Skip to main content
. 2021 Nov 12;7:e770. doi: 10.7717/peerj-cs.770

Table 3. The numbers of network layers of the model and frozen network layers when fine-tuning.

Model Quantity of total layer Quantity of frozen layer when fine-tuning
LSTM 5 3
Bi-LSTM 5 3
GRU 5 3
IndRNN 4 2
Stacked_Bi_GRU 14 10
CNN_LSTM 19 14
Deep_CNN 8 7