app.py

from flask import Flask, request, jsonify, send_file
from flask_cors import CORS
from faster_whisper import WhisperModel
from transformers import pipeline
from TTS.api import TTS
import tempfile
import os

app = Flask(__name__)
CORS(app)

# Load models
whisper_model = WhisperModel("small", device="cpu", compute_type="int8")
llm = pipeline("text-generation", model="tiiuae/falcon-rw-1b", max_new_tokens=100)
tts = TTS(model_name="tts_models/en/ljspeech/tacotron2-DDC", progress_bar=False, gpu=False)

@app.route("/talk", methods=["POST"])
def talk():
    if "audio" not in request.files:
        return jsonify({"error": "No audio file"}), 400

    # Save audio
    audio_file = request.files["audio"]
    with tempfile.NamedTemporaryFile(delete=False, suffix=".wav") as tmp:
        audio_path = tmp.name
        audio_file.save(audio_path)

    # Transcribe
    segments, _ = whisper_model.transcribe(audio_path)
    transcription = "".join([seg.text for seg in segments])

    # Generate response
    response = llm(transcription)[0]["generated_text"]

    # Synthesize speech
    tts_audio_path = audio_path.replace(".wav", "_reply.wav")
    tts.tts_to_file(text=response, file_path=tts_audio_path)

    return send_file(tts_audio_path, mimetype="audio/wav")

@app.route("/")
def index():
    return "Metaverse AI Character API running."

if __name__ == "__main__":
    app.run(host="0.0.0.0", port=7860)