The model was trained using clagator/biobert_v1.1_pubmed_nli_sts as a base and the smm4h dataset from 2017 from subtask 3.

See here for the scripts and datasets.

Sarker, Abeed (2018), “Data and systems for medication-related text classification and concept normalization from Twitter: Insights from the Social Media Mining for Health (SMM4H)-2017 shared task”, Mendeley Data, V2, doi: 10.17632/rxwfb3tysd.2

Acc@N denotes the accuracy taking the top N predictions of the model into account, not just the first one.