Transformers-sklearn: a toolkit for medical language understanding with transformer-based models

Table 7 The experimental results of transformers-sklearn and transformers in four medical NLP tasks (mode_type = “albert”)

Name	Score		Second		Lines of code		Pre-trained model
Name	Ours	Transformers	Ours	Transformers	Ours	Transformers	Pre-trained model
TrialClassification	0.7142^a	0.4504^a	1062	1068	38	246	albert_chinese_base
BC5CDR	0.8422^a	0.8523^a	444	492	41	309	albert-base-v2
DiabetesNER	0.6196^a	0.6436^a	1122	1253	63	309	albert_chinese_base
BIOSSES	0.1892^b	0.4394^b	12	11	41	246	albert-base-v2

^aThe value of Macro F1, where the bolded indicates the best performance
^bThe value of Person correlation, where the bolded inidcates the best performance

ISSN: 1472-6947