Skip to main content
. 2021 Jun 23:1–11. Online ahead of print. doi: 10.1007/s13369-021-05810-5

Table 3.

Comparative Analysis Of Models

MODEL Dataset EM F1
GPU version of BERT(with sklearn wrapper) SQuAD 1.1 dev 81.2 88.6
BERT for QA SQuAD 1.1 dev 81.3 88.7
Distil BERT for QA with Knowledge Distillation SQuAD 1.1 dev 80.1 87.5
GPU version of BERT(with sklearn wrapper) CORD-19 79.3 86.4
BERT for QA CORD-19 81.5 88.3
Distil BERT for QA with Knowledge Distillation CORD-19 80.6 87.3