Lwerewolf

4 ms·

1.
by Lwerewolf 5d ago
This one is 2x m5 max, so ~1.2TB/sec.
2.
by Lwerewolf 16d ago
There are resident systems, and they're of exactly the type listed in the article (i.e. water heaters). Residential... area heating pumps, so to say, are moving towards propane (afaik). I've been looking into getting a propane mon
3.
by Lwerewolf 23d ago
I came to this relatively late (2013-ish, windows kernel development, scouring OSDev, etc) so I thought that was always the right name for them. Prior experience was mostly... higher-level langs.
4.
by Lwerewolf 1mo ago
As others have mentioned, nothing's stopping any other major provider from offering it. Given its popularity, you can guess how that'll develop. So, overall, irrelevant.
5.
by Lwerewolf 1mo ago
A regular yearly replay is in order for me, too. Already done with SC2... just SC1/WC3/DoW(2)/RA2(mental/flippedmissions/etc)/RA3(uprising)/HOMM3/etc/etc/etc. Ah well.
6.
by Lwerewolf 1mo ago
The model weights are MIT licensed.
7.
by Lwerewolf 1mo ago
Welcome to the club. If you're _really_ competitive in cs2, I'd swap out to a 9800x3d setup, but it's still a maybe. Very little reason to upgrade right now other than to run LLMs.
8.
by Lwerewolf 1mo ago
You need latency for token parallelism, not bandwidth. Hence actual RDMA that bypasses the software TCP stack (ROCe or whatever).
9.
by Lwerewolf 1mo ago
You just need social stigma on bad tires. It's the single most important thing on your vehicle, for literally all the reasons out there.
10.
by Lwerewolf 1mo ago
I literally had Fable tell me that I've checked out a nonexistent upstream commit of LMDB - two days ago. Sol called its BS, of course.
11.
by Lwerewolf 1mo ago
I mean, a big RAG setup with a smart LLM ought to do it, right? Or even just locally provided knowledge database - given the size of LLMs, what's a clone of wikipedia and whatever else you'd need?
12.
by Lwerewolf 1mo ago
Never been easier to make one.
13.
by Lwerewolf 2mo ago
Poolside's stuff (US) is pretty good.
14.
by Lwerewolf 2mo ago
You don't have to. Chatgpt-sol-high already does that for me. Could be the extra instructions or the base system prompt or whatever, point is - it already does it.
15.
by Lwerewolf 2mo ago
The MI350p exists and should run a decent quant (say, the ~96GB antirez mix) well, but you can get two rtx pro 6000s for one of these, or 8x (actually more) r9700 + probably the gear to run them, etc. Otherwise, you can probably buy one of
16.
by Lwerewolf 2mo ago
Almost reads like state policy-induced FOMO.
17.
by Lwerewolf 2mo ago
Just tried the preview on my little test codebase and a "check this out and tell me what you think" prompt used over double the tokens of the previous iteration, but it was a lot more eager as well. Kind of reminds me of the new l
18.
by Lwerewolf 2mo ago
I've had it run to ~400k when debugging "obscure" (to it) sequences. Wouldn't recommend more.
19.
by Lwerewolf 2mo ago
Might wanna try it now, seems to have been largely fixed. Check huggingface threads and reddit. Works for me, very memory hungry and PP speed drops off a cliff around 200k context (~30tok/s decode and ~40tok/s pp - like... hope it
20.
by Lwerewolf 2mo ago
...they'd messed something up in that quant, apparently fixed promptly. ds4 now supports it as well and it's... well, context-limited (<=250k on a 128gb machine) but it positively flies on an m5 max - the 60tok/s decode &#
21.
by Lwerewolf 2mo ago
Tool calling in pi completely broke with the updated q4 gguf (spinquant-less). Guessing it'll take some time.
22.
by Lwerewolf 2mo ago
Right: https://huggingface.co/poolside/Laguna-S-2.1-FP8/discussions... Wait time.
23.
by Lwerewolf 2mo ago
Well, just ran said gguf on the GeneralsX codebase with a pretty open-ended "Explain this codebase to me, and the general game loop." prompt, and... Let me also look at the GameLogic::update() to see the rest of the update flo
24.
by Lwerewolf 2mo ago
Deleted earlier, didn't see you post, pasting here: /* Just started testing with the gguf (with gpu offload, m5 max 128gb), q4_k_m, running seemingly well. Speed is initially slightly faster than antirez/ds4 - decode tok&#x2F
25.
by Lwerewolf 2mo ago
This: https://github.com/Blaizzy/mlx-lm/tree/pc/add-lg ...and this is what I should probably wait for (not sure why it's in vlm): https://github.com/Blaizzy/mlx-vlm/tree&#x
26.
by Lwerewolf 2mo ago
nvfp4 mlx, literally barebones pi. edit: on bigger tests, got it to loop pretty easily unfortunately, probably local settings.
27.
by Lwerewolf 2mo ago
Almost like a built-in heavyweight harness.
28.
by Lwerewolf 2mo ago
Testing it now. At the very least, competitive with DS4-Flash indeed. On my small (and per Sol's words, _very_ semantically dense) C test codebase, it found things that only gpt-5.2 managed to find back in the day, but also made a stup
29.
by Lwerewolf 2mo ago
Pretty sure this might be a duplicate. Regardless, tried the 1bit bonsai 27b gguf three different ways - their llama.cpp fork (prism, was it) on 2 machines (1255u/16g, m5 max/128g) and the web (on the 1255u). With llama.cpp it wor
30.
by Lwerewolf 2mo ago
Not too sure about the engines as of late. Bigger and heavier vehicles - yes, but still mostly rural/highway, and at way lower speeds than the EU. Overall, IMO it's just the distances involved.

More ›