GoogleyWoog

Yes. The Llama 70B derived models, as well as Mixtral 8x7B and the new Mistral Medium 70B are competitive with ChatGPT 3.5. Most of them can do 16,000 token context similar to ChatGPT as well.

You only NEED 40GB of free RAM to run them at decent quality, but it’s slow.

With a 24GB GPU like a 3090 or 4090 you can run them at a reasonable speed with partial GPU offload. About 1-2 words per second. I run 70Bs in this manner on my computer.

With two 24GB GPUs you can run them very fast, like ChatGPT.

There’s of course a whole world in between as well, but those are the rough hardware requirements to match ChatGPT in a self-hosted sort of way. There’s also a new thing people are doing where they add layers from one model onto another one, like a merge but keeping >50% of the original layers from each model. “Goliath 120B” and the like, which is made from 2 different 70Bs. They’re even better but it’s a bit beyond reasonable consumer hardware for now.

Yeah I’ve always thought mandatory reserve training seemed like a good idea. Give everyone some basic life skills, team building, athleticism, and show them how shitty war would be to encourage peace.

deleted by creator

I hope to one day have enough free time to want to watch someone play video games unedited for hours at a time, let alone be willing to pay money for the privilege.

Cauliflower.

I don’t think the west was any different. Very dry winter, only a few decent snow falls and they were gone pretty fast.

Crawling low

That’s what I’m wondering too… Gboard with no network permission seems pretty ideal.

I’m not a secret agent or criminal, I just want an alternative to Google/Microsoft’s suite that doesn’t blatantly harvest all my information and try to sell me things based on it.