Extend documentation with simple example on how to run prediction on a .wav file

[emma3826]

would be great to have a simple example on how to upload and run prediction on a wav file as part of ESC50: Environmental Sound Classification.ipynb

something like this:
learn_inf = load_learner('model name', cpu=True)
audio = AudioTensor.create('file name')
with out_pl_env: display(Audio(content))
pred,pred_idx,probs = learn_inf.predict(audio)
lbl_pred_env.value = f'Prediction: {pred}; Probability: {probs[pred_idx]:.04f}'

same example as in the fastai documentation, only adapted for audio use case

[mogwai]

Feel free to create a pr for this @emma3826