import streamlit as st
import whisper
import tempfile
from pydub import AudioSegment

# Define available models
available_models = ["tiny", "base", "small", "medium", "large"]

st.title("Voice Recognition App")

st.write("Upload an audio file and choose a Whisper model to transcribe it to text.")

# Model selection dropdown
model_choice = st.selectbox("Choose a Whisper model", available_models)

# Load the selected Whisper model
st.write(f"Loading {model_choice} model...")
model = whisper.load_model(model_choice)
st.write(f"{model_choice} model loaded successfully.")
# File uploader for audio file
uploaded_file = st.file_uploader("Choose an audio file", type=["wav", "mp3", "m4a"])

if uploaded_file is not None:
    # Save the uploaded file temporarily
    with tempfile.NamedTemporaryFile(delete=False) as temp_file:
        temp_file.write(uploaded_file.read())
        temp_file_path = temp_file.name

    # Convert audio file to a format supported by Whisper (if necessary)
    audio = AudioSegment.from_file(temp_file_path)
    temp_wav_path = tempfile.mktemp(suffix=".wav")
    audio.export(temp_wav_path, format="wav")

    st.audio(uploaded_file, format="audio/wav")

    st.write("Transcribing audio...")

    # Transcribe audio using Whisper model
    result = model.transcribe(temp_wav_path)
    st.write("Transcription:")
    st.write(result["text"])