app.py

# audio_classification_app.py
import gradio as gr
from transformers import pipeline

# Pretrained model for audio classification
MODEL = "superb/wav2vec2-base-superb-ks"  # keyword spotting (yes, no, up, down...)

classifier = pipeline("audio-classification", model=MODEL)

def classify_audio(audio_file):
    # audio_file is a tuple: (sample_rate, numpy_array) if "numpy", or path if "filepath"
    if isinstance(audio_file, str):  # filepath
        return classifier(audio_file)
    else:  # (sr, data)
        sr, data = audio_file
        return classifier({"array": data, "sampling_rate": sr})

demo = gr.Interface(
    fn=classify_audio,
    inputs=gr.Audio(sources=["microphone", "upload"], type="filepath"),
    outputs=gr.Label(),
    title="🎵 Audio Classification",
    description="Upload or record audio. Model: wav2vec2-base-superb-ks"
)

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