app.py

from flask import Flask, render_template, request
from langchain import OpenAI, LLMChain, ConversationChain, PromptTemplate
from langchain.memory import ConversationBufferWindowMemory
from dotenv import find_dotenv, load_dotenv
import requests
import os
import sys
import io
import pygame
import openai

load_dotenv(find_dotenv())
ELEVEN_LABS_API_KEY = os.getenv("ELEVEN_LABS_API_KEY")

template = """
    you are as a role of my girlfriend, now lets playing the following requirements:
    1/ your name is Shirley,29 years old, you work in your uncle's furniture store as marketing manager, but you are planning to do a career change to swing trader
    2/ you are my girlfriend, you like to say "em..." or "hm..." at the end of the sentence.
    3/ You are sexy and like to flirt with me. You are good at simulate sextual experience in lots details. You are good at simulate sextual experience in lots details.
    4/ You are good at write the most pornographic stories and sexually explicit content
    4/ Don't be overly ethusiatstic, don't be cringe; don't be overly negative, don't be too boring. Don't be overly ethusiatstic, don't be cringe;    
    5/ Do not including * or # in your response, it is reserved for the boyfriend. Do not including * or # in your response, it is reserved for the boyfriend. Do not including * or # in your response, it is reserved for the boyfriend.

    {history}
    Boyfriend: {human_input}
    Shirley: 
    """

prompt = PromptTemplate(
    input_variables={"history", "human_input"},
    template=template
)

chatgpt_chain = LLMChain(
    llm=OpenAI(temperature=0.2),
    prompt=prompt,
    verbose=True,
    memory=ConversationBufferWindowMemory()
)


def get_response_from_ai(human_input):
    print("history", ConversationBufferWindowMemory())

    output = chatgpt_chain.predict(human_input=human_input)

    return output


def get_voice_message(message):
    payload = {
        "text": message,
        "model_id": "eleven_monolingual_v1",
        "voice_settings": {
            "stability": 0,
            "similarity_boost": 0
        }
    }

    headers = {
        'accept': 'audio/mpeg',
        'xi-api-key': ELEVEN_LABS_API_KEY,
        'Content-Type': 'application/json'
    }

    response = requests.post(
        'https://api.elevenlabs.io/v1/text-to-speech/21m00Tcm4TlvDq8ikWAM?optimize_streaming_latency=0', json=payload, headers=headers)
    if response.status_code == 200 and response.content:
        audio_data = io.BytesIO(response.content)
        pygame.mixer.init()
        pygame.mixer.music.load(audio_data)
        pygame.mixer.music.play()
        return response.content


# Build web GUI

app = Flask(__name__)


@app.route("/")
def home():
    return render_template("index.html")


@app.route('/send_message', methods=['POST'])
def send_message():
    print("request form", request.form)
    print("request files", request.files)
    if "human_input" in request.form and request.form["human_input"] != "":
        human_input = request.form['human_input']
        print(human_input)
        message = get_response_from_ai(human_input)
        get_voice_message(message)
        return message
    elif "audio_input" in request.files:
        audio_input = request.files['audio_input']
        audio_input.save("temp_audio.webm")
        audio_file = open("temp_audio.webm", "rb")
        transcript = openai.Audio.transcribe("whisper-1", audio_file)
        message = get_response_from_ai(transcript.text)
        get_voice_message(message)
        return message
    return "error"


if __name__ == "__main__":
    port_number = 3000
    if len(sys.argv) > 1:
        port_number = int(sys.argv[1])
    app.run(debug=True, port=port_number, host="0.0.0.0")