Raise more informative error when processing stereo tracks

[edward-io]

Thanks for converting crepe to PyTorch!

When running python -m torchcrepe --audio_files=..., I encountered the error:

  File "~/miniconda3/lib/python3.11/site-packages/torchcrepe/core.py", line 733, in resample
    audio = audio.detach().cpu().numpy().squeeze(0)
            ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
ValueError: cannot select an axis to squeeze out which has size not equal to one

Investigating the error, audio has shape torch.Size([2, 10674560]). Raising an exception when the input audio track is stereo (or downmix to mono) could be useful to other users of the project.

Thanks!