Mistral-NeMo-Minitron-8B-Base-Minivoc-32k-v0.1a
I am trying to find a speculation decoding draft model for Mistral Large 2, but there is a problem: so far, I can't find a modern small model with the correct vocab size of 32k. Except for Kaitchup's, from the looks of it. There are probably different models that has the right specs, but for now, this will have to do.
kaitchup/Mistral-NeMo-Minitron-8B-Base-Minivoc-32k-v0.1a
kaitchup/Llama-3.1-8B-Minivoc-32k-v0.1a
Thank you for looking at this request. :)
Sure, they are both queued and should be done within an hour or so, if all goes well. You can keep them coming, too :) Next time, though, please provide a url to the repository :^)
Unfortunately, he wants to extort money for access. I don't think I want to support his business model of using free model data and then "reselling" it. Hope he dies painfully. What a great way to end up on my blacklist (kaitchup, not you).
More than fair. That sort of thing is just bad for our hobby.