app.py

import gradio as gr
from huggingface_hub import InferenceClient
import speech_recognition as sr
from gtts import gTTS
import tempfile
import os
import sys

# Check for Hugging Face API token
HF_TOKEN = os.getenv("HF_TOKEN")
if not HF_TOKEN:
    print("Error: HF_TOKEN environment variable not set. Please set it with your Hugging Face API token.")
    sys.exit(1)

# Initialize Hugging Face client
try:
    client = InferenceClient(model="mistralai/Mixtral-8x7B-Instruct-v0.1", token=HF_TOKEN)
    print("Successfully initialized InferenceClient with Mixtral-8x7B-Instruct-v0.1")
except Exception as e:
    print(f"Failed to initialize InferenceClient: {e}")
    sys.exit(1)

# Speech-to-Text Function
def speech_to_text(audio_path):
    if not os.path.exists(audio_path):
        return "Error: Audio file not found."
    recognizer = sr.Recognizer()
    with sr.AudioFile(audio_path) as source:
        audio_data = recognizer.record(source)
    try:
        text = recognizer.recognize_google(audio_data)
        print(f"Speech-to-Text Output: {text}")
        return text
    except sr.UnknownValueError:
        return "Could not understand the audio."
    except sr.RequestError as e:
        return f"Speech recognition service error: {e}"

# Text-to-Speech Function
def text_to_speech(text):
    try:
        print(f"Text-to-Speech Input: {text}")
        tts = gTTS(text)
        temp_file = tempfile.NamedTemporaryFile(delete=False, suffix=".mp3")
        temp_file.close()
        tts.save(temp_file.name)
        print(f"Audio file generated at: {temp_file.name}")
        return temp_file.name
    except Exception as e:
        print(f"Text-to-Speech Error: {e}")
        return None

# Chatbot Response Function
def respond(message, history, system_message, max_tokens, temperature, top_p):
    messages = [{"role": "system", "content": system_message}]
    
    # Build message history
    for user_msg, bot_msg in history:
        if user_msg:
            messages.append({"role": "user", "content": user_msg})
        if bot_msg:
            messages.append({"role": "assistant", "content": bot_msg})
    
    messages.append({"role": "user", "content": message})
    response = ""

    try:
        for message in client.chat_completion(
            messages=messages,
            max_tokens=max_tokens,
            stream=True,
            temperature=temperature,
            top_p=top_p,
        ):
            # Check for valid token content to avoid TypeError
            token_content = message.choices[0].delta.content
            if token_content is not None:
                response += token_content
        if not response:
            return "Error: Empty response from the model."
        print(f"Chatbot Response: {response}")
        return response
    except Exception as e:
        print(f"Chatbot Error: {e}")
        return f"Error generating response from the chatbot: {e}"

# Voice-to-Voice Functionality
def voice_to_voice(audio, history, system_message, max_tokens, temperature, top_p):
    # Convert user voice to text
    user_message = speech_to_text(audio)
    if user_message.startswith("Error") or user_message.startswith("Could not understand"):
        return user_message, history, None

    # Get chatbot response
    response_text = respond(user_message, history, system_message, max_tokens, temperature, top_p)
    if response_text.startswith("Error"):
        return response_text, history, None

    # Update chat history (truncate to last 5 messages)
    history.append((user_message, response_text))
    history = history[-5:]

    # Convert chatbot response to audio
    audio_file = text_to_speech(response_text)
    if not audio_file:
        return "Failed to generate audio response.", history, None

    return response_text, history, audio_file

# Gradio Interface
def main_interface():
    with gr.Blocks() as demo:
        gr.Markdown("# Voice-to-Voice Chatbot")
        system_message = gr.Textbox(
            value="You are a friendly and helpful chatbot.", 
            label="System Message", 
            lines=2
        )
        max_tokens = gr.Slider(
            minimum=1, maximum=2048, value=512, step=1, label="Max New Tokens"
        )
        temperature = gr.Slider(
            minimum=0.1, maximum=4.0, value=0.7, step=0.1, label="Temperature"
        )
        top_p = gr.Slider(
            minimum=0.1, maximum=1.0, value=0.95, step=0.05, label="Top-p (Nucleus Sampling)"
        )
        audio_input = gr.Audio(
            sources=["microphone"], type="filepath", label="Speak Your Question"
        )
        response_output = gr.Textbox(label="Chatbot Response")
        audio_output = gr.Audio(label="Response Audio", type="filepath")
        history_state = gr.State([])

        gr.Button("Submit").click(
            fn=voice_to_voice,
            inputs=[audio_input, history_state, system_message, max_tokens, temperature, top_p],
            outputs=[response_output, history_state, audio_output],
        )

    return demo

if __name__ == "__main__":
    demo = main_interface()
    demo.launch()