Table 1. Document types included in the Norwegian clinical corpus used for the continuous pretraining of NorDeClin-BERT-base (NorDeClin Bidirectional Encoder Representations from Transformers).

Document type	Number of files	Size
Anesthesia	46,310	94.8 MB
Treatment	29,919	49.3 MB
Discharge summaries	586,637	1.6 GB
Ergotherapy	33,220	38.4 MB
Pharmacy	3484	4.6 MB
Physiotherapy	69,324	80.4 MB
Individual plan	558	1.4 MB
Admission records	248,208	779,2 MB
Laboratory	66	53.8 kB
Surgery	313,795	446.8 MB
Summary records	5710	9.2 MB
Radiology	63,734	30.1 MB
Somatic care	110,248	211.3 MB
Nursing	299,212	220.7 MB
Training dataset (no duplicates)	1,670,464	3.2 GB