4 ms·
It absolutely does not say that. It in fact provides benchmarks that show it under performing them. Not great to blindly trust benchmarks, but there are no cla
by jasonjmcghee 3y ago
It absolutely does not say that. It in fact provides benchmarks that show it under performing them.
Not great to blindly trust benchmarks, but there are no claims it will outperform GPT-4 or Opus.
It was a checkpoint, so it's POSSIBLE it COULD outperform.