Last time I ran rpc-server, it kept crashing on me badly. I’ve seen some updates here and there in the release notes, but are they actively keeping it working now? I don’t want to switch my tooling over again for something that isn’t rock solid.
The problem I have with vLLM is that it’s built for the big guys as you put it, which means you need the hardware the big guys have or you have to start making trade offs. Llama.cpp was built first as a CPU inference platform. They added in GPU offloading later, that’s kind of why some of the flags names can seem backwards.
I’m sure that explains it. I didn’t realize programming.dev blocked hexbear. Kind of stupid honestly. It’s a great community but has it’s quirks for sure.
There’s another comment in this thread by user unperson that I was building on, hence referring to their comment. You asked where m3u8 came from, it’s another comment in this thread. Since it’s a derogatory term, it’s possible you have filtered their comments.
Edit: why are you assuming I am trying to offend anyone?
These are the episodes that we’re talking about, where we have had repeated, direct, violent attacks on federal law enforcement in order to veto the ballot box with violence. And the leftist justifies this violence to himself by saying that they alone are the arbiters of what is right or wrong – not the law, not history, not civilization, not public opinion, and certainly not God. The leftist assumes for himself judge, jury, executioner.
There’s also a big brigading problem with going against the “common knowledge” of Lemmy. Brave can do no good. Reverse proxy on the internet and you’re secure. Etc.
That your comment is downvoted and barely debated speaks volumes to Lemmy as actual discourse.
Yea I was just reading through that. Thanks for the direct link. He conveniently leaves out any details of the size of the “LLM” he “trains”. Seems like a cool science experiment but a huge leap to extrapolate the premise from.
How is this even legal.
Ansible
Last time I ran rpc-server, it kept crashing on me badly. I’ve seen some updates here and there in the release notes, but are they actively keeping it working now? I don’t want to switch my tooling over again for something that isn’t rock solid.
The problem I have with vLLM is that it’s built for the big guys as you put it, which means you need the hardware the big guys have or you have to start making trade offs. Llama.cpp was built first as a CPU inference platform. They added in GPU offloading later, that’s kind of why some of the flags names can seem backwards.
Does antirez have new quants for ds4 ready?
I’ve been hoping they would release a refreshed 27b model. 3.6-27b has been my daily workhorse for personal projects.
Yea I am pretty sick of that shit too. That’s honestly why I have stayed on mine for so long.
https://hexbear.net/u/unperson
I’m sure that explains it. I didn’t realize programming.dev blocked hexbear. Kind of stupid honestly. It’s a great community but has it’s quirks for sure.
There’s another comment in this thread by user unperson that I was building on, hence referring to their comment. You asked where m3u8 came from, it’s another comment in this thread. Since it’s a derogatory term, it’s possible you have filtered their comments.
Edit: why are you assuming I am trying to offend anyone?
Feed the m3u8 unperson mentioned into yt-dlp.
It’s always projection with these clowns.
There’s also a big brigading problem with going against the “common knowledge” of Lemmy. Brave can do no good. Reverse proxy on the internet and you’re secure. Etc.
That your comment is downvoted and barely debated speaks volumes to Lemmy as actual discourse.
Thanks for pointing out the actually definition of security by obscurity. Popularity has nothing to do with it.
How does this stop them? They can easily create alt accounts and vote just as much as before.
Ah, yes! Light gray text on white background. Perfectly legible.
Literal 1984 shit. Taking one of the most extreme examples of hit happening and denying it completely. So sick of this shit.
It would be news if he didn’t.
Pretty sure they’ve already been doing just that.
Yea I was just reading through that. Thanks for the direct link. He conveniently leaves out any details of the size of the “LLM” he “trains”. Seems like a cool science experiment but a huge leap to extrapolate the premise from.