From: Survey of transformers and towards ensemble learning using transformers for natural language processing
Model
BERT
RoBERTa
rouge1
0.1406
0.2864
rougeL
0.1097
0.2306