app.py

from fastapi import FastAPI, Request, WebSocket, WebSocketDisconnect
from fastapi.responses import StreamingResponse, JSONResponse
import outetts
import io
import json
import base64
# Initialize the interface
interface = outetts.Interface(
    config=outetts.ModelConfig.auto_config(
        model=outetts.Models.VERSION_1_0_SIZE_1B,
        # For llama.cpp backend
        #backend=outetts.Backend.LLAMACPP,
        #quantization=outetts.LlamaCppQuantization.FP16
        # For transformers backend
        backend=outetts.Backend.HF,
    )
)

# Load the default speaker profile
speaker = interface.load_default_speaker("EN-FEMALE-1-NEUTRAL")

app = FastAPI()

@app.get("/")
def greet_json():
    return {"Hello": "World!"}

@app.websocket("/ws/tts")
async def websocket_tts(websocket: WebSocket):
    await websocket.accept()
    try:
        while True:
            # Empfange Text-Chunk vom Client
            data = await websocket.receive_text()
            # Status: Warming up
            await websocket.send_text(json.dumps({"generation_status": "Warming up TTS model"}))
            output = interface.generate(
                config=outetts.GenerationConfig(
                    text=data,
                    generation_type=outetts.GenerationType.CHUNKED,
                    speaker=speaker,
                    sampler_config=outetts.SamplerConfig(
                        temperature=0.4
                    ),
                )
            )
            # Status: Generating linguistic features
            await websocket.send_text(json.dumps({"generation_status": "Generating linguistic features"}))
            # Stream audio chunks
            for chunk in output.stream(chunk_size=4096):
                audio_b64 = base64.b64encode(chunk).decode("ascii")
                await websocket.send_text(json.dumps({
                    "data": {
                        "audio_bytes": audio_b64,
                        "duration": None,
                        "request_finished": False
                    }
                }))
            # Final event
            await websocket.send_text(json.dumps({
                "data": {
                    "audio_bytes": "",
                    "duration": None,
                    "request_finished": True
                }
            }))
    except WebSocketDisconnect:
        pass
    except Exception as e:
        await websocket.send_text(json.dumps({"error": str(e)}))