main.py

from fastapi import FastAPI, WebSocket
from fastapi.responses import FileResponse
from fastapi.staticfiles import StaticFiles
from transformers import pipeline
from kokoro import KPipeline
import asyncio
import numpy as np

app = FastAPI()
app.mount("/static", StaticFiles(directory="static"), name="static")

# Initialize pipelines
llm = pipeline("text-generation", model="HuggingFaceTB/SmolLM2-360M-Instruct")
tts = KPipeline(lang_code='a', model='shahid202/Kokoro-82M-TTS')

@app.get("/")
async def get_index():
    return FileResponse('static/index.html')

@app.websocket("/ws/chat")
async def websocket_endpoint(websocket: WebSocket):
    await websocket.accept()
    while True:
        try:
            user_msg = await websocket.receive_text()
            # Generate LLM response
            output = llm(f"User: {user_msg}\nBella:", max_new_tokens=30, clean_up_tokenization_spaces=False)
            full_text = output[0]['generated_text']
            text = full_text.split("Bella:")[-1].strip()
            
            # Stream audio chunks
            for _, _, audio in tts(text, voice="af_heart", speed=1.0):
                await websocket.send_bytes(audio.astype(np.float32).tobytes())
                await asyncio.sleep(0.01) # Small buffer delay for stability
        except Exception as e:
            print(f"Error in WebSocket: {e}")
            break