Skip to main content
. 2023 May 29;2023:4300408. doi: 10.1155/2023/4300408

Table 4.

BLEURT evaluation based on 800 test data.

Base model (LSTM) −1.528
Popular hashtags with transformer −1.302