app.py

import gradio as gr
from langchain_groq import ChatGroq
from langchain.schema import SystemMessage, HumanMessage
import requests
import tempfile
import time

# Configuration of the Groq model
groq_api_key = "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"  # Replace with your actual Groq API key
llm = ChatGroq(api_key=groq_api_key, model_name="llama3-70b-8192")  # Corrected model name

# ElevenLabs API key and voice ID
XI_API_KEY = "sk_b254c267851485b60d23fb2e15fa8fde9f5fbc0d835127e2"  # Replace with your ElevenLabs API key
VOICE_ID = "iYwRDEf2D1WyqRRecXPA"  # Replace with your voice ID

def translate_and_speak(user_input, target_language):
    try:
        start_time = time.time()  # Start total processing time

        # Generate translation using Groq model
        translation_start = time.time()
        system_prompt = f"You are expected to translate the user input exclusively into {target_language} without adding anything else."
        messages = [
            SystemMessage(content=system_prompt),
            HumanMessage(content=user_input)
        ]
        response = llm.invoke(messages)
        translation_end = time.time()

        # Check if the response is valid
        if not response or not hasattr(response, 'content'):
            raise ValueError("Invalid response from the translation model.")

        generated_text = response.content.strip()

        # Use ElevenLabs API to generate speech
        tts_start = time.time()
        url = f"https://api.elevenlabs.io/v1/text-to-speech/{VOICE_ID}"
        headers = {
            "Accept": "audio/mpeg",
            "Content-Type": "application/json",
            "xi-api-key": XI_API_KEY
        }
        data = {
            "text": generated_text,
            "model_id": "eleven_multilingual_v2",
            "voice_settings": {
                "stability": 0.75,
                "similarity_boost": 0.75
            }
        }
        tts_response = requests.post(url, json=data, headers=headers)
        tts_end = time.time()

        if tts_response.status_code == 200:
            # Save audio to a temporary file
            with tempfile.NamedTemporaryFile(delete=False, suffix=".mp3") as fp:
                fp.write(tts_response.content)
                audio_file = fp.name
            end_time = time.time()

            # Calculate processing times
            translation_time = translation_end - translation_start
            tts_time = tts_end - tts_start
            total_time = end_time - start_time

            # Prepare timings information
            timings_info = f"Translation time: {translation_time:.2f} seconds\n"
            timings_info += f"Text-to-Speech time: {tts_time:.2f} seconds\n"
            timings_info += f"Total processing time: {total_time:.2f} seconds"

            return generated_text, audio_file, timings_info
        else:
            error_message = f"Text-to-Speech API Error: {tts_response.status_code} - {tts_response.text}"
            return error_message, None, None
    except Exception as e:
        # Return the exception message
        error_details = f"An error occurred: {str(e)}"
        return error_details, None, None

# Create Gradio interface
iface = gr.Interface(
    fn=translate_and_speak,
    inputs=[
        gr.Textbox(lines=2, placeholder="Enter text to translate...", label="Input Text"),
        gr.Dropdown(
            choices=["Spanish", "French", "German", "Italian", "Chinese", "Japanese"],
            value="Spanish",
            label="Target Language"
        )
    ],
    outputs=[
        gr.Textbox(label="Translated Text"),
        gr.Audio(label="Spoken Audio", autoplay=True),
        gr.Textbox(label="Processing Times")
    ],
    title="Multilingual Text Translator and Speech Synthesizer",
    description="Translate text into the selected language and listen to the spoken audio.",
    allow_flagging="never"
)

# Launch the app
iface.launch()