Mistral-7B-OpenOrca should be made available instead of current Zephyr-7b

#285
by muzammil-eds - opened

Hi,
The results of Open-Orca/Mistral-7B-OpenOrca are currently better than all other <30b models and it is comparable to llama 2 70B, so there should be Open-Orca/Mistral-7B-OpenOrca instead of HuggingFaceH4/zephyr-7b-alpha,

image.png

Thanks for your Consideration.

Yes, I have already checked these results; therefore, we should prefer to use the Open-Orca/Mistral-7B-OpenOrca Model.

Sign up or log in to comment