How to apply SpecAugment when fine tuning Whisper model.

[amitliron]

According to the paper (Robust Speech Recognition via Large-Scale Weak Supervision):
whisper (Large V2) was trained the with SpecAugment.

I want to finet tune whisper on my data (Fine-Tune Whisper For Multilingual ASR with 🤗 Transformers).

1. Is it right to add augmentations (i.e SpecAugment) when fine-tuning the model ?
2. How can I add SpecAugment augmentations while fine-tuning whisper model ?