app.py

import os
import whisper
from gtts import gTTS
import gradio as gr
from groq import Groq

# Load the Whisper model for speech-to-text
whisper_model = whisper.load_model("large")

# Initialize Groq client for text generation
GROQ_API_KEY = "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"
client = Groq(api_key=GROQ_API_KEY)

# Function to convert text to speech using gTTS and return the audio path
def text_to_speech(text):
    tts = gTTS(text=text, lang='en')
    response_audio_path = "response.mp3"
    tts.save(response_audio_path)
    return response_audio_path

# Function to transcribe audio to text using Whisper
def transcribe_audio(audio):
    print("Transcribing audio...")
    result = whisper_model.transcribe(audio)
    return result["text"]

# Function to get a response from the Groq API using LLaMA model
def get_response_from_groq(input_text):
    chat_completion = client.chat.completions.create(
        messages=[{"role": "user", "content": input_text}],
        model="llama3-8b-8192",
    )
    response = chat_completion.choices[0].message.content
    return response

# Gradio function to handle chatbot interaction
def chatbot(audio):
    # Step 1: Transcribe audio to text using Whisper
    user_input = transcribe_audio(audio)
    print(f"User said: {user_input}")
    
    # Step 2: Get chatbot response from Groq (LLaMA model)
    response = get_response_from_groq(user_input)
    print(f"Chatbot response: {response}")
    
    # Step 3: Convert response text to speech and return audio
    response_audio_path = text_to_speech(response)
    
    # Ensure the audio file exists and return text + audio response
    if os.path.exists(response_audio_path):
        return response, response_audio_path
    else:
        return response, None  # Fallback in case the audio file wasn't generated

# Gradio interface for the chatbot
interface = gr.Interface(
    fn=chatbot,
    inputs=gr.Audio(type="filepath"),  # Use microphone input without the 'source' argument
    outputs=["text", gr.Audio(type="filepath")],  # Text and audio output
    live=True,
    title="Voice-Enabled Chatbot",
    description="Speak into your microphone, and the chatbot will respond with both text and audio."
)

# Launch Gradio interface
interface.launch(share=True)