import gradio as gr
import requests
import numpy as np
import io
import wave

# Function to send audio to Groq API and get transcription
def transcribe(audio_data):
    # Convert the NumPy audio array to bytes
    audio_bytes = io.BytesIO()
    
    # Convert NumPy array to WAV format (use appropriate rate, channels, etc.)
    with wave.open(audio_bytes, "wb") as wf:
        wf.setnchannels(1)  # Mono channel
        wf.setsampwidth(2)  # 16-bit audio
        wf.setframerate(16000)  # Assuming 16kHz sample rate
        wf.writeframes(audio_data.tobytes())
    
    audio_bytes.seek(0)  # Rewind to the beginning

    # Groq API endpoint for audio transcription
    groq_api_endpoint = "https://api.groq.com/openai/v1/audio/transcriptions"
    
    # Replace 'YOUR_GROQ_API_KEY' with your actual Groq API key
    headers = {
        "Authorization": "Bearer YOUR_GROQ_API_KEY",
    }
    
    # Prepare the files and data for the request
    files = {
        'file': ('audio.wav', audio_bytes, 'audio/wav'),
    }
    data = {
        'model': 'whisper-large-v3-turbo',  # Specify the model to use
        'response_format': 'json',          # Desired response format
        'language': 'en',                   # Language of the audio
    }

    # Send audio to Groq API
    response = requests.post(groq_api_endpoint, headers=headers, files=files, data=data)

    # Parse response
    if response.status_code == 200:
        result = response.json()
        return result.get("text", "No transcription available.")
    else:
        return f"Error: {response.status_code}, {response.text}"

# Gradio interface
iface = gr.Interface(
    fn=transcribe,
    inputs=gr.Audio(source="microphone", type="numpy"),  # Changed to numpy
    outputs="text",
    title="Voice to Text Converter",
    description="Record your voice, and it will be transcribed into text using Groq API."
)

iface.launch()