { loader: "worker-loader" },
});
}
config.module.rules.push({
test: /\.js$/,
loader: "@open-wc/webpack-import-meta-loader",
});
return config;
},
};
export default nextConfig;

View File

@ -9,6 +9,7 @@
"dependencies": {
"@ai-sdk/mistral": "^1.2.3",
"@mistralai/mistralai": "^1.5.2",
"@radix-ui/react-accordion": "^1.2.3",
"@radix-ui/react-checkbox": "^1.1.1",
"@radix-ui/react-collapsible": "^1.1.3",
"@radix-ui/react-dialog": "^1.1.6",
@ -26,6 +27,7 @@
"autoprefixer": "10.4.20",
"class-variance-authority": "^0.7.1",
"clsx": "^2.1.1",
"kokoro-js": "^1.2.0",
"lucide-react": "^0.486.0",
"next": "latest",
"next-themes": "^0.4.3",
@ -38,6 +40,7 @@
"zod": "^3.24.2"
},
"devDependencies": {
"@open-wc/webpack-import-meta-loader": "^0.4.7",
"@types/node": "22.10.2",
"@types/react": "^19.0.2",
"@types/react-dom": "19.0.2",
@ -46,6 +49,7 @@
"tailwind-merge": "^2.5.2",
"tailwindcss": "^4.1.0",
"tailwindcss-animate": "^1.0.7",
"typescript": "5.7.2"
"typescript": "5.7.2",
"worker-loader": "^3.0.8"
}
}

View File

@ -0,0 +1,49 @@
console.log("Initializing Kokoro TTS Worker");
import { KokoroTTS } from "https://cdn.jsdelivr.net/npm/kokoro-js@1.2.0/+esm";
async function detectWebGPU() {
try {
const adapter = await navigator.gpu.requestAdapter();
return !!adapter;
} catch (e) {
return false;
}
}
// Device detection
const device = (await detectWebGPU()) ? "webgpu" : "wasm";
self.postMessage({ status: "device", device });
console.log(`Detected device: ${device}`);
// Load the model
const model_id = "onnx-community/Kokoro-82M-v1.0-ONNX";
const tts = await KokoroTTS.from_pretrained(model_id, {
dtype: device === "wasm" ? "q8" : "fp32",
device,
});
console.log("Kokoro TTS model loaded successfully");
self.postMessage({ status: "ready", voices: tts.voices, device });
console.log("Available voices:", tts.voices);
// Listen for messages from the main thread
self.addEventListener("message", async (e) => {
const { text, voice } = e.data;
try {
// Generate speech
const audio = await tts.generate(text, { voice });
// Send the audio file back to the main thread
const blob = audio.toBlob();
self.postMessage({
status: "complete",
audio: URL.createObjectURL(blob),
text,
});
} catch (error) {
self.postMessage({ status: "error", error: error.message });
}
});

View File

@ -1,6 +1,6 @@
{
"compilerOptions": {
"target": "es5",
"target": "ESNext",
"lib": ["dom", "dom.iterable", "esnext"],
"allowJs": true,
"skipLibCheck": true,
@ -23,6 +23,12 @@
"@/*": ["./*"]
}
},
"include": ["next-env.d.ts", "**/*.ts", "**/*.tsx", ".next/types/**/*.ts"],
"include": [
"next-env.d.ts",
"**/*.ts",
"**/*.tsx",
".next/types/**/*.ts",
"public/kokoro-worker.js"
],
"exclude": ["node_modules"]
}