trained_models.md

Here are the links to our trained models for the ACL'23 paper

WebQSP

BERT-base: https://1drv.ms/u/s!AuJiG47gLqTzoWW_IFTA1YPLa3nX?e=EQlSWa

T5-base: https://1drv.ms/u/s!AuJiG47gLqTzoWbazzBjbwFi4T1L?e=EdM5aI

T5-large: https://1drv.ms/u/s!AuJiG47gLqTzoW4WrPfhC0X-ojWb?e=wMYZ32

GraphQ

BERT-base: https://1drv.ms/u/s!AuJiG47gLqTzoWJod9Ty_Rx0YPJ3?e=AwENDA

T5-base: https://1drv.ms/u/s!AuJiG47gLqTzoWSI0lY2lU-TT9cW?e=1XV9fp

T5-large: https://1drv.ms/u/s!AuJiG47gLqTzoWh-djrPjWhM94bN?e=aFiEiN

DeBERTaV3-base: https://1drv.ms/u/s!AuJiG47gLqTzoWPTZf5Y5qfNDGE8?e=f2NukN

GrailQA

BERT-base: https://1drv.ms/u/s!AuJiG47gLqTzoWBTKyR_bWsPuHNc?e=qpzBna

T5-base: https://1drv.ms/u/s!AuJiG47gLqTzoWF95it8IHwE3WE9?e=r9PCC5

To use these models for prediction, download them and replace [output_dir]/model.tar.gz in README.md with the link to the target model.