about the dataset

#2
by abdoubtmne - opened

Hello Pierre,
I'm interested in your model. I noticed that for the first version, you used the abisee/cnn_dailymail dataset and translated it to French using the T5-base model. For the second version, did you use the same dataset, or did you combine it with another one?
Additionally, if you don't mind, I have a suggestion: How about performing both summarization from text and elaboration a text from short texts or ideas simultaneously?

Sign up or log in to comment