Fine Tune Whisper with wav files

[pnandhini11]

We are collecting real time audio files for our work which are in wav file. We are trying to fine tune #whisper model using those wav files. Most of the examples are using mp3 files downloaded from huggingface dataset. Any help or pointers regarding how to prepare the dataset with wav file would be of very useful for us.

Started by referring Audio File preparation documentation in Huggingface:

https://huggingface.co/docs/datasets/audio_dataset
https://huggingface.co/docs/datasets/audio_process

Thanks in advance!