Closed VP007-py closed 3 years ago
Thanks for filing the first issue! Sorry for the delayed response (didn't have e-mail notifications on 😓).
The issue is GPT-2 is an autoregressive LM and gives true log-likelihood scores. You need to use LMScorer, not MLMScorer. The README was unclear; my mistake.
I've updated the README and added pre-emptive error messages to MLMScorer, LMScorer, etc.; hope this helps.
@JulianSlzr Thanks for the fix and kudos to your awesome work !
I tried scoring sentences with the models mentioned here . Every model works fine except for
gpt2-117m-en-cased
andgpt2-345m-en-cased
. The following error pops upAny fixes ?