[Feature] Align Existing Transcript and Waveform Audio Labelling

[NicholasTheGreat]

Hi I would love to be able to import an existing transcript, perform wav2vec2 forced alignment and visualise the result in waveform. It would also be great to be able to click parts of the waveform and label them as words or sounds such as [laughing], then have the word and segment level timestamps updated accordingly in the transcript.

P.S. Importing and exporting in CSV format with the columns, start, end, speaker, text would also be much appreciated.