app.py

import gradio as gr
from transformers import pipeline

# Load models
transcriber = pipeline("automatic-speech-recognition", model="openai/whisper-base")
summarizer = pipeline("summarization", model="facebook/bart-large-cnn")

# Function to process audio
def process_audio(audio_file):
    # Step 1: Transcribe audio
    transcription = transcriber(audio_file)["text"]
    
    # Step 2: Summarize transcription
    summary = summarizer(transcription, max_length=50, min_length=10, do_sample=False)[0]["summary_text"]
    
    return transcription, summary

# Gradio Interface
interface = gr.Interface(
    fn=process_audio,
    inputs=gr.Audio(source="upload", type="filepath", label="Upload Audio File"),
    outputs=[
        gr.Textbox(label="Full Transcription"),
        gr.Textbox(label="Summary")
    ],
    title="Audio Transcription and Summarization",
    description="Upload an audio file to get a full transcription and a brief summary of its content."
)

# Launch the interface
interface.launch()