Trying to use Onsets and Frames

[Trey Harrison]

I've tried to use https://piano-scribe.glitch.me/ and "Onsets and Frames" here: https://colab.research.google.com/notebooks/magenta/onsets_frames_transcription/onsets_frames_transcription.ipynb

But unfortunately neither one is working for me. The google colab gets this error during model initialization:

ImportError: This version of TensorFlow Probability requires TensorFlow version >= 2.9; Detected an installation of version 2.8.2. Please upgrade TensorFlow to proceed.

And the glitch site just stalls forever on "Transcribing".

Seems like there might be a way to deal with the TensorFlow error but as a noob I'm not sure how...

Is there any other easy way to try audio to midi using these models?

Thanks

Trey