app.py

import gradio as gr
import os
import librosa
import soundfile as sf
import tempfile
from transformers import pipeline

# Load summarization model (replace with your preferred one)
summarizer = pipeline("summarization", model="facebook/bart-large-cnn")

# Load Whisper or any ASR model of your choice
import whisper
asr_model = whisper.load_model("base")  # You can change this to 'medium' or 'large'

def transcribe_and_summarize(audio_file):
    try:
        # Convert audio to suitable format
        with tempfile.NamedTemporaryFile(suffix=".wav", delete=False) as tmp_wav:
            data, sr = librosa.load(audio_file, sr=16000)
            sf.write(tmp_wav.name, data, sr)

            # Transcribe audio
            result = asr_model.transcribe(tmp_wav.name)
            transcription = result["text"]

            # Summarize
            summary = summarizer(transcription, max_length=120, min_length=30, do_sample=False)[0]["summary_text"]

        return transcription, summary
    except Exception as e:
        return f"Error during processing: {str(e)}", ""

# Gradio UI
demo = gr.Interface(
    fn=transcribe_and_summarize,
    inputs=gr.Audio(type="filepath", label="Upload MP3 File"),
    outputs=[
        gr.Textbox(label="Transcription"),
        gr.Textbox(label="Summary")
    ],
    title="🎧 MP3 to Transcript & Summary",
    description="Upload an MP3 file and get the transcript with an audio summary.",
)

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