[nmitchko]

This is a pretty useless post. You could also follow the same 1000x tutorials about llama and use the already uploaded hugging face formats that are on hugging face...

Here are some actually useful links

https://blog.ovhcloud.com/fine-tuning-llama-2-models-using-a...

https://huggingface.co/meta-llama/Llama-2-70b-hf

https://huggingface.co/meta-llama/Llama-2-7b-hf

[onlypositive]

Is it really "useless" if I didn't even know about llama? And look, now I have 3 more links to dive into.

This is the opposite of useless.

[mciancia]

Well it's quite possible that it is useless for you since you didn't hear about llama by now ;)

[hoten]

I guess this was meant as a tongue-in-cheek joke comment, but it comes off as needless gatekeeping.

[onlypositive]

I dunno what's worse, the pointless commentary, needless gatekeeping, the superfluous white knighting or the fact we're getting upvotes for all this nonsense.

[DigiDigiorno]

Three links is not a large number of links, AND you don't "have" to dive into every link in lists of "useful links".

That said, I don't think op is useless.

[alfalfasprout]

Honestly HN has turned into reposting random how-tos about the latest hype LLMs.

[g42gregory]

If that's what people like, why is that a concern?

[hamilyon2]

HN hyped machine learning long before the current craze cycle

[IAmNotACellist]

And it's probably already in TextGen and better UIs for serious stuff than yet another Jupyter notebook.

[3abiton]

I still don't get why it's useless. Maybe less useful, but useless seems to be a hyperbole.

[nmitchko]

For each step in the article:

1. You should really download the huggingface hosted model.

2. Why convert the model if meta already hosts it here: https://huggingface.co/meta-llama

3. This step completely glossed over the hardware requirements. Doesn't explain any of the instructions needed in the finetuning process.

4. "Now run the model". What? What about the model precision, hosting, standard open source tools, etc,etc...

[jeremycarter]

Thanks!