Table 8 Comparison of run-time (in days) statistics of BioALBERT versus BioBERT
From: Benchmarking for biomedical natural language processing tasks with a domain specific ALBERT
Model | Training time (in days) |
|---|---|
\({\text {BioBERT}}_{{Base2}}\) | 23.00 |
\({\text {BioBERT}}_{{Base1}}\) | 10.00 |
\({\text {BioALBERT}}_{{Base1}}\) | 3.00 |
\({\text {BioALBERT}}_{{Base2}}\) | 4.08 |
\({\text {BioALBERT}}_{{Large1}}\) | 2.83 |
\({\text {BioALBERT}}_{{Large2}}\) | 3.88 |
\({\text {BioALBERT}}_{{Base3}}\) | 4.02 |
\({\text {BioALBERT}}_{{Base4}}\) | 4.45 |
\({\text {BioALBERT}}_{{Large3}}\) | 4.62 |
\({\text {BioALBERT}}_{{Large4}}\) | 4.67 |