bofenghuang

/

asr-wav2vec2-ctc-french

like 11

Hi @bofenghuang !

I want to reproduce the finetuning part of the model using similar datasets you used, but loading the audio in the batches (not at the beginning like mentioned here: https://www.kaggle.com/code/kingabzpro/fine-tuning-xlsr-wav2vec2-for-wolof-asr-with/notebook). I don't know if you used the same training script or did you use another script.

Thank you