The notebook executes fine without any errors when you “Run All”. Once you upload any .wav file and execute the code, an output csv file is generated where the first column is a timestamp in seconds, the second column is the estimated frequency in Hz, and the third column is a value between 0 and 1 indicating the model's voicing confidence (i.e. confidence in the presence of a pitch for every frame). You can also see the audio and frequency plots at the end.
Notebook Credit 🌟