First link: https://github.com/ggerganov/llama.cpp

Which in turn has the following as the first link: https://arxiv.org/abs/2302.13971

Is it really quicker to ask here than just browse content for a bit, skimming some text or even using Google for one minute?

>Is it really quicker to ask here than just browse content for a bit, skimming some text or even using Google for one minute?

I don't know if it's quicker but I trust human assessment a lot more than any machine generated explanations. You're right I could have asked ChatGPT or even Googled but a small bit of context goes a long way and I'm clearly out of the loop here -- it's possible others arrive on HN might appreciate such an explanation or we're better off having lots of people making duplicated efforts to understand what they're looking at.

Well, I'm saying if you just followed the links on the submitted page, you'd reach the same conclusion but faster.

You gave an awesome answer in 2 minutes! Might be faster than reading!

If you cannot click two links in a browser under two minutes, I'm either sorry for you, or scared of you :)

llama is a text prediction model similar to GPT-2, and the version of GPT-3 that has not been fine tuned yet.

It is also possible to run fine tuned versions like vicuna with this. I think. Those versions are more focused on answering questions.

>I can't tell from the Gist alone

Literally the second line: "llama is a text prediction model similar to GPT-2, and the version of GPT-3 that has not been fine tuned yet"

I'm sorry! I added this improvement based on that persons question!

Sorry -- I missed that. I'll delete my comments -- obviously I'm just an idiot asking dumb questions that have no value to anybody. I thought I read through it.

not at all, your question was really good so I added the answer to it to my gist to help everyone else. Sorry for the confusion I created by doing that!