TalTal

Hi @alisawuffles, thanks for you novel work! I have a question about the evaluation indicator. When you evaluate models on triviaqa datasets, you calculate the accuracy as follows: test_df['output'] =...