Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
algo_trader
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
11 ms
·
91.
▲
by
algo_trader
2y ago
> a analytics SaaS which displays, among many other pieces of data, Does yelp (or others) allow the restaurant to modify the menu/offer coupons in a dynamic way? (i.e. change multiple times per hour, update with in minutes) thanks
92.
▲
by
algo_trader
2y ago
> and steel has a weight problem and a major corrosion problem if i want a salt water low tech floating shed (not mission critical), can you make it work with galvanized/re-painting schedule? or is marine grade steel a must ? (man,
93.
▲
by
algo_trader
2y ago
You are not wrong AlphaZero-type systems were unable to significantly improve over near-optimal solvers such as z3.
94.
▲
by
algo_trader
2y ago
What level of mechanisation is used? And what level of automation do you think can take this x10 ?
95.
▲
by
algo_trader
2y ago
Are you using code-completion-in-the-IDE ? which tool ?
96.
▲
by
algo_trader
2y ago
For practical systems, do you think improving the compiler toolchain is a good direction ? For example, speeding up compilation of 100s of similar code candidates, doing the same for eval, etc
97.
▲
by
algo_trader
2y ago
For practical systems, to improve synthesis/codegen, do you think working on the compile/eval toolchain is a good direction ? For example, if we can cache/re-use the compiler run on previous codegen candidates, to speed up co
98.
▲
by
algo_trader
2y ago
Are DC costs more manageable at medium voltages? If i really wanted 11kV MVDC is this a feasible spec for an industrial site?
99.
▲
by
algo_trader
2y ago
what would you recommend as a maximum-2B model for coding/reasoning domain ?
100.
▲
by
algo_trader
2y ago
Apparently, container optimization was unsolved for very large fleets. I did not know this. Google OR improves existing solutions by 10%-20% utilization which is incredible.
101.
▲
Mathematical Optimization for Cargo Ships
(research.google)
12 points
by
algo_trader
2y ago
|
1 comments
102.
▲
by
algo_trader
2y ago
Hailo seems to be accelerating (ha! pun!) with funding, hw and sw Suprised by the price tag. The board is probably <$10. 15tops should not cost $50. I wonder whats the price for 1000 of these units. (hailo also has bigger units)
103.
▲
by
algo_trader
2y ago
Great looking widgets and graphs. What tool is used for these ?
104.
▲
by
algo_trader
2y ago
Any good write up on Hailo and similar ? BTW, is it possible to have 10 of these connected to a single board/cpu ?
105.
▲
by
algo_trader
2y ago
> Ocean carriers already optimize Of course. But slow sailing is a co-ordination problem. You need an enforcer and a penalty system. > Sofar Ocean .. does it in a particularly interesting way What is the "particularly interesting
106.
▲
by
algo_trader
2y ago
> export both the coal and the iron ore when shipping finished steel is obviously about half the weight Wait until you hear about 2% lithium ore shipped 5000km for refining ..
107.
▲
by
algo_trader
2y ago
In general, are the per-layer KV caches all independent of each other? If we split a model layer-wise for inference across X GPUs, does the cache split as well?
108.
▲
by
algo_trader
2y ago
Is the Plandex Cloud still active ? I dont see a link to it on the github/website. DO i have to self host right now ? Good luck with this
109.
▲
by
algo_trader
2y ago
Always shocking to see how the sausage is made. Some of these would have to be a. foreign agents and/or b. co-financed by groups connected to extreme middle east organizations.
110.
▲
Activist Groups Trained Students for Months Before Campus Protests
(wsj.com)
3 points
by
algo_trader
2y ago
|
1 comments
111.
▲
by
algo_trader
2y ago
If we had a magical (fast) oracle for grading responses, have people done search/expert iteration for LLMs? Specifically for codegen, i am playing with an iterative interpreter that can quickly (re)evaluate a tree of similar responses
112.
▲
by
algo_trader
2y ago
Try Reinforcement Learning for Combinatorial Optimization: A Survey or a more recent and spirited discussion on reddit.. https://old.reddit.com/r/reinforcementlearning/comments/196i...
113.
▲
by
algo_trader
2y ago
> it might be desirable in the future to have a repository of analyzed code. We did a code verifier back end for a small app store. Were mostly MScs + some Phd consulting. The world is indeed waiting for a CloudOptimizingCompiler (CoCo?
114.
▲
by
algo_trader
2y ago
What is the cheapest/simplest configuration for an agricultural PV-to-always-on-e-motor (1MW)? We can serialize the PVs to get 1000v? And then feed that directly to a (suitable) drive without an inverter? Possibly even a DC motor? Thi
115.
▲
by
algo_trader
3y ago
Somewhat off topic: Is there a similar "magic value card" for low memory (2GB?) 8-bit LLMs? Since memory is the expensive bit, surely there are low cost low memory models?
116.
▲
by
algo_trader
3y ago
> I've invested in over 100 startups through syndicates on AngelList Good on you ;) Are these startups there for the convenience, or are these slightly less connected founders?
117.
▲
by
algo_trader
3y ago
Did you export your subscription or queue from GP? They promised this in Dec23. I was a late comer to podcasts, and i just liked the clean GP design.
118.
▲
by
algo_trader
3y ago
> JetBrains’s remote SSH dev; Does that require a static ip on remote? I have skimmed the Jetbrain FAQ [1] and it says "no relay servers are involved" [1] https://www.jetbrains.com/help/idea/faq-about
119.
▲
by
algo_trader
3y ago
Do you have thoughts on Patras as a less crowded alternative for English-speaking nomads and setting up a remote office? I guess the weather is also sunny and beachy ?
120.
▲
by
algo_trader
3y ago
The website indicates each Tensix core ~1TOP theoretical. So not amazing. It will have to compete on price/memory in the crowded inference space.
More ›