WanderPanda

6 ms·

1.
by WanderPanda 4mo ago
It highly depends on the task. For math and coding, sure. But for knowledge tasks GPT-4 is wayy better than even SOTA ~100B models. For my knowledge test cases the lines get blurry at >400B
2.
by WanderPanda 5mo ago
I applaud that you recently started providing the KL divergence plots that really help understand how different quantizations compare. But how well does this correlate with closed loop performance? How difficult/expensive would it be t
3.
by WanderPanda 5mo ago
I would be really interested in a podcast with the CEO where he goes a bit into the trade-offs of backwards and forwards compatibility. I can not imagine that their planning was so immaculate that there aren't any regressions that a cl
4.
by WanderPanda 5mo ago
This is so true! Shows a lack of care that usually doesn’t stop at just the naming
5.
by WanderPanda 6mo ago
They are heavily post-trained on code and math these days. I don‘t think we can infer that much about their behavior from just the pre-training dataset anymore
6.
by WanderPanda 7mo ago
Amazing work and people should really appreciate that the opportunity costs of your work are immense (given the hype). On another note: I'm a bit paranoid about quantization. I know people are not good at discerning model quality at th
7.
by WanderPanda 8mo ago
I find it hard to trust post training quantizations. Why don't they run benchmarks to see the degradation in performance? It sketches me out because it should be the easiest thing to automatically run a suite of benchmarks
8.
by WanderPanda 10mo ago
Wait but the one you linked seems to be pneumatically driven, while the op one is an actual combustion engine, right?
9.
by WanderPanda 10mo ago
Small feedback if any of the Antigravity people read here: "Fast" is not a great name for the "eager" option (vs. "Planning") because "Fast" is associated with "dumb" in LLMs (fast/flas
10.
by WanderPanda 10mo ago
SWIFT is Belgian, though?
11.
by WanderPanda 10mo ago
Mechanically sure, but I still feel way safer when a Tesla (of any kind) is approaching me as a pedestrian or bicyclist than any other vehicle (except maybe Waymo) because I know they will alert the driver and brake if necessary. Any other
12.
by WanderPanda 11mo ago
Makes sense! I like that you guys are more open about it. The other labs just drop stuff from the ivory tower. I think your style matches better with engineers who are used to datasheets etc. and usually don't like poking a black box
13.
by WanderPanda 11mo ago
Damn TIL, I always used > Cursor: disable completions and forgot to turn it on again I need to try snooze then!
14.
by WanderPanda 11mo ago
Why did you stop training shy of the frontier models? From the log plot it seems like you would only need ~50% more compute to reach frontier capability
15.
by WanderPanda 11mo ago
Until it isn't
16.
by WanderPanda 11mo ago
Did you check out the STM32N6? It apparently has an h264 encoder
17.
by WanderPanda 11mo ago
Amazing: (Mar 5 2022) TinyGL 0.4.1 is out (Changelog) (Mar 17 2002) TinyGL 0.4 is out (Changelog) "our plans are measured in centuries"
18.
by WanderPanda 11mo ago
I have a strong Tinnitus on one ear after an ear surgery for 8 years now. And I usually don‘t notice it for months at a time, even though it is there all the time (thanks for reminding me :p) So it’s not as bad as it might feel in the begin
19.
by WanderPanda 11mo ago
I think this is the frontier when it comes to "unstructured": https://youtu.be/nmEy1_75qHk They for sure did not anticipate that the user would backflip into their robot and knock it (and himself) out :D
20.
by WanderPanda 11mo ago
Theoretically, when the market offers me an order book and I take offers on one or the other side that should be totally fair? I think until execution/fill the information should be totally between me and the exchange and no one else,
21.
by WanderPanda 1y ago
Alarm is a good example of an “output only” task. The more inputs that need to be processed the less a pure chatbot interface is good (think lunch bowl menus, shopping in general etc.)
22.
by WanderPanda 1y ago
Did they still not release "Bring your own subscription" "login with ChatGPT" and letting people apply their subscription/quota to other apps/services? There are so many use-cases where someone builds a usefull
23.
by WanderPanda 1y ago
It was 4x over the original version IIRC so should be ~ 2x over the previous
24.
by WanderPanda 1y ago
Imagine regulators doing their job for once and creating a clean regulation that removes the uncertainty about the liability for such releases. Such that they can just slap Apache or MIT on it and call it a day and don't require to col
25.
by WanderPanda 1y ago
I think modularization of templates is really hard. Best thing I can think of is a cache e.g. for signatures. But then again this is basically what the mangling already does anyways in my understanding.
26.
by WanderPanda 1y ago
you know whats a glorified Markov chain? The Universe
27.
by WanderPanda 1y ago
For me it all made sense when I heard that IQ/g-factor basically vanishes in the absence of time pressure (heard if from Richard Haier on Lex). For a very narrow range of professions, like ATCs, time is absolutely critical but for most
28.
by WanderPanda 1y ago
Is Whisper still SOTA 3 years later? It does not seem there is a clearly better open model. Alec Radford really is a genius!
29.
by WanderPanda 1y ago
At this point, I'm grateful and in awe that it runs reliably at all. I can easily imagine a case where the same or more resources are spent, and the outcome is that still nothing runs. At some point, the rot in institutions can not be
30.
by WanderPanda 1y ago
Same experience. They should really store these blobs centrally under a hash and link to them from the venvs

More ›