transcription.py

import io
import groq
import numpy as np
import soundfile as sf


def transcribe_audio(audio, api_key):
    if audio is None:
        return ""

    client = groq.Client(api_key=api_key)

    audio_data = audio[1]  # Get the numpy arry from the tuple
    buffer = io.BytesIO()
    sf.write(buffer, audio_data, audio[0], format='wav')
    buffer.seek(0)

    bytes_audio = io.BytesIO()
    np.save(bytes_audio, audio_data)
    bytes_audio.seek(0)

    try:
        # Use Distil-Whisper English powered by Groq for transcription
        completion = client.audio.transcriptions.create(
            model="distil-whisper-large-v3-en",
            file=("audio.wav", buffer),
            response_format="text"
        )
        return completion

    except Exception as e:
        return f"Error in transcription: {e}"


def generate_response(transcription, api_key):
    if not transcription:
        return "No transcription available. Please try speaking again."

    client = groq.Client(api_key=api_key)

    try:
        completion = client.chat.completions.create(
            model="llama3-70b-8192",
            messages=[
                {"role": "system", "content": "You are a helpful assistant."},
                {"role": "user", "content": transcription}
            ]
        )
        return completion.choices[0].message.content

    except Exception as e:
        return f"Error in response generation: {e}"


def process_audio(audio, api_key):
    if not api_key:
        return "Please enter your Groq API key.", "API key is required."
    transcription = transcribe_audio(audio, api_key)
    response = generate_response(transcription, api_key)
    return transcription, response