githubEdit

🎞️Measuring model quality

We can use the perplexityarrow-up-right example to measure perplexity over a given prompt(lower perplexity is better)

Last updated

Was this helpful?