I wanted to know that if I want to fine-tune for translation instead of transcription, I only need to change the task in the WhisperAudioProcessor and the metrics (using BLEU instead of WER) in the compute metrics function right?
Do I need to make any other changes to modify the code for translation to English?
I’ve done a project where I fine-tuned Whisper-Tiny for translation tasks, and it gave good results. You can take a look at my repo to see the steps I used, which might help you solve the issue you’re facing.
MediBeng-Whisper-Tiny Github repo: