Most models seem very reluctant to do that. (Historically, rendering full 3D was also easier than rendering wireframes. Art imitating life.)
565 karma · joined January 15, 2018
Security, algorithms and algo trading since 19XX..
Most models seem very reluctant to do that. (Historically, rendering full 3D was also easier than rendering wireframes. Art imitating life.)
Do retails get/pass on to you hourly/second-ly aggregate data or what?
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
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, its expensive... )
AlphaZero-type systems were unable to significantly improve over near-optimal solvers such as z3.
And what level of automation do you think can take this x10 ?
For example, speeding up compilation of 100s of similar code candidates, doing the same for eval, etc
For example, if we can cache/re-use the compiler run on previous codegen candidates, to speed up compilation of the next candidate sniplet
If i really wanted 11kV MVDC is this a feasible spec for an industrial site?
Google OR improves existing solutions by 10%-20% utilization which is incredible.
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)
What tool is used for these ?
BTW, is it possible to have 10 of these connected to a single board/cpu ?
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" way? It seems like a "basic" optimizer over known data
Wait until you hear about 2% lithium ore shipped 5000km for refining ..
If we split a model layer-wise for inference across X GPUs, does the cache split as well?
I dont see a link to it on the github/website. DO i have to self host right now ?
Good luck with this
Specifically for codegen, i am playing with an iterative interpreter that can quickly (re)evaluate a tree of similar responses
or a more recent and spirited discussion on reddit..
https://old.reddit.com/r/reinforcementlearning/comments/196i...
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?!) - there are so many possibilities for shared compute, library caching, and machine learning.
I will PM you, just to keep the dream alive.
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?
This cuts out half the components compared to an EV drive train, since we have much simpler cooling/packaging/response demand?
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?
Good on you ;)
Are these startups there for the convenience, or are these slightly less connected founders?
I was a late comer to podcasts, and i just liked the clean GP design.
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-remote-develop...
I guess the weather is also sunny and beachy ?
It will have to compete on price/memory in the crowded inference space.