hermes - 💡(How to fix) Fix Please add SWE benchmark to show how good this approach is relative to the vanilla models(or to the simple /goal agent) [1 participants]

Hi, Could you please add benchmark results, especially the SWE-bench, showing the scores for the vanilla models used and the hermes-agent versions of them? Let's see how good the hermes-agent really is. Since your name says 'Research', I think you already have these numbers, so, why don't you publish them?