TimoImhof's picture
Update README.md
21c4731
metadata
datasets:
  - squad
metrics:
  - squad
language:
  - en

This is the "roberta-base" model with a question answering head, trained on the "squad" dataset. It serves as baseline model to compare with.

It was evaluated on the squad validation dataset and the achieved score was : {'exact_match': 85.87511825922422, 'f1': 92.12966408818717}