Table 3 The hyper-parameters of the prosed model.

From: Enhancing medical text classification with GAN-based data augmentation and multi-task learning in BERT

Hyperparamete

Value

Hyperparameter

Value

Maximum text length

512

Number of encoder module

2

batch size

32

Encoder of DMT-BERT

256

learning rate

2e-5

Disease classifier/ Predictor

768-256-4/256-2

Epochs

200

Dropout

0.2

AG module

50

AD module

50

Back to article page