Fig. 2
From: Few-shot relation classification by context attention-based prototypical networks with BERT
Losses of models on the training set.. Shows the change curves of different models during the training phase
Fig. 2 | EURASIP Journal on Wireless Communications and Networking