Labyricorn/voicebox
1
0
Fork 0
mirror of https://github.com/jamiepine/voicebox.git synced 2026-09-16 13:20:39 -07:00
Code Issues Packages Projects Releases Wiki Activity

feat: add LuxTTS as second TTS engine with multi-engine support

Browse Source 
Introduce LuxTTS (ZipVoice) alongside Qwen TTS, enabling users to choose
between engines at generation time. LuxTTS offers fast, English-focused
voice cloning at 48kHz with ~1GB VRAM.

Backend:
- Add LuxTTSBackend with encode_prompt/generate_speech integration
- Multi-engine registry (get_tts_backend_for_engine) replacing singleton
- Engine-prefixed voice prompt cache keys to avoid collisions
- Engine field on GenerationRequest (default 'qwen' for backward compat)
- Engine dispatch in /generate and /generate/stream endpoints
- LuxTTS in model status, download, and delete maps

Frontend:
- TTS Engine selector dropdown in GenerationForm (Qwen TTS / LuxTTS)
- Conditionally hide Model Size and Delivery Instructions for LuxTTS
- Engine field added to TypeScript types and Zod schema
- LuxTTS section in Model Management page
This commit is contained in:
James Pine
2026-03-13 00:21:43 -07:00
parent 6359dee406
commit d46eb5bcc6
10 changed files with 584 additions and 121 deletions
Download Patch File Download Diff File Expand all files Collapse all files
app/src/components/Generation/GenerationForm.tsx
+74 -49
View File
@@ -76,29 +76,56 @@ export function GenerationForm() {
)}
/>
<FormField
control={form.control}
name="instruct"
render={({ field }) => (
<FormItem>
<FormLabel>Delivery Instructions (optional)</FormLabel>
<FormControl>
<Textarea
placeholder="e.g. Speak slowly with emphasis, Warm and friendly tone, Professional and authoritative..."
className="min-h-[80px]"
{...field}
/>
</FormControl>
<FormDescription>
Natural language instructions to control speech delivery (tone, emotion, pace).
Max 500 characters
</FormDescription>
<FormMessage />
</FormItem>
)}
/>
{form.watch('engine') !== 'luxtts' && (
<FormField
control={form.control}
name="instruct"
render={({ field }) => (
<FormItem>
<FormLabel>Delivery Instructions (optional)</FormLabel>
<FormControl>
<Textarea
placeholder="e.g. Speak slowly with emphasis, Warm and friendly tone, Professional and authoritative..."
className="min-h-[80px]"
{...field}
/>
</FormControl>
<FormDescription>
Natural language instructions to control speech delivery (tone, emotion,
pace). Max 500 characters
</FormDescription>
<FormMessage />
</FormItem>
)}
/>
)}
<div className="grid gap-4 md:grid-cols-3">
<FormField
control={form.control}
name="engine"
render={({ field }) => (
<FormItem>
<FormLabel>TTS Engine</FormLabel>
<Select onValueChange={field.onChange} defaultValue={field.value}>
<FormControl>
<SelectTrigger>
<SelectValue />
</SelectTrigger>
</FormControl>
<SelectContent>
<SelectItem value="qwen">Qwen TTS</SelectItem>
<SelectItem value="luxtts">LuxTTS</SelectItem>
</SelectContent>
</Select>
<FormDescription>
{field.value === 'luxtts' ? 'Fast, English-focused' : 'Multi-language'}
</FormDescription>
<FormMessage />
</FormItem>
)}
/>
<FormField
control={form.control}
name="language"
@@ -124,29 +151,6 @@ export function GenerationForm() {
)}
/>
<FormField
control={form.control}
name="modelSize"
render={({ field }) => (
<FormItem>
<FormLabel>Model Size</FormLabel>
<Select onValueChange={field.onChange} defaultValue={field.value}>
<FormControl>
<SelectTrigger>
<SelectValue />
</SelectTrigger>
</FormControl>
<SelectContent>
<SelectItem value="1.7B">Qwen TTS 1.7B (Higher Quality)</SelectItem>
<SelectItem value="0.6B">Qwen TTS 0.6B (Faster)</SelectItem>
</SelectContent>
</Select>
<FormDescription>Larger models produce better quality</FormDescription>
<FormMessage />
</FormItem>
)}
/>
<FormField
control={form.control}
name="seed"
@@ -170,11 +174,32 @@ export function GenerationForm() {
/>
</div>
<Button
type="submit"
className="w-full"
disabled={isPending || !selectedProfileId}
>
{form.watch('engine') !== 'luxtts' && (
<FormField
control={form.control}
name="modelSize"
render={({ field }) => (
<FormItem>
<FormLabel>Model Size</FormLabel>
<Select onValueChange={field.onChange} defaultValue={field.value}>
<FormControl>
<SelectTrigger>
<SelectValue />
</SelectTrigger>
</FormControl>
<SelectContent>
<SelectItem value="1.7B">Qwen TTS 1.7B (Higher Quality)</SelectItem>
<SelectItem value="0.6B">Qwen TTS 0.6B (Faster)</SelectItem>
</SelectContent>
</Select>
<FormDescription>Larger models produce better quality</FormDescription>
<FormMessage />
</FormItem>
)}
/>
)}
<Button type="submit" className="w-full" disabled={isPending || !selectedProfileId}>
{isPending ? (
<>
<Loader2 className="mr-2 h-4 w-4 animate-spin" />