mobiuslabsgmbh

/

Llama-3.1-70b-instruct_4bitgs64_hqq

like 31

Thank you! Is there a detailed benchmark of this model ?

I've seen this model here, but if one compares among several benchmarks, one might also spot another one.
The guy knows what he's doing, so I felt like suggesting it here.

https://huggingface.co/spaces/DontPlanToEnd/UGI-Leaderboard

I mean this here doesn't look bad, though I cannot guarantee if the numbers are right

of cause, now that there is the new model by Nous research, you might wanna consider that, too:

https://huggingface.co/collections/NousResearch/hermes-3-66bd6c01399b14b08fe335ea

Here it is https://huggingface.co/mobiuslabsgmbh/Hermes-3-Llama-3.1-70B_4bitgs64_hqq
I don't have enough time and resources to run a detailed eval for this model though, hope it works just fine!