app.py

import gradio as gr
from datasets import load_dataset, Audio
import torch
from transformers import pipeline

pipeline = pipeline("audio-classification", model="DanielDBGC/my_awesome_lang_class_mind_model")

def predict(input_sound):
    dataset = load_dataset("PolyAI/minds14", name="en-US", split="train")
    dataset = dataset.cast_column("audio", Audio(sampling_rate=16000))
    sampling_rate = dataset.features["audio"].sampling_rate
    audio_file = dataset[0]["audio"]["path"]
    predictions = pipeline(audio_file)
    return {p["label"]: p["score"] for p in predictions} 

gradio_app = gr.Interface(
    predict,
    inputs= gr.Audio(label="Record or upload someone speaking!", sources=['upload', 'microphone'], type = 'numpy'),
    outputs=[gr.Label(label="Result", num_top_classes=3)],
    title="Guess the language!",
)

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