biobert-v1.1_pub_section

• original model file name: textclassifer_biobert-v1.1_pubmed_20k
• This is a fine-tuned checkpoint of dmis-lab/biobert-v1.1 for document section text classification
• possible document section classes are:BACKGROUND, CONCLUSIONS, METHODS, OBJECTIVE, RESULTS,

metadata

training_metrics

• val_accuracy: 0.8522772192955017

• val_matthewscorrcoef: 0.8009328246116638

• val_f1score: 0.8517481088638306

• val_cross_entropy: 0.4344026446342468

• epoch: 12.0

• train_accuracy_step: 0.8203125

• train_matthewscorrcoef_step: 0.7453048229217529

• train_f1score_step: 0.8245896100997925

• train_cross_entropy_step: 0.480397492647171

• train_accuracy_epoch: 0.8297363519668579

• train_matthewscorrcoef_epoch: 0.7703952193260193

• train_f1score_epoch: 0.8274592757225037

• train_cross_entropy_epoch: 0.5001224875450134

• test_accuracy: 0.8441678881645203

• test_matthewscorrcoef: 0.7905130982398987

• test_f1score: 0.8435087203979492

• test_cross_entropy: 0.4557005763053894

• date_run: Apr-22-2022_t-14

• huggingface_tag: dmis-lab/biobert-v1.1