Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
MrDrMcCoy
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
4 ms
·
61.
▲
by
MrDrMcCoy
2mo ago
Perfectly acceptable by modern quality standards.
62.
▲
by
MrDrMcCoy
2mo ago
Docker containers aren't very secure because the daemon runs as root by default. Rootless container runtimes are plenty secure, especially if the user that runs the container is unprivileged. Easiest thing in the world with Podman.
63.
▲
by
MrDrMcCoy
2mo ago
We got rid of all our offices, whole company is remote now.
64.
▲
by
MrDrMcCoy
2mo ago
Especially the eggplant.
65.
▲
by
MrDrMcCoy
2mo ago
There already are a lot of forks for things they decline to implement. TurboQuant, ROCmFPX, and more. I need to set up an agent that will loop on merging them.
66.
▲
by
MrDrMcCoy
2mo ago
Fact. My company's largest partner is CoreWeave, and convincing leadership that we could run it ourselves on partner discounted hardware for a lot less money has gone nowhere.
67.
▲
by
MrDrMcCoy
2mo ago
+1 on therapeutic ketamine. Got me out of a bad spot, and felt better long after it was over. You can get it by mail in the US from Mindbloom, who offered excellent care despite being remote in my case.
68.
▲
by
MrDrMcCoy
2mo ago
Phishing implies exploitable data collection. Is that happening here?
69.
▲
by
MrDrMcCoy
2mo ago
Looks like a Wireguard stunnel replacement, which is very useful!
70.
▲
by
MrDrMcCoy
2mo ago
I want to know where the public water fountains went. Those are arguably a safety feature.
71.
▲
by
MrDrMcCoy
2mo ago
Oh, nice!
72.
▲
by
MrDrMcCoy
2mo ago
I would love to have a Fairphone, but sadly I live in the USA.
73.
▲
by
MrDrMcCoy
2mo ago
It used to be common to give babies a tiny amount of brandy to sooth them. Second hand smoke is a thing.
74.
▲
by
MrDrMcCoy
2mo ago
I've wanted to do something similar for a while now: set up a camera in a lumber mill that broadcasts a constant live stream of wood grain. Never the same pattern twice, which could be used for art and entropy. Sadly, I don't know
75.
▲
by
MrDrMcCoy
2mo ago
The US is not better in those regards, and is more likely to fall at this point. That would leave the EU as the last best place in the world to live.
76.
▲
by
MrDrMcCoy
2mo ago
Certainly not above reproach, and there are reasons to have beef with them (I know I do). That said, they're doing some aspects of modern democracy better than the US, and we really ought to borrow each other's best ideas.
77.
▲
by
MrDrMcCoy
2mo ago
I have never seen a server board that offers 8x 16 lane slots. Being stuck with 7 slots means if I want an even 8 GPUs, I'd need to bifurcate one slot into x8 x8, which sacrifices a precious PCIe4 bandwidth. On the chassis side, not ha
78.
▲
by
MrDrMcCoy
2mo ago
I gave you a direct link to where you can get a download, and explained why I chose to link you where I did. I can't post screenshots here drawing you a map of how to use a website. I don't work for Huggingface or any AI company a
79.
▲
by
MrDrMcCoy
2mo ago
It's not the pricing that bothers me, it's the inflexibility of using fully-integrated vendors like Dell/HP over custom, vs standards compliant vendors like Supermicro/ASRock. The difference for me lies in the ability to
80.
▲
by
MrDrMcCoy
2mo ago
> "I went to youtube, gmail, ycombinator, deliveroo, then I went to another site I randomly chose, because they're the best, duh" The path I described is all within HuggingFace. > And How am I supposed to know that arri
81.
▲
by
MrDrMcCoy
2mo ago
I broadly knew that about temperature, but lack the background in machine learning/statistics to differentiate top-n-sigma from top-k/top-p.
82.
▲
by
MrDrMcCoy
2mo ago
> Excuse me, but thats a direct link you've just sent. I asked where I can find the links. I like to believe in the source of truth. You asked where to find the GGUF files of this model for direct download and I provided it. Almost
83.
▲
by
MrDrMcCoy
2mo ago
Here: https://huggingface.co/unsloth/Qwen3.8-27B-GGUF/tree/main
84.
▲
by
MrDrMcCoy
2mo ago
What on earth are you talking about? llama.cpp != Ollama. You can (and should) just use llama.cpp directly. Upstream llama.cpp can take the shorthand huggingface path and automagically download it into a cache folder as part of the launch.
85.
▲
by
MrDrMcCoy
2mo ago
For those us us who don't know, what do those parameters do and why are they better?
86.
▲
by
MrDrMcCoy
2mo ago
Official llama.cpp releases ship with huggingface support. If you don't want to download it yourself, you can just use the `repo/model:quant` convention and it will handle downloading locally for you.
87.
▲
by
MrDrMcCoy
2mo ago
That is obvious. Less obvious is what context shifting does and whether you want it or not.
88.
▲
by
MrDrMcCoy
2mo ago
For the things it does, what tool is better than ffmpeg? Really struggling to see the cargo cult angle. Similar for llama.cpp, as it is literally the only framework I can get to run on my multi-Radeon rig. It is the most portable runtime ou
89.
▲
by
MrDrMcCoy
2mo ago
Your disdain for the EU is genuinely puzzling to me.
90.
▲
by
MrDrMcCoy
2mo ago
Please don't tell me what I need. Real 8 slots would have saved me from having to bifurcate slot 7 in my 8 GPU AI rig.
More ›