Table 3: Global performance (all PHI categories) on the test set. Metrics are reported on the macro-average level.

Model	Precision	Recall	F 1
BiLSTM	0.8154	0.7949	0.805
BiLSTM-CRF	0.8391	0.818	0.8284
C2V-BiLSTM-CRF	0.7925	0.3183	0.4542
Transformer	0.5027	0.6345	0.561
Transformer-CRF	0.6068	0.5843	0.5953
Transformer-BiLSTM	0.7259	0.6865	0.7056