why the original model (“csebuetnlp/mT5_multilingual_XLSum”) gave good summary when testing it on arabic samples, but when testing the finetuned model on XLSUM dataset for 10 epochs, it didn’t give good results. although the rouge measures are good during finetuning, also the training and validation loss are close to zero. please could anyone help me?, it is urgent problem.
New contributor
hiba younis is a new contributor to this site. Take care in asking for clarification, commenting, and answering.
Check out our Code of Conduct.