app.py

import whisper
import gradio as gr
import json
import tempfile
import os

print("Loading Whisper model...")
# Load model when space starts
model = whisper.load_model("base")
print("Model loaded successfully!")

def transcribe_audio(audio_file):
    """
    Simple function to transcribe audio to text
    """
    try:
        print(f"Processing audio file: {audio_file}")
        
        # Transcribe the audio
        result = model.transcribe(audio_file)
        
        print("Transcription completed successfully!")
        
        # Return as JSON string
        return json.dumps({
            "text": result["text"],
            "language": result["language"],
            "status": "success"
        })
        
    except Exception as e:
        print(f"Error: {str(e)}")
        return json.dumps({
            "error": str(e),
            "status": "failed"
        })

# Create a simple web interface
demo = gr.Interface(
    fn=transcribe_audio,
    inputs=gr.Audio(sources=["upload"], type="filepath", label="Upload Audio File"),
    outputs=gr.Textbox(label="Transcription Result", lines=5),
    title=" Whisper Transcription API",
    description="Upload an audio file to convert it to text using OpenAI's Whisper model.",
    examples=[
        ["https://huggingface.co/datasets/Narsil/asr_dummy/resolve/main/1.flac"],
        ["https://huggingface.co/datasets/Narsil/asr_dummy/resolve/main/2.flac"]
    ]
)

# Launch the application
if __name__ == "__main__":
    demo.launch(debug=True, share=True)