import streamlit as st
from transformers import pipeline
import requests

# Get the Hugging Face API Key from the user
my_key = st.text_input('Enter your Hugging Face API Key', type='password')

# Set the API URL for Whisper model
API_URL = "https://api-inference.huggingface.co/models/openai/whisper-large-v3-turbo"
headers = {"Authorization": f"Bearer {my_key}"}

# Function to send the file to the API and get the response
def query(filename, file_data):
    response = requests.post(API_URL, headers=headers, files={'file': file_data})
    return response.json()

# Streamlit UI elements for file upload
st.title("Whisper API Audio Transcription")
st.markdown("Upload an audio file and get transcription results")

# File uploader widget
uploaded_files = st.file_uploader("Choose an audio file", type=["mp3", "wav", "flac"], accept_multiple_files=True)

if uploaded_files:
    results = {}
    for uploaded_file in uploaded_files:
        st.write(f"Processing file: {uploaded_file.name}")
        
        # Send the file to Hugging Face API
        output = query(uploaded_file.name, uploaded_file)
        
        # Store the result
        results[uploaded_file.name] = output

    # Display the results
    st.write("Results:")
    for file, result in results.items():
        st.write(f"**Results for {file}:**")
        st.json(result)
else:
    st.write("Please upload an audio file.")