Skip to main content
. 2024 May 16;3:e52095. doi: 10.2196/52095

Table 2.

Descriptive statistics of training sets and model performance.

Descriptive statistics Value, range Value, mean (SD)
Number of disclosure statements 1-200 100.0 (57.42)
Number of tokens 4-1402 712.9 (405.94)
Number of sentences 5-1031 525.2 (294.13)
Entities per sentence 0.771-1.72 1.34 (0.14)
RoBERTa_base F1-score 0.43-0.94 0.81(0.13)
GatorTron_base F1-score 0.37-0.92 0.84 (0.13)
RoBERTa_large F1-score 0.44-0.96 0.84 (0.14)
GPT-2_large F1-score 0.30-0.72 0.58 (0.12)